Research: The Platform for Innovation, Competitiveness and Growth by T Carter Comms

Research: The Platform for Innovation, Competitiveness and Growth A Compilation of Working Papers by OECS Scholars

Copyright © 2020 All rights reserved. No part of this publication may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by reviewers, who may quote brief passages in a review. ISBN (978-976-635-175-5) (Print edition) ISBN (978-976-635-174-8) (ebook/Electronic edition) Research provided by the Ministry of Finance, Ministry of Education, Ministry of Economic Development, National Integrated Planning & Programme Unit, Department of Sustainable Development, Government of Saint Lucia; Sir Arthur Lewis Community College; the Eastern Caribbean Central Bank; and the Caribbean Development Bank. Designed by Tahira Carter, Senior Communications Specialist, OECS Commission Published by the Communications Unit of the OECS Commission media@oecs.int Visit www.oecs.int

Research: The Platform for Innovation, Competitiveness and Growth A Compilation of Working Papers by OECS Scholars

Contents 6

Forewords

An Investigation of Commercial Bank Deposits and Credit in Saint Lucia

Investigating the Development Impact of Inward Migrant Remittances on the Economies of the OECS

The Impact of VAT Reduction on Food Prices in Saint Lucia

Determinants of the Use of Electronic Payments in the ECCU - Panel Data Evidence

Assessing the Long Term Planning Requirements for the Water Sector in Saint Lucia

100

Climate Change and its Impact on the Agriculture Sector in Saint Lucia

120

Does the Price of Fuel in Saint Lucia Mimic International Fuel Price Developments?

140

An Investigation of Public Sector Project Implementation in Saint Lucia

154

Taxes and Demand for Intra-Regional Travel

166

Male Academic Underachievement in Tertiary Education

186

The Effects of Tactile Learning Strategies on Attitudes of Form 4 CCSLC Mathematic Students

210

Measuring Vulnerability: A Multidimensional Vulnerability index for the Caribbean

246

To Have or Not to Have Private Health Insurance Coverage?

268

The Impact of Motivation on Implementing Innovation in Organization

Forewords Dr. Didacus Jules Director General, Organisation of Eastern Caribbean States

Developing a Culture of Research in Policy Formulation in the OECS The Pillars of Regional Integration in the Organisation of Eastern Caribbean States include: ● The free movement of people (and all attendant rights associated with this free movement) ● The free circulation of goods and capital within an Economic Union ● The free exchange of services When the Research & Policy Unit of the Ministry of Finance of Saint Lucia invited the OECS Commission to participate in their Research Symposium, we not only embraced the idea but were thrilled to learn that research was being given that level of encouragement and visibility. Even more exciting was the fact that many of the papers being presented in this Symposium (contained in this publication) were highly relevant to discussions, programmes or initiatives being contemplated or executed by the OECS at a regional level. Our approach to creating new initiatives invariably – especially since the COVID-19 pandemic – involves the establishment of virtual Working Groups drawing on the best expertise in the OECS and its Diaspora to shape these initiatives. This approach ensures that the agenda is rooted in a shared vision, that stakeholders are integrally engaged from the inception and that our chances of successful implementation are greater. We proposed to the Saint Lucia Research & Policy Unit that we use their initiative as an OECS best practice and work jointly to encourage similar work in all of the Member States of the OECS so that we create a capacity – not just within Ministries of Finance or Planning, but a national capacity – for research and data driven policy formulation. The publication of this body of research is only the first step in that direction. The objective is to work towards the establishment of an OECS Research Network that will facilitate: ● Multi-island collaboration on research relevant to the developmental urgencies of Member States ● Exchanges of expertise within the OECS Member States ● Collaboration with Caribbean professionals working in multilateral and international institutions (joint authoring, peer review, mentoring) ● Seeking research attachments and internships with institutions, universities, and agencies relevant to the research agenda ● Annual publication of the research findings ● Presentation and incorporation of this research in the design of national and regional initiatives We are excited about this collaboration agenda and urge professionals in the public and the private sectors, graduate students and experts in the Diaspora to join us in this exhilarating journey.

6 | Research: The Platform for Innovation, Competitiveness and Growth

Dr. Veronica Simon Head, University of the West Indies Open Campus, Saint Lucia

"In small island states such as ours, with very limited financial resources, we cannot afford to continue to create experimental policy." The Research and Policy Unit of the Department of Finance must be commended on the publication of its first working paper series in partnership with the Organisation of Eastern Caribbean States (OECS). This is indeed a significant milestone in the unfolding of what began in 2017 as a forum for exposing and encouraging the investigation and discussion of pertinent local issues through annual symposia. This initiative to document formally, the rich information emanating from the symposia, is critical to establishing a solid reservoir of information that can be easily accessed not only by scholars, but most importantly, by local policy makers. As I noted in my Keynote address at the Unit’s 2018 symposium, in small island states such as ours, with very limited financial resources, we cannot afford to continue to create experimental policy. Research is intrinsically connected to development by creating the setting for change; therefore, we need to construct strong, clearly defined research frameworks which buttress the policy implementation process and facilitate lasting change. Such frameworks can only be developed out of a deep understanding of the interplay and dialogue on the ground. One must identify, establish and validate the shape of one's own space in order to facilitate meaningful philosophical approaches to development. Small island developing states, in particular, must consider the relevance of policy to their idiosyncratic environments. We should be ever mindful of the need to maintain a decolonising agenda which would reduce intellectual dependency and influence our ability to negotiate more effectively with external agencies. Strengthening local research through documentation such as this publication, provides the discursive space within which we can negotiate projects that can aptly support the implementation of grounded policies for growth and development in a small island context. Indigenous research continues the ongoing decolonising process that countries like ours must accelerate. It establishes that the stories told from this part of the globe are of no less value to international discussion than those from any other geographical location. Useful policy can only be informed from the ground up and ought not to be constructed entirely from a theoretical space devoid of locally informed narrative. Hence, a publication such as this one, is critical to the in-depth clarification of issues arising from a national context, even while relating and contributing to the global discourse. Research is an essential part of life, whether social, political, academic, or simply every day. It informs decisions, actions, philosophies and attitudes; therefore, research which illuminates one’s own context is invaluable. This publication is the first step towards building a comprehensive collection of OECS research and certainly paves the way for the compilations to follow.

A Compilation of Working Papers by OECS Scholars

Embert St Juste Former Head of the Research and Policy Unit, Ministry of Finance Government of Saint Lucia

The efforts of the Research and Policy Unit of the Department of Finance, Saint Lucia, and the Commission of the Organisation of Eastern Caribbean States (OECS) in publishing the first working paper series must be commended, particularly given the disruptive effects of COVID-19. As the former head of the unit for almost 9 years, I have always championed the need for collaborative policy-based research spanning various fields, but with economics and finance at the core. I am of the firm belief that evidenced based research is key to policy making and that policy makers should insist that key policy decisions should be based on proper empirical analysis and research, given what is at stake. On this basis, as head of the Unit, I insisted that the staff be assigned responsibilities for completing at least one research paper as part of the unit’s annual work plan. Collaboration both within the Unit, across departments and agencies and even externally were encouraged. The results of our efforts were galvanized when we made regular presentations of policy briefs to the Economic Policy Committee, chaired by the Permanent Secretary, in 2014 and 2015 on various policy initiatives. A myriad of policy-based topics was presented ranging from ideas on how to develop the cocoa industry in Saint Lucia to options for reducing the fiscal deficit. The policy briefs were imbued with practical ideas focusing on workable solutions. After each presentation, vigorous discussions followed, not only critically analyzing the paper but also positive ideas were offered on further improving the paper. In 2016, the Unit was invited by the UWI Open Campus to participate in a research symposium, in which a number of presentations were made including by the staff of the Research and Policy Unit. Our research initiatives were exposed even more when the staff were invited to participate in the Central Bank of Barbados Annual Research Workshop in 2016 and 2017. It should be noted that the Saint Lucia Department of Finance was the only finance department represented. The promotion of productivity through hosting of the annual Productivity Awareness Week has provided an excellent platform for the Unit to showcase its research activities. The collaboration with the National Productivity Council began in 2018 and ever since, a wide cross section of research topics has been presented to the public, policy makers, students and other stakeholders. This year’s theme “Building Economic and Social Resilience through Digital Transformation”, was aptly chosen given the structural reforms that are urgently required to mitigate the long-term effects of COVID-19 on the economic and social fabric of our country. While it is true that COVID has disrupted many lives and has setback the economic and social achievements of many countries globally, greater opportunities have emerged to leverage the investments made in digital technologies to transform the way we work, through greater efficiencies.

8 | Research: The Platform for Innovation, Competitiveness and Growth

We in Saint Lucia have made important strides in making use of technology for improving the operations of the public service, for example online payment of government services. The COVID experience should propel us to further accelerate the digital transformation, not only to avoid human transmission of the virus (e.g avoiding queuing for renewal of drivers’ licenses), but also to boost productivity in the public and private sectors. The pandemic has exposed the underlying vulnerabilities of the structure of our economy and should provide an opportunity to develop appropriate strategies to strengthen resilience by further advancing our technological transformation. Our education system is one case in point and needs urgent reforms to revolutionize the way our children learn. The need to achieve better educational outcomes in keeping with the requirements of the global economy has been recognized and our digital transformation initiatives must be congruent with this imperative. I wish to encourage our young researchers to continue undertaking research in their various disciplines and to seek exposure for their work with a view to further improvement. This is a very useful way to learn and grow as young professionals. The publication of the first working paper series is a very significant milestone in the quest for continued advancement of our research efforts. I commend this collaborative effort of the Department of Finance and the OECS Commission and wish a successful launch of this working paper series.

A Compilation of Working Papers by OECS Scholars

Abbreviations AIC

Akaike Information Criteria

ARDL

Autoregressive Distributed Lag

BMCs

Borrowing Member Countries

CCSLC

Caribbean Certificate of Secondary Level Competence

CDB

Caribbean Development Bank

CDP

Constituency Development Programme

CIF

Cost, Insurance and Freight

CPI

Consumer Price Index

CSEC

Caribbean Secondary Education Certificate

CSO

Central Statistical Office

CVSS

Common Vulnerability Scoring System

CXC

Caribbean Examination Council

DCVM

Dara Climate Vulnerability Monitor

ECCB

Eastern Caribbean Central Bank

ECCU

Eastern Caribbean Currency Union

ECLAC

Economic Commission for Latin America and the Caribbean

EM-DAT

Emergency Events Database

EVI

Economic Vulnerability Index

FAO

Food and Agriculture Organization of the United Nations

FDI

Foreign Direct Investment

10 | Research: The Platform for Innovation, Competitiveness and Growth

FPE

Final Prediction Error

FGLS

Feasible Generalised Least Squares

FSRA

Financial Services Regulatory Authority

GDP

Gross Domestic Product

GLS

Generalized Least Squares

GOSL

Government of Saint Lucia

IICA

Inter-American Institute for Cooperation on Agriculture

ITRC

Infrastructure Transitions Research Consortium

ICR

Implementation Completion Report

IMF

International Monetary Fund

LCC

Low-Cost Carrier

LIAT

Leeward Islands Air Transport

LPG

Liquefied Petroleum Gas

MCP

Mean Caribbean Posting

MTEF

Medium-Term Expenditure Framework

MVI

Multidimensional Vulnerability Index

NAO

National Authorization Office

NHS

National Health Service

NIPP

National Integrating and Planning

NISMOD

National Infrastructure Systems Model

OECS

Organisation of Eastern Caribbean States

OLS

Ordinary Least Squares

A Compilation of Working Papers by OECS Scholars

PCA

Principal Component Analysis

PCU

Project Coordination Unit

PHI

Private Health Insurance

PMCC

Pearson's Product Moment Correlation Coefficient

PMDU

Project Management and Delivery Unit

POS

Point of Sale

RGDP

Real Gross Domestic Product

SAMOA

SIDS Accelerated Modalities of Action

SAP

Structural Adjustment Programme

Schwarz Information Criteria

SDF

Special Development Fund

SDG

Sustainable Development Goals

SIDS

Small Island Developing States

SISI

Strategic Imports Sub-Index

SITC

Standard International Trade Classification

TAM

Technology Acceptance Models

TFCs

Taxes, Fees and Charges

ULG

Unleaded Gasoline

United Nations

UNCDP

United Nations Committee for Development Policy

UNCTAD

United Nations Conference on Trade and Development

UNDESA

United Nations Department of Economic and Social Affairs

12 | Research: The Platform for Innovation, Competitiveness and Growth

UNDP

United Nations Development Programme

UNOHRLLS -

United Nations Office of the High Representative for Least Developed Countries, Landlocked Developed Countries and Small Island Developing States

VAT

Value Added Tax

VRCP

Vulnerability and Resilience Country Profile

World Bank

WEO

World Economic Outlook

WHO

Worls Health Organization

WTI

West Texas Intermediate

A Compilation of Working Papers by OECS Scholars

1 An Investigation of Commercial Bank Deposits and Credit in Saint Lucia Kimbert Evans

14 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT This paper empirically examines the determinants of commercial bank deposits in Saint Lucia using data over the period 1995-2018 to identify the factors which have been contributing to Saint Lucia’s constant deposit growth. Using the ARDL bounds test approach to assess cointegration, results indicate that in the short-run, GDP and inflation rates are cointegrated with deposits rates while in the long-run no cointegration exists between the independent variables and deposits. Employing the Toda-Yamamoto test to assess causality among deposits and the other variables, results indicate that in the long-run, corporate income tax, GDP, inflation, remittances and interest rates all positively and significantly influence deposit rates in Saint Lucia. These findings suggest that in the short-run, Saint Lucian policy makers can influence deposit rates through growth enhancing policies. However, in the long-run policies aimed at attracting remittances as a special interest rate on bank deposits for persons who live abroad or policies aimed at encouraging investments with remitted funds may be particularly important for the Saint Lucian economy. This study has implications for development and formulation of monetary policy for Saint Lucia and similar small open economies. INTRODUCTION Saint Lucia’s financial sector, as governed and regulated by the Eastern Caribbean Currency Union’s (ECCU)1 central bank, and the Financial Services Regulatory Authority (FSRA) comprises of five commercial banks2. In addition, non-banking financial institutions including credit unions, offshore banks, insurance companies and development banks have been established since the financial inclusion thrust. This development of the financial sector in Saint Lucia, has resulted in continued growth in commercial bank deposits. Saint Lucia’s commercial bank’s stock of deposits moved from EC$2.2 billion in 2004 to EC$3.7 billion in 2012 and a further increase to EC$4.3 billion in 2018 representing an average growth of EC$145 million per annum. Bank credit also grew from EC$1.8 billion in 2004 to EC$4.4 billion in 2012 then fell to EC$3.4 billion in 2018, average growth of EC$117 million per annum (see Figure A of Appendix). The continued strong deposit growth relative to the slow or falling credit growth is hypothesized to pose a multitude of concerns relating to bank profitability and sustainability. In the international literature, deposit is believed to be a function of two main factors (see Athukorala and Sen, 2001; Finger and Hesse, 2009 and Ferrouhi and Lehadiri, 2014); an individual’s ability to have money to deposit and an individual’s perception of the future. However, an individual’s ability to deposit money is dependent on the individual’s wealth and economic fundamentals. Proxies for wealth and economic fundamentals are economic activity as measured by real GDP and remittances as measured by western union and money gram inflows. An individual’s perception of the future, the main determinant of macroeconomic stability is proxied by inflation. In addition, the extent to which firms are profitable also influence how much money persons and companies have and therefore the ability to deposit. A proxy for profitability of firms is corporate income tax. The available literature also speaks to branch network, bank density, per capital income as other factors that influence saving growth. However, the available literature does not shed much light on the determinants of bank deposits in Small Island Developing States (SIDS) nor does the statistics highlight the specific linkages between deposits and credit demand. The first part of this paper therefore aims to fill the gap in the literature by empirically investigating commercial bank deposits in Saint Lucia. Understanding the dynamics of commercial bank deposit growth will be useful for the development and formulation of policy in Saint Lucia and similar small open economies. This paper is divided into ten sections. Section 2 provides the motivation and objectives of this empirical investigation. Section 3 reviews the existing literature. Section 4 provides the descriptive statistics. Section 5 presents a comparison of the variables, section 6 describes the data, section 7 outlines the model, and section 8 presents the methodology. Section 9 outlines and discusses the results of the paper. Finally, section 10 provides conclusions and policy recommendations.

A Compilation of Working Papers by OECS Scholars

OBJECTIVES / MOTIVATION The general objective of this study is to investigate commercial bank deposit growth in Saint Lucia. Specific objectives of the study are to: I. Identify and estimate a model to determine the factors which contribute to commercial banks deposit growth in Saint Lucia; II.

Evaluate the magnitude of each determinant of deposit growth in Saint Lucia; and,

III.

Draw relevant policy recommendations.

LITERATURE REVIEW Many studies exist on the determinants of commercial bank deposits and the implications of increasing deposits in the banking system in both developed and developing countries. Early researchers such as Soyode and Oyejide (1975) used stepwise multiple regression technique to examine the factors which influence Commercial bank deposits in Nigeria. The findings suggested a positive and significant relationship between branch network3 and savings growth in Commercial banks in Nigeria therefore highlighting the importance of branch network in understanding deposit growth in Nigeria. Other researchers such as Srinivasan and Meyer (1986) assessed deposit growth in countries like India, Pakistan, Nepal and Sri Lanka using Generalized Least Squares (GLS) technique. Using this alternative approach, the results suggests that per capita income, bank density and real rate of interest were the main determinants of increasing deposits in those countries. The findings also suggest that individuals would prefer bank deposits relative to other financial assets or physical assets as gold or inventory if the interest on bank deposits is sufficiently attractive. Further to the work of Srinivasan and Meyer (1986), findings from the research by Athukorala and Sen (2001) states that favourable interest rates, rate of economic growth, spread of banking facilities and inflation were positively correlated with increasing deposit rates. More recent studies as Lomuto (2008), used Ordinary Least Squares (OLS) technique and time series data from 1968-2006 to identify the determinants of Kenyan Commercial bank deposit growth. The results points to lagged commercial bank deposits, deposit rates, nominal exchange rate, investment income ratio, number of cheques cleared, real GDP, ratio of monetary GDP to total GDP and Structural Adjustment Programmes (SAPs) all significantly influencing Commercial bank deposit growth in Kenya. Applying the same modelling technique using different factors, Adem (2008) and Ferroukhi (2017) assessed the determinants of bank deposits in Ethiopia and Morocco respectively. Their empirical findings suggest that branch opening and individual remittances influenced Commercial Bank of Ethiopia’s deposit growth while bank size and interest rate on deposits determined bank deposit growth in Morocco. Therefore, in the international literature, rate of economic growth, personal remittances, inflation and interest rates are main determinants4 of deposit growth in developed and advanced economies. Accordingly, it is important to ascertain the main determinants for deposit growth in Small Island Developing States (SIDS) as Saint Lucia. Deposits in Saint Lucia have been on an upward trajectory for the past decade and a half with an average growth of EC$145 million per annum. This paper therefore attempts to investigate the factors which influences deposit growth in Saint Lucia using the Autoregressive Distributive Lag (ARDL) bounds test. Following the approach by Chaudhry et al. (2014) and (Pitonakova (2016), the paper uses the (ARDL) method and contributes to the existing literature, by utilizing an approach which is rigorous and able to explicitly model the changes in variance over time in the assessed variables.

16 | Research: The Platform for Innovation, Competitiveness and Growth

DISCRIPTIVE STATISTICS During the period 1995-2000 Saint Lucia was at a distinct advantage as the economy was largely dependent on the agriculture sector as their main income earner with preferential access to the European Union for sale of bananas, their main agricultural produce. Therefore, during this period the economy experienced high economic growth resulting in the increase in disposable income and commercial bank deposits. Data shows that during this period real GDP growth rates averaged 2.0 percent while commercial bank deposits averaged EC$1.4 billion. while bank loans averaged EC$1.3 billion. Additionally, inflation rate averaged 1.0 percent and remittances averaged EC$63.0 million. Growth rates of both inflation and remittances were consistent at 1.0 percent. The end of the preferential period brought about a decline in the performance of the agriculture sector and the introduction of a more service-based economy. This period also brought about increases in foreign direct investment from an average of EC$0.72 million between 1995 and 2000 to an average of EC$1.18 million in the period 2001 to 2006. Average loans and real GDP increased to EC$1.9 billion and EC$2.7 billion respectively with average growth rates of 8.0 and 2.0 percent respectively. Personal remittances and CIT also increased to EC$69.76 million and EC$52.99 million respectively with growth rates of 3.0 and negative 2.0 percent, respectively. It is expected that this growth in the economy will cause an increase in commercial bank deposits and the data shows that commercial bank deposits did increase to an average of EC$2.2 billion with an average growth rate of 8.0 percent. The Saint Lucian economy in the period 2007 to 2012 and 2013 to 2018 experienced a recession and low growth but average deposit levels continued its upward trajectory to EC$3.4 billion and EC$3.9 billion respectively. Average loans and CIT were affected therefore decreasing from EC$3.9 billion to EC$3.7 billion and from EC$94.39 million to EC$78.18 million. Growth rates for all variables decreased or remained constant but growth of commercial bank deposits in Saint Lucia continued upward. DATA COMPARISON FIGURE 1: GROWTH RATES OF DEPOSITS AND REAL GDP

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

A Compilation of Working Papers by OECS Scholars

The existing literature points to the expectation that as GDP increases deposit rates will rise as increased activity will result in circulation of funds in the economy. Accordingly, for the period 1995 to 2018, growth rates of deposits and real GDP seem to have followed a one-to-one relationship with a lag. As deposits increased from 1997 to 1998, real GDP also increased but with a small lag. As deposits fell from 1998 to 2001, real GDP also fell but with a larger lag. This one-to one relationship continued till the end of 2018 but with slight variations between the variables for the differ-nt years. FIGURE 2: GROWTH RATES OF DEPOSITS AND PERSONAL REMITTANCES

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

In addition, the empirical studies have suggested that as personal remittances increase commercial bank deposits are expected to increase. Data shows that growth rates of deposits and personal remittances followed a one-to-one relationship. As deposits grew from 1997 to 1998, personal remittances also rose and as deposits fell immediately thereafter from 1999 to 2000, personal remittances also fell. For the period 2002 to 2006, commercial bank deposits increased sharply and personal remittances also increased but at a slower rate. This trend is consistent with economic theory which states that as remittances increase commercial bank deposits will increase.

FIGURE 3: GROWTH RATES OF DEPOSITS AND CORPORATE INCOME TAX

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

18 | Research: The Platform for Innovation, Competitiveness and Growth

Growth rates of deposits and corporate income tax, on the other hand, did not follow that one-to-one relationship. For the period 1999 to 2018, growth rates of deposits remained relatively flat while growth rates of Corporate Income tax fluctuated.

FIGURE 4: GROWTH RATES OF DEPOSITS AND SAVINGS RATE

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

The available literature on deposit growth states that as interest rates decrease, deposits rates will fall as persons may be inclined to invest money in other financial instruments. Data shows that the savings rate moved from 4.0 percent in 2000 to 2.0 percent in 2018. The deposit rate however, has been fluctuating moving from 6.0 percent in 2000 to 3.0 percent in 2018. Therefore, for the last nineteen years the savings rate in Saint Lucia has been decreasing however deposit interest rates have been fluctuating. FIGURE 5: GROWTH RATES OF DEPOSITS AND INFLATION

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

A Compilation of Working Papers by OECS Scholars

Economic theory states that as inflation increases, persons will choose to either reduce consumption and increase savings or maintain consumption and spend more therefore reducing savings. Therefore, inflation is expected to significantly affect deposit growth. Data shows that for the period 1995 to 1997, growth rates of both deposits and inflation in Saint Lucia decreased with inflation falling at a quicker rate with a lag. For the period 2002 to 2003, growth rates of deposits and inflation moved in opposite directions, deposits increased but inflation decreased.

FIGURE 6: GROWTH RATES OF DEPOSITS AND LOANS

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

It is expected that as deposit rates increase, loan rates will also increase. Data shows that growth rates of deposits and loans followed a one-to-one relationship with a lag. For the period 1995 to 1996, deposits decreased slightly from 9.0 percent to 6.0 percent and loan rates also increased slightly from 9.0 percent to 11.0 percent. Thereafter, from 1998 to 2002, deposit growth decreased from 13.0 percent to 1.0 percent and loan growth also decreased but from 1999 to 2003 from 9.0 percent to negative 4.0 percent. Subsequently, both deposits and loans fluctuated till the end of the period. DATA This research employs annual data spanning from 1995 to 2018 sourced from the ECCB. It has been suggested empirically, that three thematic areas influence deposit growth in commercial banks; the state of the economy, the stability of the economy and monetary policy. Therefore, based on the existing literature and on data availability, real gross domestic product (GDP), corporate income tax and personal remittances were selected as proxies for the state of the economy, inflation rate as the proxy for the stability of the economy and interest rate as the proxy for monetary policy. Commercial bank deposits were used as the dependent variable and real GDP, inflation rate, interest rate, personal remittances and corporate income tax as our explanatory variables. All variables were converted to natural logarithms therefore all variables can be interpreted in growth terms. MODEL SPECIFICATION To assess cointegration among our variables, the autoregressive distributed lag (ARDL) method was selected as the principal approach. This ARDL method, as compared to other cointegration techniques, does not require the series to be of the same order of integration. The ARDL approach allows the regressors to be stationary at level I(0) or at first difference I(1). The ARDL equation is specified as follows; Equation 1

20 | Research: The Platform for Innovation, Competitiveness and Growth

where: DEP is commercial bank deposits in natural logarithms; RGDP is real gross domestic product in natural logarithms; INF is inflation rate in natural logarithms; IR is interest rate in natural logarithms; REM is personal remittances in natural logarithms; CIT is corporate income tax in natural logarithms; . represents the coefficients on the respective variables; and, ε represents the residuals. If DEP increases with RGDP, INF, IR, REM or CIT, then is expected to be significant and positive. DATA To assess cointegration, long-run relationships and causality among the variables, this paper uses a three-step approach. The first step is the stationarity test using the Augmented Dickey Fuller (ADF) test (1979), Phillips-Perron (PP) test (1988) and the Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test (1992). The second step is the test for cointegration using the bounds test and if there is cointegration among the variables the third step will be to test for causality using the Toda-Yamamoto (T-Y) test. COINTEGRATION To test for cointegration among the variables, this paper uses the ARDL bounds test as introduced by Pesaran and Smith (1995) and further developed by Pesaran et al. (2001). According to Pesaran et al. (2001) and further supported by Nkoro and Uko (2016), the ARDL bounds test is preferable when dealing with variables which are integrated of different orders5 and is robust when applied to small sample sizes6. Another advantage of the ARDL model is its ability to simultaneously estimate both the short run and the long run coefficients (Sami and Kreishan, 2012). Consequently, this study therefore estimates the following equation:

Equation 2

CAUSALITY To assess long-run causality, this study adopts the T-Y approach. According to Kumar (2014), this approach is applicable regardless of the order of integration of the variables or whether or not there is cointegration among the variables. In the first step, the optimal lag length of the VAR is estimated, then the VAR is segmented with d additional lags. In the final step, the equation below is estimated and examined for significance of the coefficients, exclusive of the additional lags, to estimate causality from one variable to the next. For example, to determine whether GDP influences DEP, we test the null hypothesis H0: all ϕ1i=0 for all i = 1…k. This statistic follows a X2 distribution.

Equation 3

A Compilation of Working Papers by OECS Scholars

22 | Research: The Platform for Innovation, Competitiveness and Growth

EMPIRICAL RESULTS AND DISCUSSION STATIONARY TESTS The first step of this empirical analysis is to assess the stationary properties of all variables to determine whether they are all of the same order of integration or whether they are of mixed orders of integration. Specifically, for the ARDL bounds test, variables should not be integrated of order I(2)7. This stationary test is done using the Augmented Dickey-Fuller (ADF) test (1979), the Phillips-Perron (PP) test (1988) and the Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test (1992) and presented in Table 1 below. TABLE 1: UNIT ROOT TESTS AT LEVEL AND 1ST DIFFERENCE OF VARIABLES Variables

ADF

KPSS

Results

Deposits

Level -2.26

Level -2.54

Level 0.69

I(1)

1st Difference -3.33*

1st Difference -3.32*

1st Difference 0.08*

Level -1.36

Level 1.36

Level 0.51

1st Difference -4.12***

1st Difference -4.13***

1st Difference 0.06***

Level 0.94

Level -0.94

Level 0.69

1st Difference -4.52***

1st Difference -0.08***

Inflation rate

Level -4.27***

Level 0.27***

I(0)

Remittances

Level -4.38***

Level -1.59

Level 0.66

I(1)

1st Difference -2.54

1st Difference -4.38***

1st Difference 0.09***

Level -0.56

Level -0.51

Level 0.62

1st Difference -4.93***

1st Difference -4.95 ***

1st Difference 0.07 ***

Corporate Income Tax GDP

Savings rate

I(1)

Note: The null hypothesis for ADF and PP is variable has a unit root (non-stationary) and the null hypothesis for KPSS is variable is stationary. *** and * denote significance at 1% and 10% respectively. Source: Authors’ estimation.

The results indicate that only inflation is stationary at level terms which suggest that it is integrated of order I(0) and deposits, corporate income tax, GDP, remittances and interest rate are stationary after first differencing (integrated of order I(1)). Since none of the variables are I(2) or higher, valid inferences can be made from the ARDL bounds test. CO-INTEGRATION TESTS The ARDL bounds test uses lag lengths to determine the optimal lag structure of the ARDL model in equation (2) above. Therefore, a maximum of 1 lag was inputted into the model and the model automatically selected optimal lags of 1 for DEP and GDP and 0 for CIT, INF, REM and IR. In all cases, as supported by Akaike Information Criteria (AIC), the automatically selected optimal lag lengths ruled out the existence of non-normality, serial correlation, heteroskedasticity and instability in the model. Using these optimal lag lengths, the ARDL bounds test results are presented in table 2.

A Compilation of Working Papers by OECS Scholars

TABLE 2: ARDL BOUNDS RESULTS FOR COINTEGRATION

7.06***

Critical Values at 1% Level of Significance

Critical Values at 5% Level of Significance

Critical Values at 10% Level of Significance

I (0)

I (1)

I (0)

I (1)

I (0)

I (1)

3.29

4.37

2.56

3.49

2.20

3.09

Note: *** denotes significance level at 1%. Source: Authors’ estimation.

The results show that the computed F-statistic (7.06) is greater than the critical values at 1%, 5% and 10% respectively. Therefore, this implies that there is cointegration among the series in the model. The existence of cointegration aids in analysing the short-run and long-run relationship of the factors which are significant estimates for growth of deposits in Saint Lucia. FACTORS AFFECTING GROWTH OF DEPOSITS Based on automatically selected ARDL (1, 0, 0, 1, 0, 0) model8, the results of the short-run and longrun estimates which affect growth of deposits in Saint Lucia are presented in table 3 and 4 below. Accordingly, inflation and GDP were revealed as significant in the short-run but insignificant in the long run. Conversely, corporate income tax, remittances and savings rate were revealed as insignificant in both the short and long run. TABLE 3: SHORT-RUN ESTIMATES Variable

Coefficient

D(LNCIT)

-0.007

D(LNINF)

Std. Error

t-Statistic

Prob.

0.027

-0.249

0.807

1.105

0.236

4.675

0.000***

D(LNGDP)

0.527

0.196

2.694

0.017**

D(LNREM)

0.040

0.085

0.469

0.646

D(LNIR)

-0.088

0.061

-1.459

0.165

CointEq(-1)

-0.082

0.010

-8.446

0.000***

Note: ***, * denotes significance level at 1% and 10% respectively. Source: Authors’ estimation. TABLE 4: LONG-RUN ESTIMATES Variable

Coefficient

Std. Error

t-Statistic

Prob.

LNCIT

-0.372

0.546

-0.681

0.506

LNINF

13.897

16.279

0.854

0.407

LNGDP

0.739

3.289

0.225

0.826

LNREM

0.030

1.361

0.022

0.982

LNIR

-1.184

1.319

-0.897

0.383

7.820

65.522

0.119

0.907

Note: Authors’ estimation. 24 | Research: The Platform for Innovation, Competitiveness and Growth

The results show that none of the variables9 are significant10 in the long run. Therefore, according to the ARDL model, in the long-run, neither corporate income tax, inflation, GDP, remittances or savings rate are suitable estimates to determine growth of deposits in Saint Lucia. However, the ARDL model states that in the short run, a 1% change in inflation causes a 1.10% increase in deposits. Supporting this study’s results, is Alfaro et al. (2004) and Athukorala and Sen (2004) which indicated that inflation rate has a significant and positive impact on deposit growth in Turkey11 and India, respectively. According to the ARDL results, real GDP is another important variable which significantly affects deposit growth in Saint Lucia in the short-run. Its partial elasticity is 0.53 indicating that a 1% change in real GDP, in the short-run, causes deposits to increase by 0.53%. This result confirmed that as real GDP of Saint Lucia grows, individuals will have access to more money therefore they will deposit more. This positive relationship between real GDP and deposits is consistent with the empirical works of Athukorala and Sen (2001) and Lomuto (2008). Other variables such as corporate income tax and interest rates on bank deposits generated negative coefficients of 0.007 and 0.088 respectively, suggesting that corporate income tax and interest rates may negatively affect deposit growth in Saint Lucia however the probabilities for both variables were greater than 0.05 indicating that those probabilities are insignificant. Therefore, according to the ARDL results above, it can be concluded that in the short-run, corporate income tax and interest rates are not suitable estimates for deposit growth in Saint Lucia12. Conversely, the results for remittances were positive but also insignificant, indicating that in the short-run, similar to corporate income tax and interest rates, remittances are not significant estimates of deposit growth in Saint Lucia. Supporting this result is Hassan et al. (2016), which concluded that interest rate does have insignificant effects on commercial banks deposits in Nigeria and Mushtaq and Siddiqui (2017) which also concluded that interest rate has an insignificant effect on savings in Islamic countries. The error correction term, as shown in Table 3 above (CointEq (-1)) was negative and significant therefore confirming the existence of cointegration among the variables in the model. The negative sign indicates a move back towards equilibrium if there is a shock in the system and the coefficient value of 0.082 showed that in the short run deviations from the long-run equilibrium are corrected at 0.08% every year. Finally, the stability test results of the cumulative sum of recursive residuals (CUSUM) and the cumulative sum of squares of recursive residuals (CUSUMSQ) were analysed for evidence of structural breaks in the model. The results, as shown in Figure 7 and Figure 8 below, illustrate that the CUSUM and CUSUMSQ estimate lie in the 95% confidence interval indicating that the model was correctly specified and stable. This provides evidence that structural breaks are not present and long run estimates sufficiently capture the long-run relationship among the variables. FIGURE 7: CUMULATIVE SUM (CUSUM)

Note: Authors’ estimation. A Compilation of Working Papers by OECS Scholars

FIGURE 8: CUMULATIVE SUM OF SQUARES (CUSUM)

Note: Authors’ estimation.

TODA – YAMAMOTO CAUSALITY TEST According to the unit root tests, the maximum order of integration among the variables is one (d=1). The Final Prediction Error (FPE), Akaike Information Criteria (AIC), Schwarz Information Criteria (SC) and Hannan-Quinn Information Criteria (HQ) also states that the optimal lag length is 2 (k=2). Therefore, equation three (3) is estimated with three (3) lags. The T-Y causality results are presented in table 5 below. The results indicate that all variables are statistically significant. Accordingly, all variables from the three thematic areas13 positively contribute to growth of deposits in Saint Lucia.

TABLE 5: T-Y CAUSALITY TEST Variable

Chi-sq

LNCIT

26.117

0.000***

LNGDP

19.001

0.000***

LNINF

32.072

0.000***

LNREM

27.356

0.000***

LNSR

25.269

0.000***

Note: *** indicate significance at the 1%. Source: Authors’ estimation.

26 | Research: The Platform for Innovation, Competitiveness and Growth

Prob.

CONCLUSION AND POLICY RECOMMENDATIONS This paper examined the factors which influence deposit growth in Saint Lucia for the period of 19952018. The linkages between deposits and other variables in Saint Lucia is important for policy makers to understand what has been the driving force of the constant growth in Saint Lucia deposit. For this purpose, the data was first examined to determine whether the series had unit root. Then a co-integration analysis was done followed by a causality test. The results of the various stationary tests showed the variables were integrated of order I(0) and I(1), therefore, the co-integration relationships were examined using the recently developed ARDL model and the causality among the variables was examined using the T-Y approach. Overall, the findings of this paper indicate that in the short-run GDP and inflation have positive and significant impact on deposit growth in Saint Lucia. Therefore, GDP and inflation are significant shortrun estimates for deposit growth. This indicates that in the short-run, as GDP and inflation increases, deposits at commercial banks will also increase. However, in the long run, none of the variables proved to be significant estimates of deposit growth in Saint Lucia. This indicates that, in the long-run, an increase in GDP can cause persons to have more disposable income but will not cause an increase in deposits. Similarly, in the long-run, growth in CIT, inflation, remittances and interest rates on deposits will not cause growth in deposit rates. Conversely, the T-Y approach, an alternative approach in assessing the long run estimates, suggested that in the long-run all variables played a significant role in contributing to the constant deposit growth in Saint Lucia. Indeed, results of this study do provide supportive evidence that the state of the economy, the stability of the economy and monetary factors are suitable thematic areas to explain the constant deposit growth in Saint Lucia. The findings of this paper have several policy implications. According to Lomuto (2008) a policy measure which increases GDP will ensure a boost in deposit growth. Thus, the government in partnership with the private sector can boost deposit growth through growth enhancing policies as tax holidays and tax rebate incentives for manufacturing firms. Boadi (2015) also suggested that the regulatory supervisors can improve macroeconomic measures like inflation to improve deposit rates. The findings also indicate that Saint Lucia should continue policies aimed at promoting economic growth which influences growth in corporate income tax receipts for higher deposit growth. Accordingly, policies aimed at attracting remittances or policies encouraging investments with remitted funds may be particularly important for the Saint Lucian economy.

NOTES The ECCU is made up of six independent states; Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, Saint Lucia and St. Vincent and the Grenadines. 1

Two out of the five are locally incorporated; 1st National Bank and Bank of Saint Lucia while the other three are registered branches of multinational financial institutions; CIBC First Caribbean International Bank, RBC and Bank of Nova Scotia. 2

Branch network is the dispersion of bank branches to boost availability of bank products and services (Adeyinka, 2013). 3

Those variables were consistent in the international literature.

The variables used in this study are integrated of order 0 and 1.

See also Acaravci and Ozturk (2012) and Gasmi and Laourari (2017).

The existence of I (2) variables in the model will render the F-statistic invalid (Gasmi and Laourari, 2017).

A Compilation of Working Papers by OECS Scholars

The optimal model was automatically selected using the ARDL method.

Corporate Income Tax, Inflation rate, GDP, Remittances and Interest rate.

Significant variables have probability values between 0.00 and 0.10.

Inflation captures the degree of macroeconomic volatility therefore causing a positive impact on deposits (Alfaro et al. 2004). 11

Contrasting this result is the empirical work of Ferrouhi (2017) which indicated that interest rates positively affects deposit rates in Morocco as well as the work of Adem (2008) which indicated that remittances from Ethiopia’s diaspora was one of the main determinants of commercial bank deposits in Ethiopia. 12

The state of the economy, the stability of the economy and monetary policy.

REFERENCES Acaravci, A., & Özturk, İ. (2012). Foreign direct investment, export and economıc growth: empirical evidence from new EU countries. Adem, S. B. (2015). Determinants of Commercial Bank Deposits in Ethiopia: A Case of Commercial Bank of Ethiopia. Degree of Master Thesis, Addis Ababa University. Adeyinka, S. (2013). Capital adequacy and banks’ profitability: An empirical evidence from Nigeria, American International, Journal of Contemporary Research, (3) 10-14. Alfaro, L., Chanda, A., Kalemli-Ozcan, S., & Sayek, S. (2004). FDI and economic growth: the role of local financial markets. Journal of international economics, 64(1), 89-112. Athukorala, P. C., & Sen, K. (2004). The determinants of private saving in India. World Development, 32(3), 491-503. Boadi, E. K., Li, Y., & Lartey, V. C. (2015). Determinants of Bank Deposits in Ghana: Does Interest Rate Liberalization Matters? Modern Economy, 6(09), 990. Chaudhry, I. S., Riaz, U., Farooq, F., & Zulfiqar, S. (2014). The monetary and fiscal determinants of national savings in Pakistan: An empirical evidence from ARDL approach to co-integration. Pakistan Journal of Commerce and Social Sciences (PJCSS), 8(2), 521-539. Dickey, D. A. and W. A. Fuller (1979), “Distribution of the Estimators for Autoregressive time series with a unit root”, Journal of the American Statistical Association, 74(366a), 427-431. Ferrouhi, E. M. (2017). Determinants of Bank Performance in a Developing country: Evidence from Morocco. Organizations & Markets in Emerging Economies, 8(1). Finger, M. H., & Hesse, H. (2009). Lebanon-determinants of commercial bank deposits in a regional financial center (No. 9-195). International Monetary Fund. Gasmi, F. and I. Laourari (2017), “Has Algeria suffered from the Dutch Disease? Evidence from 1960–2013 data”, Toulouse School of Economics (TSE), Working Paper No. 17-780. Hassan, O. M. (2016). Effect of Interest Rate on Commercial Bank Deposits in Nigeria (2000–2013). In First

28 | Research: The Platform for Innovation, Competitiveness and Growth

American Academic Research Conference on Global Business, Economics, Finance and Social Science, New York, NY, USA, May (pp. 25-28). Hendrickson, M., Katojumuise, K., McBain, H., & Pérez, E. (2002, March). Monetary, fiscal policy and economic performance in a monetary union: the case of the Eastern Caribbean Currency Union. In the conference on “Towards Regional Currency Areas”, Santiago, Chile (pp. 26-27). Lomuto, J. K. (2008). “Determinants of Kenyan Commercial Banks Deposit Growth” (Doctoral dissertation, Master‟ s Thesis, University of-Nairobi, Nairobi). Kumar, R.R. (2014), “Exploring the nexus between Tourism, Remittances and Growth in Kenya. Quality & Quantity, 48(3), 1573-1588. Kwiatkowski, D., Phillips, P. C., Schmidt, P., & Shin, Y. (1992). Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root? Journal of econometrics, 54(1-3), 159-178. Mushtaq, S., & Siddiqui, D. A. (2017). Effect of interest rate on bank deposits: Evidence from Islamic and nonIslamic economies. Future Business Journal, 3(1), 1-8. Nkoro, E., & Uko, A. K. (2016). Autoregressive Distributed Lag (ARDL) cointegration technique: application and interpretation. Journal of Statistical and Econometric Methods, 5(4), 63-91. Pesaran, M. H., & Smith, R. (1995). Estimating long-run relationships from dynamic heterogeneous panels. Journal of econometrics, 68(1), 79-113. Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of applied econometrics, 16(3), 289-326. Phillips, P.C. and P. Perron (1988), “Testing for a Unit Root in Time Series Regression”, Biometrika, 75(2), 335-346. Pitoňáková, R. (2016). Determinants of Household Bank Deposits: Evidence from Slovakia. Journal of Economics, Business and Management, 4(9), 528-533 Sami, J., & Kreishan, F. (2012). FDI and export-led growth in Jordan: evidence from cointegration and causality test. Economics Bulletin, 32(2), 1-18. Soyode, A., & Oyejide, T. A. (1975). Branch network and economic performance: A case study of Nigeria’s commercial banks. Nigerian Journal of Economic and Social Studies, 17(2), 119-133. Srinivasan, A., & Meyer, R. L. (1986). An Empirical Analysis of Rural Deposit Mobilization In South Asia (No. 2142-2018-6110).

A Compilation of Working Papers by OECS Scholars

APPENDIX FIGURE A: DEPOSITS AND LOANS IN SAINT LUCIA (BILLION)

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

FIGURE B: GDP GROWTH RATES IN SAINT LUCIA (%)

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB). 30 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Kimbert Evans

Economist, Department of Finance Government of Saint Lucia

Mr. Evans is an Economist at the Research and Policy Unit of the Department of Finance. He is currently responsible for debt, monetary and health sectors and has been responsible for the manufacturing and tourism sectors. Mr. Evans is a graduate of the University of the West Indies, St. Augustine, Trinidad and Tobago with a Master of Science degree in Economics and a Bachelor of Science degree in Economics and Management. Prior to working with the Government of Saint Lucia, Mr. Evans worked at StateTrust and Trust International Bank and Republic Bank formerly known as The Bank of Nova Scotia. Mr. Evans’ research publications focus on Foreign Direct Investments (FDI) and Economic growth in Small Island Developing States (SIDS) as well as Determinants of Credit Growth in Saint Lucia.

A Compilation of Working Papers by OECS Scholars

Investigating the Development Impact of Inward Migrant Remittances on the Economies of the OECS Tommy Descartes

32 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT Globally migrant remittance to home country is becoming increasingly an area of interest among development economists. However, there is little evidence indicating that remittances impact national output. This paper uses a panel of six countries from the Organisation of Eastern Caribbean States (OECS) to assess the extent that remittances impact Gross Domestic Products, given that the region has a large diaspora. Using the Common Mean Group Estimator, the paper found that remittance does impact the ECCU as a bloc, however, the results are less convincing at the individual country level. INTRODUCTION In 2018, total immigrant remittances were estimated at US$689.4 Billion, equivalent to 0.8 per cent of global Gross Domestic Product, and has exceeded all other forms of financial flows to developing countries. Of this amount an estimated US$529.3 went to low- and middle-income countries. Such developments in the trajectory of remittances over the years have led to a burgeoning of research on the impact of remittances at both the macro and micro levels. Additionally, the United Nations as part of its Sustainable Development Goals 2030 agenda has earmarked remittances as a crucial driver to development finance and poverty eradication. Moreover, remittance has outperformed most other financial flows with the exception of Foreign Direct Investment (FDI), which is also true in the Organisation of Eastern Caribbean States (OECS) case. (Ratha, 2003) in the World Bank Development Finance report elucidate on several facets of the remittance phenomena. Among the key points the report outlined was the fact that remittances were a more stable source of development finance and foreign exchange compared with other sources and exhibited strong counter-cyclical characteristics, whereas other foreign financial flows were generally pro-cyclical in nature. Further, (Ratha, 2003) advocated that remittance reaches its targeted recipient better than any other form of aid and is largely insulated from the possibilities of corruption. Additionally, it was noted that remittance flows to receiving countries in economic crisis did not wane during crises, if anything, the reverse was true, attributed primarily to a phenomena known as home-bias. Like most developing countries, the Eastern Caribbean Currency Union (ECCU) members receive substantial remittance from its diaspora annually. Estimates for 2018 suggest that total remittance for the region was valued to the tune of US$226.4 million averaging 3.5 percent of regions annual GDP, which reflects an increase of 97.6 percent over 1990 remittance receipts of $95.0 million. It is worth noting that it is widely accepted that due to the relative underdevelopment of financial systems in developing countries, data on remittance is most likely under reported. Further, as depicted in graph 1, similar to global trends, remittance remained relatively stable pre, post and during the financial crisis for the region as seen in Panel (C), a phenomenon which was highlighted by (Moore & Greenidge, 2008). Additionally, a closer look at graph 1 shows that remittances have been on an upward trajectory and exhibit less volatility than FDI and ODA. Despite the obvious emergence of remittances in the global financial system, with particular importance for developing countries, clear evidence of its impact on the macro-economic performance of countries has been less prominent. Several reasons have been proffered to explain this remittance growth puzzle. The first is that the increase in remittances is illusory at best reflecting improvements in the coverage and measurement of remittances, which has been facilitated by the increase in financial and wire transfer technology which has increased central banks’ ability to monitor data. A second reason put forward is that of the inextricable link between migration and remittances, as such there is a trade-off that takes place. Hence, countries lose human capital in the process which will impact its growth potential. This is accentuated further if the education level of migrants is higher than that of their home country, as is the case for the OECS and the Caribbean region, which raises the question of the brain drain. There is also the view that remittance is inversely related to labour market participation, as households receiving remittances may refrain from participating in job training and may not form part of the labour force, which A Compilation of Working Papers by OECS Scholars

will negatively impact output. Notwithstanding these concerns of remittances lack of impact on GDP, it has been clearly proven that remittance increases the wellbeing of receiving households and might translate in poverty reduction, improve access to health and education, and as such might be acting as a buffer in many countries.

GRAPH 1: OECS GDP, REMITTANCES, FDI AND ODA (1986-2016)

It is these developments and the growing economic challenges that the ECCU region is presently faced with, that has motivated this research on the impact of remittance flows on growth in the region. The remainder of the paper is as follows; section 2 reports on the stylized facts on remittances and migration in the OECS, in section 3, the paper presents a brief review of the literature on remittances, both theoretical and empirical. The fourth section lays out the data used along with the econometric methodology and estimation strategy for the paper, while section 5 puts forward briefly the results and conclusions emanating from this paper.

34 | Research: The Platform for Innovation, Competitiveness and Growth

STYLIZED FACTS: CARIBBEAN MIGRATION AND REMITTANCES MIGRATION Historically, migration has always been part of the Caribbean and OECS’s way of life. Estimates from the most recent censuses in the six OECS member countries suggested that the total population is estimated at 605,997. However, according to the World Bank migration database, at the end of 2018 the Organisation of Eastern Caribbean States had a total of 373,019 living outside of the region equivalent to 61.6 percent of the selected OECS population. Of that total, 160,136 reside in the United States, followed by Canada (39,440) and the United Kingdom at 36,850. Anecdotally it is generally believed that OECS migration is primarily for economic reasons, where locals make a decision to migrate in search of better paying jobs. Additionally, there is also the view that Caribbean migrants migrate to access opportunities to improve their human capital via higher education. Recent studies on migration in Latin America and the Caribbean have noted differences in both the host country distribution and the education level of Caribbean migrants when compared with the Latin American region. Migrants from the region, although preferring to migrate to the United States, have a diverse mix of host destinations, which is different for the Latin American region which is generally US concentrated.

GRAPH 2: NO. OF OECS MIGRANTS BY MAJOR HOST COUNTRY

Another very interesting feature of the prolife of Caribbean migrants noted by the IMF is that they are generally more educated on average relative to the sending country, with almost half of Caribbean migrants having a college education. Further the Caribbean migrant’s educational attainment also outperforms that of Latin America. Invariably higher levels of education ought to translate into better employment and earning potential. Unlike Latin American migrants who are generally employed in low-skilled jobs (construction, transport, food preparation etc.), Caribbean migrants are employed in administration, sales, management and the health sector. This clearly raises the question of brain drain.

A Compilation of Working Papers by OECS Scholars

GRAPH 3: NO. OF MIGRANTS AND % OF HOME POPULATION BY COUNTRY (2018)

Graph 3 above shows the number of individuals in the selected OECS countries that are living overseas along with the percentage of the home country’s population that is abroad. The graph shows that the Dominica has the largest diaspora at 73,955, followed by Antigua & Barbuda (73,491), while St. Kitts and Nevis has the small population 40,612 living abroad. In terms of percentage of home population, Dominica has the highest at 106.6 percent, followed by Antigua & Barbuda at 72 percent. Graph 4 gives a more comprehensive view of the flow of migrants from the OECS to major Receiving Countries. Similar to graph 2, it highlights the fact the United States is the host country of choice for most OECS countries, in particular Antigua and Barbuda, Dominica and Grenada and to a lesser extent the other OECS countries. In terms of the United Kingdom, the distribution of OECS nationals who have migrated to the UK is generally evenly distributed, whereas St. Vincent and the Grenadines and Grenada contribute the largest share of OECS national residing in Canada and Trinidad and Tobago. Also, Dominican and Saint Lucian comprise the largest number of OECS citizens residing in France, which is suggestive of the French Creole heritage of both these countries. Other countries such as US Virgin Island (Antigua and Dominica) and Barbados (Saint Lucia and St. Vincent and the Grenadines) to have sizable OECS citizens.

36 | Research: The Platform for Innovation, Competitiveness and Growth

GRAPH 4: MIGRATION FLOW FROM OECS TO RECEIVING COUNTRY (2018)

Source: Author’s calculation.

REMITTANCES In terms of inward remittance flows as of 2018, Grenadians remit the most estimated at US$48.4 million, followed by Dominica (US$46.4 million) and St. Vincent and the Grenadines (US$41.7 million). As it pertains to percentage of GDP, Dominica has the highest at 9.6 percent, followed by St. Vincent and the Grenadines at 5.0 percent, with Saint Lucia having the lowest at 1.8 percent. Graph 6 goes a bit further and shows where these remittances emanate from. As is expected, the largest amount of remittances is derived from the United States with Dominica and St. Vincent and the Grenadines received the largest share of US remittances. This is then followed by Canadian remittances flows of which St. Vincent and the Grenadines benefits an estimated fifty percent of total. While remittances from the United Kingdom are equally distributed between Saint Lucia, St. Vincent and the Grenadines and Dominica. At the regional level, the majority of the remittance flows from Trinidad and Tobago goes to St. Vincent and the Grenadines, while Barbados outward remittance flows is mainly to Saint Lucia and St. Vincent and the Grenadines. While the OECS relies heavily on the US diaspora for remittances, this presents a concentration risk, which may translate into a sudden stop of remittances flow in the event the US encounters a short contraction in economic activity.

A Compilation of Working Papers by OECS Scholars

GRAPH 5: REMITTANCES VALUE AND % GDP (2018)

GRAPH 6: REMITTANCE FLOW FROM SOURCE COUNTRY TO THE OECS (2018)

Source: Author’s calculation.

38 | Research: The Platform for Innovation, Competitiveness and Growth

GRAPH 7: REMITTANCE PER NO. OF MIGRANTS IN USD

Graph 7 depicts the remittance per no. of migrants for each OECS country. It notes that St. Vincent and the Grenadines (svg) has the highest at $796.76 USD, followed by Saint Lucia (lca) at $717.01 USD. What this shows is that while Dominica (dom) and Antigua and Barbuda (ant) have the largest diaspora outside the region, they are remitting significant lower relative to other ECCU countries with smaller diaspora, and raise questions with regards to what extent the remittance from these countries are compensating for the sizeable lost in human capital through brain drain. BRIEF REVIEW OF LITERATURE There are several definitions for remittances proposed by various agencies including the International Monetary Fund and the World Bank; however, (Kapur, 2003) defines remittances as financial flows arising from the cross-border movement of nationals. The theoretical body of literature on remittance behaviour is inextricably linked to theories of migration, and any assessment of one will inevitably result in the overlap of the other, primarily because one’s decision to remit is based on a previous decision to migrate. However, theories of remittances are quite parsimonious when juxtaposed with migration theories. Despite this close link between the two, for brevity purposes this paper will delve only into theories pertaining to remittances. MOTIVATION FOR REMITTING The first research attempts at theorizing remitter motive for remitting was conducted by (Lucas & Oded, 1985) in their groundbreaking work carried out in Botswana. Before formally conducting empirical research work in Botswana Lucas et al. developed a theory or remittances which is widely accepted today. (Lucas & Oded, 1985) posited that “Pure Altruism” is the main motive for remitting. The theory is based on the care migrants have for the members of households left behind and that remitters receive utility from the utility recipients receive from receiving remittances. However, there is also ongoing discussion which suggests that remitters may not just be purely altruistic but may actually be using remittance as a form of insurance, family loan agreements, and may be using remittances as an enforcement mechanism to state their claim to family inheritance. A Compilation of Working Papers by OECS Scholars

40 | Research: The Platform for Innovation, Competitiveness and Growth

EMPIRICAL RESEARCH ON REMITTANCE ON GROWTH The impact of remittances has been very topical among development economists with both varying and sometimes conflicting results. Table 1 below provides a summary of a selected few empirical work done and its impact on growth. However, there has been a growing body of literature which suggests that remittances may have severe adverse implications for growth. According to (Acosta, Lartey, & Mandelum, 2007) the increase in remittances may have led to the Dutch Disease in El Salvador as households receiving unearned income from remittances substituted their remittance income in place of labour income while simultaneously increasing consumption primarily toward non-tradeable sector. Additionally, critics of remittances such as (Chami, Fullenkamp, & Jahjah, 2003), have highlighted that given that a remittance transaction takes place with asymmetric information which results in a moral hazard in the use of these funds, remittances consequential will have negative implications for growth.

TABLE 1: FINDINGS FROM SELECTED EMPIRICAL RESEARCH

Researcher

Model

Key Findings

(Shera & Meyer, 2013)

Panel Data – Fixed Effect and Random Effects model

The model found a positive impact, with a 1% increase in remittances increasing GDP by 0.14 percent.

Buch and Kuckulenz (2004)

Panel of 87 developing countries

(Chami, Fullenkamp and Panel of 113 countries over Jahjah, 2003) 29 years using fixed and (Siddiqui, 2016)

random panel techniques.

The model found that the moral hazard associated with remittances can be severe enough to reduce economic activity.

Using a CGE model for Pakistan

Found positive impact on macroeconomic variables.

DATA ANALYSIS In order to investigate whether remittance may have any impact on growth, a theoretical growth account model must be used and augmented for remittances in order to assess the impact of remittances on growth of the income of receiving countries. The Neo-Classical Growth Model historically has been used as a base theoretical model for conducting empirical work to understand growth with various augmentations. In deriving the models to be used in this paper, reference was made to (Barro, 2003) list of variables in his work on the “Determinants of Growth in a Panel of Countries”, although the full array of variables were not included due to data unavailability issues. The variables selected captures human capital, proxied by education expenditure (both public and private) and population variables; physical capital is captured by gross fixed capital formation, inflation and real effective exchange are used as proxies for macroeconomic stability, terms of trade is used to assess degree of openness, initial income per capita is also included, and finally, official development assistance (ODA), Foreign Direct Investment (FDI) and Inward Remittances are used as proxies for foreign financial flows. See appendix. The database comprises annual data which were obtained from several sources, mainly the World Bank and the Eastern Caribbean Central Bank. The number of observations span twenty-eight years from 1986 to 2013 for six ECCU countries, namely Saint Lucia, Grenada, St. Vincent and the Grenadines, Dominica, Antigua and Barbuda and St. Kitts and Nevis. The panel is strongly balanced with 168 observations with thirteen variables. A Compilation of Working Papers by OECS Scholars

TABLE 2: DESCRIPTIVE STATISTICS OF THE EASTERN CARIBBEAN CURRENCY UNION MEMBER STATES

ECONOMETRIC METHODOLOGY The main econometric models utilized in this paper are the Mean Group Estimator (Pesaran and Smith 1995) and the Common Correlated Mean Group Estimator (Pesaran 2006). The MGE model was preferred over other methods because it is used for macro-panel data with t>n, i.e. there are more time observations than cross-sectional variables, which is the case in this paper. Additionally, the mean group estimator allows for heterogeneous slope coefficients across members and also makes consideration for cross country correlation, which would be useful to see the impact of remittance for individual countries. Further, it allows for the inclusion of trend coefficient which can be used to capture total factor productivity, although this is not a central focus of this paper. The Common Correlated Mean Group Estimator was also chosen on the basis that it has been suggested as one of the three methods for dealing with crosssectional dependence according to (Pesaran 2002) and (Eberhardt and Teal 2009). 42 | Research: The Platform for Innovation, Competitiveness and Growth

Further to income, educational attainment appears to improve the odds of coverage. Interacted with income, we see that an individual who earns the mean income and has post-secondary education has higher odds of having insurance coverage than an individual who also earns the mean income but without post-secondary education. One’s income and or educational attainment are not factors that can be altered on a large scale in a short or even medium term time horizon. This therefore implies that with Saint Lucia’s existing status quo, at least 65.0 percent of the population will remain without health coverage. In such circumstances improving access to health interventions of this cohort would require some subsidization of cost. A surprising finding of our study was that locality i.e. whether someone lived in a urban or rural setting was not a significant contributor to health insurance. The prevalence of health insurance coverage in the two groups was similar to their prevalence in the study i.e. one third of survey respondents were rural while a third of those who had health coverage hailed from a rural area. Furthermore, there was not statistically difference in earnings between the two areas. This is an important finding for policy makers who may have intuitively thought that rural areas had higher incidences of no health coverage and consequently may have been inclined to focus on that demographic rather than the urban one. This is not the case. The finding that self-employed people are less likely to have health coverage with an average rate of less than 10.0 per cent (see figure 6) suggests that the group has to be a key demographic which policy makers focus on. Self-employed persons with employees had one of the highest average incomes of approximately $9,000 monthly (see table 2) but still had low health insurance take up rates.

yit = xit βi + μit

Equation 1

xit= a2i + λift + Yigt+ μit

Equation 2

μit= a1i + λift+ εit

Equation 3

where yit and xit are observables and βi are country specific slopes on individual the observables. NEO-CLASSICAL REMITTANCE AUGMENTS GROWTH MODEL This paper uses a similar Neo-Classical remittance model to that of (Shera & Meyer, 2013) in their assessment of the impact of remittances on Albania, using a panel of twenty-one (21) countries, using the fixed effect approach. Like (Shera & Meyer, 2013) this paper specifies a simple log linear Cobb-Douglas production function in assessing the GDP per capita sensitivity to annual remittance flows and other wellaccepted determinants of economic growth in the literature. The model equation is specified as follows below in equation (4).

In GDPPCit = β1 + β2 In REMit + β2 In GCFit + β3 In POPit + β4 In ODAit + β5 In FDI it + β6 In CONit + β7 In TOTit + β8 In REERit + β9 In EDU_EXPit β10 INFit+ β11 In IIPCit + εit

Equation 4

A Compilation of Working Papers by OECS Scholars

where the subscripts i and t represent country and time period respectively. All variables have been transformed in logarithm form for ease of estimation, with the exception of inflation which are already in growth rates. Equation 5 is similar to 4, however, it excludes consumption. This second equation was estimated on the basis that many empirical work have found that the effect of remittances is not directly correlated with income per capita but rather is passed-through the consumption channel, given the high consumption component of remittance receipts. In GDPPCit = β1 + β2 In REMit + β2 In GCFit + β3 In POPit + β4 In ODAit + β5 In FDI it + β6 In TOTit +

β7 In REERit + β8 In EDU_EXPit β9 INFit+ β10 In IIPCit + εit

Equation 5

ESTIMATION AND DIAGNOSTICS In estimating the panel dataset, several diagnostic checks were conducted. These included residual diagnostic checks, stationarity checks; cross-sectional dependence tests and models were all estimated using robust options to account for possible heteroscedasticity issues. STATIONARITY TEST First, given the number of observations is twenty-eight (28), all the variables were assessed independently for the presence of unit roots which would result in spurious regression if not dealt with. Three of the four first generation Panel Unit Root Tests were employed with most suggesting that variables were stationary at levels. Table 1 below provides the results for all variables.

TABLE 3: TEST FOR STATIONARITY

Levin-Lin-Chu

Im-Pesaran-Shin

Hadri

Adj T-Stat

P-values

Z-t-tilde-bar P-values

Statistic

P-values

lnGDP_PC

-4.3641

(0.0000)***

-2.8208

(0.0024)***

23.5784

(0.0000)***

lnCON

-2.5583

(0.0053)***

-0.3689

(0.3561) --

15.6028

(0.0000)***

lnGCF

-3.4828

(0.0002)***

-2.4156

(0.0079)***

7.7268

(0.0000)***

lnPOP

-4.1192

(0.0000)***

1.5942

(0.9442)---

39.9613

(0.0000)***

lnFDI

-2.7588

(0.0029)***

-3.1139

(0.0009)***

5.0857

(0.0000)***

lnREM

-2.6839

(0.0036)***

-2.2318

(0.0128)***

4.9678

(0.0000)***

lnTOT

-1.8150

(0.0348)**

-1.4330

(0.0759)**

6.3847

(0.0000)***

44 | Research: The Platform for Innovation, Competitiveness and Growth

Levin-Lin-Chu

Im-Pesaran-Shin

Hadri

Adj T-Stat

P-values

Z-t-tilde-bar P-values

Statistic

P-values

-0.9602

(0.1686)---

-1.6002

(0.0548)*

24.7084

(0.0000)***

lnEDU_EXP -0.7671

(0.2251)---

1.4385

(0.9248)---

28.8472

(0.0000)***

lnIIPC

-4.0076

(0.0000)***

-3.8702

(0.0001)***

22.4447

(0.0000)***

INF

-5.6951

(0.0000)***

-7.0183

(0.0000)***

-0.6927

(0.7557)---

lnODA

-3.1996

(0.0007)***

-4.6117

(0.0000)***

0.3837

(0.3506)---

lnREER

N.B. *,**,***, indicates statistical significance at the 10%,5% and 1% levels respectively. P-values are presented in parentheses. Additionally, once the models were estimated, the (Pesaran 2003) post-estimation CADF panel unit root test in the presence of cross section dependence test was employed for both the Mean Group and the Common Correlated Mean Group Estimator, with both indicating model stationarity. CROSS-SECTIONAL DEPENDENCE A major draw-back to the MGE model is that by allowing each country to differ from each other, could lead to the possibility of outliers emerging, emanating from several factors primarily from omitted variable bias. To deal with this drawback vce robust are used to estimate each model. TABLE 4: TEST FOR CROSS-SECTIONAL DEPENDENCE PESARAN CD-TEST CD-Test

P-value

Corr

Abs-Corr

Mean Group Estimator

-0.03

0.974

-0.002

0.184

Common Correlated EMG

-1.47

0.142

-0.072

0.169

Null Hypothesis of Cross-Sectional Independence CD~N (0, 1)

From table 3 above the Pesaran CD Test strongly rejects the alternative hypothesis of cross-section dependence across the panel of countries. Given that the models all passed both the stationarity and cross-sectional dependence diagnostic checks the results were accepted as valid. ESTIMATION AND RESULTS The paper uses a Neo-Classical Remittance Augments Growth Model to estimate the impact of remittance flows on GDP per capita. Two versions of the growth accounting equation were used, one included consumption while the other excluded consumption on the premise that remittance might indirectly affect income per capita through the consumption channel. As such, four (4) models were estimated; (1) MGE with consumption, (2) MGE without consumption, (3) CCEMG with consumption and (4) the CCEMG without consumption. Of the models estimated, only the model 1 (the Mean Group Estimator – with consumption was positively significant at the five percent significance level for the region as a whole, A Compilation of Working Papers by OECS Scholars

suggesting that a 1 percentage change in remittance flows will positively impact income per capita by 0.0108 percent on average in the ECCU region. At the individual country level only two of the six countries registered positive significance of remittances on income per capita, for both with and without consumption models; namely St. Vincent and the Grenadines and St. Kitts and Nevis. In the case for St. Kitts and Nevis a 1 percentage change in remittances is estimated to positively impact income per capita by 0.059 percent and 0.080 percent respectively in the consumption and without consumption model. While for St. Vincent and the Grenadines a 1 percent increase in remittance receipts is estimated to impact income per capita positively by 0.145 percent and 0.092 percent respectively in the consumption and without consumption model. With regards to the other countries although the results were insignificant the signs varied across countries, with remittance having positive sign under both models for Antigua and Barbuda and Dominica, while results were mixed for Grenada being negative under with consumption model and positive under the with consumption model. In the case of Saint Lucia remittances were negatively insignificant for both models.

TABLE 5: ESTIMATION RESULTS

MGE_1 lngdp_pc lnCON lnGCF InPOP lnFDI lnODA lnREM lnTOT InREER InEDU_EXP lnF lni_GDP_PC

MGE_2

CCEMG_1

CCEMG_2

lngdp_pc

0.421***

0.572**

-6.82

-2.61

0.175***

0.122*

0.115*

0.0186

-4.29

-2.35

-2.01

-0.29

-0.362

-0.717

-4.004

-13.46

(-0.41)

(-0.51)

(-1.29)

(-1.60)

-0.0074

-0.0137

0.00386

0.0549

(-0.84)

(-0.85)

-0.25

-2.29

-0.00728

0.000

0.00646***

-0.00581

(-1.65)

(-0.01)

-6

(-0.71)

0.0108*

0.009

0.0284

0.0923

-2.06

-0.83

-1.72

-1.89

0.223*

0.0379

0.317***

0.165

-2.54

-0.59

-7.37

-2.01

-0.0667

-0.373*

-0.269

-0.952*

(-0.46)

(-2.55)

(-1.09)

(-2.64)

0.0742

0.133

0.0249

0.12

-0.58

-1.62

-0.29

-1.28

-0.00202

-0.00386

0.000321

-0.000514

(-1.84)

(-1.92)

-0.18

(-0.32)

0.114**

0.540**

-0.0952

0.115

-3.09

-3.91

(-0.45)

-0.41

46 | Research: The Platform for Innovation, Competitiveness and Growth

__000007_t _cons N

-0.0172**

0.00281

-0.0301

0.052

(-2.80)

-0.74

(-0.73)

-1.57

6.195

12.67

-29.11

235.1

-0.62

-0.8

(-0.59)

-1.63

168

t statistics in parentheses =”* p<0.05

** p<0.01

*** p<0.001”

CONCLUSION Motivated by the little that is understood about the impact of remittance flow on the Eastern Caribbean Currency Union, coupled with the noticeable steady rise in remittances to the region, this paper saw it fitting to investigate the relationship between remittance and income per capita in the ECCU region. The paper drew on previous empirical work done by (Shera & Meyer, 2013) et al, it however differs from its predecessor in that it uses the Mean Group Estimator and the Common Correlated Effects MG Estimator, whereas most previous research used the fixed or a random effect model. Additionally, the paper conducts this analysis for a currency union with a fixed exchange rate. The research found a positive and significant effect of remittances for the ECCU bloc as a whole, although results suggested varying magnitudes and significance across countries. Notwithstanding, the paper can be seen as a springboard for further analysis to be conducted on the dynamic of remittances and macroeconomic variables in the ECCU. REFERENCES Acosta, P. A., Lartey, E. K., & Mandelum, F. S. (2007). Remittances and the Ductch Disease. Federal Reserve Bank of Atlanta. Barro, R. (2003). Determinants of Growth in a Panel of Countries. Annals of Economics and Finance, 231-274. Chami, R., Fullenkamp, C., & Jahjah, S. (2003). Are Remittances Flows a Source of Capital for Development? Washington DC: International Monetary Fund. Kapur, D. (2003). “Remittances: The New Development Mantra? G-24 Technical Working Group. Lucas, R. E., & Oded, S. (1985, October). Motivation to Remit: Evidence from Botswana. Journal of Political Economy, No.5, 901-918. Retrieved from http://www.jstor.org/stable/1833062 Moore, A., & Greenidge, K. (2008). Determinants of Volatility of Remittances in the Caribbean. Central Bank of Barbados Annual Review Seminar. Rapoport, H., & Docquier, F. (2005). The Economics of Migrant’s Remittances. IZA. Ratha, D. (2003). Worker’s Remittances an Important and Stable Source of External Development Finance. Global Development Finance, 157-175. Shera, A., & Meyer, D. (2013). Remittances and their impact on Economic Growth. Periodica Polytechnica, 3-19. Siddiqui, R. (2016). Pakistan: Migration, Remittances and Development. Research Gate. Wooldgridge, J. M. (2006). Introductory Econometrics A Mordern Approach. Mason OH: Thompson South Western. A Compilation of Working Papers by OECS Scholars

APPENDIX TABLE A: VARIABLE DESCRIPTION

Variable Name

Variable

GDP per capita

GDP_PC

Net Inflows Foreign Direct Invest

Apriori Expectation

Data Source World Bank Dataset

FDI

Positive

World Bank Dataset

Net Inflow of Remittances

REM

Positive

United Nation Dept. of Population

Official Development Assistance

ODA

Positive

World Bank Dataset

Trade Openness

TRO

Positive

Eastern Caribbean Central Bank

Consumption Expenditure

CON

Positive

Eastern Caribbean Central Bank

Expenditure on Education

EDU_EXP

Positive

Eastern Caribbean Central Bank

Real Effective Exchange Rate

RER

Negative

World Bank Dataset

Gross Capital Formation

GCF

Positive

World Bank Dataset

Terms of Trade (Export/Import)

TOT

Positive

Eastern Caribbean Central Bank

REER

Negative

World Bank Dataset

Population Growth

POP

Positive

World Bank Dataset

Initial Per-Capita Income

IIPC

Positive

World Bank Dataset

Real Effective Exchange Rate

FIGURE A: GRAPH OF RESIDUAL PLOTS

48 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Tommy Descartes

Chief Economist, Department of Economic Development, Transport and Civil Aviation, Government of Saint Lucia

Tommy Descartes is the Acting Chief Economist in the Department of Economic Development, Transport and Civil Aviation in Saint Lucia. He previously worked as an economist in the Research and Policy Unit of the Department of Finance and as the Deputy Director for Social Research in the Ministry of Social Transformation in Saint Lucia. Mr. Descartes holds an MSc in Econometrics and Economics from the University of Nottingham in the United Kingdom as well as a BSc in Economic and Management from the University of West Indies. His areas of interest are applied development economics and econometrics.

A Compilation of Working Papers by OECS Scholars

The Impact of VAT Reduction on Food Prices in Saint Lucia Bill Monrose and Janai Leonce

50 | Research: The Platform for Innovation, Competitiveness and Growth

INTRODUCTION Value Added Tax (VAT) was first introduced in Saint Lucia in 2012 at an initial rate of 15% which was applied to the purchase of most goods and services from business which are registered for VAT with the Inland Revenue Department. Money collected through taxes accounts for 26% of Saint Lucia’s Gross Domestic Product (GDP). This range is relatively high compared to the 22-24 percent range from other Caribbean countries. These taxes left businesses and investors with less disposable income to operate. Furthermore, according to the Central Development Bank (CDB) note stimulating consumption vs investment, Saint Lucia’s economy is essentially driven by consumption, which accounts for roughly 82% of GDP. For these reasons, in 2017, the Saint Lucian government found it suitable to reduce VAT to 12.5%. A reduction in VAT may cause prices to fall which may allow consumers to have more disposable income. Lower prices foster greater consumption and ultimately stimulate growth. Therefore, theoretically, it could also be speculated that the reduction in VAT would have decreased the prices of taxable goods and services on the consumer market. Looking at the Consumer Price Index (CPI) of Saint Lucia, the prices of all goods and services (excluding education and utilities) rose when VAT was first introduced to the country in October 2012. Based on the monthly CPI for 2012 – from September to October – the average price of food and non-alcoholic beverages increased by 8%. While prices continued to fluctuate from this period, there was a very slight decrease of 2% from January to February 2017 for the cost of the same commodity after VAT was decreased to 12.5%. In products where VAT is applicable, there was an almost insignificant decrease in costs. For example, the index of alcoholic beverages, tobacco and narcotics decreased from 100.14 to 99.84 during that same period. However, it is important to note that beyond February 2017, there were fluctuations in prices, which seemed almost identical to the fluctuations before this period. This may indicate that there was little to no change in prices after VAT was reduced. Considering that the main motive of reducing VAT was to promote greater consumption due to lower prices, and ultimately stimulate growth, then it would be useful to investigate the effectiveness of this venture. The paper intends to determine the extent to which the reduction of VAT to 12.5% had an effect on the consumer prices of food and beverages in Saint Lucia. The paper further aims to distinguish which types of goods saw increases and decreases in prices. REVIEW OF LITERATURE Consumers may notice that prices of certain goods rise or fall periodically over a period of time. Generally, the final prices of consumer goods are each influenced by various factors. The supplier takes into consideration the initial price of raw materials to produce the item. Furthermore, this price may be increased to accommodate the profit that the seller wants to garner. In addition to these foundational influences, other indirect factors come into play in the determination of prices. According to an article issued by the Australian Government entitled “Analyse pricing influences”, the level of demand plays a significant role in the determination of prices, as a greater demand for a product would influence an increase in the price. Additionally, the level of demand ties into competition within the market. An introduction of new suppliers of a product would insinuate more competition and hence, a lesser demand and decrease in price. Lastly, government regulations may alter prices through the imposition of excise taxes. It is a common concept that various taxes affect the price of certain goods where the tax may be imposed on either the buyer or seller. If the tax is imposed on the seller, then the tax would be included into the price for consumers. On the other hand, if the tax is imposed on the consumer, then the tax is charged to the buyer on top of the price. The seller’s main concern is the amount paid by the consumer minus the tax. Transitionally, the buyer is mainly concerned with the total price that they are paying for the item. A Compilation of Working Papers by OECS Scholars

It should be noted that a reduction in tax, however, does not necessarily mean the supplier will carry it over to the consumer price unless the price is controlled by the government. According to the paper, “Effects of reducing the value added tax rate on food in Latvia”, ideally, a reduction in VAT should promote an increase in food sales due to decreased prices. It follows that the government would be at a slight loss of revenue, however, total societal gain would increase. Looking at inflation within Latvia, it was predicted that reducing VAT would result in an overall negative inflation for products such as animal produce, fruits and vegetables and other foods. Therefore, ultimately consumers would be left with more disposable income due to the reduction in food prices, in this case, and thus there would be more economic growth due to consumption. To determine whether VAT had any effect on inflation, it may be useful to carry out a t-test. In this case, more specifically, an independent t-test will be utilized, which is generally used to determine whether the mean difference between two unrelated populations is statistically significant. With an independent t-test, we measure the same variable twice (between the two periods in this case). Also, we fail to reject the null hypothesis granted that the mean difference between the two periods is equal to 0, or in other words: H0: ud=0, where ud represents the mean difference. Otherwise, we reject the null hypothesis. It has been noted that currently, no study was found to address whether VAT reduction had an impact on consumer prices within the Caribbean. DATA AND STYLIZED FACTS STATISTICAL DESCRIPTION Table 1 below contains statistical information on a few products from the largest weighted subgroup under food and beverages. before and after VAT was reduced from February 1st 2017. In our case, flour and cornflakes were our largest weighted. For further clarity, the variable ‘n’ represents the total number of price values used for the particular good. TABLE 1: STATISTICAL DATA BEFORE AND AFTER VAT REDUCTION BETWEEN LARGEST WEIGHTED FOOD ITEMS Before VAT Reduction Variable

Mean

Std. Dev.

Min

After VAT Reduction Max

Mean

Std. Dev.

Min

Max

People's Choice Enriched Flour 2kg

25.00

2.87

0.03

2.82

2.93

25.00

3.04

0.05

2.89

3.07

Easy Bake 1kg

25.00

4.21

0.12

4.02

4.64

25.00

3.92

0.16

3.77

4.56

Easy Bake - Whole Wheat 1kg

25.00

4.70

0.37

4.35

5.54

25.00

4.38

0.09

4.21

4.52

Counter Flour 2lbs

25.00

2.00

0.00

1.99

2.00

23.00

2.00

0.01

1.98

2.04

Cornflakes (Kellogg's) 12 oz

25.00

13.64

0.35

13.12

14.30

25.00

19.25

4.68

13.08

23.99

Cornflakes (Sunshine) 12oz

25.00

7.46

0.04

7.37

7.57

22.00

7.17

1.26

1.90

9.41

Cornflakes (IGA) 18 oz

25.00

8.89

0.09

8.73

8.99

25.00

8.79

0.26

7.78

9.12

Cornflakes (Universal) 12oz

25.00

7.08

0.04

7.03

7.16

25.00

6.95

0.20

6.04

7.12

Cornflakes (Universal) 18oz

25.00

10.20

0.14

10.07

10.56

25.00

9.94

0.16

9.27

10.09

Table 2 contains statistical information on both VATable and non-VATable goods under food and beverages, before and after VAT was reduced from February 1st 2017. For further clarity, the variable ‘n’ represents the total number of price values used for the particular good. 52 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 2: STATISTICAL DATA BEFORE AND AFTER VAT REDUCTION BETWEEN VATABLE AND NONVATABLE FOOD ITEMS Before VAT Reduction n

Mean

Std. Dev.

Min

After VAT Reduction Max

Mean

Std. Dev.

Min

Max

Non-VATable VATable

Macaroni Swiss Pasta 400g

25.00

3.14

0.05

3.02

3.22

25.00

3.25

0.05

3.14

3.31

Nido Powdered Milk 1800g

25.00

20.27

1.22

18.67

21.91

25.00

36.98

16.20

19.26

55.00

Carib Pearl Long Grain Parboiled 400g

25.00

1.94

0.13

1.84

2.37

25.00

1.89

0.01

1.87

1.90

Chicken Mixed Parts 1lb

25.00

4.51

0.18

4.25

4.75

25.00

4.46

0.14

4.22

4.69

Chicken Neck Fresh or Frozen 1lb

25.00

2.75

0.08

2.58

2.86

25.00

2.56

0.16

2.32

2.77

White Sugar 2lbs

25.00

2.23

0.05

2.18

2.30

25.00

2.19

0.02

2.14

2.21

Brown Sugar 2lbs

25.00

1.94

0.02

1.90

1.97

25.00

1.91

0.02

1.87

1.95

Morning Coffee [150g]

25.00

3.51

0.09

3.36

3.68

25.00

3.55

0.03

3.47

3.64

[Frozen Pork - 1lb]

25.00

8.82

0.25

8.29

9.19

25.00

8.34

0.39

7.69

8.90

(Lamb Neck - 1lb]

25.00

9.44

0.50

8.73

10.17

25.00

9.52

0.50

8.73

9.90

(Farmers Choice - 200g]

25.00

10.96

0.06

10.69

11.05

25.00

11.37

0.19

11.03

11.60

(The Chef Streaky Bacon - 200g)

25.00

8.74

0.17

8.32

8.97

25.00

8.63

0.10

8.45

8.75

Barons Ketchup 397g

25.00

4.09

0.05

4.05

4.22

25.00

4.01

0.15

3.76

4.31

baronbla~85g

25.00

6.09

0.17

5.39

6.28

25.00

7.06

1.81

4.25

8.64

Table 3 below contains statistical information on both VATable and non-VATable goods un der food and beverages, before and after VAT was reduced from February 1st 2017. For further clarity, the variable ‘n’ represents the total number of price values used for the particular good. TABLE 3: STATISTICAL DATA BEFORE AND AFTER VAT REDUCTION BETWEEN VARIOUS FOOD ITEMS Before VAT Reduction n

Mean

Frontera Merlot 750ml

25.00

28.89

Guinness Stout

25.00

Piton Beer

Std. Dev.

After VAT Reduction

Min

Max

Mean

8.23

19.29

37.81

25.00

21.98

4.64

0.01

4.61

4.66

25.00

3.45

0.05

3.37

3.52

Heineken Beer

25.00

4.65

0.04

4.57

Bounty 750ml

25.00

23.89

0.85

Ferands Strawberry

25.00

14.65

ferrands Vanilla

25.00

ferrands Chocolate Irish Potatoes

Std. Dev.

Min

Max

0.59

21.38

24.70

4.64

0.08

4.47

4.73

25.00

3.45

0.01

3.44

3.49

4.69

23.00

4.50

0.11

4.31

4.62

22.43

25.58

25.00

25.34

0.32

24.74

25.90

0.64

13.53

15.57

22.00

14.66

0.09

14.53

14.98

14.66

0.47

14.24

15.23

25.00

14.58

0.00

14.57

14.58

25.00

15.99

0.63

14.59

16.66

25.00

16.24

0.16

15.75

16.48

25.00

1.58

0.09

1.45

1.76

25.00

1.47

0.11

1.31

1.67

A Compilation of Working Papers by OECS Scholars

EXEMPTED AND ZERO-RATED GOODS The line graph illustrates the trend followed by Easy Bake’s flour before and after the reduction of VAT from 15% to 12.5%. As observed, there is a decrease in the CPI at the point of VAT reduction. However, this decrease seems to be very slight and steady from December 2016. This food item is regarded as a zero-rated good. Therefore, VAT reduction should have no drastic impact on this item

FIGURE 1: THE PRICE OF EASY BAKE’S 1KG FLOUR BEFORE AND AFTER VAT REDUCTION

The CPI of a pound of Chicken Neck seemed to be unaffected by the reduced VAT rate. Notwithstanding, this was expected considering that all raw chicken parts are VAT exempted. Different Fluctuations within the line graph could be attributed to various other factors such as changes in the CIF value of Chicken Neck and other duties.

FIGURE 2: THE PRICE OF 1LB OF CHICKEN NECK BEFORE AND AFTER VAT REDUCTION

54 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 3: THE PRICE OF NIDO’S 900G PACKAGE OF POWDERED MILK BEFORE AND AFTER VAT REDUCTIONS

The figure to the left shows a representation of the CPI of Nido’s 900-gram package of milk. Note that there was not a decrease but rather, there was a slight increase when VAT was reduced. VAT IMPOSED GOODS This graph shows that while there was a slight decrease, there was not any significant change in the CPI of Kellogg’s 12-ounce cornflakes at the point VAT was reduced.

FIGURE 4: THE PRICE OF KELLOGG’S 12OZ PACKAGE OF CORNFLAKES BEFORE AND AFTER VAT REDUCTION

A Compilation of Working Papers by OECS Scholars

It can be seen that there was a significant reduction in the CPI of IGA’s 18-ounce package of cornflakes when VAT decreased to 12.5%. There is an apparent and sudden negative veering of the line at the point when the tax was reduced.

FIGURE 5: THE PRICE OF IGA 18OZ PACKAGE OF CORNFLAKES BEFORE AND AFTER VAT REDUCTION

FIGURE 6: THE CPI OF UNIVERSAL’S 18OZ PACKAGE OF CORNFLAKES BEFORE AND AFTER VAT REDUCTION

Again, we can see here from the illustration above that the gradient of the line representing Universal cornflakes’ CPI experienced a sudden steep reduction when VAT was reduced.

56 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 7: THE CPI OF 1LB OF LAMB NECK BEFORE AND AFTER VAT REDUCTION

Based on the graph above, lamb neck prices experienced a sudden decrease when VAT was reduced.

FIGURE 8: THE CPI OF MORNING COFFEE’S 150G PACKAGE BEFORE AND AFTER VAT REDUCTION

The price of Morning Coffee’s 150-gram package remained fairly constant at the point when VAT reduced. Furthermore, there was a temporary increase in CPI beyond February 2017. This suggests that VAT reduction had no direct impact on the price of this item. The CPI of Piton Beer vividly represents the expected outcome of VAT reduction in which there was a steep reduction in its CPI at that point where the tax was reduced – as seen from Figure 9.

A Compilation of Working Papers by OECS Scholars

FIGURE 9: THE PRICE OF PITON BEER BEFORE AND AFTER VAT REDUCTION

FIGURE 10: THE PRICE OF BARON’S KETCHUP BEFORE AND AFTER VAT REDUCTION

Figure 10 above shows price movement for the 397-gram package of Barons Ketchup before and after the VAT reduction from 15% to 12.5%. As observed, there were few fluctuations before and after the reduction. There was also a steep decrease in its price at the moment when VAT was reduced.

58 | Research: The Platform for Innovation, Competitiveness and Growth

METHODOLOGY This paper refers to a broad group of data concerning price indices of various goods and services offered within Saint Lucia. The CPI basket has 12 different categories which accounts for over 2000 items in total. It should also be noted that VAT was implemented on a large subset of those items. However, for the purposes of this paper, we will look at the largest weighted indices in order to make inferences about the effect that the reduction in VAT had on the prices of goods and services. In this case, the largest weighted indices were food and non-alcoholic beverages, utilities – comprising housing, water, electricity, gas and other fuels – and lastly, transport. Within these sub-components, the largest weighted items were analyzed. For example, under food and non-alcoholic beverages, bread and cereal products were analyzed as those two sub-groups were the largest weighted indices within the group. Transport and utilities will be analyzed in the same manner by looking at their two largest weighted sub-groups. The mean price of these goods throughout the period will be calculated before and after VAT was reduced in 2017. In order to better measure the difference between the prices before and after 2017, it was useful to calculate the percentage increase, or decrease – denoted by a negative value. Considering that these sub-groups also form an umbrella for many items produced by various suppliers, graphs will be constructed for certain items which had the most interesting changes in prices. This will aid in the analysis of the price movements before and after the VAT reduction. Additionally, we will also make use of an independent t-test to determine whether the mean prices before and after VAT reduction had a significant difference. For the purposes of this paper, we will utilize the t-test in three (3) scenarios. Scenario One (1) includes variables from the largest weighted goods in the food basket. In our case, we have products of flour and cereal. The results obtained for the largest weighted goods should be representative of the overall impact of VAT reduction on food prices. The test would be done for 2 pairs of periods – six (6) months before and after February 2017, and then 2 years before and after the same aforementioned date. Scenario Two (2) includes the VATable goods and non-VATable goods (zero-rated and VAT exempted). This was done to compare the results of the t-test on VATable goods to the results of the non-VATable goods. Through this scenario, we will be able to determine the extent to which the reduction in tax impacted the prices of both VAT-imposed and VAT-free goods. Like Scenario One (1), the test would be done for 2 pairs of periods – six (6) months before and after February 2017, and then 2 years before and after the same aforementioned date. Scenario Three (3) includes VATable, zero-rated and exempted goods, like Scenario Two (2). In this scenario, however, we compare the trends of forecasted and actual prices of goods. Population one (1) will consist of actual prices from January 2016 until December 2017 – approximately one year before and after the VAT reduction. Population two (2) will contain actual prices from January 2016 to January 2017 (one year before VAT reduction), and then forecasted prices from February 2017 to December 2017 (approximately one year after VAT reduction). The forecasted prices in this case would be based on previous price movements and trends before the reduction in VAT. If the trend followed by actual prices were statistically and significantly similar to the trend of forecasted prices, then this may infer that VAT reduction had no measurable effect on the prices of the item or variable. However, the opposite is true if the two trends were statistically different from each other, and we infer that VAT reduction did impact food prices. RESULTS Viewing the data, it was clear that certain products of flour and cereal saw decreases in prices after VAT was decreased in 2017. However, it was important to note that these items saw decreases in different ways – which may be due to flour being a zero-rated good while the cornflakes were VAT-able. Looking A Compilation of Working Papers by OECS Scholars

at the line graphs plotted for various products, it was apparent that the majority of products which are imposed with VAT experienced a steep negative veering in the gradient of the line at the moment VAT was reduced – which indicated that there was a sudden decrease in the price of the product. Fig. 5 captures this visualization in the case of IGA cornflakes, where there was steep negative movement in its price at the moment when VAT was reduced in 2017. This differs from the slight and less dramatic decrease or increase experienced by the tax-exempted and zero-rated goods (see figs. 1-3). Considering that these two subgroups were regarded as the largest weighted food products, ideally, the trends followed by them should be similar for the other subgroups under food and beverages. Looking at Easy Bake’s one-kilogram package of flour, the statistical mean price – initially being 4.2072 – slightly reduced to 3.924 (see fig. 1). However, this reduction happened subtly over the period – there was no sudden decrease at the time when VAT was decreased. On the other hand, the price of the 18-ounce package of Universal cornflakes remained above 10.0 from 2015 until it finally saw a decrease to 9.940, specifically on February 1st, 2017 – the date when the VAT reduction to 12.5% became effective. This trend was further exemplified by the majority of products under this sub-group. Apart from the largest weighted products, other food items were analyzed in order to determine whether this trend was truly representative of most products under food and beverages. Food items such as lamb neck saw a sudden decrease in prices from February 1st 2017 (see fig. 7). This trend seemed to be common amongst most VAT-imposed items tested for the purposes of this paper – including beverages such as Piton beer, Heineken and the Guinness stout. There were few exceptions where there was not any notable difference between prices shortly before and after VAT reduction. In the case of Morning Coffee’s 150-gram package, the price appeared to remain fairly constant even after the point of the tax reduction. An instance such as this can be explained by many factors. Looking to make the most profit as possible, suppliers generally pass on added expenses and taxes – including Value Added Tax (VAT) to the consumer. Therefore, the presence of VAT may increase the price of a product. However, the supplier may opt not to pass the reduced tax to the consumer, which leads to added profit to the seller. In these cases, the final price remains relatively constant even after a reduction in tax. Based on the graph plotted for Kellogg’s cornflakes, prices decreased slightly at that point. However, by contrast, it was not significant compared to the decrease in price experienced by other VAT imposed food items. We now look at the results obtained from the t-test. Scenario one (1) included our largest weighted goods from the food basket. The t-test was used to determine whether the mean price before and after the VAT reduction were different and statistically significant. The t-test was done for two periods – 6 months before and after VAT reduction and then two years before and after the same event. Our null hypothesis stated that the statistical mean price difference between the two periods (before and after VAT reduction) is equal to zero and is insignificant. Looking at Easy Bake’s whole wheat flour, with a significance level of 0.05, we obtained a p-value of 0.1013 when testing 6 months before and after VAT reduction. Thus, we failed to reject the null hypothesis and deduce that the prices did not change significantly after VAT was reduced. On the other hand, over a four-year period (two years before and after the VAT decrease), we obtained a p-value of 0.0003 and thus rejected the null hypothesis. This indicates that while prices saw changes over the overall period of four years, those changes cannot be attributed to the decrease in VAT since our test showed that there were no changes six months before and after the tax reduction. This result was expected since flour (Easy Bake whole wheat) is considered to be a zero-rated good, which means that there is no VAT imposed on products of its nature. In the case of Sunshine cornflakes, we obtained a p-value of 0.0006 within the one-year period (six months before and after February 2017) and rejected the null hypothesis. Therefore, within that period, there was a significant difference between the prices before and after the VAT reduction which hints that the VAT decrease may have impacted the price of Sunshine cornflakes during that period. This was expected since cornflakes is a VAT-able good. However, obtaining a p-value of 0.2883 for the fouryear period means that we fail to reject the null hypothesis. Hence, the prices two (2) years before and after the reduction in VAT were not significantly different. This indicates that while prices decreased briefly around the VAT reduction, the decreased prices were not sustained but rather increased to levels 60 | Research: The Platform for Innovation, Competitiveness and Growth

relatively equal to the prices before the date when VAT reduced. Scenario two (2) involves the use of the t-test on both VATable and non-VATable goods (zero-rated and exempted). Like scenario one (1), the t-test was used for two periods – one year (6 months before and after VAT reduction and then four years (two years before and after). Our null hypothesis stated that the statistical mean price difference between the two periods (before and after VAT reduction) is equal to zero and is insignificant. In the case of Swiss’ Macaroni Pasta, during the one-year period, we obtained a p-value of 0.4230 which meant that we failed to reject the null hypothesis. However, for the four-year period, we reject the null hypothesis after yielding a p-value of 0.0000. This result was also similar to the t-test results of white sugar and brown sugar, where we obtained p-values of 0.1372 and 0.1925 respectively for the one-year period and thus we failed to reject the null hypothesis on both occasions. However, for the four-year period, we reject the null hypothesis after obtaining p-values of 0.0001 and 0.0000 respectively. This indicates that there was no change in prices immediately after VAT reduced, however, prices did change over the overall period of four years which may be due to other factors apart from VAT reduction. In the case of the chicken mixed parts, for the one-year period, we failed to reject the null hypothesis (p-value: 0.8053) and we also failed to reject the null hypothesis in the four-year period (p-value: 0.2650). This indicates that the price remained constant throughout the entire period and was not affected by the reduction in VAT. In the examples mentioned above for scenario two thus far, it can be observed that we rejected the null hypothesis for the one-year period in all cases which was indeed expected since these goods belonged to the zero-rated and VAT-exempted goods category. Looking at our VATable goods, we observe a shift in the results. Choosing random and varied goods, we look at Frontera Merlot grape wine and Baron’s black pepper. We rejected the null hypothesis for both products when looking at the one-year period after obtaining p-values of 0.0040 and 0.0028 respectively. We also rejected the null hypothesis on both occasions when looking at the four-year period from obtaining p-values of 0.0003 and 0.0000. Therefore, this indicated that prices did change as a result of VAT reduction. Furthermore, this price change was sustained beyond the period after the VAT reduction. However, looking at Baron’s ketchup, we fail to reject the null hypothesis for the one-year period (six months before and after the VAT decrease) after receiving a p-value of 0.5978, while we reject the null hypothesis for the four-year period (two years before and after the VAT decrease). This indicates that there was no significant difference between the mean prices shortly before and after the VAT reduction. Looking at fig. 10, we can see that there was a sudden decrease at the point when VAT was reduced (February 2017). However, the result from the t-test could be attributed to the fluctuations before and after the VAT reduction. These fluctuations may have caused the means before and after the reduction in VAT to seem relatively equal, regardless of the decrease seen at the point where VAT was reduced. Lastly, looking at Ferrands vanilla ice-cream as a last example, we reject the null hypothesis for the oneyear period but fail to reject the null hypothesis for the four-year period. Therefore, we deduce that the VAT reduction had an effect on its price. However, the effect was only temporary since the price increase back to levels similar to the prices before the reduction in VAT. Finally, a forecasted trend of the prices of certain food products one year after the VAT reduction was determined and analyzed in order to compare with actual price values of the period from the beginning of 2016 until the end of 2017, after VAT was reduced. Two populations of unequal variances were evaluated. Both populations shared the prices throughout 2016 as the commonality. However, population 1 contained forecasted values for the prices from February 2017 until December 2017, while population 2 had actual price values for that period. In this t-test, the null hypothesis claims that these two populations are statistically equal and hence, the mean difference is insignificant. This would infer that the actual prices should follow the forecasted trend of the prices after VAT was reduced in 2017. Looking at the results of the test, we failed to reject the A Compilation of Working Papers by OECS Scholars

null hypothesis in the case of Easy Bake’s flour and Nido’s powdered milk since we obtained p-values of 0.2181 and 0.6858 respectively. This indicates that we failed to reject the null at all levels of significance. This was the expected result since these products are regarded as zero-rated goods. Therefore, there would be no significant difference between the trends of the forecasted data and the actual data, since VAT reduction had no effect on these goods particularly. In contrast, we were able to reject the null hypothesis in the cases of Universal cereal and Piton Beer, since the values 0.0056 and 0.0180 were yielded respectively as our p-values. Therefore, we can conclude that the two populations are statistically different. In this case, the trend of the actual price valued veered away from the trend of the forecasted values after VAT was reduced. This gives evidence that VAT indeed had an effect on the prices of these products, given that they are both applicable to be imposed by VAT. CONCLUSION From the results of this paper, it was indeed evident that VAT reduction in February 2017 had a measurable impact on the prices of VATable food and beverages. While the prices of certain VATable goods remained constant beyond the reduction in VAT – which may be attributed to other factors – the majority of goods saw decreases in their prices. It was also observed that the majority of non-VATable goods (zero-rated and exempted) did not seem to experience any drastic changes in their prices at the moment when VAT decreased. This was expected since goods of this category are not subjected to VAT and thus should not be affected by a decrease in the tax. POLICY RECOMMENDATIONS While this paper analyzed whether goods saw changes in prices due to VAT reduction in 2017, it may be useful to determine which other factors caused these price movements. A change in the price of an item could be due to changes in the CIF value of the good when initially purchased without taxes. Furthermore, other duties and taxes may influence the price of the particular good. Therefore, to have a complete understanding as to whether VAT reduction really had an impact on the prices of goods, it may be necessary to look into this area. Furthermore, this paper only focuses on food and beverages. Consequently, the conclusions drawn from this paper can only be applied to this sub-group. However, it may be beneficial to broaden the range of items to other large-weighted groups such as utilities and transport, in order to have a more generalized idea of how VAT reduction impacted the prices of all goods and services in Saint Lucia. With this information, it may be possible to determine whether the initiative was successful in promoting economic growth within the country through the availability of more disposable income. REFERENCES Department of Industry. “Analyse Pricing Influences.” Australian Government; Department of Industry, Innovation and Science, Australian Government, Department of Industry, Innovation and Science, 23 July 2018, www.business. gov.au/products-and-services/pricing/analyse-pricing-influences. “HOME.” IDRE Stats, stats.idre.ucla.edu/stata/output/t-test/. Moody's Analytics. “Saint Lucia - Economic Indicators.” Saint Lucia | Economic Indicators | Moody's Analytics, www. economy.com/saint-lucia/indicators#ECONOMY. Nipers, Aleksejs, and Irina Pilvere. “Assessment of Value Added Tax Reduction Possibilities for Selected Food Groups in Latvia.” Proccedings of International Scientific Conference "RURAL DEVELOPMENT 2017", 2018, doi:10.15544/rd.2017.048. The Inland Revenue Department of Saint Lucia. Introduction to VAT, irdstlucia.gov.lc/index.php/ vat/179-vat-getting-started-introduction-to-vat. 62 | Research: The Platform for Innovation, Competitiveness and Growth

About the Authors

Bill Monrose

University Student University of the West Indies

Janai Leonce

Chief Economist, Department of Finance Government of Saint Lucia

Mr. Monrose is a student at the University of the West Indies (UWI) Cave Hill Campus, in Barbados, and is pursuing a Bachelor of Science in Mathematics with a minor in Spanish. In 2019, during the summer vacation, he was given the opportunity through the Summer Employment Program in Saint Lucia to work as an intern at the Department of Finance, where he was assigned to the Research and Policy Unit. During his tenure as an intern, he successfully produced a paper which bears relevance to “The Impact of VAT Reduction on Food Prices in Saint Lucia”, alongside Chief Economist of the Research and Policy Unit, Janai Leonce.

Mr. Janai Leonce, is currently the Chief Economist at the Research and Policy Unit in the Department of Finance (Saint Lucia). Mr. Leonce holds a master’s degree from Durham University in Finance and Investment and a bachelor's in Economics and Management from the UWI (St. Augustin). Prior to working for the Government of Saint Lucia, Mr. Leonce worked at the Eastern Caribbean Central Bank as an Economist and also lectured at the Sir Arthur Lewis Community College (Saint Lucia). Mr. Leonce is the author of several working papers covering topics on unemployment and gender wage discrimination. He has also written several articles in local newspapers on issues relating to productivity on behalf of the National Competitiveness and Productivity Council (NCPC) and Sir William Arthur Lewis.

A Compilation of Working Papers by OECS Scholars

4 Determinants of the Use of Electronic Payments in the ECCU - Panel Data Evidence Martina Regis 64 | Research: The Platform for Innovation, Competitiveness and Growth

DISCLAIMER The ECCB strongly supports academic freedom and a researcher's right to publish and encourage such activity among its employees. However, the ECCB does not endorse the views contained in an employee's publication or guarantee the technical accuracy. The views and opinions expressed in this paper are solely those of the author (s) and do not necessarily state or reflect those of the ECCB.

ABSTRACT The regional and global payment system is being rapidly transformed, reflective of significant developments in technological innovation. The new digital forms of payment may either replace or complement traditional physical sources of payment such as cash and cheques. Inspired by the global and regional interest by central banks in financial innovation, this paper seeks to empirically investigate the macroeconomic factors which may have affected the use of electronic payments in the ECCU over the period 2005 to 2016. Using panel data estimation techniques, this paper seeks to assess the impact of macroeconomic and infrastructural factors which determine the dynamics of cards and electronic payments, as measured by card and digital payments volumes per person. The econometric results suggest that income level and available payments technology, proxied by the Point-of-Sale penetration rate, had a consistently positive and significant impact on the use of these payment instruments in the ECCU during the period. Conversely, cash and cheque usage did not indicate the conventional substitution effect. ATM terminals did not show any significant effect on card payments. These findings may have important policy implications for the future of cash in the ECCU, and highlight the need for more micro-level research in this area, to better understand the challenges to greater adoption of electronic payments. INTRODUCTION The past two and a half decades have been characterized by rapid innovation in the development of payment technologies and solutions (Liu et al., 2015). This is reflected in the rapid growth in e-commerce, the proliferation of debit and credit cards for retail payments, and the burgeoning use of smartphones and mobile wallets. More recently, the attention has shifted to innovations in digital currencies. In particular, innovation in cryptography has facilitated the development of alternatives to traditional currencies such as distributed ledger technology which has led to interest by governments and the private sector to assess the scope for currency innovation (Runnemark et al., 2015; Cameron, 2016). This spread has been so extensive that authors have predicted that these financial innovations may eventually replace central bank money (Friedman, 1999; King, 1999) or lead to a cash-free society (Arvidsson, 2018). There have been many benefits proposed for shifting to electronic payments. These benefits include enhancing speed, security and efficiency, which are characteristic of electronic transactions, while offering greater transparency in the source and estimation of funds (VISA, 2017). A digital economy is posited to limit the informal economy in many developing economies and to help reduce the likelihood of corruption (Wald, 2018 in Capgemini, 2018), by reducing the use of cash in payment transactions1. The study estimates that emerging countries may lose a total of $110b a year from continued transactions in cash. The research initiated by VISA (2017) on 100 cities has also approximated that the digitization transformation could save countries $470b annually from persons devoting less time to banking, transit and retail transactions as well as from the reduction in crime. These advantages also extend to trade, competitiveness and business development, by promoting efficiency and speed in making payments (World Bank, 2014) and by improving global logistics and global e-commerce for many SMEs (McKinsey, 2018). The benefits for banks are also tremendous and include a reduction in their transaction costs, while providing them with greater options to operate across borders (Chavan, 2013). In fact, empirical research (see Humphrey et al., 2003) suggests that replacing cash with non-cash methods may reduce the overall cost of a country’s payment system by approximately 1.0 per cent or more of its GDP annually, once the transaction costs are absorbed. Overall, electronic payments have the potential to reduce the A Compilation of Working Papers by OECS Scholars

costs of handling cash, reduce the informal economy and boost government revenue. Specific to the ECCU, the high incidence of crime has raised the risk of handling cash by commercial banks and retailers. It has also increased the cost of managing cash for the Central Bank across the eight geographically-distinct countries - through its production, distribution, security and transportation. These risks have motivated the ECCB and other central banks to explore and assess the adoption of viable non-cash alternatives. As the global payments environment evolves, consumers in the Eastern Caribbean Currency Union (ECCU) are also adjusting to the adoption of new forms of payments, including cards and other forms of electronic payment2. By the end of 2016, credit and debit card transactions in the ECCU amounted to $2.7b from a total $793.2m in 2005, or an annual average growth rate of 11.8 per cent, reflecting increasing use of card payments for retail purchases (see Figure 1). Meanwhile data showed that the value of currency in circulation3 grew at an average annual rate of 4.3 per cent over the same period, FIGURE 1: TOTAL CARD TRANSACTIONS AND CURRENCY IN CIRCULATION

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

Despite such rapid innovations in payment technologies and its wide-ranging impact for both advanced and developing economies, there has been limited research highlighting this transition across countries, particularly developing countries. Most of the research has been focused on developed countries (Humphrey et al. 1996; Guariglia and Loke, 2004; and Goczek and Witkowski, 2015). Yet, this area has important implications for many central banks, which are considering the future of cash and which are launching research and experiments in the area of digital currencies. Accordingly, there is a strong need to understand the factors that determine the adoption of non-cash payments and to determine if this trend has implications for the use of cash. The ECCU region provides a useful research experiment, as it comprises developing countries which are also members of a currency union. The main research questions to be considered in this paper are: what are some of the macroeconomic and technological factors that determine the adoption of electronic payments in the ECCU? Have these trends resulted in the reduction of cash over time? Tackling these issues may be difficult given the challenges in differentiating between the micro-level (individual-level) and macro-level factors in payment choices. Nevertheless, it is necessary to understand these interactions as part of the process of adjusting to this payment transition. 66 | Research: The Platform for Innovation, Competitiveness and Growth

Against this background, the research aims to analyse the determinants of electronic payments in the ECCU. The scope of the study is restricted to the volume of digital and card payments. Given the significance of electronic payments to growth, the study attempts to fill an important gap in empirical literature in the ECCU. The remainder of the paper is organized as follows. Section 2 provides a brief review of the relevant literature while Section three presents some stylized trends in cash and electronic payments. Section 4 explains the econometric approach used to derive the findings which are presented in Section 5, while Section 6 concludes and presents some considerations for policy.

LITERATURE This section presents an overview of the various strands of the extant literature that seeks to consider some of the factors that may affect non-cash payments adoption. It combines issues related to the digitization of financial services, psychological aspects of technological adoption and the drivers of noncash payments using both macro-level and micro-level data. Relevant literature on non-cash payments typically involves a review of the theoretical and empirical literature on the determinants of money demand. Several theories have been proposed related to the demand for money including by Fisher (1911), Keynes (1956), Friedman (1956) and Baumol and Tobin. Money was seen as being influenced by several factors such as the price level, income and interest rates, which was proxied as the opportunity costs of holding money. However, recent trends related to declining interest earned on bank accounts, the possibility of theft and technological advances in payment systems have muted some of the traditional factors. Consequently, many authors have begun to incorporate new payment technologies as important variables in money demand models (See Amromin and Chakravorti, 2007) or have focused on the specific factors that determine non-cash use and incorporated cash proxies as important variables (for example Goczek and Witkowski, 2015). Two of the earliest forays into the impact of financial technology on currency were undertaken by Boeschoten (1992) for the Netherlands; and Duca and Whitesell (1994) for the United States. Using both micro and macro-econometric level surveys, Boeschoten found that financial advances such as ATMs and cards have reduced the demand for currency in the Netherlands. Given the nature of individual choices in this area, it is useful to note that a significant amount of the work in this area focused on survey data and analysed the effects of consumers’ social and demographic characteristics (Stavins, 2001; Stix, 2004; Borzekowski et al. 2006; Nasri, 2011). Some of the notable studies which consider non-survey macro-level data include Humphrey et al. (1996); Guariglia and Loke, 2004 and Goczek and Witkowski, 2016) who study the specific factors that determine non-cash use. The limited literature conducted in this area, has mainly focused on advanced economies. One of the earlier and most referenced works is the cross-country study by Humphrey et al. (1996) which considers the factors which determine the migration to non-cash payment instruments. Using aggregate data on fourteen developed countries over the period 1987-1993, Humphrey et al. (1996) analyze the effects of economic and institutional determinants on the use of five types of paper and electronic payment instruments. The explanatory variables considered were the prices of each payment instrument, real GDP, number of Point-of-Sale (POS) and ATM terminals, lagged volumes of the other payment instruments, cash holdings per person, as well as institutional variables such as violent crime, and concentration of the banking system. The results showed that the price variable had limited effect on non-cash usage, while POS and ATM terminals were positively related to the use of debit card usage. Another recurring theme in the literature is the degree of substitution between cash and electronic forms of money. Noting the lack of data on cash transactions, Snellman et al. (2001), used a panel of ten European countries over the period 1989–1996 to analyse the extent to which non-cash payments have been replaced with cash transactions at the Point-of-Sale) POS) terminals by using a Gompertz- curve. The authors found that the increased use of debit and credit cards was one of the key factors in the substitution away from cash, and that differences across countries were largely linked to differences in A Compilation of Working Papers by OECS Scholars

their rate of implementation of card payment technology. However, this perceived substitution effect has not always prevailed. A number of studies have shown despite the latest trends in non-cash instruments, the use of cash has remained surprisingly resilient in consumer payment behavior globally, particularly for certain types of transactions (see for example Kosse, 2014; Bagnall et al. 2014, van der Cluijsen, 2014 and Wakamori and Welte, 2017). As data on financial technologies such as ATMs and POSs have expanded in recent years, research focused specifically on the determinants of non-cash use has also developed. Guariglia and Loke (2004) and Goczek and Witkowski (2016) are the two notable studies in this area. Using dynamic panel data methods, the authors were able to explore the impact of past habits through lags in the dependent variables. The former used panel data from 15 countries over the period 1990 – 1998 to assess the factors, which drive non-cash payment volume and value. They found that volume proved more useful than the value of payments and that past habits were a key factor in the intensity of use4. Similarly, Goczek and Witkowski (2016) considered both micro and macro-level data to assess card payments. At the macro level, the authors employed System GMM (Bond and Blundell) and considered both social as well as macroeconomic factors including general trust, interest rate, consumption, past habits as well as the penetration of ATMs and POS terminals in explaining the non-cash payments. In discussing the adoption of new financial instruments, a more general understanding of the factors that drive the adoption and diffusion of technology may also be warranted. Noting the extensive literature developed in this area, Lai (2017); and Patel and Connolly (2007) presented a comprehensive review of the models and concepts that may help explain technological adoption by individuals. The research originates from a variety of fields including sociology, psychology and information system (Patel and Connolly, 2007). The papers reviewed some of the key models in this area including the Theory of Diffusion of Innovation (Rogers, 1995), Theory of Reasonable Action (TRA) by Fishbein and Ajzen (1975) as well as two strands of the Technology Acceptance Models (TAM) by Venkatesh and Davis (2000) among others. The foundation of most of these theories is the Diffusion of Innovation, which suggests that the spread of a new idea or product is determined by several factors such as relative advantage, compatibility, complexity, trialability and observability (Patel and Connolly, 2007). The theory also suggests that there are several stages to diffusion, with some groups being more likely to adopt innovations (early innovators) than others (the late majority or laggards). Noting the extensive research on age differences on the use and adoption of technology, Morris and Venkatesh (2000) investigated age differences in the adoption and sustained usage of technology in the workplace using longitudinal field investigation of one organization. Their work suggests that age was a critical factor in the adoption of technology, even while noting the limitations of the research. Notwithstanding, the research may present a useful first step in linking demographic factors such as age to the decisions regarding the adoption of banking sector innovation. While this study focuses on the macro-level variables, micro-level studies offer valuable insights on the factors that drive payment instrument adoption. These micro-level studies on consumer payment behavior have blossomed in the past fifteen years (van der Cruijsen, 2016). These studies have highlighted the importance of social and demographic factors such as age and education (Klee (2008); Schuh and Stavins (2011) and Bagnall et al. (2014), habit and persistence (Wakamori and Welte, 2014/7, van der Cruijsen, 2017) as well as transactions characteristics such as transaction value, overall acceptance and security (Huynh et al. 2014; Fung et al. 2015 and Bangall et al., 2017) as important determinants of consumer payment behavior. Using a few of the macro-level variables identified in the literature, this study will explore some of the determinants for the use of non-cash payments in the ECCU member countries.

68 | Research: The Platform for Innovation, Competitiveness and Growth

TRENDS IN CASH AND NON-CASH PAYMENTS IN THE ECCU This section analyses the trends in cash and non-cash payments in the ECCU over the period under study. The data include volume and value of various non-cash payment transactions for the eight member states for the period 2005 to 2016. Collectively, these payments include card payments, direct debit and standing orders5. Figure 2 shows the growth rates in the value of the major payment instruments in the ECCU. The largest expansion was in card payments with an average of 11.8 per cent annually over the period, followed by other electronic payments6, while the value of cheques issued in the ECCU declined. The volume of payments shows similar trends, as the card payments transactions grew at an annual average rate of 25.2 per cent, while the volume of cheques issued fell by 1.8 per cent (Figure 3).

FIGURE 2: TRENDS IN VALUE PAYMENT INSTRUMENTS IN THE ECCU

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

FIGURE 3: TRENDS IN THE VOLUME OF PAYMENTS IN THE ECCU

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

A Compilation of Working Papers by OECS Scholars

Given the untraceable nature of currency, it is not possible to determine precisely how much cash is being used for making payments. The value of the currency in circulation is therefore used as a proxy. Data show that despite growth in the use of electronic payments, the value of currency in circulation in the ECCU grew at an average annual rate of 4.3 per cent during the period. At the end of 2016, total value of currency in circulation was $1.12b which was equal to approximately $1,779 in currency per person in the ECCU or 6.1 per cent of GDP (Figure 4). The trends in the volume of card payments may be due to the increasing availability of infrastructure, such as point-of-sale (POS) terminals. These terminals are increasingly used by larger, more-established merchants, with a slow uptake by smaller merchants. This trend is observed in the positive correlation between POS terminals and non-cash transactions and conversely a negative correlation with cash in circulation to GDP (see Appendix I). To facilitate global comparison, the volume of non-cash payments is often scaled to GDP or population. The country-level analysis showed marked differences in the intensity of cash and non-cash use. In keeping with recent trends in electronic payments, the volume of these payments per inhabitant has increased dramatically across the member states over the period, with the lowest level recorded in St Vincent and the Grenadines of 6.1 to a high of 61.0 in Dominica in 2016 (See Figure 5). The differences across ECCU member-countries provide support to the hypothesis that the structure of the economy may not be an important factor in non-cash payments (Bech et al., 2018).

FIGURE 4: TRENDS IN CURRENCY IN CIRCULATION IN THE ECCU

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

70 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 5: DIGITAL PAYMENTS PER PERSON

Source: Author computed with data obtained from the Eastern Caribbean Central Bank (ECCB).

FIGURE 6: CASH IN CIRCULATION TO GDP

Unlike non-cash payments, cash payments are difficult to measure given its characteristic as an anonymous payment instrument. Currency in circulation to GDP is therefore often used as a proxy for cash intensity. The cash intensity ratio in the ECCU has been within a range of 5.5 to 6.3 per cent over the period 1990 to 2016. Figure 6 presents a country-level breakdown which indicates that cash demand increased in five of the eight member territories over the period, suggesting that despite the strides made in non-cash payments, cash use has not declined comparatively. A Compilation of Working Papers by OECS Scholars

DATA DESCRIPTION AND EMPIRICAL SPECIFICATION A review of the literature revealed a number of possible macro-level and micro-level factors that may influence the use of digital payments across countries, including social, demographic, and technological. Accordingly, six main indicators were considered for analysis, taken from the ECCB’s database as well as macroeconomic data on seven7 of the eight member states of the ECCU (Anguilla, Antigua and Barbuda, Dominica, Grenada, St Kitts and Nevis, Saint Lucia and St Vincent and the Grenadines). Data from the seven countries span the period 2005 to 20168. All variables are annual time series data, and the sample period and the variables all chosen based on data availability and the previously discussed literature. For cross-country comparison, payment data are calculated as a percent of the population. The models were estimated for two alternative dependent variables - the number of card payments per person and the number of digital payments per person - which were used to proxy the trends in the use of electronic payments in the ECCU. The potential determinants of electronic payments included: GDP per capita, Point of Sale diffusion ratio, the number of cheques issued per person and cash in circulation as a proportion of M19. The study uses a panel-data linear regression model with fixed effects as seen in Equation below:

Yi,t = αi+βj Xj,i,t + + ui,t Where Y is the dependent variables expressing the two measures of digital payments while α represent the country-specific intercepts. Xj are the potential determinants of electronic payments, βj are the coefficients of the independent variables, and ui,t are the idiosyncratic errors, while i is the country and t is the time period. To determine a suitable procedure for estimation and to avoid problems of spurious regressions, unit root tests are conducted to assess the time series properties of the data being used. The first generation panel unit root test of Levin, Lin and Chu (2002) was used. The results of the unit root tests are shown in Table 1.

TABLE 1: PANEL UNIT ROOT TEST RESLTS VARIABLES Cards Payments Volume Per Capita

LLC -1.478*

GDP Per Capita

-5.300***

Point-of-Sales Terminal Per 1,000 Pop.

-2.926***

Cheque Volume Per Capita Cash in Circulation: M1

-2.060** -3.528***

Currency with the Public: M1

-1.833**

Digital Payments Volume Per Person

3.770***

Note: *** p<0.01, **p<0.05, *p<0.01

72 | Research: The Platform for Innovation, Competitiveness and Growth

One of the challenges in the analysis for small, developing countries like the ECCU, with limited data, is model selection. Panel data techniques often help to expand the database. In the panel approach however, it is useful to begin by testing the presence of fixed effects against random effects. Accordingly, this test was conducted by using the Hausman tests, in which the null hypothesis is that a random effects model is preferred. Following this, diagnostic tests were also conducted for all models. These include Modified Wald tests to determine heteroskedacticity; a Wooldridge test (2002) to test for serial correlation and the Pesaran’s (2004) test to determine the presence of cross-sectional dependence. Cross sectional dependence may occur from spillover effects across ECCU member countries which may be possible given the presence of similar commercial banks and sharing a common central bank. In analyzing the possible determinants of electronic payments used in the region, the paper uses several panel data methods. The standard fixed effect (FE) model serves as a baseline model. In addition, to alleviate the potential challenges of heteroscedasticity, autocorrelation and cross-sectional dependence in the data, the Feasible Generalised Least Squares (FGLS) estimator10 of Parks (1967) and Kmenta (1986) and the pooled OLS regression with Driscoll-Kraay (DK) standard errors are used. Generally, the DK approach helps to correct for heteroscedasticity in the presence of correlations across countries. The results of the FGLS and DK approaches will therefore form the basis for the analysis. The study presents two simple estimation models to analyze the hypothesis, drawing largely from Guariglia and Loke (2004); and Goczek and Witowski (2016). In the first model, the dependent variable is the number of card payments per person. This payment measure is regressed on selected variables including: real GDP per capita – a proxy for income level in each country; currency in circulation to M1, and POS diffusion ratios. The latter serves as a proxy for available technology, and is useful as it has made the use of cards for payments more convenient. The baseline model used for the empirical analysis can be depicted as follows:

Yit = β1 + β2 GDPPCit + β3 POSit + β4 CICit + uit where: Yit = measures the annual number of card payments per capita in country i in year t. In light of the micro and macro literature, a proxy for income was also included in the study. Noting previous results from survey data on U.S. payment habits, Humphrey et al. (2006) noted that individuals with higher real income use non-cash payment instruments more frequently than cash. To address this, the paper used GDP per capita, which measures the log of real per capita GDP for each of the seven countries, and proxies the economic condition. It is expected that higher levels of income per capita may be positively linked to the greater use of non-cash payment instruments. In this study, the point-of-sale diffusion ratio was used (POS) and defined as the number of point of sale (POS) terminals per 1,000 inhabitants. The number of POS terminals may be viewed as a proxy to represent the general availability of the infrastructure necessary to make card payments. Bolt et al. (2008) found that a key factor in the adoption of non-cash use and adoption was availability of POS terminals. A positive sign is expected for this coefficient11. The ratio of cash to M1 was also included to determine whether cash serves as a substitute for noncash payments. While cash payments are difficult to measure in the absence of survey data, the paper follows the example of other studies in using currency in circulation as a proportion of M1 as a proxy12. Accordingly, this variable is expected to have a negative relationship with digital payments, assuming the substitution effect holds. In further exploring the potential of other forms of technology on electronic payments, the paper included a proxy for ATM penetration as another variable in the model, which was defined by the number of ATM A Compilation of Working Papers by OECS Scholars

terminals per inhabitant. It is expected that ATM terminals may influence the use of cards as they are an important medium for the use of cards, by offering greater speed and convenience for deposits, withdrawals and transfers than traditional bank teller services. Previous studies such as Humphrey at al. (1996), Takala and Viren (2007)13, and Guariglia and Loke 2004 considered the role of ATMs in non-cash use. The results were ambiguous regarding whether ATMs encouraged the use of cards or encouraged cash in making payments. Specifically, the use of ATM’s may either encourage households to hold more debit cards, but may similarly encourage more cash withdrawals at the ATM, which could ultimately lead to fewer card payment transactions. A similar approach was taken for the second regression, which captures total digital payments volume per capita as the dependent variable. The general model can be depicted as follows:

Yit = β1+ β2 GDPPCit +β3 CHQit+β4 CICit + uit A variable to capture cheque volume per person was included, which was used to capture the possible substitution effect of cheques with electronic payments, particularly for large-valued transactions. It is expected that cheque volumes may be negatively associated with the use of digital payments. Although most of the previous literature on this subject included social, demographic or psychological variables, such as proportion of youth, security and trust factors, these potential social and demographic factors were not included in the models due to the unavailability of data14. Fixed effects models typically result in the violation of several data assumptions such as heteroskedasticity and cross-sectional dependence. To improve the value of making inferences, regression model’s assumptions are violated which is common with fixed effects, Feasible Generalised Least Squares (FGLS) and Driscoll-Kraay techniques have often been used. The FGLS is able to account for heteroscedasticity as well as dependence in the residuals of time series cross-section models while supporting small scale panels. Additionally, as part of the robustness checks and to ensure greater validity of the results, the two models are also estimated with Driscoll and Kraay (DK) standard errors for fixed effects, which, similar to FGLS, is often recommended when data exhibit cross-sectional dependence. Given that the countries share similar banks and payment options, patterns of dependence are likely to be exhibited among countries in the dataset. Another rationale for using DK-FE, is linked to the possibility of what Hoechle (2007) described as the incorporation in panel regressions of social norms and psychological behaviour patterns as unobservable common factors. Both procedures may be useful but the FGLS is deemed to be more appropriate.

RESULTS AND DISCUSSION This section presents the findings from the two main models related to the factors that determine card payments and overall electronic payments volume in the ECCU. The results of the Hausman tests suggest that fixed effects are more appropriate for all of the models considered. In all of the models, the Hausman’s statistic χ2 proved to be highly significant and supports the rejection of the null hypothesis, which suggests that the Fixed Effect is the preferred model (see results tables). Additionally, the diagnostic checks reveal that the data show a combination of heteroskedasticity, autocorrelation and cross-sectional dependence (see regression tables). These results confirm the need for robust standard errors such as FGLS and DK to alleviate these problems. Tables 2 reports the results of the regression related to card payments volume using per capita GDP, POS diffusion rate and currency in circulation. Column 1 presents estimates for the baseline Fixed Effects, while Columns 2 and 3 present the DK standard errors and the FGLS method which allow for heteroscedasticity and cross-dependence. Since the FGLS is more flexible to small datasets, it was deemed to be most suitable among the three estimation techniques used, and so its results are interpreted. 74 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 2: REGRESSION RESULTS FOR CARD PAYMENT VOLUME PER PERSON VARIABLES GDP per capita

(1) Fixed Effects

(2) Driscoll-Kraay

(3) FGLS

1.059**

0.190**

(0.393)

(0.086)

0.651*

0.235***

(0.151)

(0.336)

(0.078)

-0.089

-0.037

(0.176)

(0.095)

(0.063)

1.059*** (0.203)

Point of Sales per '000 pop. Currency with the Public: M1 Observations R

0.651***

0.520

R2 - Adj

0.461

F-stat

P Number of groups

0.0160

0.912 7

R2 - Within

0.520 22.40

Hausman Test = 21.19 (0.000) Heteroskedasticity test X2 = 4227.53

Pr > X2 = 0.000

Autocorrelation Test F(1,6) = 3.357

Pr > F = 0.107

Cross-sectional dependence test = 7.553

Pr = 0.000

The results of the first model suggest that income per capita and the POS diffusion ratio positively influence card payments volume in the ECCU during the time period, based on all three techniques. Specifically, the FGLS results suggest that increases in the general income level and the more POS terminals (other things being equal) would result in a greater number of card payments. The results are not surprising, possibly reflecting the fact that higher income levels increase the purchasing power of consumers and greater consumption more generally, which would be reflected in more frequent card payment transactions. Additionally, technological innovation and availability, proxied by the diffusion rate of POS terminals, had a consistently positive and highly significant impact on the number of card transactions. The coefficient on cash with the public, which proxies the use of cash in the economy, was negative across all three procedures but was not statistically significant, possibly suggesting that the expected substitution effect was weak during the period. This result may also be likely if there are merchants who do not have POS terminals (due to cost or low volumes of transactions), if there are minimum purchases on card payments or if the habit of using cash may be so entrenched that it may take longer to switch to other payment forms.

A Compilation of Working Papers by OECS Scholars

TABLE 3: REGRESSION RESULTS FOR CARD PAYMENT VOLUME (CURRENCY WITH PUBLIC) VARIABLES

(1) Fixed Effects

GDP per capita

(2) Driscoll-Kraay

(3) FGLS

1.093***

0.167*

(0.389)

(0.088)

0.656*

0.193***

(0.151)

(0.338)

(0.066)

0.021

0.019

(0.128)

(0.089)

(0.050)

1.093*** (0.194)

Point of Sales per '000 pop. Currency with the Public: M1 Observations R

R2 - Adj

0.656***

0.518 0.459

R Within

0.403

F-stat

0.0122

0.915

P Number of groups

7 14.15

Hausman Test = 21.35 (0.000) Heteroskedasticity test X2 = 6144.68

Pr > X2 = 0.000

Autocorrelation Test F(1,6) = 3.358

Pr > F = 0.117

Cross-sectional dependence test = 7.709

Pr = 0.000

Notes: The models are estimated by Fixed Effects, Driscoll-Kraay robust S.E., and Feasible Generalised Least Squares (FGLS). Robust standard errors are reported in parenthesis below each coefficient estimate. The p-values for Hausman test are in parenthesis. A constant term was estimated but it is not reported. *,**,*** denote significance at 10%, 5% and 1% levels As a robustness check, the model is re-estimated using an alternative variable for cash use: the ratio of currency with the public to M1. The results are presented in Table 3, applying the three panel procedures identified previously. Results from the FGLS procedure show that the expected sign and the significance of the coefficients remain broadly similar to the previous regression. GDP per capita and the POS diffusion rate remain significant with the expected positive sign, with POS diffusion rate. The declining significance of the income variable may be attributable to the growing prevalence and frequency in the use of cards to make payments. The coefficient on the cash ratio is positive, but is not statistically significant. The insignificant relation may be related to the potential complementary role of cash with other payment forms, which may support the view that electronic payments do not fully substitute for cash. The findings from the two models suggest that as more merchants begin to introduce POS terminals, it may encourage additional cards transactions by the public. This supports the findings of Guariglia and Loke (2004) who found that the effect of POS terminals was quite significant and larger in the volume regressions compared with value of payments. Practically, this implies that the greater availability of these terminals would enhance the ease and frequency of using cards in making payments. With respect to the limited role of cash as a substitute to card payments, the findings are in line with the findings of Snellman et al. (2000) who noted that the nature of substitution of card payments for cash depend on the stage of each country in its substitution process. Accordingly, it can be deduced that the ECCU may simply be at a lower stage of the transition. In the alternative model which includes ATM penetration variables, the coefficients on GDP per capita and POS terminals variables are positive and statistically significant. However, in the case of ATM variables, the coefficient 76 | Research: The Platform for Innovation, Competitiveness and Growth

was positive and significant from the Fixed Effect and DK procedures, but this relationship failed to hold when the preferred FGLS procedure was applied. This may suggest that the availability of ATM may have encouraged cash withdrawals or may not be widespread enough to induce a change in the number of card transactions. Authors such as Stix (2003) have noted the ambiguous role of ATMs in the demand for cash (and by extension electronic payments). TABLE 4: REGRESSION RESULTS FOR CARD PAYMENT VOLUME (CURRENCY IN CIRCULATION)) VARIABLES GDP per capita

(1) Fixed Effects

(2) Driscoll-Kraay

(3) FGLS

1.423***

0.438***

(0.439)

(0.147)

0.716**

0.293**

(0.305)

(0.289)

(0.128)

-0.249

-0.274

(0.329)

(0.610)

(0.185)

-0.311

-0.311**

-0.083

(0.310)

(0.121)

(0.139)

1.423*** (0.356)

Point of Sales per '000 pop. Cheque volume per capita Currency in Circulation: M1 Observations

0.716**

0.403

R Adj

0.321

R2 - Within

0.403

F-stat

0.000

0.835

Number of groups

0.00122 7

7 31.54***

X2 Hausman Test = 35.33 (0.000) Heteroskedasticity test Chi2 = 9243.78

Pr > X2 = 0.000

Autocorrelation Test F(1,6) = 95.469

Pr > F = 0.000

Cross-sectional dependence test = 3.599

Pr = 0.000

ELECTRONIC PAYMENTS Tables 3 and 4 present the results from the regression using electronic payments volume as the dependent variable. Cheque payments volume is introduced in this model as it was considered as a key substitute for electronic payments in the ECCU such as payroll transfers and standing orders. Similar to the model of card payments, the coefficients on per capita income and POS diffusion ratio retain their positive and statistically significant association. In the case of POS terminals, the positive and significant coefficient could suggest that although many electronic payments (such as standing orders, direct debits) are not linked to POS terminals, it may suggest that greater accessibility to POS terminals may encourage residents to use other forms of electronic payments more generally. A Compilation of Working Papers by OECS Scholars

78 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 5: REGRESSION RESULTS FOR ELECTRONIC PAYMENT VOLUME (CURRENCY WITH THE PUBLIC) VARIABLES GDP per capita

(1)

(2)

(3)

Fixed Effects

Driscoll-Kraay

FGLS

1.526***

0.341**

(0.436)

(0.139)

0.735**

0.261**

(0.306)

(0.300)

(0.112)

-0.240

-0.118

(0.331)

(0.617)

(0.162)

-0.071

0.050

(0.226)

(0.085)

(0.062)

1.526*** (0.342)

Point of Sales per '000 pop. ATM per person Currency with the Public: M1 Observations R

R2 - Adj

0.735**

0.395 0.312

R Within

0.395

F-stat

0.000

0.860

Number of groups

0.006 7

7 15.81**

Hausman Test = 8.36 (0.079) Heteroskedasticity test X2 = 4161.89

Pr > X2 = 0.000

Autocorrelation Test F(1,6) = 85.399

Pr > F = 0.000

Cross-sectional dependence test = 2.625

Pr = 0.009

A Compilation of Working Papers by OECS Scholars

CONCLUSION AND POLICY CONSIDERATIONS The region has witnessed a proliferation of non-cash payment instruments in the last decade – reflected by the widespread use of debit and credit cards, direct debits, electronic transfers and mobile wallets. The aim of the paper was to provide an empirical approach to determine the factors that may help drive card and electronic payments volume in the ECCU. The study is slightly different from other studies in this area by considering fewer variables given data unavailability and idiosyncrasies of the ECCU financial system15. This discussion comes at a time when the region is pursuing ways to capitalize on the efficiency gains from the widespread use of these payment technologies. An understanding of these factors may therefore be useful for policy-making and in making projections regarding the production and issuance of currency. The study modeled the volume of payments as a function of income, the POS diffusion ratio, currency in circulation and cheque transactions. The study covered the period 2005 to 2016 for a panel of seven ECCU countries excluding Montserrat. The results show the positive and significant effect of income and POS terminal diffusion on electronic payments in the ECCU while highlighting that the cash holding did not necessarily fall, using two proxies for cash. The findings suggest that the increase in card payments is driven, in part, by greater availability of financial infrastructure (proxied by point-of-sale terminals) and per capita income. The positive relationship between total card transactions and POS penetration would suggest that increasing the availability of new payment infrastructure may result in greater use of these technologies. Importantly, the weak relationship with currency in circulation suggests the possible absence of a substitution effect between the two forms of payment. This may be due to the use of ATMs, which may have facilitated greater use of cash or the fact that cash continues to be used for smaller transactions, or by small businesses. This was evident in the alternative specification which indicated that the association between the availability of ATM machines and the number of card transactions was not statistically significant. In presenting the findings, there are limitations to the study which merit discussion. The period of study was quite limited, and may not fully reflect more recent trends. Additionally, the research question can be addressed using both micro-level and macro-level data. In the absence of micro-level data in the ECCU, the research only considers the macro-level variables while acknowledging that some of the factors that affect consumer payment decisions may be individual in nature. This shortcoming however should not minimize the value of macro level factors. However, the micro-level data from surveys may lend themselves to more in-depth analysis on the specific reasons for the persistent use of cash in the ECCU as well as the inclusion of social and demographic factors such as age and income which were excluded from this analysis. Indeed, the study by Bagnall et al. (2014) reiterates that the persistent use of cash despite the popularity of electronic alternatives highlights an urgent need for more contemporary studies on consumer payments. With respect to the policy considerations, it is noted that despite the significant strides and potential benefits from using electronic payments, there may still be some reluctance on the part of agents in using these instruments. Electronic payments have generated security and privacy concerns among some individuals. Credit card fraud and infringements on privacy are often considered a negative result of the increase in the use of card payments. Another concern often cited by countries in embracing digital payments, is the challenge it presents for vulnerable groups, particularly the elderly, the poor and illiterate (Chikalipah, 2017)16, who may not have ready access to the technology. Older persons may find these new technologies difficult to use or understand, while others may simply prefer the simplicity, spending control and lower costs of cash relative to payment cards (Arango et al., 2012). These findings suggest that the region may need to craft specific strategies to reduce the use of cash. Given the significance of POS terminals, consideration may be given to expanding the existing (and new) payment technologies and infrastructure such as the introduction of additional POS terminals by smaller merchants, which may allow the digital economy to thrive. Unfortunately, the relatively high fees related to these electronic payments may be viewed as a hindrance to greater use of these instruments in the ECCU by both merchants and citizens. The payment of public sector services17 is one of the primary methods driving the persistence of cheques and cash. This may be reduced with the mandatory acceptance of cards for paying taxes or the use of direct account

80 | Research: The Platform for Innovation, Competitiveness and Growth

transfers for remitting funds for government services. In doing so, there is a need for enhancing awareness of governments, merchants and citizens on the value of moving towards a society with less cash payments, such as speed, efficiency while enhancing trade and interconnectivity. Despite the efficiency gains in moving to electronic payments, it is critical that security and privacy must not be jeopardized. This may require the establishment of a proper regulatory environment to facilitate data security and minimize fraud which would help build trust and encourage greater use of these instruments.

NOTES Cash is anonymous and untraceable, and may be vulnerable to suspicious transactions relative to electronic payments. 1

Generally, electronic payment system refers to an electronic means of making payments for goods and services procured online or in physical locations and facilitates for the secure processing of transactions in real time. 2

Cash in circulation is used here as a proxy and should not be interpreted as the level of cash payments in the ECCU. 3

Given the important findings from the aforementioned research, this study opted to focus on the volume of payments instead of the value of payments. 4

Card payments include credit and debit card transactions, while electronic payments would include card payments as well as direct debits and standing orders made via electronic means by commercial banks. 5

Other electronic payment transactions include direct deposits, automated deposits of payroll, standing orders.

Data for Montserrat was incomplete, so Montserrat was excluded.

The limited time period was somewhat mitigated by the use of panel data techniques.

While interest rates were considered and have been used in the previous literature, given the sticky nature of interest rates in the ECCU, the interest rate variable was omitted given the limited practical value. 9

While the two methods control for the aforementioned data challenges, the FGLS is seen as more appropriate for smaller samples. 10

Previous studies Humphrey at al. and Guariglia and Loke (2004) considered the role of ATMs in non-cash use.

Other variables used in previous studies include currency in circulation to GDP.

Takala and Viren (2007) take a slightly different approach by considering how ATMs have contributed to the use of cash. 13

Considerations may be given to proxies such as homicide rate per 1,000 for crime and the level of non-performing loans for trust. 14

For instance, the sticky nature of interest rates.

Chikalipah (2017) found that illiteracy was one of the main reasons for financial exclusion in Sub-Saharan Africa.

Such as the payment of taxes, licenses, goods and services.

A Compilation of Working Papers by OECS Scholars

REFERENCES Akinyemi, I.O., Asani, E. and Adigun, A.A., 2013. An investigation of users’ acceptance and satisfaction of the e-banking system as a panacea towards a cashless economy in Nigeria. Journal of Emerging Trends in Computing and Information Sciences, 4(12), pp.954-963. Arango, C., Huynh, K.P., Fung, B. and Stuber, G., 2012. The changing landscape for retail payments in Canada and the implications for the demand for cash. Bank of Canada Review, 2012(Autumn), pp.31-40. Arellano, M. and Bond, S., 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The review of economic studies, 58(2), pp.277-297. Arvidsson, N., 2018. The future of cash. The Rise and Development of FinTech: Accounts of Disruption from Sweden and Beyond, pp.85-98. Bagnall, J., Bounie, D., Huynh, K.P., Kosse, A., Schmidt, T., Schuh, S.D. and Stix, H., 2014. Consumer cash usage: a crosscountry comparison with payment diary survey data. Bech, M.L., Faruqui, U., Ougaard, F. and Picillo, C., 2018. Payments are a-changin’ but cash still rules. BIS Quarterly Review, March. Berger, A.N., 2003. The economic effects of technological progress: Evidence from the banking industry. Journal of Money, credit and Banking, pp.141-176. Boeschoten, W.C., 1998. Cash management, payment patterns and the demand for money. De economist, 146(1), pp.117-142. Borzekowski, R., Elizabeth, K.K. and Shaista, A., 2008. Consumers' use of debit cards: patterns, preferences, and price response. Journal of Money, Credit and Banking, 40(1), pp.149-172. Payment, S.O., Clearing and Settlement Systems In the CPMI Countries. Available at http://www.bis.org/ Bulo, D.C., 2018. Payment System Innovations and Economic Growth: An Empirical Study of the BRINCS Countries. International Journal of Innovative Research and Development, 7(3). Capgemini and BNP Paribas (2017), World Payments Report 2017 available at https://www.capgemini.com/fr-fr/wp-content/ uploads/sites/2/2017/10/world-payments-report-2017_year-end_final_web-002.pdf Chavan, J., 2013. Internet banking-benefits and challenges in an emerging economy. International Journal of Research in Business Management, 1(1), pp.19-26. Chikalipah, S., 2017. What determines financial inclusion in Sub-Saharan Africa? African Journal of Economic and Management Studies, 8(1), pp.8-18. Cuthbertson, K., 1991. Modelling the demand for money. Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, pp.319-340. Duca, J.V. and Whitesell, W.C., 1995. Credit cards and money demand: A cross-sectional study. Journal of Money, Credit and Banking, 27(2), pp.604-623. Fishbein, M. and Ajzen, I., 1975. Belief, attitude, intention and behavior: An introduction to theory and research. Friedman, B.M., 1999. The future of monetary policy: the central bank as an army with only a signal corps? International finance,

82 | Research: The Platform for Innovation, Competitiveness and Growth

2(3), pp.321-338. Goczek, L. and Witkowski, B., 2016. Determinants of card payments. Applied Economics, 48(16), pp.1530-1543. Guariglia, A. and Loke, Y.J., 2004. What determines the value and volume of noncash transactions? Evidence from a panel of European and North American countries. Applied economics, 36(4), pp.291-303. Hasan, I., De Renzis, T. and Schmiedel, H., 2012. Retail payments and economic growth. Bank of Finland research discussion paper, (19). Hasan, I., Schmiedel, H. and Song, L., 2012. Returns to retail banking and payments. Journal of Financial Services Research, 41(3), pp.163-195. Hasan, I., De Renzis, T. and Schmiedel, H., 2013. Retail payments and the real economy. Humphrey, D.B., Pulley, L.B. and Vesala, J.M., 1996. Cash, paper, and electronic payments: a cross-country analysis. Journal of Money, Credit and Banking, 28(4), pp.914-939. Humphrey, D.B., Sato, S., Tsurumi, M. and Vesala, J.M., 1999. The evolution of payments in Europe, Japan, and the United States: lessons for emerging market economies. The World Bank. Humphrey, D.B., Kim, M. and Vale, B., 2001. Realizing the gains from electronic payments: Costs, pricing, and payment choice. Journal of Money, Credit and Banking, pp.216-234. Humphrey, D.B., Pulley, L.B. and Vesala, J.M., 2000. The check's in the mail: Why the United States lags in the adoption of cost-saving electronic payments. Journal of Financial Services Research, 17(1), pp.17-39. Humphrey, D., Willesson, M., Bergendahl, G. and Lindblom, T., 2006. Benefits from a changing payment technology in European banking. Journal of Banking & Finance, 30(6), pp.1631-1652. Huynh, K.P., Schmidt-Dengler, P. and Stix, H., 2014. The role of card acceptance in the transaction demand for money. Jonker, N., Kosse, A. and Hernández, L., 2012. Cash usage in the Netherlands: How much, where, when, who and whenever one wants? (No. 1002). Netherlands Central Bank, Research Department. King, M., 1999. Challenges for monetary policy: new and old. Quarterly Bulletin-Bank of England, 39, pp.397-415. Klee, E., 2008. How people pay: Evidence from grocery store data. Journal of Monetary Economics, 55(3), pp.526-541. Kmenta, J. and Balestra, P., 1986. Missing measurements in a regression problem with no auxiliary relations. Advances in econometrics, 5, pp.289-300. Lai, P.C., 2017. The literature review of technology adoption models and theories for novelty technology. JISTEM-Journal of Information Systems and Technology Management, 14(1), pp.21-38. Legris, P., Ingham, J. and Collerette, P., 2003. Why do people use information technology? A critical review of the technology acceptance model. Information & Management, 40, pp.191-204. Levin, A., Lin, C.F. and Chu, C.S.J., 2002. Unit root tests in panel data: asymptotic and finite-sample properties. Journal of econometrics, 108(1), pp.1-24. Liu, J., Kauffman, R.J. and Ma, D., 2015. Competition, cooperation, and regulation: Understanding the evolution of the mobile payments technology ecosystem. Electronic Commerce Research and Applications, 14(5), pp.372-391. Manyika, J., Lund, S., Bughin, J., Woetzel, J.R., Stamenov, K. and Dhingra, D., 2016. Digital globalization: The new era of global flows (Vol. 4). San Francisco, CA: McKinsey Global Institute.

A Compilation of Working Papers by OECS Scholars

Morris, M.G. and Venkatesh, V., 2000. Age differences in technology adoption decisions: Implications for a changing workforce. Personnel psychology, 53(2), pp.375-403. Nasri, W., 2011. Factors influencing the adoption of internet banking in Tunisia. International Journal of Business and Management, 6(8), p.143. Patel, H. and Connolly, R., 2007. Factors influencing technology adoption: A review. In 8th International Business Information Management Conference, Dublin, Ireland. Vol. 15, p. 2018. Parks, R.W., 1967. Efficient estimation of a system of regression equations when disturbances are both serially and contemporaneously correlated. Journal of the American Statistical Association, 62(318), pp.500-509. Rogoff, K.S., 2017. The curse of cash: How large-denomination bills aid crime and tax evasion and constrain monetary policy. Princeton University Press. Runnemark, E., Hedman, J. and Xiao, X., 2015. Do consumers pay more using debit cards than cash? Electronic Commerce Research and Applications, 14(5), pp.285-291. Skinsley, C., 2016. Should the Riksbank issue e-krona? Speech at Fintech Stockholm, November 2016 Snellman, J.S., Vesala, J.M. and Humphrey, D.B., 2001. Substitution of non cash payment instruments for cash in Europe. Journal of Financial Services Research, 19(2-3), pp.131-145. Stavins, J., 2001. Effect of consumer characteristics on the use of payment instruments. New England Economic Review, February. Stix, H., 2004. How do debit cards affect cash demand? Survey data evidence. Empirica, 31(2-3), pp.93-115. Takala, K. and Viren, M., 2007. Impact of ATMs on the Use of Cash. Available at SSRN 1017750. Van der Cruijsen, C., Hernandez, L. and Jonker, N., 2017. In love with the debit card but still married to cash. Applied Economics, 49(30), pp.2989-3004. Venkatesh, V., Morris, M.G. and Ackerman, P.L., 2000. A longitudinal field investigation of gender differences in individual technology adoption decision-making processes. Organizational Behavior and Human Decision Processes, 83(1), pp.33-60. Visa, Roubini ThoughtLab Consultants, 2017. Cashless Cities: Realizing the Benefits of Digital Payments. Available at https:// usa.visa.com/dam/VCOM/global/visa-everywhere/documents/visa-cashless-cities-report.pdf Von Kalckreuth, U., Schmidt, T. and Stix, H., 2014. Choosing and using payment instruments: evidence from German microdata. Empirical Economics, 46(3), pp.1019-1055. Wakamori, N. and Welte, A., 2017. Why do shoppers use cash? Evidence from shopping diary data. Journal of Money, Credit and Banking, 49(1), pp.115-169.

84 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX FIGURE A: CARD TRANSACTIONS AND POINT OF SALES

FIGURE B: CASH IN CIRCULATION AND POINT OF SALES

A Compilation of Working Papers by OECS Scholars

TABLE 6: REGRESSION RESULTS FOR CARD PAYMENT VOLUME (ATM PER PERSON) VARIABLES

(1)

(2)

(3)

Fixed Effects

Driscoll-Kraay

FGLS

0.978***

0.179**

(0.278)

(0.089)

0.500***

0.201***

(0.215)

(0.070)

0.043***

-0.001

(0.007)

(0.010)

(0.004)

-0.016

0.026

(0.125)

(0.076)

(0.050)

GDP per capita

0.978*** (0.211)

Point of Sales per '000 pop.

0.500*** (0.094)

ATM per person

0.043***

Currency with the Public: M1 Observations R

R2 - Adj

0.657 0.640

R Within

0.657

F-stat

0.000

0.967

Number of groups X

0.004 7

7 14.88***

86 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Martina Regis

Economist Eastern Caribbean Central Bank

Martina Regis is currently an Economist in the Research Department at the Eastern Caribbean Central Bank (ECCB), with several years of professional experience in research and public policy. Her research interests span several macroeconomic issues including fiscal policy, regional financial integration, financial market development as well as public policy. Prior to joining the Research Department, Martina was a member of the ECCB’s Financial and Enterprise Development Department, where she was responsible for conducting research, and providing policy advice and technical assistance on regional money and capital market development. Martina holds an M.Sc. from the University of Birmingham and a B.Sc. from the University of the West Indies.

A Compilation of Working Papers by OECS Scholars

5 Assessing the Long Term Planning Requirements for the Water Sector in Saint Lucia Alice Providence and Gemma Edwin

88 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT This paper analyzes the long term planning requirements for the Water Sector in Saint Lucia based on UN projected population growth and Tourism demand projections using the National Infrastructure Systems Model (NISMOD). The National Infrastructure Systems Model is a tool developed by the UK Infrastructure Transitions Research Consortium which utilizes an evidence-based approach to assess long-term investment strategies which are required for infrastructural sectors as well as climate change risk and vulnerability in cross-sectoral infrastructure networks. NISMOD comprises a series of analysis which are needed to provide evidence to support infrastructure investments and policy decisions in Saint Lucia, and provides cross sectoral initiatives and investment strategies to sustain the Water Sector in Saint Lucia. INTRODUCTION The increase in the world’s population and the resulting impact on the ability of water resources to meet this growing demand is one of the direst issues currently facing the world. Sustainable goal number 6 advocates for the “availability and sustainable management of water and sanitation for all.” According to the United Nations World Water Development Report 2019, Water consumption has been increasing worldwide by approximately 1% per year since the 1980’s and this is primarily as a consequence of socio-economic developments and population growth. Economic developments through industrialization have also resulted in a corresponding increase in water consumption which threatens the availability of water resources to meet the increased pressures resulting from vast increases in demand. Undoubtedly, as supported by SDG goal number 6 water availability is pivotal for promoting sustainable development, livelihoods, and the maintenance of the world’s ecosystems. Goswami (2017) posits that the management of water resources contributes to greater certainty and efficiency in productivity across all economic sectors and contributes to the health of the nation. The rise in global temperatures have now become an imminent threat and this is manifested through exacerbating dry seasons and increases in the frequency and intensity of flooding occurrences worldwide, thus compromising water supply globally. Within the Caribbean region, Saint Lucia and its regional counterparts experience water supply deficits and this is further compounded by climate change impacts. The geographical location and size of small island developing states (SIDS) makes them particularly vulnerable to natural hazards which threatens economic resiliency and human survival resulting in negative spillovers for all social and economic sectors within the region. It is anticipated that the water deficit gap will widen due to several factors including increased tourism arrivals, economic activity, industrialization and population growth. Inefficient water infrastructure, limited fiscal space, lack of implementation of land use policies and high levels of non-revenue water (NRW) are also contributing factors which exacerbate water supply deficits. The Saint Lucia Sectoral Adaptation Strategy and Action Plan for the Water Sector (2018-2028) states that over the years the overall growth of the tourism sector and increases in the population have resulted in marked increases in surface water consumption. This study will therefore investigate the water demand projections for Saint Lucia and prescribed policy interventions which are required to ensure that Saint Lucia plans ahead to minimize water insecurities in the future. To do this, population and tourist projections will be utilized using the National Infrastructure Systems Model. The National Infrastructure Systems Model was developed by the UK Research Transitions Consortium and has been utilized to guide investments in Infrastructure in the United Kingdom. This paper will present the case for Saint Lucia focusing solely on the water Sector using an evidence-based approach. MOTIVATION/OBJECTIVE The general objective of this study is to assess the planning requirements required for the sustainability of the water sector in Saint Lucia. A Compilation of Working Papers by OECS Scholars

The Specific objectives of the study are to: 1. Utilize the NISMOD model which will assess the water needs for Saint Lucia. 2. Identify how assumptions of increased population growth and tourism arrivals will increase the demand for water. 3. Depict an imminent future water supply deficit based on the inability of Saint Lucia’s current water supply to meet future increased demand. 4. Draw relevant policy recommendations for the sustainability of the water sector to sustain livelihoods and socio-economic development. LITERATURE REVIEW One continuous challenge in planning infrastructure investments is the level of complexity associated with the interacting systems with general considerations for the risk to these and connected infrastructures. Several modeling approaches have been utilized prior to the NISMOD in an attempt to inform infrastructure planning with consideration for the interdependent nature of the critical infrastructure. Empirical approaches sought to assess historical data on critical infrastructure’s failures and accidents to determine a frequency and significance of failure pattern. (McDaniels et al. 2007 – empirical framework/ Wallace et al. 2003 – managing disruptions). This empirical based risk analysis in quantifying the consequences of interdependent failures would thus provide alternatives to minimize risk. These approaches however could not account for intangible interdependencies, were only able to assess after a significant failure therefore limiting the ability to model and plan for new technologies or new threats / disasters, and index calculations were dependent on data from varying databases without a standardized data collection methodology. (Ouyang 2014 – review of modeling and simulation). Agent based approaches to infrastructural interdependencies sought to take a bottom-up approach in modeling the behaviors of decision makers and system participants, utilizing discrete-event simulations in capturing varying interdependencies among critical infrastructural systems. (Brown 2001- Assessing Infrastructure) Through this data modeling approach, what-if scenarios can be assessed as well as the effectiveness of different control strategies. Modeling expected behavior proposes several challenges chief of which is the possibility of an infinite number of permutations which leave assumptions open to scrutiny and prove to be unjustifiable theoretically or statistically (Ouyang 2014). The economic theory-based approach adopted Leontief’s proposed input-output economic model in the assessment of the interdependencies between critical infrastructure. The resultant Input-output Inoperability Model (IIM) defined inoperability as the inability of an infrastructural system to perform its intended function (Haimes & Jiang 2001 – Leontief based model risk). In so doing the model calculated the output of the risk of failure at which point the infrastructure would become inoperable. The Computable General Equilibrium (CGE) model (Rose & Liao 2005) built on this component allowing the substitution of inputs in response to market changes based on the inclusion of economic resilience derived from the producer’s production functions. This modeling method proved too high-level and difficult to calibrate in areas where relevant data was limited. The additional dependence on external sources of derivation of elasticity values to adequately assess the resiliency component of the model further limited the accuracy and availability of such data for the analysis. These methodologies focused on short term, single sectors or singular scenario analysis. Further, these models were unable to account for the changing capacities of existing infrastructure to meet future 90 | Research: The Platform for Innovation, Competitiveness and Growth

demands through a cross-sectoral analysis of infrastructural services. METHODOLOGY The National Infrastructure Systems Model (NISMOD) is a systematic methodology used to analyze the performance of National infrastructure systems. National Infrastructure Systems Modelling (NISMOD) capability, which has been developed by the Infrastructure Transitions Research Consortium (ITRC) - a UK based research consortium, led by the University of Oxford. This model was designed to provide useful insights into the cross-sectoral performance of Saint Lucia’s infrastructure systems. This includes, with respect to: 1. The structure and function of current infrastructure systems. 2. The characterization of future infrastructure needs by developing scenarios of the future. 3. The development of a future infrastructure-related vision for Saint Lucia. 4. The identification of strategic alternatives to meet future needs. 5. Analysis of the performance of strategic alternatives. 6. The assembly of adaptive pathways of infrastructure investments and policies. DATA ● United Nations projected population growth data. ● Government of Saint Lucia; Ministry of Tourism projections for tourism arrivals. ● Water and Sewage Company Inc. water production and consumption data. UN PROJECTED POPULATION GROWTH DATA Drivers of infrastructure needs: Residential population forecast Residential population growth is a driver which places equal demand on infrastructure across all different sectors. Saint Lucia residential population forecasts were obtained from the World Population Prospects: The 2019 Revision by the United Nations. The scenarios are depicted below as high (high fertility rate), moderate (moderate fertility rate) and low (low fertility rate).

A Compilation of Working Papers by OECS Scholars

FIGURE 1: SAINT LUCIA POPULATION FORECAST (2016-2050)

PROJECTIONS FOR TOURISM ARRIVALS Drivers of infrastructure needs: Tourism arrivals forecast Future tourism growth in Saint Lucia was assumed to result from the expansion of the Hewanorra International Airport and the expansion of the sea port in Castries to accommodate quantum class vessels which would simultaneously increase tourist arrivals. A moderate growth scenario, low growth rate corresponding to a base scenario of minimal growth (0.1%), and a high growth rate in line with the average Caribbean market growth (4.4%). Historically, statistics have shown that each stay-over tourist stays for an average of 9 nights (2010- 2017 Annual Trend Fact Sheet). In order to arrive at peak calculations, the peak tourist month (December) has been utilized.

FIGURE 2: SAINT LUCIA STAY OVER TOURISTS FORECAST (2016-2050)

92 | Research: The Platform for Innovation, Competitiveness and Growth

A Compilation of Working Papers by OECS Scholars

Three infrastructure-led cruise-ship scenarios were utilized as a Base, Medium and Optimistic case based on berth enhancement and tourist infrastructure for the cruise port until 2042. Based on historical trends each cruise ship tourist is anticipated to stay for an average of 1 day. The peak tourist arrival month December was utilized to arrive at peak calculation figures. Other air and sea terminals (George F.L Charles Airport, and Rodney Bay, Marigot Bay, Soufriere Marinas) assume no significant expansion in capacity.

FIGURE 3: SAINT LUCIA CRUISE SHIP TOURIST FORECAST (2016-2050)

For input into the NISMOD, the total medium scenario resident and tourist population projections were utilized.

FIGURE 4: SAINT LUCIA COMBINED RESIDENT AND TOURIST FORECAST (2016-2050)

94 | Research: The Platform for Innovation, Competitiveness and Growth

WATER PRODUCTION AND CONSUMPTION DATA In order to assess the water supply capacity data was obtained for all water treatment plants and raw intakes in Saint Lucia. According to a World Bank 2014 report, the Theobalds treatment plant supplies approximately 65% of the island’s total water supply. Other water treatment plants and raw water intakes on the island account for the remaining 35% of water production.

TABLE 1: WATER SUPPLY DATA

Sector

Type

Treatment Plant

Unit

Capacity

water

infrastructure

Theobalds TP

m3/year

11,615,259.95

water

infrastructure

Other WTP

m3/year

10,151,737.20

water

infrastructure

Raw Intakes

m3/year

204,096.71

Total

m^3/year

21,971,093.86

TABLE 2: WATER CONSUMPTION DATA

Sector

Type

Unit

Year

Capacity

water

Domestic

m3/year

2018

4,968,303.92

water

Commercial

m3/year

2018

1,282,517.84

water

Government

m3/year

2018

1,140,667.22

Hotel

m3/year

2018

1,518,516.50

Boats

m3/year

2018

92,866.55

NRW (real)

m3/year

2017

6,072,846.41

NRW (apparent)

m3/year

2017

5,385,354.36

Total

m^3/year

2017

20,461,072.81

Due to the unavailability of water demand data, it was assumed that Water demand equated consumption in order to perform the analysis. Real and apparent non-revenue (NRW) figures were also utilized to account for demand.

A Compilation of Working Papers by OECS Scholars

FUTURE NEEDS: WATER SECTOR Future needs for the water sector for Saint Lucia were based on the sum of resident and tourist needs using the values obtained for total water demand in the country and assumptions from other literature to inform growth projections. Per capita usage was derived by dividing total consumption (domestic, commercial and government) by the 2018 population of Saint Lucia of 181,889. Annual water capacity needs were therefore calculated by multiplying the per-capita value and per-tourist values by the respective low, moderate and high scenarios for future residential and tourist growth.

TABLE 3: LOW, MODERATE AND HIGH DATA SCENARIOS

Scenario 2016

2017

2018

2019

low

98,561

100,982

102,175

moderate

98,642

101,204

high

98,769

101,472

2020

2025

2030

2035

102,665 103,123

104,939

106,063

105,687

102,575

103,981 105,371

112,040

118,233

121,759

103,002

105,305 107,618

119,139

130,408

138,483

The figure below shows water demand forecasts denoted by the red line and supply capacity highlighted by the darker blue and light blue areas on the graph below. From the graph it can be seen that at roughly the year 2027, water demand will increase exceeding the current supply capacity of water on the island given the assumed increases in population growth and tourist arrivals. These projections are based on all water treatment plants (WTPs) operating at full capacity.

FIGURE 5: NISMOD OUTPUT

96 | Research: The Platform for Innovation, Competitiveness and Growth

However, much of Saint Lucia’s rural communities experience decreases in water availability during the dry months (December – May). During this period, river intake decreases by approximately 40% on average; coupled with the high rate of non-revenue water loss; water availability is decreased by approximately 60%. The figure below highlights the issues of water security during the dry months.

FIGURE 6: NISMOD OUTPUT

Utilizing the modeling capacities of the NISMOD, an assessment of confirmed and proposed future infrastructural investments in the water sector were introduced to the model to assess these projects ability to meet the current and projected demand primarily in the dry season. The figure below outlines the findings of this assessment which highlighted the need for more aggressive interventions in order to address the current and impending water security issues in Saint Lucia. CONCLUSION AND POLICY RECOMMENDATIONS The National Infrastructure Systems Model has shown that the annual demand for Water in Saint Lucia will exceed the available supply of water by 2027. This will be further expedited by climate related events and non-climatic drivers including increases in non-revenue water and inefficiencies within the water distribution system in Saint Lucia. A National Infrastructure Strategy (NIS) approach which entails effluent reuse, increased storage, demand efficiencies, desalination, and improved water transmission are recommended as Long-Term Planning requirements for the Water Sector. The implementation of NIS approach to water infrastructure would further enable Saint Lucia to make considerable strides towards accomplishing Sustainable Development Goal Target 6.1: universal and equitable access to safe and affordable drinking water for all; 6.4: substantially increase water-use efficiency across all sectors and ensure sustainable withdrawals and supply of freshwater to address water scarcity and substantially reduce the number of people suffering from water scarcity.

A Compilation of Working Papers by OECS Scholars

FIGURE 7: NISMOD OUTPUT

Meeting future water needs in Saint Lucia during the dry season highlights the importance of capital investments, energy-intensive desalination, demand-side interventions and policies to reduce water use and increase the efficiency of the water supply network. A focus on water use demand reductions will maintain a sustainable supply in years to come. Climate change adds further risks and uncertainties to the global picture requiring the adoption of adaptive management in water resources based on monitoring and reevaluation. Although climate change may be perceived as a long-term problem, it needs to be addressed now as decisions today affect society’s ability to adapt to increasing variability in tomorrow’s climate. If we are to balance freshwater supply with demand and protect the integrity of aquatic ecosystems, a fundamental change in current wasteful patterns of consumption and high levels of wastage post production is needed. Recognition of the links between rapidly growing populations and shrinking freshwater supplies is the essential first step in making water use sustainable. The following policies are therefore recommended for the Water Sector for improved performance and sustainability. ● Implementation of Land Use policies to protect water supply sources through land conservation for the preservation of watershed/water intake areas. ● Incentives to encourage rain water harvesting. The proposed rainwater harvesting initiative will reduce dependence on the utility’s water distribution system by enhancing resilience during periods of water intermittency and shortages, during dry spells. ● Enforcement of regulations for water abstraction for the maintenance of water levels and stream flows. ● Improved governance and maintenance of critical water infrastructural assets. ● Effluent reuse from wastewater. ● Required capacity increases through desalination or storage. ● Enhanced enabling environment and improved behaviour for water-related climate adaptation action.

98 | Research: The Platform for Innovation, Competitiveness and Growth

Sustainable Development Goal 1.4 highlights the importance of access to basic services as an equal right recognizing that access to resources enables all residents to improve their livelihoods. Focused efforts on water security through the implementation of the recommended policy changes and an aggressive National Infrastructure Strategy are necessary to ensure a more water secure Saint Lucia.

REFERENCES Brown T, Beyeler W, Barton D. (2004). Assessing infrastructure interdependencies: the challenge of risk analysis for complex adaptive systems. International Journal of Critical Infrastructure 2004; 1(1):108–17. Government of Saint Lucia (2011). Critical Infrastructure and Key Resources: Saint Lucia Profile. Haimes, Y. Y., & Jiang, P. (2001). Leontief-based model of risk in complex interconnected infrastructures. Journal of Infrastructure systems, 7(1), 1-12. McDaniels, T., Chang, S., Peterson, K., Mikawoz, J., & Reed, D. (2007). Empirical framework for characterizing infrastructure failure interdependencies. Journal of Infrastructure Systems, 13(3), 175-184. Ouyang, M. (2014). Review on modeling and simulation of interdependent critical infrastructure systems. Reliability engineering & System safety, 121, 43-60. Peters, E. J. (2015, May). Wastewater reuse in the Eastern Caribbean: a case study. In Proceedings of the Institution of Civil Engineers-Water Management (Vol. 168, No. 5, pp. 232-242). Thomas Telford Ltd. Rose, A., & Liao, S. Y. (2005). Modeling regional economic resilience to disasters: A computable general equilibrium analysis of water service disruptions. Journal of Regional Science, 45(1), 75-112. Tapper, R., Hadjikakou, M., Noble, R., & Jenkinson, J. (2011). The impact of the tourism industry on freshwater resources in countries in the Caribbean, Mediterranean, North Africa and other regions. Report to the Travel Foundation. DESA, U. (2010). United Nations Department of Economic and Social Affairs/Population Division (2009b): World Population Prospects: The 2008 Revision. Internet: http://esa. un. org/unpp (gelesen am 16. Wallace, W. A., Mendonça, D., Lee, E., Mitchell, J., & Chow, J. (2001). Managing disruptions to critical interdependent infrastructures in the context of the 2001 World Trade Center attack. Impacts of and Human Response to the September 11, 2001 Disasters: What Research Tells Us. Wasco by Santander Investment Services (2016). Project for the Private Sector Participation Transaction in Saint Lucia’s Water Utility. Diagnostic of the Utility. World Bank (2014). Request for expression of interest for selection. Auditor Hewanorra International Airport and George F.L. Charles Airport World Bank (2014). Saint Lucia Flood Event of December 24-25, 2013 Amsweb. Saint Lucia Business Focus, (2015, September 22). Redeveloping Hewanorra International Airport. Retrieved from http://businessfocusstlucia.com/ redeveloping-hewanorra-international-airport/ SLASPA (n.d.). Why Redevelop HIA. Retrieved from http://www.slaspa.com/ content Pages/ view/why-redevelop-hia Airports Worldwide, (2010). Hewanorra Intl Airport. Retrieved from http://www. airportsworldwide.com/saint_lucia/ hewanorra_intl_saint_lucia.php Peter, J. Ministry of Infrastructure, Ports, Energy and Labour (2018, January 19). Berth expansion to aid cruise

A Compilation of Working Papers by OECS Scholars

visitor increase. Retrieved from http://infrastructure.govt.lc/news/berth-expansion-to-aid-cruise-visitor-increase Annual Trend Fact Sheet SLTB (2010-2017) Caribbean Development Bank, CDB, (2014, December 15). Grant to further develop Vieux Fort, Saint Lucia water supply systems: Caribbean Development Bank. (2014, December 15). Retrieved from https://www.caribank.org/ newsroom/news-and-events/cdb-grant-further-develop-vieux-fort-st-lucia-water-supply-systems Government of Saint Lucia (2017). Economic and Social Review 2018. Retrieved from https://www.finance.gov.lc/ resources/view/2076 Government of Saint Lucia (2018). Economic and Social Review 2019. Retrieved from https://www.finance.gov.lc/ resources/view/2089 Saint Lucia Tourist Board, (2016). Tourist Arrivals Report. Retrieved from http://www.caribbeanhotelandtourism. com/wp-content/uploads/2017/02/St.-Lucia-Year-to-date-Stay-Over-Arrivals-December-2016.pdf Saint Lucia’s National Adaptation Plan (NAP) 2018-2028. (2019, January 7). Retrieved from https://www. latinamerica.undp.org/content/rblac/en/home/library/ environment_energy/saint-lucia-s-national-adaptation-plan-nap--2018-2028.html UK transitions Research Consortium (2019). The next generation of national infrastructure planning. Retrieved from https://www.itrc.org.uk/

100 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Alice Providence

Economist, National Integrated Planning and Program Unit Ministry of Finance, Government of Saint Lucia

Alice Providence has been an Economist with the National Integrated Planning and Program Unit (NIPP) of the Ministry of Finance, Job Creation, External Affairs and the Public service for approximately two years. She is responsible for providing policy advice and recommendations to the Director of The NIPP on Integrated Infrastructural Planning in Saint Lucia. Ms. Providence has more than nine years working experience in the Banking sector, having commenced her tenure as a Customer Service Representative and subsequently assuming the role of Operations Officer within the Operations Unit of the Bank of Saint Lucia. Ms. Providence earned a Master’s degree in Project Management as an OAS Scholar at the University for International Cooperation in Costa Rica. She possesses a Bachelor’s degree in Economics with Minors in Finance and Human Resource Management from the University of the West Indies St. Augustine Campus, Trinidad. She also possesses various training certificates in banking related courses. Ms. Providence is passionate about the socio-economic development of Saint Lucia and the region as a whole and hopes that her work in Economics can contribute to development through the translation of research and theory into policy to improve standards of living for the betterment of all.

A Compilation of Working Papers by OECS Scholars

101

6 Climate Change and its Impact on the Agriculture Sector in Saint Lucia Petriana Daniel

102 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT This paper empirically examines the determinants of commercial bank deposits in Saint Lucia using data The Agricultural sector has been a large contributor to economic growth and development within the Caribbean region, however, its contribution to GDP has been on a downward trajectory. The recent phenomena of increasing frequency of natural disasters due to climate change has impacted economies worldwide. There is the assumption that climate related disasters adversely impact the agricultural sector in particular. This paper uses the Autoregressive Distribution Lag (ARDL) model to evaluate the determinants of the agriculture sector’s performance in Saint Lucia and the Hodrick Prescott filter to compare the sector’s performance relative to potential. The findings suggest that socio economic, infrastructural, institutional and technological factors are significant to agricultural production. However, environmental factors, measured by disasters and rainfall, are only significant in the short run. In addition, the Hodrick Prescott filter indicates that the average annual production is below potential. Policies should then be geared at building resilience and improving the institutional and socio-economic factors to improve agriculture production in Saint Lucia. BACKGROUND/ INTRODUCTION Climate change refers to significant changes in global temperature, precipitation, wind patterns and other measures of climate that occur over several decades or longer (University of California, Davis, 2019). It has become a global concern because of the increase in severity and frequency of natural disasters, reduction in production of agricultural products as well as destruction of indigenous species of both plants and animals. The Food and Agriculture Organization (FAO) estimates the global losses to the agriculture sector as a result of natural disasters amounts to approximately USD $ 100 Billion from 2004 to 2014 (Food and Agriculture Organization of the United Nations, 2016). The Caribbean region has been impacted by this recent phenomenon due to the myriad of natural events occurring as hurricanes, storms, droughts, floods and change is ocean temperature. In the most recent past, countries such as Dominica, Grenada and St. Vincent and the Grenadines have suffered major damages from natural disasters ranging from floods, drought, hurricanes etc. For example, in 2017 Hurricane Maria hit Dominica with damages amounting to an estimated $179.62 million USD which required an estimated total recovery effort of 88.5 Million USD. Likewise, Saint Lucia was hit by Hurricane Tomas in October 2010 resulting in damages of approximately $60 million USD. The frequency of these events in the region more so in the case of Saint Lucia has impacted productive sectors as agriculture due to the loss in land, the damages of crops and the cost to rehabilitate farms. Saint Lucia’s agriculture sector contribution to total GDP has been on a downward path, moving from 11.6% in 1995 to 5.92% in 2001, and has continued to decline registering a contribution of 1.9% to GDP in 2018. Natural disasters have become more catastrophic over the years and is expected to worsen in the future. It therefore remains a threat to the Caribbean region that is neither distant nor conceptual- the signs are apparent and the effects are inevitable. Saint Lucia as well as the rest of the region remains vulnerable to natural disasters, because of its geographic location, the size and the economy's reliance on the agriculture sector. Hence, the survival of the agriculture sector in Saint Lucia and across the region is essential for food security. It is against this backdrop that the paper seeks to assess the impact of climate change related disasters and other variables as institutional, infrastructural, technological and socio-economic factors on the agriculture sector in Saint Lucia. An Autoregressive Distributed Lag (ARDL) co-integration technique will be used to assess the determinants of agriculture production and the effects of climate change related events on the sector in Saint Lucia. Kumar and Preitte in 2013 used regression to discuss the impact of climate variation on agricultural productivity and food security in rural India (Kumar & Pritee, 2013). The paper then uses a Hodrick Prescott filter to assess the sectors potential GDP to then measure its impact relative to potential. This paper is therefore structured as follows; section 3 highlights the expectations and predictions of natural disasters. Section 4 then discusses the literature surrounding the topic while section 5 assesses A Compilation of Working Papers by OECS Scholars

103

the stylized facts and the trends based on data compiled. The methodical approach and the results are detailed in section 6. Section 7 discusses the findings, government’s climate change resilience financing and specific recommendations (processes and procedures) to build resilience in agriculture. EPISODES OF NATURAL EVENTS IN THE REGION At a category 3 hurricane Ivan in 2004 hit Grenada leaving the agriculture sector with virtually complete loss of the entire year’s crop of nutmeg. Prior to Ivan, Grenada, was the second largest nutmeg producer in the world, suffered an overall 85 percent damage of which 60 percent were completely destroyed. The other two main export crops of which bananas were completely wiped out while 60 percent of the cocoa trees were damaged (Latin America and the Caribbean Hazard Risk Management Unit, 2005). In July of 2005, 10 months later, the passage of Hurricane Emily, a category 1 hurricane further weakened output in the economy and the agriculture sector (Organization of Eastern Caribbean States, 2005). Dominica suffered similar episodes as the economy was impacted by Tropical Storm Erika in August of 2015 and hurricane Maria in September of 2017. Post tropical storm Erika losses and damages to the agriculture sector was estimated to be $45.7 million USD. Flooding and landslides attributed to the significant level of damage. Agriculture is a major source of employment in Dominica and contributes to 17 percent (93.4 million) to the country’s GDP (Government of the Commonwealth of Dominica, 2015). In September of 2017 Hurricane Maria hit Dominica causing severe damage to a recovering agriculture sector with landslides and floods attributing to most of the damage. Losses and damages were estimated to be in the amount of 179.62 million USD which required an estimated total recovery effort of 88.5 Million USD (Government of the Commonwealth of Dominica, 2018). The details of the disaster on the other Caribbean territories are not absent from Saint Lucia. Saint Lucia, in October of 2010 was impacted by Hurricane Tomas, with total damage of 60 million USD. Forestry, a subsector in agriculture, suffered loss and damages estimated by 37.0 percent, bananas 36.0 percent, banana infrastructure 17.4 percent and other crops 8.0 percent. Bananas, the main export crop, were heavily impacted with an estimated 21.6 million USD in loss and damages. An unexpected Christmas Eve through occurred in 2013 curtailing the recovery efforts from Tomas. Total damages and losses estimated to be 12.9 million USD. The damages and losses amounted to 27 percent of total GDP contribution from the sector and posed an exogenous shock which affected 11.0 percent of the population who work within the sector (Government of Saint Lucia; World Bank, 2014). Tropical Storm Matthew, in 2016, crippled the recovery efforts from Hurricane Tomas thus affecting output. The sub sector most impacted was bananas with an estimated direct damage between 70 percent primarily due to toppling, snapping and flooding of fields which lead to a decrease in exports and domestic food supply. Approximately 2 years later Tropical Storm Kirk hit the island adversely impacting the banana subsector as 65.6 percent of its production. Bananas used to be the main contributor to GDP however in 2018 the sector contributes 0.8% of the total GDP. The statistics show that Saint Lucia experienced approximately 27 natural disasters with a total economic loss of over US$2.5 billion over the past 35 years. Hence, the data highlights the frequency of natural disaster episodes in the Caribbean region and in Saint Lucia. It further highlights that the magnitude of these events affects economic growth more so the performance of the agriculture sector.

104 | Research: The Platform for Innovation, Competitiveness and Growth

LITERATURE REVIEW A wealth of research exists on climate change related disasters and its impact on agriculture and food security. (Quasem, Leal, Mari de la Trinxeria, Jaafar, & Ghan, 2011) in a journal article indicated that a direct relationship exists between agriculture and climate change. The literature suggests that the agriculture sector remains the most vulnerable given its worldwide distribution and dependency on climate and environmental factors. The FAO has contributed to this literature by a series of investigations conducted to determine the effects on climate related disasters on agriculture. A recent article undertaken in 2016 by the FAO on “Damage and losses from climate-related disasters in agricultural sectors”, suggests that there has been a significant upswing in disasters worldwide for the past three decades. Climate related events such as floods, storms and droughts are notably the most frequent and have been associated with huge economic losses. The article further highlighted that from 1980 to 1990 a total of 149 climate related disasters were recorded globally with an estimated economic damage of $14billion USD. Further increases of disasters and economic loss were recorded from 2004 to 2014, of 332 disasters with an associated cost of $100 billion USD (Food and Agriculture Organization of the United Nations, 2016). (Mahato, 2014) in his investigation on global climate change and its impact on food production highlighted that the final yield of most crops is reduced as a result of a reduction in the duration of crops which is brought about by increased seasonal temperature. The study also estimated that productivity within the agriculture sector worldwide is projected to be between 3 and 16% by 2080 while production in the agriculture sector in the developing countries are projected to decline from an average of 10 to 25% by 2080. This is mainly due to average temperatures which are now above crop tolerance levels (Mahato, 2014). (Kumar & Gautam, 2014) in an investigation on climate change and agriculture in India supported the existing literature that the agriculture sector remains highly prone to climate related events and has a direct bearing on billions of lives. Their study stated that quality and quantity of the water resource and crop productivity are two main threats posed by climate change on the agricultural sector. It has been established that in countries like India rainfall has a direct relationship with crop production. A warmer climate will result in increased evaporation of surface moisture, affecting ground water level which is a direct determinant of the frequency of droughts and or floods. The study further pointed that variations in precipitation as well as higher temperature negatively impact production patterns of crops thus leading to a decrease in productivity (Kumar & Gautam, 2014). Studies on the Caribbean region on climate change speaks to the vulnerability and susceptibility of the islands to disaster related events due to its geographic location. During the active hurricane season in 2004 damages to the agriculture sector in the Caribbean accounted for 35.2% of total damages. In addition, an increase in temperature due to climate change within the Caribbean ocean surfaces has affected coral reefs and fish production. Studies point to the degree of the impact of climate change on agriculture production is dependent on the location and the farming system employed. (Solomon, et al., 2007) in the study on climate change and agriculture in the Caribbean region stated that the impact on agriculture production is dependent on other factors and its interaction with weather; topography, soil types, water availability, kind of crops, livestock, species of trees used by the farmers in their agro-ecosystems. A reduced amount of rainfall, increased floods, temperature extremes, all impact food security negatively as these are factors limit agriculture production. Inter-American Institute for Cooperation on Agriculture (IICA) in a 2017 report on Climate Change and Agriculture anticipates that temperature in Saint Lucia will trend upward between 2.4°C and 3.3°C with an increase in the magnitude and frequency of storms in addition to drought and water shortages. Sea levels are projected to rise while storm surges continue to threaten the coastal agricultural lands. As a A Compilation of Working Papers by OECS Scholars

105

result of unpredictable rainfall patterns, flooding will lead to soil erosion and degradation which would result in crop loss due to increased temperatures (1.25 – 1.75 degrees Celsius in the wet season and 1.25-2.5°C in the dry season). Most farms depend on rain to thrive, changes in the temporal distribution would lead to loss of production as the rains do not arrive as expected. Impact on production will vary as decreases are expected in short term crops while bananas would increase. Livestock sector will be negatively impacted as increase in temperature would lead to heat stress which would contribute to a decline in meat and milk production, parasite stock, lower fertility and high calf mortality (Inter-American Institute for Cooperation on Agriculture, 2017). Based on the paper by IICA and the expected damages it is necessary to gain an understanding of the extent of impact on climate change on the agriculture sector in Saint Lucia and to build resilience. Brzabih, Stage and Chambwera (2010) in investigation in Tanzania indicated that although agricultural productivity is projected to decline the negative impact can be limited. Reduction in land productivity can be replaced with increased use of capital and labour hence increasing overall productivity levels. Government policies need to be geared towards giving farmers the opportunity to invest in autonomous climate adaptation, and also policies that lead to overall improvement of performance in the economy. The above-mentioned literature confirms that agricultural production is highly influenced by climate change. The impact can depend on factors as rainfall and temperature. These can either be positive however have been found to be mostly negative with the severity of the effects of climate related disasters concentrated within specific geographic locations that are vulnerable to climate change. Developing countries are also increasingly vulnerable as production lacks adequate resources for the adaption and substituting of the various factors for production. In this context, this paper seeks to add to the existing literature by investigating the extent to which climate related disasters and other factors impact the agriculture sector in Saint Lucia using the Auto Regressive Distributed Lag (ARDL) model and Hodrick Prescott techniques. The Autoregressive Distributed Lag (ARDL) model is used in similar papers by Dumrul & Kilicarshan 2017 to assess the impact of climate change on agriculture production in Turkey. The ARDL model is a more robust approach and can be used for small data samples which is suitable for this research that has a sample size of 35 years. This model can be applied to estimate the long-run and short-run coefficient simultaneously (Dumrul & Kilicarslan, 2017). The Hodrick Prescott Filter (HP Filter) is the most widespread instrument used for trend estimation in economics as it offers the advantage of yielding estimates for the most current trend. Acharya and Bhatta in 2013 used the HP Filter to smoothen trend-stationary variables before including it in the regression equation. The HP Filter minimizes the variance of the old series around the new one, subject to a penalty constant that constrains the second difference of the smoothed series (Acharya & Bhatta, 2013).

DESCRIPTIVE STATISTICS OF MAJOR STUDY VARIABLES Determinants of Agricultural Growth In order to assess the impact of climate related disasters on climate change it is essential to examine the other variables which impacts agriculture production. The determinants fall under five main headings and they are: institutional, infrastructural, technological, environmental and socio-economic factors.

106 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE1: AGRICULTURAL LAND (SQUARE MILES)

INSTITUTIONAL FACTORS Land holdings Credit to the agriculture sector and land holding are two main indicators used as institutional factors that influence the performance of agricultural growth. The data suggest that total size (square miles) of agricultural land holdings in Saint Lucia have been on a downward path, moving from 212.0 (sqm) in 1983 to 106.0 (sqm) in 2016. A reduction of land holding can influence the output produced by the sector. Agriculture percentage of GDP as such in 1983 recorded growth levels of 9.4 percent. It can be noted that growth spurts of 10.8 percent and 11.8 percent were recorded in 1985 and 1988 respectively. This performance has been on a downward trend since the early 1990s.

FIGURE 2: AGRICULTURE, FORESTRY, AND FISHING, VALUE ADDED (% OF GDP)

A Compilation of Working Papers by OECS Scholars

107

Credit to the sector Credit and access to finance is a vital prerequisite for the survival of an economy and farmers. In order to facilitate investment and meet capital requirements, credit to the agriculture sector ought to be readily available, easy to obtain, adequate and most importantly timely. Credit can either be from an institutional or private entity with credit terms being short, medium or long term. Saint Lucia’s credit to the agriculture sector has been volatile however it has been on a decline from 2009.

FIGURE 3: CREDIT TO AGRICULTURAL SECTOR (MILLIONS)

Infrastructural factors Agricultural machinery and markets are indicators used to capture infrastructural factors for agricultural production. Infrastructural facilities include structures like tractors, pump sets, good rural-urban road network and developed agricultural marketing. However, the agricultural sector in Saint Lucia is not highly dependent on mechanization as farmers are more rural and utilize customary approaches as traditional farming which are labour intensive. However, this variable is essential since it is utilized at various stages of production. One of the largest international markets for exports of Saint Lucia’s agricultural produce (mainly bananas) is the United Kingdom. Over the years, market share has been on the decline since the loss of Saint Lucia’s preferential treatments at the Lomé Convention. Trinidad and Tobago and Barbados are two of Saint Lucia’s largest regional markets for agricultural products. To capture the performance and demand of the market, real GDP was used in the absence of other variables that better captures market demand and competing products.

108 | Research: The Platform for Innovation, Competitiveness and Growth

Technological factors Technological factors take into consideration inputs (seeds, fertilizers and pesticides/chemicals) that form the basis of production. Availability of inputs as seeds and fertilizers directly impact yields. Fertilizers are important to the initial stages of plant growth. Fertilizers when applied to the plant tissues or soil supplies the plant with nutrients which are crucial to plant growth and increase production. In addition, pesticides are the mixture of substances which helps in preventing, destroying or controlling the pests of unwanted species on plants which affect production. The availability of these variables is important in agriculture production yield. Socio-economic factors The two socio-economic factors considered in the model are population and real GDP. The domestic population currently consumes the agriculture production and therefore captures local demand. In addition, increasing performance of the economy suggests increased demand for agriculture output and therefore are important variables to the agriculture production model. Environmental Factors Favorable weather conditions play a pivotal role in production. Temperature, rainfall and the disaster episodes in Saint Lucia are used to measure weather conditions.

FIGURE 4: TEMPERATURE FOR VIEUX FORT AND GEORGE F. L .CHARLES

Temperature over the assessed period has been increasing in Saint Lucia. This is in support with the various literature which suggest higher global temperatures. Rainfall recorded in north (George F. L. Charles) and the south (Vieux Fort) of the island shows there exist variations in rainfall with peak period in 2010/11.

A Compilation of Working Papers by OECS Scholars

109

These variables highlighted can influence the performance of the agricultural production in Saint Lucia. In order to measure the linkages that exist amongst the variables several empirical tests will be carried out to explore the impact of the explanatory variables on the dependent variable.

FIGURE 5: RAINFALL FOR VIEUX FORT AND GEORGE F. L. CHARLES

DATA AND METHODOLOGY The model uses annual data of agricultural production (bananas, eggs, fish), average rainfall, temperature, economic cost (GDP of Saint Lucia) obtained from the World Bank and Research and Policy Unit. The data on production are obtained from past economic reviews prepared and published by the research and Policy Unit Department of Finance, agriculture publications (written by the Ministry of Agriculture Department of Statistics). Rainfall and temperature data are obtained from the Saint Lucia Meteorological Office that covers collections from the north and south of the island (Vieux Fort and Castries). GDP data are obtained from the Central Statistics office and Eastern Caribbean Central Bank database and the World Economic Outlook (WEO). The annual data for population, inputs, machinery, agriculture GDP in current prices, land holdings, credit to the sector were obtained from the database of the World Bank. The data on the occurrences of disasters was obtained from the emergence events database (EM-DAT) website. The span of the data set is covered from 1983 to 2018 with the reason of availability of unbroken data set for production and GDP. HYPOTHESIS The main hypothesis is to assess the determinants of agriculture production while highlighting the effects of climate related conditions on agriculture production in the short and long run. Production is mainly affected by the major factors as environmental, socio-economic, infrastructural, institutional and technological.

110 | Research: The Platform for Innovation, Competitiveness and Growth

In this context, agriculture production can be defined as: Agriculture Production = Environmental + Socio-economic + Infrastructural + Institutional + Technological AGDP t = α +β1 + β2 + β3 + β4 + β5+ εt

Equation 1

TABLE 1: THE DETERMINANTS OF AGRICULTURAL PRODUCTION Variables

Sub-categories

Agriculture GDP

Dependent Variable

Bananas, Fish, Eggs,

Environmental

β1

Temperature, Rainfall, Disaster

Socio-economic

β2

Real GDP (SLU), population

Infrastructural

β3

Markets, Machinery

Institutional

β4

Credit to agricultural sector, Land holdings

The ARDL method is used to assess the impact of climate change related natural disaster on the agricultural sector. This method is used to determine the strength of the relationship between the dependent variable which is production and a series of changing variables, known as the independent variables. This econometric model regresses environmental, socio-economic, infrastructural, institutional and technological factors. The regression has been performed using the statistical package EViews. The mathematical representation of the ARDL approach is as follows:

∆AGDPt = α + β0 AGDPt-1 + β1 ENVt-1 + β2 SOCIO-ECOt-1 + β3INFt-1+ β4 INSTt-1 + β5TECt-1+ Σpi=1α1i ∆AGDPt-1 + Σpi=1 λ1i ∆ENVt-1+ Σpi=1 €1i ∆SOCIO-ECOt-1+ Σpi=1 ₳1i ∆INFt-1+ Σpi=1 ₴1i ∆INSTt-1+Σpi=1 ₵1i ∆TECt-1 + ε1t

Equation 2

The ADRL is based on two steps. One is to determine the existence of a long run cointegrating relationship amongst the variables by using the F-statistics and by comparing them with the critical values set out by Pesaran et al. (2001). According to Pesaran et al. (2001) there are two types of critical values: lower bounds and upper bounds. The I(0) variables are referred to as lower bound critical values while the I(1) variables are referred to as upper-bound critical values. The F-statistics when calculated if higher than the upper bounds it means the null hypothesis of no cointegration is rejected. If the F-statistics is below the lower bound, the null hypothesis of cointegration cannot be rejected, this indicates the absence of a long run equilibrium relationship. If it is between lower and upper bounds a conclusive inference cannot be made without knowing the order of integration of the underlying repressors. In order to investigate the short run relationship between the variables, the error correction model based on the ARDL approach is as follows:

A Compilation of Working Papers by OECS Scholars

111

∆AGDP= γ0 + i=1 γ1 ∑∆AGDP + γ2 ∑∆ENVt-i + γ3∑∆SOCIO-ECOt-i + γ4∑∆INFt-i + γ5∑∆INSTt-i + y5∆TECt-I + Ɛt yECMt-1+Ɛ

Equation 3

The Hodrick Prescott (HP) Filter is a technique which is commonly used in macroeconomic analysis and is named after economist Robert Hodrick and Edward Prescott who in the 1990s made it popular. Hodrick was an economist who specialized in international finance and Prescott along with another economist won the Nobel Memorial Prize for research in the field of macroeconomics. The HP filter is a data smoothing technique which is applied during this analysis to smoothen the fluctuations of short-term variances as this technique minimizes the variance of old series around the new series.

RESULTS The statistical approach is applied to the variables to test the stationarity of each variable. The Augmented Dicky-Fuller (ADF), the Phillips-Perron (PP) and the Kwiatkowski–Phillips–Schmidt–Shin (KPSS) were used to test the unit roots for each variable. The unit root test, in table 2, was applied on the variables in level and first difference forms as shown in the table above. The results indicate that the order of integration is I(0) and I(1). Average rainfall, agriculture GDP, population and disaster are I(0) variables while real GDP, agriculture inputs, agriculture land, agriculture machinery and credit to the agriculture sector are I(1) variables. Average temperature was excluded from the model as it is an I(2) variable as such cannot be used in the ARDL model. The markets were also excluded since the majority of Saint Lucia’s agricultural production is consumed domestically.

TABLE 2: TESTING UNIT ROOT Variables

ADF

KPSS

Results

Bananas

levels -0.359

levels -0.987

levels 0.588

I (1)

1st difference -8.706 ***

1st difference -8.745***

1st difference 0.239***

levels -0.945

levels -0.948

levels 0.666

1st difference -3.788***

1st difference 0.0957***

Disaster

levels -5.550***

level -7.610***

levels 0.010***

I (0)

Eggs

levels -0.372

levels -0.728

levels 0.664

I (1)

1st difference -8.854***

1st difference -11.305***

levels -1.403

levels -1.433

levels 0.594

1st difference -5.256***

1st difference -5.236***

1st difference 0.201***

Rainfall GFL

levels -6.512***

levels -6.511***

levels 0.080***

I(0)

Rainfall VF

levels -5.261***

levels 0.211***

I(0)

Barbados RGDP

Fish

112 | Research: The Platform for Innovation, Competitiveness and Growth

I (1)

1st difference 0.485 2nd difference 0.246*** I(1)

SLU RGDP

Temperature GFL

Temperature VF

TT GDP

UK RGDP

Average Temperature

levels -2.356

levels -2.708

levels 0.702

1st difference -5.307***

levels -2.969

levels -2.890

1st difference -7.296***

1st difference -17.755***

levels -3.080

level -2.917

1st difference -7.316***

1st difference -14.270***

level -0.659

level -0.407

level 0.636

1st difference -3.414***

1st difference -3.367***

1st difference 0.193***

level -0.617

level -0.407

level 0.713

1st difference -3.661***

1st difference -3.674***

1st difference 0.698***

level -2.992**

level -2.861*

level 0.612

I(1)

1st difference 0.372 2nd difference 0.229*** level 0.576

I(2)

1st difference 0.456 2nd difference 0.500 level 0.603 1st difference 0.418 2nd difference 0.367 I(1)

I(1)

1st difference 0.438 2nd difference 0.500 Average Rainfall

level -5.957***

level -5.959***

levels 0.071***

Log average Temperature

level -2.979***

level -2.847*

level 0.130 1st difference 0.438 2nd difference 0.500

I(0)

I(2)

Agriculture GDP

level -3.704***

level -3.669***

level 0.103***

I(0)

Agriculture Inputs

level -2.245

level -2.213

level 0.571

I(1)

1st difference -5.838***

1st difference -5.946***

1st difference 0.212***

level -0.973

level -0.921

level 0.653

1st difference -2.739*

1st difference -4.751***

1st difference 0.218***

level 0.050

level -0.530

level 0.0451

1st difference -3.685***

1st difference -3.741***

1st difference 0.261***

level -2.519

level -1.934

level 0.290***

I(1)

1st difference -4.472***

1st differ3nce -4.308***

level -3.770***

level -3.212**

level 0.709823

I(0)

Agriculture Land

Agriculture machinery

Credit to Sector

Population

I(1)

1st difference 0.470 2nd difference 0.088*** Notes: ***, **,* denotes significant levels at 1%, 5% and 10%

A Compilation of Working Papers by OECS Scholars

113

TABLE 3: ARDL (1,0,0,1,1,0,1,0,1) COINTEGRATION TEST RESULTS Test Statistic

Value

F-Statistic

4.2996737

Critical Value Bonds Significance

I0 Bounds

I1 Bounds

10%

1.85

2.85

2.11

3.15

2.33

3.42

2.62

3.77

Note: k shows the number of explanatory variables

The F-statistics values are above the critical values therefore this indicates that there is a long-run relationship between the variables in the model. Long term coefficients calculated according to the estimation results of the ARDL (1,0,0,1,1,0,1,0,1) model.

TABLE 4: LONG-RUN ARDL ESTIMATES

Dependent Variables of Agriculture GDP Regressor

Coefficient

t-statistic

Probability

lagri_land_sm

0.124

0.542

0.593

lagri_inputs

-0.298

-1.999

0.059*

lagri_mach

-0.125

-4.226

0.000***

lav_rain

-4.991

-3.479

0.002***

lreal_gdp

2.822

5.357

0.000***

disaster

-0.088

-1.618

0.121

lcre_sec

0.024

2.039

0.054*

lpop

-6.265

-3.841

0.001***

Diagnostic Test Statistic t-statistic (Probability) R

0.9043

Adj.R2

0.851

F-Statistic

15.273

Durbin-Watson

2.144

Serial Correlation

0.757 (0.604)

Normality

1.749 (0.417)

Heteroskedasticity

0.816 (0.723)

Note: ***, **,* denotes significant levels at 1%, 5% and 10%

114 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 5: SHORT-RUN ARDL ESTIMATES

Dependent Variables of Agriculture GDP Regressor

Coefficient

t-statistic

Probability

D(lagri_land_sm)

0.109

0.349

0.730

D(lagri_inputs)

-0.350

-4.407

0.0008***

D(lagri_mach)

0.064

1.498

0.149

D(lav_rain)

-0.165

-3.099

0.005***

D(lreal_gdp)

3.114

8.794

0.000***

D(disaster)

-0.051

-3.001

0.007***

D(lcre_sec)

0.029

3.145

0.005***

D(lpop)

6.452

3.072

0.006***

cointEq(-1)

-0.973

-9.44

0.000

Note: ***, **,* denotes significant levels at 1%, 5% and 10%

In the long and short run, the results indicate that agriculture land was positive and insignificant. Agriculture inputs were negative and significant in the long run. A 1% change in agricultural inputs causes a negative 0.29 percent change in agricultural production. However, in the short run agriculture inputs had negative impact on production while being highly significant which suggest that 1% change in agriculture input results in a decline of 0.35% in production. Agricultural machinery was significant in the long run however insignificant in the short run. The negative coefficient in the long run indicates that a 1% change in agricultural machinery results in a decline of 0.12% in agriculture production. Average rainfall was negative and significant in the long run which denotes that a 1% change in rainfall results in a decrease of 0.499% in agriculture production. However, in the short run average rainfall was highly significant. A 1% change in average rainfall will lead to a decline in production of 0.17%. This indicates that rain is an important factor to agriculture production however agriculture production declines with increasing rainfall. Saint Lucia’s real GDP has a positive impact on agricultural production and is highly significant in both the long run and the short run. A 1% change in real GDP causes agricultural production to increase by 2.82 percent in the long run while in the short run a 1% change in real GDP will lead to an increase in production 3.1%. Disaster has a negative and insignificant impact on agriculture production. According to the results, the long run disaster does not affect agriculture production. Contrary, in the short run, disaster has a negative and highly significant impact on production. With a 1% change in disaster, production decreases by 0.05%. Credit to the sector has a positive and significant impact on agriculture production in both the long run and the short run. In the long run, a 1% change in agricultural credit causes an increase of 0.02% in agriculture production. In the short run credit to the agriculture sector in Saint Lucia is highly significant with a 1% change in credit results in a 0.03% increase in production. A Compilation of Working Papers by OECS Scholars

115

Population has a negative impact and is extremely significant in the long run. A 1% change in Saint Lucia’s population causes a 6.26% decline in agriculture production. However, in the short run the population is positive and highly significant. A 1% change in population will lead to a 6.45 % increase in production. Diagnostic Test The robustness of the ARDL model was confirmed by the following test. As highlighted below the models pass all diagnostic tests. Serial Correlation Test The Breusch-Godfrey serial correlation LM test was used to examine whether there was serial correlation in the residual of each of the models. The Null Hypothesis is no Serial Correlation and the alternative Hypothesis is Serial Correlation exists in the residuals. If the probability is less than 5% reject the null hypothesis, otherwise, do not reject. The results are presented in the table above. The probabilities for all the five models are greater than 5%, therefore, the Breusch-Godfrey serial correlation Lm test indicates that there is no serial correlation in the residuals. Heteroskedasticity Test This test was completed using the Breush-Pagan-Godfrey test. The Null Hypothesis is that no Heteroskedasticity exists and the Alternative Hypothesis is Heteroskedasticity exists. If the probability is less than 5% reject the null hypothesis, otherwise do not reject. In the table above the probability for all models is greater than 5%, therefore it can be concluded that there is no Heteroskedasticity in the residuals. Normality Test Jarque-Bera Normality test is used to investigate normality in the residuals of the models. The Null Hypothesis is residuals are normally distributed and the Alternative Hypothesis is residuals are not normally distributed. If probability value is less than 5%, we reject the Null Hypothesis, otherwise, do not reject. The results in the table above indicate that the probability values are all greater than 5% therefore, the Jarque-Bera Normality test stipulates that the residuals in the above models are normally distributed. Stability Test The CUSUM and CUSUM Squares test for stability of the ADRL model. It also shows that the recurrent estimates calculated from the regression equation are within the range at %% level of significance as shown in the figure below.

116 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 6: THE CUSUM TEST OF STABILITY

FIGURE 7: THE CUSUM OF SQUARES TEST

A Compilation of Working Papers by OECS Scholars

117

The diagnostic tests above ruled out the existence of all three events and confirms that the models are adequately specified. HODRICK PRESCOTT FILTER RESULTS Hodrick Prescott Filter was applied to the Agriculture GDP. This method filters the trend and cycle in a set of variables. The trend from the variable AGDP is extracted and decomposed into two different series: trend and cycle. The smothered series is compared to the actual and then graphed with the disaster events. The graphical output is presented below.

FIGURE 8: HODRICK PRESCOT FILTER FOR AGDP (MILLIONS)

For the period 1999 to 2007, with the removal of outliers, agriculture production has been performing below its potential by an average of $104.3 million. The graph points to the decline in agriculture production following a climate change related disaster. However, for tropical storm Dean in 2007, majority of the damages impacted one of the agricultural regions where production is relatively small hence, not having a sizable impact on the aggregate agriculture production of the country. CONCLUSION / RECOMMENDATIONS To date, there is major concern for the increase in climate change related events which poses a threat to the world’s food security and has prompted investigations in the performance of the sector. A number of studies have been conducted around the world on Climate change and its direct impacts on the various sectors in agriculture such as food production, fisheries, forestry and livestock, especially through low production as a result of increase in rainfall and natural disasters. 118 | Research: The Platform for Innovation, Competitiveness and Growth

This study concludes that environmental factors (rainfall, disasters) have an impact on agriculture production in Saint Lucia in the short run. More so, Hodrick Prescott filter and the figure 8 suggests that agriculture performance has been below potential and the occurrence of disaster related events impacts agriculture production, in the short run. In addition to environmental factors infrastructural, institution, technological and socio-economic factors are significant to agriculture production both in the long and short run. The effects of climate change on countries and islands varies and these variations may influence agricultural production and competitiveness. The literature speaks of different variables used to assess the effects of climate change on production. In these studies, temperature and rainfall are used as the two main indicators of climate change of agriculture production. However, the findings of rainfall and disaster were significant in explaining the impact on agriculture production. It needs to be highlighted that the results of the study suggest that in the long run, disaster is insignificant to agriculture performance and that other factors as institutions, such as access to credit, socio-economic factors and technological as inputs are highly important in determining agriculture production. In essence, climate related disaster adversely impacts the performance of the agriculture sector in the short run however, lack of access to credit, the weak performance of GDP and the availability of inputs/technology influences its performance. Hence, the weak and declining performance of the agriculture sector in Saint Lucia is influenced by other factors exacerbated by the frequency disasters as a result of climate related disasters. ● Having noted this, in order to minimize adverse effects of climate change it is important to establish strategies, policies, plans and programs to combat climate change. Policy makers need to make readily available technological and institutional factors as well as increase government expenditure on the agriculture sector to increase productivity. ● The government needs to institute a crop production cycle that can be harvested before the hurricane season commences. ● Provide incentive to farmers that are easily accessible for resource conservation and efficiency by increasing credit to the agriculture sector for transition to adaptation technologies. ● A long-term land use plan should be adapted to ensure food security and climatic resilience. Move into cultivation of tubers which grow underground that is rarely destroyed my natural disasters, or change in climatic conditions

REFERENCES Acharya, S. P., & Bhatta, G. R. (2013). Impact of Climate Change on Agriculture. Nepal: NRB Economic Review. Retrieved September 24, 2019, from https://www.nrb.org.np/ecorev/pdffiles/vol25-2_art1.pdf Dumrul, Y., & Kilicarslan, Z. (2017). Economic Impacts of Climate Change On Agriculture: Empirical Evidence From ARDL Approach For Turkey. Journal of Business Economics and Finance (JBEF), 6(ISS 4 ), 336-347. Retrieved 10 2, 2019, from https://dergipark.org.tr/en/download/article-file/397233 Food and Agriculture Organization of the United Nations. (2016). Damage and Losses from Climate-related Disasters in the Agricultural Sector. Rome: Food and Agriculture Organization of the United Nations. A Compilation of Working Papers by OECS Scholars

119

Global Facility for Disaster Reduction and Recovery (GFDRR). (2015). Dominica-Rapid Damage and Impact Assessment: Tropical Storm Erika. Dominica: Global Facility for Disaster Reduction and Recovery. Government of Saint Lucia; World Bank. (2014). Joint Rapid Damage and Assessment Needs: Flood Event of December 24–25, 2013: Washington DC: Global Facility for Disaster Reduction and Recovery. Retrieved September 21, 2019, from https://www. gfdrr.org/sites/default/files/publication/pda-2014-saint-lucia.pdf Government of the Commonwealth of Dominica. (2015). Rapid Damage and Impact Assessment Tropical Storm Erika-August 27, 2015. Washington DC: Global Facility for Disaster Reduction and Recovery. Retrieved September 21, 2019, from https:// www.gfdrr.org/sites/default/files/publication/Commonwealth%20of%20Dominica%20-%20Rapid%20Damage%20and%20 Needs%20Assessment%20Final%20Report%20.pdf Government of the Commonwealth of Dominica. (2018). Dominica: Post-Disaster Needs Assessment following Hurricane Maria September 18, 2017. Washington: Global Facility for Disaster Reduction and Recovery. Retrieved September 21, 2019, from https://www.gfdrr.org/en/dominica-pdna-hurricanemaria Inter-American Institute for Cooperation on Agriculture. (2017). Climate Change and Agriculture Saint Lucia; Policies, Strategies and Actions. San Isidro: Caribbean Climate Smart Agriculture Forum. Retrieved September 23, 2019, from http://repositorio.iica. int/bitstream/11324/7051/1/BVE18040211i.pdf Kumar, A., & Pritee, S. (2013). Impact of Climate Variation on Agricultural Productivity and Food Security in Rural India: Economics Discussion paper. Kiel Institute for the World Economy, 2013-43. Retrieved September 20, 2019, from http://www. economics-ejournal.org/economics/discussionpapers/2013-43 Kumar, R., & Gautam, H. (2014). Climate Change and its Impact on Agricultural Productivity in India. Journal of Climatology and Weather Forecasting, 2(1), 2332-2594. Latin America and the Caribbean Hazard Risk Management Unit. (2005). Grenada: A Nation Rebuilding Grenada, An assessment of reconstruction and economic recovery one year after Hurricane Ivan. Washington DC: World Bank. Retrieved September 18, 2019, from http://siteresources.worldbank.org/INTLACREGTOPHAZMAN/Resources/grenanda_rebuilding.pdf Mahato, A. (2014, April). Climate Change and its Impact on Agriculture. International Journal of Scientific and Research Publications, 4(4), 2250-3153. Retrieved September 17, 2019, from http://www.ijsrp.org/research-paper-0414/ijsrp-p2833.pdf Ncube, F., Cheteni, P., & Ncube, P. (2015). THE IMPACT OF CLIMATE CHANGE ON AGRICULTURAL OUTPUT IN SOUTH. University of Fort Hare. Alice, South Africa: Munich Personal RePEc Archive. Retrieved September 18, 2019, from https://mpra. ub.uni-muenchen.de/73489/1/MPRA_paper_73489.pdf Organization of Eastern Caribbean States. (2005). Grenada: Macro-Socio-Economic Assessment of the damage caused by Hurricane Emily July 14th 2015. Castries: Organization of Eastern Caribbean States. Retrieved September 20, 2019, from https://reliefweb.int/report/grenada/grenada-macro-socio-economic-assessment-damage-caused-hurricane-emily Quasem, A. A.-A., Leal, W., Mari de la Trinxeria, J., Jaafar, A. H., & Ghan, Z. A. (2011). Assessing the Impacts of Climate Change in the Malaysian Agriculture Sector And its Influences in Investment Decision. Middle-East Journal of Scientific Research, 1990-9233. Solomon, S. Q., Manning, M., Z, C., M. A., Tignor, M., & Miller, H. (2007). The Intergovernmental Panel on Climate Change AR4 Climate Change 2007: The Physical Science Basis. United Kingdom, New York: Cambridge University Press. University of California, Davis. (2019, September 12). Climatechange.ucdavis.edu. Retrieved from Science and Climate: https:// climatechange.ucdavis.edu/science/climate-change-definitions/

120 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Petriana Daniel

Economist, Research and Policy Unit Department of Finance, Government of Saint Lucia

Petriana Daniel, is currently an Economist at the Research and Policy Unit in the Department of Finance (Saint Lucia). Ms. Daniel holds a Bachelor’s degree from St. Georges University in Economics and Finance. Prior to working at the Department of Finance, Ms. Daniel was the Community Liaison Officer for the Mabouya Valley Development Corporation, a subsidiary of the Ministry of Agriculture and also worked as an Administrative Manager at Jutis Chambers Attorneys-At-Law. Ms. Daniel is the author of the working paper Contribution of Small Ruminants to the Agriculture Sector in Saint Lucia which was presented at the National Competitiveness and Productivity Council week of activities Research Symposium.

A Compilation of Working Papers by OECS Scholars

121

7 Does the Price of Fuel in Saint Lucia Mimic International Fuel Price Developments? Jilayne Clery-King

122 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT This paper is an investigation of the pass-through mechanism adopted by the Government of Saint Lucia (Department of Finance) for the period 2015-2019 to assess the adequacy of the approach in mirroring international oil prices. This mechanism was developed in response to the volatility in international oil prices in 2008 which led to substantial increases in fiscal cost and risk over the period. The study utilizes Pearson’s correlation analysis to measure the direction and magnitude of the linear relationship between two variables; international oil prices and domestic oil prices with an emphasis on gasoline and diesel. Findings suggest that the base CIF is a fair proxy to calculate domestic fuel prices given the strength of the Pearson’s coefficient correlation which suggest the nearness of the variables. However, the strength of the correlation (magnitude and direction) of international oil prices and the fuel price mechanism is weakened due to the inclusion of taxes, distribution margins and the policy decisions adopted by the Government of Saint Lucia. INTRODUCTION Saint Lucia, a small island developing state in the Caribbean is heavily dependent on oil and its by-products for its productive sectors such as transport, construction and manufacturing. As such, it continues to be impacted by global oil price volatility particularly in periods of soaring oil prices. These high oil prices significantly affect the economy’s gross domestic product given its dependability which continuously place a strain on government’s financial resources as government subsidizes some petroleum products to protect domestic consumers from fluctuations in oil prices. In light of this, the Government of Saint Lucia in September 2009 introduced the one-month market pass-through pricing mechanism which was implemented following a decision by the Monetary Council of the Eastern Caribbean Central Bank (ECCB). This required all member states to adopt a pass-through mechanism for fuel given the escalation of international oil prices in 2008. Also, the full subsidization of fuel products proved unsustainable in Saint Lucia’s context given its tight fiscal position. Subsequently, the Government of Saint Lucia modified the one-month pass through mechanism and adopted a three-month pass-through system on May 14, 2012. On January 28, 2013 the Department of Finance abandoned the Mean Caribbean Posting (MCP) price index for oil and utilized the base CIF and Landing Cost of fuel products to determine the final price of domestic prices. The intention of this approach was to reduce the disparity between domestic and international oil prices. This approach was further refined in February 2015 to a three-week pass-through system which is still being utilized to date, in an effort to increase the volatility in prices as well as the efficiency of the passthrough mechanism. At the core of the mechanism is an explicit fuel pricing formula, which determines domestic prices as the sum of the import price of fuel products, domestic wholesale and retail distribution margins, and fuel taxes. Domestic fuel prices are then changed at pre-specified regular intervals. As such, this current pricing methodology generates the Base Cost Insurance Freight (CIF) for all fuel products within an established reference period utilizing invoices provided by SOL and Rubis; the two importers of fuel products in Saint Lucia. Subsequently, this base CIF is used to calculate the respective excise tax and retail prices for each product. Since the Government of Saint Lucia subsidizes the 20 and 22-pound cylinder of Liquefied Petroleum Gas (LPG), the retail prices are adjusted to reflect the respective subsidy. Notwithstanding this cost-saving to consumers as a result of this subsidy, domestic consumers continuously question the rationale or methodology used to calculate fuel prices, particularly for Unleaded Gasoline (ULG) and Diesel. Saint Lucia’s fuel prices are calculated using a “three-week window prior to the effective date of the fuel price change. This implies that each price change reflects changes in the market occurring 21 days prior. In the past, it has been observed that the public compares spot prices on a particular day with the international prices for that given day. Hence, there are concerns that domestic spot prices are not consistent A Compilation of Working Papers by OECS Scholars

123

with international price movements of both products and that further amendments to the methodology is required. This paper therefore uses Pearson’s Product Moment Correlation analysis to assess the relationship between international fuel prices and domestic fuel prices. According to Hauke & Kossowski (2011) the Pearson’s Product Moment Correlation Coefficient is one of the most popular approaches used in several areas of research. The Pearson’s Coefficient is a measure of the degree (direction and magnitude) of the linear relationship between two interval ratios or continuous level variables. This will address the concerns of the public as it seeks to highlight the nearness of the two prices for each product. The research piece is structured as follows: The objective or motivations of this paper is highlighted in section two (2). Section three (3) and four (4) discuss the literature review respectively and section four (4) explains the methodology used. Following this, an assessment of the data and results is conducted in section five (5) and the conclusion and policy recommendation is discussed in section six (6). OBJECTIVES / MOTIVATION The objective of this paper is to determine whether domestic and regional prices of fuel products mirror international prices. Aims 1. To determine the extent to which the Base CIF of Rubis and Sol mirror international fuel products. 2. To determine the extent to which the domestic retail prices of fuel are consistent with international fuel price developments. 3. To determine the extent to which regional retail prices of fuel mirror international fuel products. RELATED LITERATURE Automatic Fuel Pricing Mechanisms have been implemented by many countries as a means of allowing the international fuel prices to be reflected in the domestic retail prices. This approach is widely recommended by international agencies as the International Monetary Fund and the World Bank. (Coady et al, 2012) proposed for the institutionalization of such mechanisms to protect fuel tax revenues and avoid price subsidies, to contain the volatility of fuel tax revenues, avoid reliance on an ad hoc approach to fuel pricing where governments change domestic prices at irregular intervals. The pass-through system adopted by most countries tends to consider the peculiarities present in the given period. The commonality with most countries particularly developing and emerging countries according to a study conducted by Coady et al (2011) of the International Monetary Fund (IMF) is the failure to fully pass-through increases in international fuel prices to domestic retail prices which tends to have adverse consequences for fuel tax revenues and tax volatility. Also, Reboredo (2011) stated that since international oil prices are denominated in US dollar terms, the movement in exchange rates of domestic currencies vis some vis the US dollar have an important bearing on the type and accuracy of the pass-through mechanism adopted. In this context, the exchange rate does not directly affect domestic prices given that the Eastern Caribbean dollar is pegged to the United States Dollar; the prevailing currency used in Saint Lucia. Notwithstanding these challenges, (Mandal & Bhattacharyya, 2012) stated that a large number of empirical studies suggest that domestic oil prices in most oil-importing countries generally adjust partially and with a lag in response to international prices.

124 | Research: The Platform for Innovation, Competitiveness and Growth

Despite the plethora of studies on oil prices and its impact on both oil-importing and exporting countries, most studies investigate the oil pricing pass-through effects as it relates to inflation, exchange rates and other influences of sub-sectors in the economy. This study therefore adds to the existing literature by estimating the correlation between international and domestic oil prices of Saint Lucia using a more rigorous approach. As such, the appropriate level of taxes and distribution margins and fluctuations passed-through retail prices will be investigated to explain any price variances. This paper employs Pearson’s Product Moment Correlation Coefficient (PMCC) to investigate the disparity between the two variables; international and domestic oil prices. Although this methodology has proven to be popular and widely used in scientific and other disciplines such as medicine and biology, social sciences and economics there are limited examples or literature depicting the use of this correlation statistic to explore oil prices and how they differ across geographical spaces (Pearson et al, 2019). Hence, this paper will contribute meaningfully to the literature surrounding the topic as it seeks to provide an explanation assessing the relationship between international and domestic prices using a rigorous approach. Also, given that the Pearson Correlation is a meaningful measure of strength of association, if the relationship between two variables is a linear, it will provide a sense as to the current level and impact of taxes and distribution margins and percentage of fluctuations passed on to domestic consumers (Kozak, Krzanowski, & Tartanus, 2012). EXPLANATION OF CURRENT FUEL CALCULATION PROCESS Steps 1. The CIF and US Gallons of fuel products are extracted from the commercial invoices and used to generate the Unit CIF per Imperial Gallon for each fuel product for both importers (Rubis and Sol). 2. The Unit CIF per Imperial Gallon for each fuel product from both importers are added and averaged to calculate an Average Base CIF for the respective fuel product. This Average Base CIF is then inputted into the fuel price build up template to generate the final retail prices. ●

The excise tax, service charge, and guaranteed company margins are added to the Average Base CIF to generate the final retail prices:

●

An excise tax of $2.50 as of the last fuel price change on June 12, 2017. As of July 2017, this excise tax increased by $1.50 to $4.00. The $1.50 being the fixed component while the $2.50 is variable.

●

A service charge of 6.0% on the Average Base CIF

● Importers guaranteed company margin of $1.005. This will result in the maximum wholesale price. ●

Retailers guaranteed company margin of $1.10. This will result in the maximum retail price.

●

Price cap of $12.75 increased on 23, April 2018 to $13.95

The full fuel price calculation process can be seen in the appendix.

A Compilation of Working Papers by OECS Scholars

125

STYLIZED FACTS In an effort to develop a better understanding of the dynamics in the oil industry both locally and internationally, time plots of various variables are exhibited in several graphs below for the period 2015-2019. The main variables consist of the historical movement of the WTI, domestic retail prices for both gasoline and diesel and the corresponding base CIF. The movement in international prices depicted an upward trend in international oil prices in 2015 as shown in figure 1. While it dipped in 2016, international prices increased at a progressive rate in the latter part of 2016 through to August 2018 and has been volatile up to the end of the assessed period. Conversely, fuel excise tax rates in Saint Lucia, a variable component in the calculation of domestic retail prices, remained fixed at $2.50 per imperial gallon for the period 2015-2017 and fluctuated significantly from July 2017 to June 2019. The highest excise tax recorded was $5.34 and $4.42 per imperial gallon while the lowest was $2.24 and $3.05 per imperial gallon for gasoline and diesel respectively (shown in Figure 1). Given the increase in international oil prices coupled with retail price caps of $12. 75 and $13.95 on both products, the excise tax variable was significantly below the targeted $4.00 mark in most instances. FIGURE 1: DOMESTIC EXCISE TAX (GASOLINE & DIESEL) & CORRESPONDING WTI INDEX

The time plot containing the retail prices for both fuel products, gasoline and diesel, and the corresponding excise tax rates on fuel for the given reference periods is displayed in figure two below. The trend in domestic fuel retail prices, depicted by the red and black dotted line, shows that retail prices have been on an upward trajectory from 2015 to 2019. The volatility in domestic fuel prices can be seen in the years 2015 through to the first half of 2017 before the introduction of the additional $1.50 excise tax and the first price cap of $12.75 per imperial gallon. However, given the imposition of this initial price cap ($12.75), domestic retail prices remained stable until the first half of 2018. Subsequently, the domestic retail prices were capped at a higher level of $13.95 per imperial gallon; the highest price over the period under review.

126 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 2: COMPARISON OF DOMESTIC RETAIL PRICES, WTI AND EXCISE TAX

In sum, the graphical plots above in figures 1 and 2 show mixed movements between all variables. While WTI fuel prices were volatile during the entire period, excise tax rates remained stable at $2.50 preceding the introduction of the $4.00 target. As the government aimed at maintaining price stability while receiving the highest revenues from fuel, the excise price fluctuated between $1.50 and $4.00 from 2017. Such mixed movements led to corresponding changes in the domestic retail price of fuel products over the period. However, overall domestic retail prices were more volatile in the earlier years compared to the latter part of the period (2017-2019) given the price caps. In order to effectively measure the correlation between all variables, the empirical test was conducted to find out the proportional effects of the explanatory variables on the dependent variable. METHODOLOGY This paper explored three approaches: (i) a comparison of the domestic CIF vis a vis international CIF in a given reference period; (ii) a comparison of domestic retail prices vis a vis international prices on a given day (spot prices); and (iii) a comparison of domestic retail prices against international prices within the established reference period (21 days). The study is conducted using the historical retail prices generated from the fuel price build-up which is the automatic fuel pricing mechanism employed by the Government of Saint Lucia for the period 2015 to 2019. In addition, the study utilises the West Texas Intermediate (WTI) as a proxy for international oil pricing. The respective correlation coefficients were calculated using STATA (a statistical software package) to determine the relationship that exist between domestic and international fuel prices A Pearson correlation coefficient also known as a “Product Moment Correlation Coefficient” (PMCC) is a statistical measure of the strength of a linear relationship between paired data. It is denoted that by -1≤ r ≤ 1 where positive values denote positive linear correlation and negative values denote negative linear correlation. A value of 0 denotes no linear correlation (no relationship exists). The closer the value is to 1 or –1, the stronger the linear correlation as such 1 indicates a perfect relationship while -1 indicates A Compilation of Working Papers by OECS Scholars

127

a perfect negative relationship. The following formula is used to calculate the Pearson r correlation (StatisticsSolutions, 2019):

rxy = Pearson r correlation coefficient between x and y n = number of observations

xi = value of x (for ith observation) yi = value of y (for ith observation) In this context the Pearson r coefficient and the coefficient of determination (r2) is useful in explaining to what extent changes in international oil prices (x) explain changes in the domestic fuel retail prices (y) and the degree to which other factors influence the domestic retail prices of fuel. This methodology was used because of its effectiveness in measuring the association between variables of interest because it is based on the method of covariance (StatisticsSolutions, 2019). The following assumptions were observed when undertaking the correlation analysis: 1. Linear relationship: Two variables should be linearly related to each other. This can be assessed with a scatterplot: plot the value of variables on a scatter diagram, and check if the plot yields a relatively straight line. 2. Homoscedasticity: the residuals scatterplot should be roughly rectangular-shaped. In addition, statistical hypotheses testing was employed to determine whether the result of the data set was statistically significant for each variable in the respective year. That is, the likelihood that a relationship between two or more variables is caused by something other than chance. Using this methodology, a p-value was generated, which represents the probability that random chance could explain the result. A p-value of 5% or lower is considered to be statistically significant which means it is representative. The degree of correlation coefficient (r-value) is assessed based on the following criteria (StatisticsSolutions, 2019): 1. Perfect: If the value is near ± 1, then it is said to be a perfect correlation: as one variable increases, the other variable tends to also increase (if positive) or decrease (if negative). 2. High degree: If the coefficient value lies between ± 0.50 and ± 1, then it is said to be a strong correlation. 3. Moderate degree: If the value lies between ± 0.30 and ± 0.49, then it is said to be a medium correlation. 4. Low degree: When the value lies below + .29, then it is said to be a small correlation. 5. No correlation: When the value is zero.

128 | Research: The Platform for Innovation, Competitiveness and Growth

DATA ASSESSMENT AND RESULTS Approach 1: Base CIF and WTI

(Determining the relationship between average c.i.f. and international prices) Correlation of Coefficient (r-value): measures the relationship of two variables and their interaction. Pearson’s correlation coefficient values (r-value) depict a positive relationship (direction) between the average base CIF and international prices for both products, gasoline and diesel, between 2015 and 2019. This means that the movement of the average based CIF was consistent with international fuel prices over the assessed period. Notwithstanding, the positive r-value the strength (magnitude) of the relationship varied for both fuel products. This implies that the degree of correlation between international prices and the average base CIF ranged from high to low between years. The magnitude for each variable (gasoline and diesel) will be discussed below in the respective section. Gasoline The r-values exhibited a high degree of correlation in most years for gasoline in Saint Lucia as the coefficient values lie between + 0.50 and +1. The strongest correlation was in 2016 which was +0.88. However, the relationship between the average base CIF and international prices appeared to be moderate in 2017, given a correlation coefficient of +0.39. Diesel The data revealed that a high degree of correlation exists between the international price of oil and the average domestic base CIF for diesel between 2015 to 2017 with values ranging from +0.57 to +0.91. This implies that the majority of the times domestic prices moved in tandem with international fuel prices. Conversely, 2018 and 2019 displayed a low degree of correlation below the stipulated +0.29 established in the criteria above. TABLE 1: R-VALUES FOR GASOLINE AND DIESEL

Pearson Correlation Coefficient Values (r-value) CIF-WTI Year

Reference Gasoline

Diesel

2015

0.67

0.57

2016

0.88

0.91

2017

0.39

0.65

2018

0.87

0.23

2019

0.70

0.19

Entire Period

0.82

0.85

A Compilation of Working Papers by OECS Scholars

129

Notwithstanding the variations in the size of the relationship for both products for particular years, the Pearson Correlation Coefficient Values (R-value) for the entire period exhibited a high degree of correlation with r-values of +0.82 and +0.85 for gasoline and diesel respectively, which suggests that the base CIF overall moves in tandem with international prices. Coefficient of Determination: the extent to which international oil price explains the movements in domestic prices. The coefficient of determination expressed as a percentage explains the proportion of the variance in the dependent variable (average base CIF) that is predictable from the independent variable (international prices). Over the period, 67.7 percent and 71.4 percent of the variance in domestic prices of gasoline and diesel respectively can be explained by the change in international oil prices. This means that 32.3 percent for gasoline and 28.6 percent for diesel are due to factors other than international fuel prices. As such, this coefficient determination table below (Table 2) constitutes periods of high and low coefficients of determination. The largest proportion was observed in 2016 for both products, while the lowest was recorded in 2017 for gasoline and 2019 for diesel.

TABLE 2: COEFFICIENT OF DETERMINATION

Correlation Values ^2 (coefficient of determination) CIF-WTI Year

Reference Gasoline

Diesel

2015

44.5%

32.7%

2016

77.8%

82.6%

2017

15.0%

41.6%

2018

75.7%

5.4%

2019

48.7%

3.5%

Entire Period

67.7%

71.4%

DOMESTIC RETAIL PRICES AND WTI Approach 2: Domestic retail price vs Spot WTI

(Determining the relationship between domestic retail and international prices- Spot prices) Correlation of Coefficient (r-value): measures the relationship of two variables and their interaction. A comparison of domestic retail prices and spot WTI prices depicted a positive relationship with varying degrees of correlation over the period. R- values for gasoline and diesel were 0.73 and 0.79 respectively which depicts a high degree of correlation for the entire period. However, the r-value in most years exhibited a low degree of correlation, suggesting that domestic retail prices did not move in tandem with international spot prices.

130 | Research: The Platform for Innovation, Competitiveness and Growth

Gasoline Based on the results in Table 3 below, the r-value for most years had a low degree of correlation which indicates that changes in international oil prices does not significantly explain the variance in domestic prices on a given day. Hence, a comparison between the magnitude of the spot WTI and retail prices for gasoline on the effective date of a price change, would show a disparity in prices. Moreover, in 2019, spot WTI and domestic retail prices had a negative correlation which implies that these variables moved in opposite directions, that is, when international prices on that day decrease, domestic retail prices for gasoline increase or vice versa.

TABLE 3: R-VALUES- SPOT PRICES

Pearson Correlation Coefficient Values (r-value) Price-WTI Year

Spot (as at) Gasoline

Diesel

2015

0.43

0.51

2016

0.75

0.81

2017

0.22

0.52

2018

0.37

0.02

2019

-0.22

-0.55

Entire Period

0.73

0.79

Diesel In most instances, the r-value or the Pearson correlation coefficient values ranged from high to moderate meaning the magnitude of the change in international prices were similar to changes in the domestic retail price of diesel. As shown in Table 3, the r-value was extremely low in 2018 (0.02) depicting that there was almost no degree of correlation between the two variables. In terms of direction, the independent and dependent variables moved in different directions in 2019; r-value was -0.55 (negative correlation). Coefficient of Determination: the extent to which international oil price explains the movements in domestic prices. The coefficient of determination in Table 4 implies that the variability of the dependent variable, gasoline and diesel retail spot prices for the entire period was 53.9 percent and 62.9 percent respectively. The remaining 46.1 and 37.1 percent of the variability between retail prices and WTI prices are influenced by factors other than international prices.

A Compilation of Working Papers by OECS Scholars

131

TABLE 4: COEFFICIENT OF DETERMINATION; SPOT PRICES

Correlation Values ^2 (coefficient of determination) CIF-WTI Year

Spot (as at) Gasoline

Diesel

2015

18.7%

25.9%

2016

55.7%

65.1%

2017

4.8%

27.2%

2018

13.7%

0.1%

2019

4.9%

30.5%

Entire Period

53.9%

62.9%

However, when analyzed individually, all years except 2016 exhibit low coefficients of determination for gasoline and diesel which implies that majority of times, the variability of domestic retail prices are not explained by changes in the price of international oil. Hence, changes in international oil prices on a particular day does not imply a change of a similar magnitude on that day. Likewise, comparing retail prices for an effective price change date will seldom show the same degree of change with international prices. Approach 3: Domestic retail price and WTI using reference periods

(Determining the relationship between retail and international prices-reference periods) Correlation of Coefficient (r-value): measures the relationship of two variables and their interaction. An analysis of the entire period (2015-2019) shows that there is a high degree of correlation between WTI and domestic retail prices for gasoline and diesel as per the corresponding reference period. This is represented by r-values of +0.80 and +0.81 for gasoline and diesel respectively. This implies that the magnitude of the change in domestic retail prices are mostly consistent with changes in the international market over the period. The degrees of correlation between the two variables are illustrated in Table 5 below. Gasoline Based on the results shown in Table 5, the change in methodology in 2015 to the three-week pass-through mechanism, showed that domestic retail prices were more consistent with international oil prices given the higher degree of correlation from +0.66 (prior 2015) to +0.88. However, given the policy decision to increase the excise tax from $2.50 to $4.00 in 2017 and introducing a price cap simultaneously resulted in an even larger variation in both the magnitude and direction of the change between the independent and dependent variables, that is international and domestic retail prices. Given the fluctuations in prices in the international market for the most part of 2018, domestic retail prices nonetheless remained capped at $13.95 resulting in the variability in prices.

132 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 5: DEGREES OF CORRELATION; DOMESTIC RETAIL PRICES AND WTI

Pearson Correlation Coefficient Values (r-value) Price-WTI Year

Reference (3 Week) Gasoline

Diesel

2015

0.66

0.57

2016

0.88

0.91

2017

-0.76

0.19

2018

0.25

2019

-0.31

-0.16

Entire Period

0.80

0.81

Diesel The coefficient values revealed that the domestic retail price of diesel was highly correlated to the international prices of oil in 2016 using the new methodology. However, similar to gasoline this was affected by the increase in excise tax and introduction of the fuel price cap causing the degree of correlation to be lower in subsequent periods. In 2019, there was a negative correlation between the two variables which indicates that international and domestic prices moved in opposite directions. Coefficient of Determination: the extent to which international oil price explains the movements in domestic prices. An examination of the coefficient of determination in this approach for the entire period shows that 64.5 and 66.9 percent of the variability of the dependent variable (gasoline and diesel retail prices respectively) can be explained by the independent variable (international prices). This can be seen in Table 6 below. The remaining percentages of the variability consist of other factors as: domestic taxes, wholesale and retail margins and other factors. TABLE 6: COEFFICIENT OF DETERMINATION; REFERENCE PERIODS

Correlation Values ^2 (coefficient of determination) Price-WTI Year

Reference (3 Week) Gasoline

Diesel

2015

42.9%

32.4%

2016

77.8%

82.6%

2017

57.8%

3.6%

2018

6.3%

6.1%

2019

9.4%

2.7%

Entire Period

64.5%

66.9%

A Compilation of Working Papers by OECS Scholars

133

The results also exhibit similarities and differences between the two dependent variables. A high percentage of the variability in retail prices for gasoline and diesel in 2016 can be explained by changes in the international price of oil. In the following year, 57.8 percent of changes in gasoline retail prices were accounted for by international prices; 20.0 percent lower than the preceding year. Subsequently, the coefficient of determination values for 2018 and 2019 were below 10 percent which implies that a significant proportion of the variability is unexplained. In the case of gasoline, factors may include the interplay between the fixed and variable component of the excise tax and the price cap. A similar pattern can be observed for domestic retail prices for diesel. In all years except 2016, the variability of the dependent variable that can be accounted for, was below 50 percent. This implies that a significant proportion of the variability is unexplained by the movement in international oil prices. This can also be deemed as a result of the interaction between the aforementioned variables and other factors external to the methodology adopted by the Government of Saint Lucia. IMPACT OF TAXES AND PRICE CAP ON RETAIL PRICES This section seeks to pinpoint what percentage of this divergence or variability between the international and domestic prices is as a result of the taxes and wholesale and retail margins added to the base CIF. A comparison is made between the CIF (approach 1) and the retail price derived using the reference period methodology (approach 3) to determine the impact of taxes and distribution margins on the final price of domestic fuel. Given that CIF (approach 1) represents the direct costs of imports of fuel while the reference period (approach 3) captures this CIF in addition to taxes and distribution margins. This is done by calculating the difference between the coefficient of determination of the reference period and the base CIF as shown in Table 7 below.

TABLE 7: IMPACT OF TAXES AND PRICE CAP ON RETAIL PRICES

Reference Price - CIF Gasoline

Diesel

2015

-1.6%

-0.3%

2016

0.0%

2017

42.8%

-38.0%

2018

-69.4%

0.7%

2019

-39.3%

-0.8%

Entire Period

-3.3%

-4.5%

A negative sign implies that value for approach 3 was lower than the value of the base CIF (approach 1) meaning that the correlation between international prices and domestic prices was stronger using the base CIF. The opposite is true when the values above are positive; this implies that the correlation between the international and domestic prices was stronger using the reference period. Based on the findings, most of the values in figure 9 were negative; which suggest that a high degree of the variability is explained by international prices from the point of the CIF methodology. The price differential or divergence represents the proportion of the domestic retail price that make up taxes and respective distribution margins under approach 3. Hence, the larger the difference, the greater the impact of taxes, distribution margins or policy decisions on retail prices. In this context, the difference between the two coefficients of determination (reference retail price and base CIF) is marginal for both gasoline and diesel prior to 2017. This implies that the fuel price mechanism 134 | Research: The Platform for Innovation, Competitiveness and Growth

slightly influenced the correlation (relationship) between the two variables; from the base CIF to the final retail price the mark-up was moderate. Therefore, the retail prices maintained the same direction as international prices despite the addition of taxes and distribution margins. Similarly, the negative sign for diesel in 2017 (as shown Figure 9) reveals that there was a 38 percent difference between the base CIF and retail prices which indicates the magnitude of the excise tax added by the Government of Saint Lucia. However, the opposite occurred in 2017 for gasoline, where the difference between the two was (+)42.8 percent. This may be as a result of the counter effect between the base CIF and domestic taxes; that is the low correlation between international prices and the base CIF and the government’s new policy decision to increase excise tax on fuel. Hence in that year, domestic retail prices (after taxes were included) were more consistent with international prices. Subsequently, the difference for gasoline was negative in 2018 and of a larger magnitude than 2016. This is primarily as a result of the annualized effect of the policy decision to increase excise taxes prices introduced midway in 2017 as well as the upward revision of the price cap from $12.75 to $13.95 in early 2018. However, the magnitude of the difference lessened in 2019 given the decline in international prices for the first half of 2019 compared to the previous year. Hence, this decline allowed for the collection of the $4.00 excise tax below and at the cap of $13.95 in some instances. In two instances the government, in an effort to recoup the revenue forgone in previous periods, allowed an excise tax exceeding $4.00 at the stipulated price cap. Conversely, in 2018, diesel had a positive value due to the fact that the coefficient of determination of the base CIF was lower than domestic retail prices. This lower base CIF of diesel permitted excise tax rates closer to the $4.00 target at the established price cap. Also, in that year the correlation between retail prices and international prices was stronger than the correlation between the base CIF and international prices. Hence, the gap which existed made the effect of additional taxes and price cap negligible; 0.7 percent more representative of international oil prices. The full effect of the decrease in international prices was experienced in the first half of 2019 as shown in Figure 3 below. In most instances the excise tax rate was $4.00, at and below the price cap of $13.95. This caused a slight difference of (-) 0.8 percent in the coefficient of determination between the two variables; the effect resulting from the policy decision taken by the government of Saint Lucia.

FIGURE 3: DOMESTIC RETAIL PRICES vs WTI 2015-2019 (REFERENCE PERIODS)

A Compilation of Working Papers by OECS Scholars

135

Notwithstanding, the mixed effect of the increase in excise tax and price cap, the variance between the base CIF and domestic price cap for the entire period was low for both products; (-) 3.5 and (-) 4.5 for gasoline and diesel respectively. Hence, 3.5 and 4.5 percent of the times, increase in domestic price of gasoline and diesel was caused more by domestic taxes as opposed to fluctuations in the international market. CONCLUSION The appropriate choice of smoothing mechanism employed by a country depends on the level of tradeoff between price and fiscal (tax) volatility (Coady et al, 2012). The analysis presented implications of alternative pricing mechanisms for domestic retail fuel prices and taxes. Three methodologies were considered: (i) the average based CIF, (ii) spot prices and (iii) the reference period. The simulations utilised a historical series of retail prices and West Texas Intermediate (WTI) prices. The statistical results show that the base CIF method closely mirrors international oil prices relative to the other two approaches, although for some years there were low correlations. The two other approaches (2 and 3) presented, revealed that the price mechanism using the reference period is more representative of international prices relative to the spot prices approach. The analysis further revealed that domestic prices of fuel products mirrored international prices in 2015 and 2016 prior to the increase in excise tax and price caps. However, in 2017-2019 domestic prices did not mimic international prices. This may have been as a result of policy decisions to lessen the trade-off between retail prices and tax volatility over the period. In some instances, this averted price hikes particularly when international oil price was on an upward trajectory. Conversely, there were a few instances when domestic prices moved in the opposite direction due to a policy decision to recoup the revenue forgone in the previous period. The general finding suggests that there is variance between domestic price of fuel and international oil prices. At the point of comparison of the base CIF and international prices there already exist a price differential however a significant portion of the prices explains international oil price developments. Upon further investigation, a greater deviation between international and domestic prices exists using the reference period. The magnitude of this variance is widened by components as the excise tax and distribution margins. As such, the increase in one component, the excise tax resulted in an increase in retail prices prior to the price cap. However, the introduction of the price caps ($12.75 replaced by $13.95 in 2018) minimised the effect of these increases on the final domestic price of gasoline and diesel. Therefore, despite an increase in domestic retail prices between 2017 and 2019 for both products, prices were not significantly higher than what would hold excluding the increases in the excise tax and price cap. RECOMMENDATIONS Based on the findings of the study, the following recommendations are proposed to improve the correlation between the international and domestic oil price movements. ●

Revision of the existing fuel price mechanism

The Government of Saint Lucia in an attempt to safeguard fuel tax revenues should ensure that taxes are calculated in an efficient and equitable manner to avoid higher or distortionary taxes elsewhere in the economy. The following can be done in this regard:

Excise tax

Conduct a cost-benefit analysis to identify which component of the excise tax should be fixed and

136 | Research: The Platform for Innovation, Competitiveness and Growth

which should be variable. This would lower the revenue forgone by the government when retail prices exceed the existing $13.95 cap if the full excise tax of $4.00 is factored in the calculation process. For example, in 2018 where the GOSL revenue was adversely impacted given an excise tax rate of $2.24 was collected, an amount below the stipulated excise tax of $2.50.

Subsequently, the GOSL should amend the current fuel price mechanism to:

(i) Establish the fixed and variable component (if changes are required);

(ii) Alternatively, consider a lumped approach with no variable component for which tax levels are fixed at the target benchmark level.

Price Cap

Conduct research to identify an optimal price cap. Following this, GOSL should amend the existing price cap or discontinue the price cap and allow a full pass-through. However, the preferred approach would be to allow a full pass-through in which only the excise tax component is variable; maintaining a price cap and a variable excise tax component simultaneously does not always maximise government excise revenue. There are instances where the trade-off between retail price and excise tax revenue is too high and taxing on government revenue. Hence, the objective should be to promote less government interference in the price calculation process. ● Introduce mitigating measures in the form of social programs geared at vulnerable groups According to a review conducted by Coady et al (2012), on average, a US$0.25 per liter increase in domestic prices decreases household real incomes by 5 percent, with this impact being similar across all income groups. Therefore, it is important that reform strategies include measures to mitigate this adverse impact. Coady et al (2012), proposed mitigation measure to the increases in fuel prices is the effective reallocation of some budgetary savings from reducing subsidies to education, health and infrastructure programs that benefit low- and middle-income households which was implemented in Indonesia after domestic fuel prices increased by 22- 33 percent. In the case of Saint Lucia, a similar approach can be implemented in instances where savings are realized following the full excise tax and price cap. Budgetary allocations can be made to a range of social protection programs. i. Similarly, in Ghana extra funds were made available to primary healthcare programs concentrated in the poorest areas through a community health compound scheme and significant investment in the provision of mass urban transport was expanded and expedited. A fund can be established in Saint Lucia to allow additional revenue to be transferred for these purposes.

NOTES Petroleum products include gasoline, diesel, kerosene and liquefied petroleum gas (LPG). Most of the subsidies referenced after 2015 are on the 20lb and 22lb LPG. 1

Government incurred revenue loss of $13m in 2007/08 and 2008/09 as a result of subsidizing fuel products.

The three-month pass through to generate domestic retail prices by calculating the average base CIF of monthly imports of fuel products. Following this, a price formula is used to include taxes and distribution margins. 3

LPG was subsidized on average by $7.4m for the past 4 years.

The Government of Saint Lucia increased the excise tax above the $4.00 target to offset revenue loss over the period.

A Compilation of Working Papers by OECS Scholars

137

The Excise Tax rate was increased to $4.00 from $2.50 on July 3rd 2017.

The Service charge was increased to 6 % from 5 %on July 1, 2015

The WTI is a popular indicator or benchmark used to measure the movement in the international prices of crude oil.

Spot domestic and international prices (WTI) are based on the effective date of the price change.

This approach is consistent with the methodology adopted in Saint Lucia.

Three-week fuel price calculation methodology.

The base CIF is a proxy for international prices in this methodology, derived from the invoices provided by local importers.

The average base CIF represents the cost to importers excluding import duties, taxes and distribution margins.

Currently, the fixed component is the $1.50 while the variable is $2.50.

REFERENCES Benesty, J., Chen, J., & Cohen, Y. H. (2019). Pearson Correlation Coefficient. Retrieved from SpringerLink: https://link.springer. com/chapter/10.1007%2F978-3-642-00296-0_5 Coady, D., Arze del Granado, J., & Eyraud, L. (2012). Automatic Fuel Pricing Mechanisms with Price Smoothing: Design, Implementation and Fiscal Implications. USA: International Monetary Fund. Greene, W. (2008). Econometric Analysis. USA: Pearson, Prentice Hall. Hauke, J., & Kossowski, T. (2011). Comparison Of Values Of Pearson’s And Spearman’s. Poland: Adam Mickiewicz University, Institute of Socio-Economic Geography and Spatial Management. Kozak, M., Krzanowski, W., & Tartanus, M. (2012). Use of the correlation coefficient in agricultural sciences: problems, pitfalls and how to deal with them. Poland: University of Exeter. Mandal, K., & Bhattacharyya, I. (2012). Is the Oil Price Pass-Through in India any Different? ResearchGate, 832-848. Pearson, Kendall, & Spearman. (2019). Correlations. Retrieved from Statistics Solutions: https://www.statisticssolutions.com/ correlation-pearson-kendall-spearman/ Reboredo, J. (2011). Modeling oil price and exchange rate co-movements. Journal of Policy Modeling. Stangor, C. (2012). Psychologists Use Descriptive, Correlational, and Experimental Research Designs to Understand Behaviour. Retrieved from BCCampus: https://pressbooks.bccampus.ca/introtopsychologykpu/ chapter/3-2-psychologists-use-descriptive-correlational-and-experimental-research-designs-to-understand-behaviour/ StatisticsSolutions. (2019). www.statisticssolutions.com. Retrieved from Correlation; Pearson, Kendall and Spearman: https:// www.statisticssolutions.com/correlation-pearson-kendall-spearman/

138 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX FIGURE A: CURRENT FUEL PRICING METHODOLOGY ADOPTED BY GOVERNMENT OF SAINT LUCIA

How is the retail price of an imperial gallon of gasoline & diesel calculated?

NB: Retail Price = Landed Cost + Wholesale Price + Retail Margin

A Compilation of Working Papers by OECS Scholars

139

ADVANTAGES AND DISADVANTAGES OF ADOPTING RECOMMENDATION 1 AND 2

FIGURE B: ADVANTAGES AND DISADVANTAGES TO THE COUNTRY

FIGURE C: ADVANTAGES AND DISADVANTAGES TO THE CONSUMER

140 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Jilayne Clery-King

Economist, Research and Policy Unit Department of Finance, Government of Saint Lucia

Jilayne Clery-King is an Economist attached to the Department of Finance; Research and Policy Unit in Saint Lucia. Over the past 12 years, she also served as a Research Officer and functioned in various administrative roles within the Government of Saint Lucia. She has a keen interest in research and has a penchant for reading. Mrs. Clery-King graduated with First Class honors from the University of the West Indies with a dual degree in Management and Economics. She also holds an MBA (Distinction) with a specialization in Project Management from Edinburgh Napier University.

A Compilation of Working Papers by OECS Scholars

141

An Investigation of Public Sector Project Implementation in Saint Lucia Rosemary Pierre Louis

142 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT Governments of the region are under extreme pressure to improve public sector performance in the region while containing expenditure growth. Factors such as salaries, healthcare, pension cost and interest cost add to budgetary pressures forcing governments to borrow to finance capital outlays. Locally the cost of borrowing has increased on account of higher borrowing from the commercial banks. With the cost of loans increasing governments are called to account for their achievements in project delivery. This paper shows that over the past decade as more resources are added to project implementation, the value of projects implemented have diminished. INTRODUCTION A project is a temporary endeavor with a beginning and an end designed to create a unique product, service or result. Project implementation is one of the five processes of Project Management. According to Baker (2003) project management involves project initiation processes, project planning processes, project implementation/executing processes, project controlling processes and project termination/ closing processes. In addition, successful project implementation requires that a project manager possess key qualities such as the ability to select a competent team, the ability to motivate staff to achieve the desired project outcomes and the ability to monitor and assess project activities using right project management tools (JK. Pinto). The implementation of projects, identified in the national budget, remains a weak area for many countries in the Caribbean region. J.W. Mabelebele (2006) stated in a study on the challenges of project implementation in South Africa that weak project implementation may arise from poor coordination of interdepartmental projects, power, authority, political interference, lack of project management skills and short budget cycles. Poorly managed projects can be costly especially when funded by market instruments as loans and bonds. This weak implementation is not absent from Saint Lucia and other regional states. According to the annual budget estimates, the project cost outlined for the period 2009/10 to 2018/19, resulted in an annual project implementation of 70.5 percent. The government of Saint Lucia has instituted a number of project implementation units for effective and efficient project implementation. Within the public service of Saint Lucia, the Project Coordination Unit (PCU), the Constituency Development Programme (CDP) and the National Authorization Office (NAO) are the three major project implementation units. In addition, other project implementation units are established for major projects as St Jude’s Reconstruction and VAT implementation Project Units. Furthermore, departments within the public service have fully staffed project implementation units responsible for the implementation of the department’s projects as the Departments of Infrastructure, Education and Health. The National Development Unit within the Department of Economic Development, Transport and Civil Aviation, an additional project implementation Unit, provides oversight and monitoring of project implementation processes. In addition, the unit has the responsibility for strengthening planning, donor coordination and project cycle management functions through coordinated approaches with line agencies. A country diagnostic data gathering exercise and workshop was conducted in 2017 to assess the country’s structure and gaps in Project implementation. The Project Management and Delivery Unit (PMDU) was established in 2017 as a recommendation from this diagnostic review. This unit has the responsibility of further strengthening project implementation in Saint Lucia, monitoring and reporting on key priority areas, providing performance insights and solving implementation problems in key priority areas while building capacity in delivery across the public service through better planning implementation and knowledge sharing.

A Compilation of Working Papers by OECS Scholars

143

To ensure evidence based national infrastructure is instituted the National Integrating and Planning (NIPP) unit was established in 2018. The unit was formed as a new evidence based national infrastructure planning and programme paradigm across all relevant government agencies. The NIPP was established with the aim of defining the overarching vision, strategy and roadmap for the development of the national infrastructure agenda by assessing the current and future infrastructural needs of agencies and the overall public service and ensuring that financial and capital resources align with national policies. Despite the establishment of project management and implementation units within the Public Service, successful and effective project implementation remains weak as projects go on beyond the delivery date and above budget estimates while many projects are not implemented. This study is undertaken to assess the efficiency of the public service in implementing the annual capital budget. The study uses a World Bank’s approach of operating cost relative to project cost for assessing the efficiency of project implementation in Saint Lucia The paper aims to provide suitable recommendations methods and strategies for effective project implementation in the public service of Saint Lucia. The paper is therefore structured as follows: Section 1: highlights the introductory and background to the research problem. The research problem and aims will be highlighted in section 2 while section 3 provides an insight into the objectives of the study which is followed by the literature review. Section 7 and 8 provides details on the descriptive statistics and findings of the study, respectively. The conclusion with some key recommendations will be presented in section 9.

PROBLEM STATEMENT Public sector project implementation is an integral part of the government. Most developing countries face difficulties in implementing its capital budget. Capital spending by the central government of Saint Lucia, which is used as a proxy for projects, contracted by 15.3 percent in 2018/2019 to $213.4 million, the lowest since 2008/2009. Given the economic impact of capital expenditure projects, there is a need for projects to be successfully implemented to achieve its desired objectives. Thus, this research seeks to highlight the most effective ways to improve the efficiency and effectiveness of public sector projects in Saint Lucia.

MOTIVATION / OBJECTIVES As an economist involved in supervision of project implementation, there was a concern of the poor levels of implementation of projects within the government as there was persistent downgrade of donor funded projects due to poor levels of implementation coupled with high commitment fees on undisbursed balances. In addition, the issue of poor project implementation remains a concern to the Saint Lucian public. Therefore, this research aims to: 1. Highlight the project implementation units within the Public Service of Saint Lucia 2. Estimate the operating cost of project implementation units over the research period 3. Estimate the effectiveness and efficiency of the project units using an approach as recommended by the World Bank.

144 | Research: The Platform for Innovation, Competitiveness and Growth

LITERATURE REVIEW Knutzen and Blitz (1991: 2) and Young (1996) defines a project as a set of principles, methods, tools and techniques for the effective management of objective-oriented work, in the context of a specific and unique organisational environment. Turner (1993: 8) and Wilson-Murray (1997) pointed that a project is a temporary endeavour in which human material and financial resources are organized in a novel way, to undertake a unique scope of work, of given specification, to deliver beneficial change defined by quantitative and qualitative objectives. The (National Implementation Research Network 2014). Kerzner (2003) defines implementation as a speciﬁed set of activities designed to put into practice an activity or programme of known dimensions while Venter (2005: 81) focus on the management dimension of projects, and states that resources must optimally be utilized to ensure that a project’s output adheres to time, budgetary and quality constraints. Maylor (1996: 3) and Burke (2006: 2, 3) further indicates that this includes planning, organising, directing and controlling activities in addition to motivating what is usually the most expensive resource on a project – people. There exists a wealth of literature surrounding the challenges faced with project management practices in the public service. Islam (2016) identifies some key challenges faced by managers in the public service including the inability of project managers to select team members based on qualifications and compensation systems based on longevity. He purported that most public project managers are not familiar with results-oriented project management. Kharbanda and Stallworth (1986:72) further states that the problem of implementation in developing countries is exacerbated by insufficient experience in project management by developing countries. Schultz and Slevin (1987), found that management support for projects, or indeed for any implementation, has long been considered of great importance in distinguishing between their ultimate success or failure. Modise Lucas Sefhemo (2016) in a study on project implementation in Botswana further claims that in developing countries if projects are not abandoned on the way, they are completed at an extra cost. There is no rule of thumb for assessing the effectiveness of projects as the cost of implementing the project depends on a range of factors such as complexity, labour rates and competition. However, the World Bank Guidelines on Implementation Completion Report (ICR) uses (a) Economic analysis i.e. Net Present Value and Economic Rate Return (b) various aspects of design and implementation such as delays in implementation of key activities and frequent staff turnovers to assess efficiency of projects. The World Bank recommends using a rule of thumb of 10 percent operating cost to project cost in assessing WB funded projects while the European Union uses a ratio between 15 percent to 25 percent. This approach speaks to the effectiveness of projects if operating costs of projects remain below these prudential limits. This study therefore wishes to contribute to the existing literature by estimating the effectiveness of project implementation in the public service in Saint Lucia using the World Bank methodology. This approach has been used by the World Bank to assess the effectiveness of its projects and believe it to be an effective approach. This study will therefore employ this approach to evaluate the effectiveness of project implementation in a small island developing state as Saint Lucia.

A Compilation of Working Papers by OECS Scholars

145

PROJECT IMPLEMENTING UNITS IN THE PUBLIC SERVICE OF SAINT LUCIA This section of the study provides a synopsis of the three major project implementation Units within the Public Service. Project Coordination Unit (PCU) The Government Saint Lucia is required to maintain, at all times during implementation of World Bank (WB) financed Projects, a Project Coordination Unit (PCU) with the responsibility of coordinating project related activities. Accordingly, the PCU is responsible for fiduciary management of all projects funded by the World Bank and therefore provides technical support to ensure that project activities are implemented in a timely and comprehensive manner. The responsibility for overall oversight and monitoring of the project procurement lies within the PCU, which acts as the World Bank’s main counterpart for all procurement aspects of the project and ensures that procurement under the project is carried out in accordance with the Grant Agreement and the Procurement Plan. The PCU is also responsible for the financial management and disbursement functions of the projects financed by the World Bank. Since its establishment in 2000, in Saint Lucia, the PCU has completed fifteen (15) Projects valued at USD127.9 million (Appendix 1). Currently, the PCU is responsible for oversight for eight (8) ongoing projects valued at USD 148.48 million. Community Development Programme (CDP) The Constituency Development Programme (CDP) was instituted in 2010 as an apt mechanism for harnessing the rich and varied talents at the community level in defining an indigenous, communitybased approach to social and economic development. Thus, the CDP has as its primary objective, which involves the development of social and economic programmes aimed at stimulating economic growth while promoting bi-partisanship and community participation with constituencies. The CDP facilitates consultation both within the community and among various community and government agencies and encourages broad level participation by civil society organizations (CSOs), various community groups and the Opposition. Since the introduction of the Constituency Development Programme (CDP) constituency representatives help in the creation of the framework for all communities to participate more directly in their own development, while at the same time strengthening the co-operative mechanisms that are expected to lead to greater social cohesiveness. The CDB project is funded through grants from the Republic of China (Taiwan) ROCT based on the annual programme developed by each constituency. However, each constituency has a maximum annual budget of EC$294,000. Constituencies may combine resources to purchase a common development objective. Expenditure categories applicable to the project include works, goods, consultancy services and operational costs. Procurement for the project is carried out in accordance with the Government of Saint Lucia’s procurement regulations. Under the CDP, the GOSL is responsible for counter funding to meet the operating costs of the project. Since the inception of the programme over 800 community projects have been completed valued at an estimated XCD 160 million (See appendix 2).

146 | Research: The Platform for Innovation, Competitiveness and Growth

National Authorization Office (NAO) The National Authorization Office (NAO) which is the European Union (EU) funded initiative aims to ensure the efficient and optimal utilization of EU development assistance in support of national development goals and objectives. The NAO was established in 1995 was managed by consultants appointed by the EU until 2000 when the NAO office was controlled by the government. In 2001 additional EU funding became available under a special framework of assistance funding. During the implementation of the Banana Accompanying Measures (BAM) between 2015 and 2018 the NAO office was part of the Ministry of Agriculture. Based on Saint Lucia’s allocation, Euros 750,000 was allotted to Saint Lucia for support to the NAO under the 11th European Development Fund (EDF) for the period 2018-2021. An additional Euros $15 million is allotted for employment generation initiatives by March 2020 but this is contingent on the full operationalization of the funded OK EU Hospital which is the largest EU funded project in the Eastern Caribbean estimated at Euros 36.6 million. From 2007 to present the EU has invested an estimated Euros 64.9 million in Saint Lucia. DESCRIPTIVE STATISTICS The medium-term expenditure framework (MTEF) sets the basis for the annual capital budget led by the Ministry of Finance, and draws together the contributions from various government agencies, ministers and parliamentarians. In the fiscal year 2009/10, $241.3 million of the capital budget was implemented equivalent to an implementation rate of 60 percent. Ten years later $213.4 million in projects were implemented. An estimated XCD $2,665.8 million has been spent on capital outlays at an average annual project implementation rate of 70.5 percent.

FIGURE 1: BUDGETED VERSES ACTUAL CAPITAL EXPENDITURE IN EC$ MILLIONS

The largest outlay was on Road and Infrastructure. Of the $646.5 million expended on road and infrastructure, $282.5 million was expended on improving the roads over the past decade. Notwithstanding this outlay 50 percent of Saint Lucia’s total road network are categorized as “poor” and “very poor” 1 according to the International Roughness Index (IRI). The remaining expenditure was on Hurricane Tomas Emergency Recovery, bridges and culverts, desilting and drainage works. Expenditure on Tourism related projects stood at $414.2 million on account of expenditure on Tourism Marketing ($381.2 million), Community Tourism Promotion ($4.0 million) and Technical Assistance for Eco/Agro Tourism Programme ($2.1 million). A Compilation of Working Papers by OECS Scholars

147

FIGURE 2: EXPENDITURE BY FUNCTIONAL CLASSIFICATION

Saint Lucia has 33 wellness centers and a polyclinic providing primary health care and two hospitals and two district hospitals providing secondary healthcare and a mental health facility. Based on the 2016 SLBS- HBS survey 18.0 percent of the 4,568 participants indicated that they were covered by Health Insurance which suggests that the majority of the population depends on the public health care system. This may explain the $404.2 million expended on that sector. The new national hospital project and the St Jude’s Hospital accounted for $178.1 million and $109.5 million respectively over the review period. Expenditure on other items which refers to the amount expended on energy, climate change, capacity building, telecommunications and enhancing competitiveness amounted to $320.5 million from fiscal year 2009/10 to 2018/19. Of that amount the purchase of furniture, equipment, vehicles and supplies amounted to $124.0 million over that same period. The education sector benefited from $134.2 million in investment over the period under review. The education system in Saint Lucia comprises eighty primary schools, twenty-four secondary schools, eight post-secondary schools and three main skills institutions none of which were constructed in the last decade. Repairs to these school plants amounted to $15.0 million. The Basic Education Enhancement Project ($12.7 million) and Education Enhancement through ICT ($11.5 million) were among the major projects undertaken. Given the vital role of the agricultural sector in sustaining livelihoods in the rural communities, $128.4 million was spent on the sector. Major outlays were on the National Abattoir Project ($17.9 million), Banana Commercialization and Agriculture Diversification ($7.1 million), Banana commercialization and Agriculture Diversification (7.0 million) Management of Black Sigatoka ($7.0 Million) and Expansion of Praedial Larceny Programme ($4.3 million). Of the total expenditure 51.2% was expended on economic infrastructure mirroring the number of projects implemented relating to tourism, community work, agriculture, roads and water infrastructure.

148 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 3: CAPITAL EXPENDITURE BY FUNCTIONAL CLASSIFICATION IN EC$ MILLIONS

Expenditure on General Services which refers to expenditure on law and order, disaster preparation and upkeep has been on a downward trajectory. During the triennium 2011/12 to 2013/14 the government spent over $26.8 million on refurbishment and maintenance of government buildings. Social spending2 over the period has declined on average by 13 percent.

FIGURE 4: EXPENDITURE ON SELECTED CATEGORIES IN EC $ MILLIONS

METHODOLOGY This paper aims to examine the operating expenditure of the departments responsible for implementation of projects within the public service in Saint Lucia and compare this to prudential ratios as proposed by the World Bank and the European Union to ascertain the efficiency of project implementation. In the case of the World Bank funded projects the Bank measures efficiency when preparing an Implementation Completion Report (ICR). Efficiency is basically a measure of how economically resources and inputs are converted to results3. The shortcomings in efficiency have to do with the extent to which the operation fails to achieve a higher return than the opportunity cost of capital. A Compilation of Working Papers by OECS Scholars

149

The World Bank Prudential Ratio = Operating Cost of Projects

Cost of the Projects

For the World Bank funded projects, the bank uses a rule of thumb of not more than 10% of the project cost while the European Union uses Prudential Ratio of 15 percent to 25 percent4. Operating Cost = Personal Emoluments + Travel and Subsistence allowance + Training + Telephone + Office and General Expenses The gaps in project management in the public service were identified by comparing the findings with best practices as documented in both academic and professional publications.

RESULTS

TABLE 1:RESULTS

Department

Operating Cost in EC$M (2009/10 to 2018/19)

Total

364.33

Of which: Infrastructure

111.51

Tourism

18.32

Commerce

11.49

Education

9.60

OPM

7.82

Planning

6.82

Agriculture

6.68

Finance

6.61

Economic Affairs St Jude Hospital

156.08

Reconstruction Project NAO

6.03

Economic Planning Unit

4.62

National Development Unit

4.16

Housing

3.59

Health

2.23

Cabinet

0.05

150 | Research: The Platform for Innovation, Competitiveness and Growth

Efficiency in the delivery of public sector projects refers to the entire process of turning public funds into outcomes beneficial to the society. Operating cost between 2009/10 to 2018/19 increased at an average rate of 10 percent annually. Notwithstanding this, the efficiency of project implementation declined per annum over the period under review despite the variations to project implementation in Saint Lucia. The results indicated declining levels of project implementation as additional project implementation Units were introduced. Using data from the smart stream recurrent expenditure reports, eleven (11) departmental PIUs were recorded excluding the Department of Economic Affairs with the responsibility for the three largest PIUs and the Economic Planning and National Development Units. Of the XCD$364.33 million expended in operating cost to implement the capital budget over the period 2009/10 to 2018/19, the St Jude’s Reconstruction Project recorded the highest operating cost of XCD156.1 million. The Technical Services Department (TSD), of the Department of Infrastructure through the project planning and designs, road construction and building and grounds maintenance units is responsible for the management, organizing, implementing and supervising all public road, infrastructure and buildings. The operating cost of the TSD amounted to an estimated XCD$111.51 million. During the year 2011-12 a total of $366.0 million and $344.79 million in projects were implemented with the operating cost of $30.6 million. Projects of this magnitude were implemented due to the rehabilitation works following the passage of Hurricane Tomas in 2010. In 212/13 a similar trend was realized with the following the passage of the Christmas Eve trough in 2012. In the last triennium however with operating cost in excess of $50,000 the value of projects implemented have declined.

FIGURE 5: EFFICIENCY OF PROJECT IMPLEMENTATION IN EC$M

A Compilation of Working Papers by OECS Scholars

151

CONCLUSION Based on the forgoing, the capacity of the Public Service to implement the capital budget has diminished. Anecdotal information suggests that the increasing operating cost may be attributed to lack of training in the following areas: project management, contract management and risk management. The lack of appropriate software coupled with the extensive use of manual systems may also be contributing factors hindering efficient project implementation in the public service. The inability of the technical personnel to write clear, detailed and unambiguous technical specifications in their area of expertise often requires hiring consultants to undertake such tasks which could be lengthy and costly. Further analysis would have to be undertaken to validate these findings.

RECOMMENDATIONS Having reviewed the cases of countries with best project management practices, the following are proposals to improve project management in the public service: 1. Project proposals should include terms of reference or specifications prior to approval in the capital budget upon submission during the Budget process. 2. The structure of the departments designed to implement projects should be reviewed. It is recommended that a single coordinating project unit adequately staffed for monitoring, etc.…. of project implementation. 3. Training of staff in the area of Project management is necessary for improving project implementation in the public service. 4. There is ample evidence that the South African government increasingly relies on the private sector for the provision of services (see Farlam 2005; Harris 2003). According to Browne, Nemo to, Visser and Whiteing (2003), the reduction in employment in many public sector departments has resulted in economy-wide shortages of some forms of skilled labour. Accordingly, the Government should consider allowing the private sector to coordinate project implementation while providing oversight. This is recommended due to the success of private sector project implementation.

NOTES 1

The International Roughness index is the key indicator of road surface condition

Social spending refers to expenditure on health, education, culture, housing & settlement and sports

World Bank guidelines, Implementation Completion and Results Report (ICR) for Investment Project Financing (IPF) operations. 3

This ratio does not apply to the Operations of the NAO office in Saint Lucia as the office is a creature of the Cotonou convention. 4

152 | Research: The Platform for Innovation, Competitiveness and Growth

REFERENCES Kattel, R., Cepilovs, A., Drechsler, W., Kalvet, T., Lember, V., & Tõnurist, P. (2013). Can we measure public sector innovation? A literature review. LIPSE Project paper. Jałocha, B., Krane, H. P., Ekambaram, A., & Prawelska-Skrzypek, G. (2014). Key competences of public sector project managers. Procedia-Social and Behavioral Sciences, 119, 247-256. Jung, J. Y., & Wang, Y. J. (2006). Relationship between total quality management (TQM) and continuous improvement of international project management (CIIPM). Technovation, 26(5-6), 716-722. Mabelebele, J. M. (2017, May). Prospects and challenges of implementing projects in public service South Africa. In Proceedings of the 2006 PMSA International Conference on Growth and Collaboration for a Project Management Profession (pp. 247-254). OECD (2014), Public Governance and Territorial Development, OECD Publishing. Sutterfield, J. S., Friday-Stroud, S. S., & Shivers-Blackwell, S. L. (2007). How not to manage a project: Conflict management lessons learned from a DOD case study. Journal of Behavioral and Applied Management, 8(3), 218. Van der Waldt, G. (2011). The uniqueness of public sector project management: A contextual perspective. Politeia, 30(2), 66-87. Venter, F. (2005). Project management in Ghana: expectations, realities and barriers to use. The journal for transdisciplinary research in Southern Africa, 1(1), 20. Zhu, X. (2013). PoA Implementation in Asia and Pacific. In 3rd Workshop on Enhancing the Regional Distribution of CDM Projects in Asia and the Pacific: Harnessing Opportunities for CDM.

A Compilation of Working Papers by OECS Scholars

153

APPENDIX TABLE A: PROJECTS IMPLEMENTED BY THE PROJECT COORDINATION Project (Name)

Value of Project €

Year Implemented Start date

End Date

3,999,861

2007

2010

Economic Diversification and Private Sector Development (SFA 2004)

2,981,414.00

2008

2011

Social Recovery and Human Development Project (SFA 2004)

2,760,000.00

2008

2011

Banana Commercialization and Agricultural Diversification (SFA 2004)

1,071,059.00

2008

2011

Banana Commercialization and Agricultural Diversification Programme (SFA 2005)

2,677,313.00

2008

2010

Economic Diversification Programme (SFA 2005) – Strengthening Trade through Rural Investments and Development of Entrepreneurship

1,150,000.00

2008

2010

Economic Diversification Competitiveness through linkages

1,281,302.28

2009

2012

Special Framework of Assistance (SFA) 2006 – Poverty Reduction Through Community Based Development Planning Social Recovery Programme

1,609,185.00

2010

2012

Technical Cooperation Facility (TCF)

269,630.00

2010

2011

Technical Cooperation Facility (TCF)

28,150.00

2012

2013

Banana Accompanying Measures

10,350,000.00

2013

2018

Construction of a New National Hospital

36,566,305.16

2009

2015

Supply of Equipment & Furniture for the Saint Lucia New National Hospital

8,972,578.83

2012

2016

Support to Saint Lucia Health Sector (NIP)

6,885,000.00

2013

2017

Infrastructure Rehabilitation Programme – EDF 10 – B Envelope

970,000.00

2013

2016

Selective Measures to provide Sustainable access to Safe Drinking Water in Saint Lucia (MDG Initiative)

810,000.00

2013

2016

Economic and Agricultural Diversification and Poverty Reduction through Integrated Natural Resources – (SFA 2003) Environmental Management Fund

154 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Rosemary Pierre Louis

Economist, Research and Policy Unit Department of Finance, Government of Saint Lucia

Rosemary Pierre Louis joined the Public Service in Saint Lucia in January 2009 as an Economist in the Research and Policy Unit. Since then, she has worked in a number areas including energy, debt and fiscal management. She also served as Deputy Project Manager for the World Bank funded portfolio in Saint Lucia for two years. The majority of her research has been in the areas of construction and project management.

A Compilation of Working Papers by OECS Scholars

155

9 Taxes and the Demand for Intra-Regional Travel

Javan Lewis

156 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT With the importance of the tourism market to Caribbean economies this study focuses on the relationship between intra-regional travel and taxes. Despite the concern of increasing intra-regional airfare prices as a result of high TFC’s this revenue stream is vital to the governments of the region. The paper explores how flexible can these governments be relating to TFC’s on air travel within the region. Elasticity multipliers developed by Inter Vistas Consulting Inc were used to develop the price elasticity of demand for intra-regional travel. The elasticity developed was -1% making intra-regional travel unitary price elastic. The results show that with a 5% and 10% reduction in taxes, fees, and charges will result in a 5% and 10% increase in travel inter-regional demand, keeping with economic theory of unitary elasticity. As a result, total government revenue would increase by 3.4% and 7% as per a 5% and 10% reduction in TFC’s respectively. Keeping with economic theory average hotel accommodation revenue would also increase by 5% and 10% also. With tourism being the driving force of the Saint Lucian economy, an increase in intra-regional travel/ tourism would have a ripple throughout the economy, boosting indicators such as unemployment. Caribbean visitors tend to drill down into the culture of Caribbean destinations and become brilliant ambassadors for the destination, hence driving the growth in arrivals. INTRODUCTION The Caribbean region is made of an archipelago of islands separated by spans of sea, large enough to make air transport the most practical and efficient means of intra and extra-regional travel for the majority of islands. The economies within the region are characterized by being small in terms of landmass and population relative to first world, high income countries, in addition to having resources and economic constraints. Most of these economies depended heavily on the agricultural sector as the main source of economic income until the mid-1990’s where the liberalization of the sector led to the termination of the preferential treatment agreements with Europe. As such service-based sectors, in particular, tourism, took the forefront of economic activity, resulting in increasing demand for air travel, both intra and extra regionally. Air travel within the Caribbean is provided by 10 airlines, with Leeward Islands Air Transport (LIAT), and Caribbean Airlines being the two main carriers. These two airlines provide travel to and from the following countries; Antigua and Barbuda, Aruba, Bahamas, Barbados, Bonair, British Virgin Islands, Cuba, Curacao, Grenada, Guyana, Guadeloupe, Dominica, Dominican Republic, Jamaica, Puerto Rico, Santo Domingo, St. Barths, St. Eustatius, Saint Lucia, St. Maarten, St. Vincent and the Grenadines, and Trinidad and Tobago. The airfare for travel within the region has a structure comprising three components; airline base fees, airline fixed surcharges and, the Government taxes, fees and charges (TFC’s). Costs of building, and operating airports and facilities are captured in the charges. Airline base fees which are generally the largest component and are driven by the basic costs associated with operating in the region. Taxes, included in the airfares, is a factor of government policies. These taxes are implemented to address the externalities imposed by air transport, such as climate change, or simply to raise revenue for the refurbishment/ development of an airport. The airport development charge is the largest component within the TFC’s. The rest of the TFC’s includes airport service charge, security charge, facilitation charge, passenger facility charge, travel tax, fuel surcharge, security surcharge, passenger service charge and environmental levies.

A Compilation of Working Papers by OECS Scholars

157

These costs and charges are present in the airfares to and from Saint Lucia. Saint Lucia, in 2017, reinstated its airport development tax which added to the existing airport service charge and travel tax of $USD25.00 and 7.5% of the base airfare respectively. In addition, the airport development charge was recently increased in Saint Lucia. The airport development charges are also present in other islands such as Grenada, Dominica, Guyana, and St. Kitts and Nevis. The introduction of taxes and charges by governments to improve revenue streams or deal with externalities results in higher airfares for travellers throughout the Caribbean region. It is within this context that the discussion of increasing ticket prices and its impact on intra-regional travel and tourism has been stimulated. The Government taxes, fees and charges component of the airfare in some cases are greater than the airline base fees. In 2011 a Low-Cost Carrier (LCC), REDjet, entered the regional air travel market. The primary goal of REDjet was to open the market to the substantial lower socio-economic population in the region which would be able to afford to travel, due to the availability of lower air fares. This airline was however short-lived due to the barriers caused by the regulatory framework of the air transport sector, and lack of government subsidies. The study therefore investigates the response of a reduction of taxes, fees, and charges to assess its impact on airfares in addition to the response to travellers to Saint Lucia. The impact of the reductions on government revenue and the wider economy is also assessed. The study uses elasticities from the study of air travel demand elasticities by Intervistas Consulting Inc., to measure the response of a reduction in airline TFC’s on intra-regional travel demand, government revenue and the wider economy. This paper therefore discusses the development of the tourism industry in the region in section 1, the literature surrounding the topic is discussed in section 2. Section 6 assesses the impact of intra-regional travel with the reduction of TFC’s on the tourism industry while the impact of government revenue and GDP on tourism dependent economies is assessed in section 5. The policy recommendations are then discussed in section 6. LITERATURE REVIEW Extensive research exists on air travel both internationally and regionally, in relation to taxes and intraregional travel however, empirical research surrounding the extent of the impact of a change in airfare prices is limited. Data collection and demand analysis for air transport is very complex as air travel is considered an intermediate product, in terms of service consumption, since flights are usually demanded to reach various objectives (InterVistas Consulting Inc., 2007). The study by InterVistas Consulting Inc., estimated air travel demand elasticities applicable to a wide range of air transport markets using Ordinary Least Squares (OLS) regression analysis method employing traffic to air fares, income (GDP) levels, and other variables. Traffic was regressed onto similar explanatory variables as in prior models, but also onto lagged values of traffic. Using this method elasticity estimates were developed for various geographic markets. These elasticity estimates will be used in this study to assess the impact of a reduction of TFC’s in air fare prices on the demand for intra-regional travel and Government’s revenue. Given the significance of the tourism industry to the economies of the countries across the region, and the changes that are taking place within the industry and to understand how the industry and economic growth is linked the Caribbean Development Bank embarked upon a study. This study investigated the relationship between economic growth and tourism to better assist their borrowing member states in developing strategies for enhancing the industry’s economic impact through policy (Caribbean Development Bank, 2017). This study derives estimates of economic impact using an econometric model to examine the relationship between tourism activity and GDP growth in their borrowing member states. The findings suggest that air travel is a vital component to the region’s tourism industry and that intra-regional travel contributes to the borrowing member countries’ GDP through intra-regional tourism. Policies such as taxes and the regulatory framework of the sector are ways in which the policy makers can influence intra-regional travel.

158 | Research: The Platform for Innovation, Competitiveness and Growth

(Burghouwt, Boonekamp, Zuidberg, & Spijker, 2015) assessed the economic benefits of reducing aviation taxes by taking into account the impact on society and the money and time saved by consumers because of less money being spent in airfares. An alternative approach used was a macroeconomic approach, which calculated the economic contribution of an increase in air travel and its impact on GDP and employment growth. The findings suggested that the potential benefits of economic growth for consumers and businesses would be forgone with consistent tax rates. The literature also deduces that a reduction of taxes on airfare, it is possible that unemployment and slow economic growth in Caribbean countries will continue to be significant crippling characteristics in their economies. (Soverall, 2012) studied the issues within the context that the Caribbean was the only region in the world despite being tourism dependent, without a LCC and suffered a reduction in travel while all other regions experienced growth. The research aimed to advance understanding of the need for an LCC in the Caribbean, the practical benefits to be derived from such an operation, and the applied significance for policy makers, particularly in the areas of intra-regional travel, air transport policy and Caribbean tourism in general (Soverall, 2012). This study was carried out using a mixture of qualitative and quantitative data analysis. According to (Soverall, 2012) the low-cost airfares resulted in an immediate “Redjet effect” by stimulating a significant increase in intra-regional travel which had steadily declined by over 25% since 2005. Highlighting the significant impact of lowering of taxes on the demand for intra-regional travel which would further stimulate the economic activity of countries in the Caribbean region. The study suggests that not only is there a need for a decrease in airfares for stimulating growth in intra-regional travel, but a review of the regulatory framework by economies is required. A study on the factors inhibiting intra-regional travel in the OECS carried out by (El Perial Management, 2015), approached the project primarily from the perspective of individuals travelling or potentially travelling within the region. In the case of business organizations and tourism authorities their surveys were implemented through the use of interviews that explored the answers to the questionnaire deeper. According to (El Perial Management, 2015) very small populations, small but slowly growing economies with high unemployment levels are foundation dynamics leading to the constrained travel by residents of the OECS countries. A way in which policy makers can alleviate this constraint on travel is by the reduction of airfare taxes which would result in lower airfares. El Perial Management (2015) states that both factors of high airfares and low connectivity are important however, poor connectivity appears to be a more significant constraint to intra-OECS travel. The existing literature highlights the limitation when traveling within the Caribbean region and the necessary amendments required to improve the sector’s performance. This study therefore adds to the existing literature by assessing the impact of the reduction of taxes resulting in lower airfares and how it affects intra-regional travel, the economy, and to estimate the net revenue impact to the government. METHODOLOGY To determine the economic impact and effect on government revenue of a reduction of taxes on air fares elasticity multipliers developed by Intervistas Consulting Inc. in a report on “Estimating Air Travel Demand Elasticities” will be employed. The concept of elasticity and price elasticity of demand was founded and developed by Alfred Marshall in the year 1890. Price elasticity demand quantifies the change in demand which occurs as a result of a change in the price of a good. Price elasticity demand is calculated by dividing the percentage change in quantity demand by the percentage change in the price of the good. Price elasticity allows firms to assess price changes and deprival of their optimal prices, and allows governments to assess the effect on tax incidents amongst other advantages. InterVistas Consulting Inc., estimated air travel demand elasticities applicable to a wide range of air transport markets using Ordinary Least Squares (OLS) regression analysis method employing traffic to air fares, income (GDP) levels, and other variables. In developing the elasticity multipliers, the two-stage least squares regression technique was also used since the explanatory variables were believed to be correlated with the model’s error term. In this instance the natural; logarithm of distance and the natural logarithm of fuel prices were used separately and combined as potential instrumental variables. A Compilation of Working Papers by OECS Scholars

159

TABLE 1. GEOGRAPHIC MARKET ELASTICITY MULTIPLIER

Geographic Market Electricity Multiplier

Comment

Intra North America

1.00

Most research uses US data. This is our reference point.

Intra Europe

1.40

Shorter average distances, observed use of very low fares resulting in great market stimulation. The significantly low fares in Europe (relative to North America) are consistent with higher elasticities in Europe. Traditionally the European market had high charter carrier share, which today is merely being converted to very low fare LCCs.

Intra Asia

0.95

The LCC phenomena is emerging in Asia, but modest sized middle class in many markets suggests somewhat less elastic than in North America.

Intra Sub-Sahara Africa

0.60

These economies have limited middle class, resulting in high weight on higher income individuals who are less elastic.

Intra South America

1.25

There is an emerging middle class which makes the market more elastic than sub-Sahara Africa, and LCCs are emerging in Brazil, Chile and Mexico.

Trans Atlantic (North America - Europe)

1.20

This market is often observed to have fares only slightly higher than domestic U.S. fares, consistent with high price elasticity. Market has been well developed by charter carriers, consistent with high price elasticity. Price is likely more important than frequency in this market than in domestic U.S.

Trans Pacific (North America - Asia)

0.60

TransPacific has had no charter services, and continues to have major markets (Japan, China) with less liberal pricing provisions. Some emergence of long haul LCCs (e.g. Oasis) but at present this market seems to be less elastic than domestic U.S. and than the well developed trans Atlantic, which serves a substantial middle class.

Europe - Asia

0.90

This market has marginally lower elasticities than the U.S. domestic market.

Source: InterVistas Consulting Inc.

An autoregressive distributed lag model was also developed and used with OLS regression analysis. Elasticities for different situations can be developed by selecting the relevant base elasticity and applying the relevant multipliers. The first step in deriving the elasticity for air travel in the region would be to decide on the aggregate level to be used. There are three different levels; Route/Market level, National level, and Pan-national level. At the route or market level the elasticity response is generally lower, where travellers are faced with a change in fares on all carriers serving a particular route. With fewer options of avoiding a price increase the national level would be expected to have a lower fare elasticity. At the Pan-national level the elasticity is expected to 160 | Research: The Platform for Innovation, Competitiveness and Growth

be lower since the option for avoiding an increase in fare prices would be reduced. The econometric analysis done by (InterVistas Consulting Inc., 2007), found that without the route substitution term, elasticities in the region of -0.8 were produced. The elasticity -0.8 will serve as the base elasticity and the suited multipliers for the situation will be implemented. The next variable taken into consideration for the first multiplier is that of the geographic area. Below is a table with the different geographic markets and their derived elasticities. According to the comments the most applicable market to the region is intra South America thus, utilising the elasticity multiplier of 1.25. The elasticity calculated is then used to quantify the effect of a reduction of 5% and 10% on intra-regional travel demand into Saint Lucia. Redjet served as an excellent opportunity to experience a new product based on low airline fees which increased both demand and airlift capacity. STYLIZED FACTS Tourism has been expanding in numbers internationally, with tourism receipts representing about 6 percent of international trade of goods and services, and nearly 2 percent of the world’s GDP in the years 2006-2010 (Culiuc, 2014). Similar trends have been occurring in the Caribbean region with (Caribbean Development Bank, 2017) stating that between the years 1989 and 2014 the number of stay-over tourist arrivals grew at an average rate of 2.5%. In order to grasp the initial idea and nature of the data the following statistics is provided to conceptualise the existing framework being used in the study. The concerned data and variables are exhibited in several graphs correspondingly. In recent years’ Caribbean arrivals into Saint Lucia have been on a steady increase between the years 2014 and 2018. This shows the potential growth in the Caribbean market.

FIGURE 1: CARIBBEAN ARRIVALS

Source: Author Compiled using data from CSO

Caribbean arrivals expenditure is only exceeded by the US, UK, and Canada markets over the years, despite a recent decrease in the more recent years.

A Compilation of Working Papers by OECS Scholars

161

FIGURE 2: EXPENDIURE BY MARKETS

Source: Author Compiled using data from CSO

FIGURE 3: PERCENTAGE OF TOTAL EXPENDITURE BY VISITORS

Source: Author Compiled using data from CSO

With the trend of increasing tourism arrivals over recent years’ government related revenue has been on an increase as seen in figure 4 below. This continues to show the importance of not only the tourism industry to the countries of the Caribbean but also the importance of the air transport market.

162 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 4: SAINT LUCIA TOURISM RELATED REVENUE

Source: Author Compiled using data from CSO

DATA ANALYSIS AND RESULTS Air transport services provide a vital social and economic link between people, countries and culture, to disparate island communities (Warnock-Smith, 2008). Over recent years’ Caribbean arrivals to Saint Lucia has been on a steady increase, suggesting that there is potential in the market. In order to gage the impact of price on demand for intra-regional travel the demand elasticity multipliers developed by Intervistas Consulting Inc were used to derive the elasticity for air travel in the region. The base elasticity of -0.8 was used and the geographic market multiplier of 1.25 was then applied resulting in an elasticity for air travel in the region of -1. From the year 2005 to 2018 the average annual stayover arrivals from the region were 65,064. With demand for regional travel being unitary elastic, if TFC’s were to be reduced by 5% regional stayover arrivals should increase by 5% to 68,317 according to the economic theory of unitary elasticity. In the event that TFC’s are reduced by 10% regional stayover arrivals should also increase by 10% to 71,570. This proves that there is an inverse relationship between the price of air travel and the demand for intra-regional travel. As a result, government revenue from TFC’s decline by 0.0025% and 1% if TFC’s are reduced by 5% and 10% respectively. In order to further assess the impact on government revenue the average accommodation cost ($338.29) was multiplied by the average length of stay of Caribbean visitors (7), and also the average number of visitors as a result of the reduction in the TFC’s. For a 5% decrease in TFC’s government revenue from V.A.T (10%) on accommodation charges would increase by $770,367.52 (5%). In the event of a 10% decrease in TFC’s government revenue from V.A.T (10%) on accommodation charges would increase by $1,540,735.04 (10%). Another stream of government revenue to be affected would be the 12.5% V.A.T on the total average spend of Caribbean arrivals. Average daily spend of Caribbean visitors is $308.53 and that is multiplied by the average length of stay which is 7 days. As a result of a 5% and 10% reduction in TFC’s government revenue on total average spend revenue would increase by $878,246.07 (5%) and $1,756,492.14 (10%) respectively. Overall government revenue is expected to grow by 3.4% and 7%, if TFC’s are reduced by 5% and 10% respectively.

A Compilation of Working Papers by OECS Scholars

163

164 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 5: GOVERNMENT REVENUE ANALYSIS AS PER REDUCTION IN TFC’S

Source: Author Compiled using data from CSO

With not only an increase in government revenue but also accommodation revenue by 5% and 10% as a result of a 5% and 10% reduction in TFC’s respectively, shows the direct impact on the tourism industry. With an increase in regional tourist arrivals who are the third biggest spending market the impact of the reduction could trickle down to the rest of the economy. Employment in the tourism sector can be expected to be increased due to the increase in arrivals. According to (El Perial Management, 2015) the true value of a visitor is not only limited to their expenditure at the destination but their willingness to; visit again and, to act as ambassadors for the destination hence driving growth within the segment and even others. They also went on to state that because Caribbean people tend to drill down into the culture when visiting a fellow Caribbean destination, they become brilliant ambassadors once they have had a good experience. This continues to show the great reach of the intra-regional tourism segment in terms of government revenue and its overall economic impact. LIMITATIONS OF STUDY A formal analysis of the relationship between taxes and inter-regional travel would constitute economic analysis through the use of a well-established model methodology. In the air transport sector due to high competition and the volatility airfares getting the relevant data sets of historical data proved difficult to attain posing as the only major constraint. In the attempt to derive an elasticity for intra-regional travel demand from the Internists multipliers was a constraint in itself as the Caribbean region was not included in the study. The region with the closest characteristics had to be chosen as a representative for the Caribbean region. CONCLUSION This research aimed to assess the impact that a reduction in TFC’s would have on intra-regional travel, government revenue and in general the economy. In order to do this assessment, the elasticities developed by Intervistas Consulting Inc. was utilised in closest reference to the Caribbean region and a price elasticity of -1 was produced. This price elasticity meant that intra-regional travel is unitary elastic and a change in price would result in an identical proportionate change in demand for intra-regional travel. A reduction of 5% and 10% in TFC’s were assessed and the following results were produced. Intra-regional A Compilation of Working Papers by OECS Scholars

165

travellers would grow by 5% and 10% as a result of a 5% and 10% reduction in TFC’s respectively. An immediate 0.25% and 1% decrease in government TFC revenue is experienced as a result of a 5% and 10% reduction in TFC’s respectively. This decrease in government revenue is recovered down the line in the other tourism related revenue avenues as a result of the increase in intra-regional travellers. Total tourism related revenue increases by 3% and 7% when there is a 5% and 10% reduction in TFC’s respectively. The increase in intra-regional travel and government revenue will in turn see an improvement in the country’s economy as there will be more revenue to be allocated by the government and there will also be an increase in regional stayover arrivals allowing hotels and air bnb’s the opportunity to increase profits. It was also concluded that with Caribbean visitors being more likely to drill down into the culture of the country expands the economic impact they would have on the economy. With a good experience they would become ambassadors hence driving growth in arrivals even further. Overall, a reduction in TFC’s will cause intra-regional travel to increase, allow governments to gain higher tourism related revenue, and as a result contribute even more to the economic growth of the islands’ economies. POLICY RECOMMENDATIONS Since intra-regional travel is unitary elastic, in order for Caribbean countries to take advantage of such a three-way action plan is deemed to be necessary. The first step of that plan is the reduction in TFC’s associated with airfares. Such a reduction as shown in the analysis would stimulate a proportionate increase in the demand for intra-regional travel. Step two would be to ensure that the region has sufficient aircraft resources to fulfil the anticipated increase in the demand for intra-regional travel. That would be achieved by governments cultivate the right regulatory framework and render government assistance where needed. The third and final step in this action plan brings the various tourism authorities around the region to the forefront. A revised strategic approach to appease regional travellers would have to be developed along with giving the markets higher priority.

REFERENCES Burghouwt, G., Boonekamp, T., Zuidberg, J., & Spijker, V. v. (2015). Economic benefits of reducing aviation taxes in Latin America and The Caribbean. Amsterdam: Seo Amsterdam Economics. Caribbean Development Bank. (2017). Tourism Industry Reform Strategies for Enhanced Economic Impact. Bridgetown: Caribbean Development Bank. Culiuc, A. (2014). Determinants of International Tourism. Washington: International Monetary Fund. El Perial Management, S. (2015). A Study on the Factors Inhibiting Intra-Regional Travel in the OECS. The OECS Commission. InterVistas Consulting Inc. (2007). Estimating Air Travel Demand Elasticities. IATA. Soverall, W. (2012). REDjet Airborne: Policy Implications for Intra-Regional Travel, Air Transport and Caribbean Tourism Development. Journal of Eastern Caribbean Studies, 37, 6-39. Warnock-Smith, D. (2008). The Socio-Economic Impact of Air Transport in Small Island States. Cranfield: Cranfield University.

166 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Javan Lewis

Economist, Research and Policy Unit Department of Finance, Government of Saint Lucia

Javan Lewis operates in the capacity of Economist in the Research and Policy Unit of the Department of Finance in Saint Lucia. His responsibilities include the monitoring of developments and changes within the economic sectors of Tourism, Employment and Labor, along with compiling, analyzing and interpreting sector specific data. He is also responsible for dialoguing and forming relationships with key sector stakeholders to enable the deduction of appropriate policy advice for conveyance. Mr. Lewis was a 2018 graduate of the University of the West Indies Cave Hill Campus where he completed his BSc. Studies for a double major in Management and Economics. Prior to his assumption of the position of Economist, Mr. Lewis taught Principles of Business and Economics at the Castries Comprehensive Secondary School in Saint Lucia.

A Compilation of Working Papers by OECS Scholars

167

10 Male Academic Underachievement in Tertiary Education A Saint Lucia Case Study 2020 Maria Mbinda-Lashley 168 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT The purpose of this study was to explore underachievement of male students in tertiary/higher education, and to understand the contributory factors giving rise to underachievement from the perspectives of the male students themselves. Male educational underachievement manifests itself in various forms, from educational disengagement, exclusion, truancy, and even resistance to education. Moreover, educationally underachieving students are more likely to be faced with limited opportunities to participate in both the job market and in academic advancements. While academic underachievement is a perennial problem that is experienced at the primary, and secondary levels of education, male academic underachievement is gaining momentum at the tertiary level of education. At this level, male students enter the institutions with above average grades, yet record high underachievement levels and higher dropout rates compared to female students. This research is informed by Bourdieu’s social and cultural capital concept (Bourdieu, 1986) which considers academic underachievement to happen as a result of class inequalities within the education system. A case study method of inquiry, using individual and focus group interviews allowed for multiple facets of male academic underachievement to be revealed and understood. Through the data analysis, the themes that emerged illustrated how underachieving male students conceptualize themselves as learners, and also revealed factors that contributed to the poor academic outcomes of these students. The research drew on the experiences of 30 male students and 8 teachers and through their voices, this research was able to highlight some factors, other than cognitive factors, which caused male students to academically underachieve at the tertiary level of learning. The principal findings are that young males in tertiary education are predestined to underachieve as a result of their limited composition of social, cultural and economic capital. Underachieving male students’ lack of investment in social capital facilitates their academic failure and limits their upward mobility. Moreover, the lack of financial information, access to financial support, institutional support, and teacher practices, lower student engagement and participation in tertiary education. Underachieving male students are more likely to be extrinsically motivated but are, in the absence of social networks, positively supported by their peers.

INTRODUCTION The academic performance of underachieving students, in particular, male students, has been an area of major concern within the Caribbean, raising significant media and academic debate as well as producing a multitude of complex and divergent hypotheses. Gender and its role in education has dominated a significant part of the literature on student underachievement within the English-speaking Caribbean (Miller, 1986, 2003; Figureoa, 1996; Chavannes, 1999; Parry, 2000; Bailey et al., 2002), with an emphasis on differential performance based on gender in Caribbean secondary schools. Divergent views on male academic underperformance, for instance, male marginalization within the school system (Miller, 1986, 1991,1994), the historical privileging of males (Figueroa, 1996), school and socialization factors (Evans, 1999) and financial constraints (Bailey and Brown, 1999) have all underscored the complexities of male academic underperformance, as the prevailing gender achievement gap in Caribbean academic institutions confirms the underperformance of male students at all levels of education. Within the Caribbean region, statistics for tertiary education reveal that female enrollment rates at tertiary institutions has significantly increased over male enrollment rates 3:1 (University of the West Indies Statistics, 2009). This indicates that, as a group, young men are lagging far behind in most school academic endeavours (Kafer, 2004). In particular, the gender disparity has become most apparent in A Compilation of Working Papers by OECS Scholars

169

the choice of subjects that students pursue in both post-secondary and tertiary education (Ministry of Education, 2011). Nonetheless, academic underachievement is a complex and challenging issue with far-reaching implications for individual well-being, as the vast majority who experience unsuccessful educational outcomes are faced with high skill deficiencies that do not match the needs of the changing work environment. The problem of underachievement is of particular concern in Saint Lucia because of the observable disengagement and non-participation of male students in schools and other social activities. This disengagement has been shown to have a direct correlation with the quality of life that these young males encounter after an unsuccessful educational experience which, in turn, results in lifelong negative experiences of unemployment, crime and delinquent behavior. In the Caribbean context, young adult male students with unsuccessful educational experiences are commonly stigmatised by society and by the media, as education misfits whose behaviour stems from the deliberate choices that they make (Gross, 2010). These poor choices perpetuate negative attitudes and behaviours toward school, class, and instruction (McCoach and Siegal, 2003) and challenge students’ interest and desire to engage in the learning environment. The consequences of male academic underachievement continue to extend in Saint Lucia far beyond the school system. These consequences reflect in the everyday social habits and practices of boys and young men and the impact is apparent in the socio-economic development of the country. As an ‘at risk’ group, young men are ill-prepared for the transitional experience from school into the labour market, as they are not adequately equipped with the necessary job market skills. Learners are not achieving at their full potential, and are not achieving adequate literacy and numeracy skills, which in turn “contributes to insufficient improvements in socio-economic development” (OECS, 2012 P. 12). Consequently, they are placed at a disadvantage within society thus increasing their chances of economic and social exclusion (Gillborn and Mirza, 2000). Recognizing the problems faced by male students in education, this research is placed in the context of education with reference to Saint Lucia, and takes into consideration the nature of male academic underachievement in tertiary education. While attempts have been made through Caribbean based research to identify factors affecting student performance in tertiary education, (Simmons et al, 2005; Bourner and Race, 1990; Jacobs, 2002) there has been limited research within the region that seeks to understand male academic underachievement from the perspectives of male students at the tertiary level of education (Joseph et al, 2012). With these challenges facing male students, the focus of this research is to understand the underlying factors responsible for male students’ academic underachievement in tertiary education. The goal is to provide an account of students’ experiences and perceptions of academic underachievement at the tertiary level of education. PROBLEM STATEMENT In Saint Lucia, boys’ underachievement is evident from as early as primary school, and advances rapidly at the secondary school level, (MOE, 2009 Report). More boys than girls underachieve at this level and fail to transition into tertiary institutions as a result of low performance at the secondary level completion examinations at the Caribbean Examination Council (CXC, 2012). A substantial number of these students remain unproductive and lack the necessary skills to help sustain their economic development, hence putting tremendous stress on an already struggling economy. Of those male students who successfully gain entry into the tertiary level, a significant number of them fail to complete the two-year programme at the institution for a number of reasons, most notably their inability to complete one or more of the prescribed courses required for successful completion. The Organisation of Eastern Caribbean States (OECS, 2010) Reform Strategy Report suggests that on completion of secondary education in Saint Lucia, approximately 42% of students transition into postsecondary education (tertiary education). However, inequality of access and participation exists at the level of post-secondary and tertiary education where enrolment is higher for female than male students. The low enrolment rates for male students could be attributed to a combination of factors including personal, financial and/or behavioral problems (OECS 2012). National school aggregate data in Saint 170 | Research: The Platform for Innovation, Competitiveness and Growth

Lucia also suggest that underachieving males have negative school experiences as a result of the lack of attention from teachers, consistently record low grades, suffer from teacher insensitivity to the economic and social problems faced by male students, and engage in disruptive male classroom behaviour all resulting in low academic outcomes and continuous school suspensions (2010). The concerns over the failure of male students to complete their education successfully has increased the importance of examining and understanding the underlying causes of male academic underachievement, so that appropriate learning strategies, other than cognitive tests, can be identified and implemented to change the male student educational experience and outcomes. This research therefore seeks to explore male academic underachievement at the tertiary level of education, and to understand the experiences of male students in relation to their learning environment, their perceptions of underachievement, and their perceptions of themselves as learners within their learning and social environment. PURPOSE The purpose of this research is to explore underachievement of male students at the tertiary level of education, and to investigate the contributory factors giving rise to underachievement from the perspectives of the male students themselves. A focused analysis of the findings will be used to assist educational policy-makers, educational institutions, and teachers, to develop appropriate programs to motivate, support, and engage learners in an effort to change learner behaviour and improve learner achievement. A significant amount of the research on male academic achievement in the English-speaking Caribbean region, has revolved around the concept of male marginalisation (Miller, 1986) and gender socialisation (Figueroa, 2004; Chevannes, 1999) as contexts for the poor performance of male students. However, not much consideration has been given to individual internal factors within the learning and social environment that may indicate possible reasons for underachievement. In most instances, academic achievement has been linked to test scores and examination results which generalise the performance of students, without taking into consideration how students conceptualise themselves and their learning environment. RESEARCH QUESTIONS In order to guide the research and explore the experiences of underachieving male students, the following questions were used: •

How do male students perceive their educational environment?

•

What are the factors responsible for male academic underachievement at the tertiary education level?

•

What factors do male students themselves attribute their academic underachievement?

METHODOLOGY This research utilizes a case study approach to address the research questions presented above. A case study provides contextual detail about a phenomenon and captures unique features about a particular situation, which may be lost when accessing large scale data, but can however be used to enhance the understanding of the particular situation (Cohen, Manion and Morrison, 2003). The case study allows A Compilation of Working Papers by OECS Scholars

171

for engagement within a specific context and makes for a better understanding of the realities and truths experienced by those within the context. The academic achievements of students vary in complexity. Understanding these experiences from the participants’ perspectives puts meaning to their behaviour and explains how that behaviour affects their overall achievement. This is congruent with the views presented under the constructivist epistemology that posits reality as socially constructed, and meanings emerge from engagement with the realities of the world emphasizing a diversity of interpretations that can be applied to explain a phenomenon (Guba and Lincoln, 1989). As realities become the construct of the mind, the different experiences of male students within the learning environment explain how and why they behave in certain ways; how male students experience underachievement and how they conceptualize themselves as learners. The primary methodology used was the in-depth individual and focus group interviews with students, and individual interviews with teachers. These were used to generate rich contextual data (Lincoln and Guba, 1985) on students’ experiences in tertiary education, their perceptions concerning the learning environment, their perceptions of teachers’ attitudes toward male students, and students’ views of education and career opportunities. As such, over a three-month period, 30 male students and 8 teachers were involved in the research study and Data was gathered using individual student and teacher interviews and student focus groups interviews. SIGNIFICANCE This study is unique, in that it examines an overlooked population of students who are generally considered academically successful by virtue of being able to attend tertiary/higher education. The study examines the nature of academic underachievement from the viewpoint and perspectives of the male students themselves, giving them a voice to express their own understanding of underachievement and the factors that have contributed to their underachievement. A fresh perspective on the underachievement discussion may provide insights into understanding underachievement that could change the way male students are perceived in the learning environment. This could also determine some measures of intervention that could be effectively applied to achieve educational success. This research is significant for educational policy makers as well, as it offers the opportunity to engage in dialogue with relevant stakeholders about male academic performance, and develop appropriate, practical educational policies that focus on empowering students’ success. The findings of this research will also help tertiary/higher education administrators and faculty to review teaching strategies and identify workable ways of encouraging the male student population to improve their performance. MOTIVATION AND INTEREST IN THE RESEARCH From observations and experiences within the teaching profession, too many young boys and men seem to have given up on their educational pursuits and have resigned themselves to performing and participating less in secondary and tertiary levels of education. From their behaviour and attitude, they seem to be disengaged from the reality that education is linked to their life choices, well-being, and economic sustainability, without which they will be unable to participate in productive social and economic development. Their poor educational attainment perpetuates their poor social and economic outcomes that continue through adulthood, and in many instances, has resulted in deviant behaviour, and the inability to invest successfully in productive development. As an educator within the school system, I work closely with students who have been categorized as underachievers because they have not been able to meet stipulated institutional requirements attained through standardised achievement tests that students complete at the end of primary, secondary and post-secondary education. These students, who are usually male, show signs of demotivation, 172 | Research: The Platform for Innovation, Competitiveness and Growth

disengage from anything school-related and generally do not perform at their optimal levels. Because of this disengagement, male students display attitudes and behaviours that conflict with their learning environment: they are usually unable to concentrate on academic work for long periods of time, display high levels of irritability in and outside of the classroom, and engage in behaviour that leads to violence, crime and increased instances of truancy. Such behaviours have led to national media reports suggesting that too many young men do not possess the skills required to assist them in making rational decisions, as such adversely affecting their ability to perform academically within the school environment. Experiences with these students show that most of them do not consider the long-term value and benefits of acquiring a good education, but are more focused on the financial gain derived from other forms of employment. The desire for positive academic achievement is not considered a priority for some of these students, especially if the outcomes are not seen as enhancing their physical and social status. Unlike female students, most of these young men display poor time management skills, poor study and communication skills that significantly lower their level of engagement and achievement within tertiary/ higher education. In the classroom, these young men display negative behaviours toward teachers whose teaching strategies and classroom management styles may not be reaching them effectively. This cultivates the belief that male students are being discriminated against within the classroom. Consequently, male students use that argument as a reason for their poor performance and perceived underachievement. In giving male students a voice, understanding their experiences within the learning environment; understanding how they conceptualise themselves as learners and how they respond to the learning environment, allows for a better appreciation of what affects and impacts on their learning. Through this understanding, appropriate policies and strategies can be developed to help educators and other stakeholders such as ministry officials, education policy makers, and parents, address male academic underachievement at the tertiary level. The lack of early intervention practices within the system has apparently failed this ‘at risk’ population which has been allowed to navigate the system with minimal professional help. Understanding students’ perceptions of self and their perceptions of the learning environment will provide new insights into the existing literature on why male students underachieve. Key term: Underachievement: - for the purposes of this research, underachievement is defined as the difference between individual potential and individual outcome, where students with different abilities (high or low) may at some stage of their career development not perform as expected (Smith 2005). In the context of this research, male students enter tertiary education with varying levels of achievement that are considered acceptable for their program of study, however at some stage during their career development at the tertiary level they experience challenges that impede their performance. As such they do not perform as expected, and this leads to low levels of performance and underachievement. LITERATURE REVIEW This research applies Bourdieu’s theory of social and cultural capital to understand and examine the reasons why male students underachieve academically in tertiary education. Bourdieu’s theoretical framework (1986) concerns itself with how social and educational inequalities are contextually produced within and across social fields. This is achieved through the interaction between habitus (set of dispositions which incline individuals to act and react in certain ways) and the different types of capital; social, economic and cultural capital (Bourdieu, 1986). Habitus is generated by one’s position within the social class, and it is through knowing one’s place within the social structure and internalizing it that one becomes aware of what is possible or not possible to achieve (Dumais, 2002). Operating at the level of the subconscious, Bourdieu’s concept of habitus is instrumental in helping individuals understand how to act, behave and think (Connoly, 1997). Bourdieu’s approach to social capital also explains how access to various forms of capital are accrued by membership within a network, and how selected groups within society use these networks to achieve A Compilation of Working Papers by OECS Scholars

173

capital accrues more to the higher socioeconomic class who have access to both economic and social capital. (Bourdieu, 1986). This argument, suggests that individuals from low socioeconomic backgrounds possess less social capital as a result of their social class and class habitus and emphasizes that within education, privilege and inequalities prevail as a consequence of the failure of institutions to recognize the varying cultural and linguistic competencies held by agents and their levels of familiarity with the dominant culture, referenced as cultural capital (Bourdieu, 1986). This cultural capital on the other hand can precipitate an unfair education system which is biased towards those students possessing inherited cultural capital (Leese, 2010). The impact of cultural and social capital can therefore be seen even before students enroll in the formal education system, as students are less knowledgeable, and less confident to make informed educational choice (Vryonides, 2007). That being the case, if male students lack appropriate social or cultural capital, their experience of entering new learning environments can be severely affected (Reay et al., 2002). Further, If a male student grows up in a culture or setting where educational expectations are low, this can greatly influence the student’s self-concept. Possession of the right cultural capital creates ease of transition and function in the learning environment which, in turn, leads to higher levels of academic success (Dillon, 2010). In the absence of appropriate cultural capital, students from low socioeconomic backgrounds feel alienated from the established norms of the academic institution (Aries and Seuder, 2005), which accounts for their unsuccessful academic outcomes. This highlights the presence of educational inequalities, where educational achievements are reinforced by the education system that has preference for students with the right cultural capital (Dumais, 2002). SOCIALISATION AND EDUCATIONAL UNDERACHIEVEMENT Referencing gender and education in the Caribbean, Bailey (2002) identifies three aspects to the gender socialisation paradigm, namely: gender stereotyping, inequalities in access, and patterns of curriculum participation, that form and influence students’ habitus. The nature of gender socialisation influences the education system, producing inequalities in access to education and plays a role in determining student academic success or failure. These assertions link habitus and socialisation in that socialization helps shape individual habitus over time and aligns with Bourdieu’s concept of habitus. Drawing on Bourdieu’s (1986) concept of ‘habitus’, roles within a given social group form systemic associations of a negative or positive attribution that is purely based on gender. As a result, over time, these gender roles form gender stereotypes that define the habitus through which roles, expectations, and gender responsibilities are interpreted (Eckert & Imhof, 2012). A study on gender academic differentials (Evers and Mancuso, 2006) revealed that practices within Caribbean homes shape an incongruous gender identity that influences academic underachievement. Males are not given adequate responsibilities at home that would instill discipline, hard work and focus, while girls are. Self-discipline taught to females enabled them to navigate the education system much more easily than males (Figueroa, 2004), however, the extent to which this societal silent expectation of males and manhood carried over by male students when they get into tertiary institutions of learning has not been sufficiently understood. It is, therefore, important to examine how these expectations influence male academic outcomes in tertiary education. In addition, stereotype-based expectations of males can have self-fulfilling effects on academic behaviour that create social problems by virtue of accumulated distorted beliefs over time (Buchanan and Hughes, 2009). In particular, teacher’s perceptions of students strongly influenced students’ performance in the learning environment, where the existence of a negative stereotype from teachers raised the chances of a stereotype threat, and reduced student learning, jeopardising student-teacher relationships (Woolf et al. 2008). Male students are treated differently and lower academic results are expected from them, leading to cases of self-fulfilling prophecies as envisaged in Bourdieu’s social capital theory. Though teachers may not consciously show these differences, subtle messages are conveyed in the classrooms which allow disruptive male behaviour to go unchecked (Figueroa, 2002).

174 | Research: The Platform for Innovation, Competitiveness and Growth

MASCULINITY, IDENTITY AND UNDERACHIEVEMENT The perception of masculinity in the Caribbean society has been viewed as a contributing factor to male underachievement (Jones & Myhill, 2004: Jha & Kelleher, 2006: Skelton, 2001). Within the tertiary institutions, masculinities are developed and the acceptance of those masculinities plays a very critical role in how young males access higher education and the labour market. Black male students in particular, identify the academic institution as a social institution that ‘re-inscribe[s] hegemonic power structures’ that helps construct cultural identities that go contrary to the need for academic achievement (Harper & Davis, 2012, p.106). In other words, when male students reinforce their hegemonic power structures within the academic institution through interactions with their peers, they resist participation in academic success in an attempt to align with, and protect, their cultural identities. As agents (parents, teachers, role models, and peers) constantly put masculinity under scrutiny, male students find themselves confined to masculine standards for conformity, and these standards or codes are eventually, over time, filtered into tertiary institutions of learning. These codes govern the male image in terms of authorized styles of speech, clothing, and dispositions which reinforce and sustain masculine reputations (Plummer, 2007). Similarly, as agents of socialisation, tertiary institutions serve as positions of strength for male students. Tertiary institutions are instrumental in reinforcing how males actively develop, negotiate and re-negotiate their masculinities (Swain, 2006). Tertiary institutions subtly reinforce masculinity through their structure, pedagogy and curriculum, and keep male students from achieving positive educational outcomes (Connell, 1996). PEER GROUP ASSOCIATION In tertiary educational institutions, some students form relationships with peers that promote academic engagement, and these relationships become more salient in student social settings and their performance in the learning environment, while others associate with peer groups that promote school disengagement. Based on the concept of social capital, students could either establish networks to acquire resources (Bourdieu, 1986) or establish networks to foster an academic identity among students (Coleman, 1990). Either way, both perspectives explain the impact of peers and peer groups on college students. Peer groups have the potential to mediate or moderate influence in students (Stanton-Salazar, 2004) as peer groups are in a position to either discourage or encourage academic success. Consequently, student experiences and activities within the learning environment are influenced by student memberships to certain peer groupings which have an effect on male student participation and overall school performance. The influences of peer endorsed masculinity, and the need for acceptance within the group, become a fundamental priority as young males mature and negotiate the school environment. The association with negative peer groups, in most cases, supersedes parental and school authority, and impacts negatively on male students academic and social development. It can be argued that “some of these groups become fundamental identity bearing groups who, not only impose their behaviour on the young men, but separate them competitively and conflictingly from other similar groups of young men” (Bailey, Branch, McGarrity & Stuart (1998), cited in Edmound-Woods (2007:59). Since some peer groups are marginalised in the learning environment, they essentially promote oppositional school identity which encourages members to resist schooling practices (Gandara, O’Hara, and Gutierrez, 2004). Male virtues, for example strength and independence, once considered positive values, have become masculine vices in the form of aggression, macho bad boy image, gangster speech and detachment. These are images associated with the ideology of the peer groups, transferred into the schools through the attitude and behavior of boys. What male students need most is positive role models who can help channel their energies into more positive outcomes (Coard, 2006).

A Compilation of Working Papers by OECS Scholars

175

SOCIOECONOMIC STATUS Studies on socioeconomic status show that family socioeconomic status has an important effect on students’ academic achievement; low socioeconomic status is believed to negatively affects student academic achievement as it prevents access to vital resources that students may need to engage in the learning environment and leads to high levels of stress within the home environment (Jaynes, 2002). In examining the various dimensions of socioeconomic status, parents face the greatest challenge in providing educational stability for their children. This presents the argument that parental involvement, and parental financial stability, impact student academic achievement (Desforges, 2003). The deficits in social capital prevent access to vital resources that students may need to engage in the learning environment, leading to high levels of stress within the home environment (Jaynes, 2002). The interplay between social, cultural and economic constructs and academic resources has a significant impact on male students’ engagement in tertiary levels of education. As literature has shown, a relationship exists between socioeconomic status and student academic performance (Jeynes, 2002; Considine and Zappala, 2002), where the socioeconomic status of parents affects the academic performance of students. Money and other economic forms of capital exacerbate inequalities in tertiary education, as the challenge facing most students is access to financial resources (Archer et al.2007). Students from low socioeconomic backgrounds lack the network to seek adequate compensating employment and, secondly, they do not have the time to engage in adequate employment due to schooling demands. They, therefore, do not have the economic capital to invest in their education. In other words, students lack the ability to purchase advantage as students do who are from higher socioeconomic backgrounds, as such male students coming from a background of inadequate resources are placed at a disadvantage because they are unable to acquire the tools for upward mobility (Archer et al. 2007). Students should therefore carefully consider their financial status when making a decision to enroll in tertiary education (Wray et al., 2014). Students also experienced financial problems because of the lack of information and guidance in attaining financial assistance (Hurtado, Laird and Perorazio, 2003). Due to parental socioeconomic status and background, students may not be equipped or have any knowledge of how or where to get financial assistance. While institutions have the resources necessary to assist students to obtain such knowledge, the onus remains on the students to seek and apply for financial support in order for them to persevere in their college education (Hurtado et al., 2003). Economic capital, in the form of scholarships or bursaries awarded to students, assists in student retention and participation in tertiary/higher education, however Students who received scholarships had a heightened sense of belonging and confidence in the academic processes (Reed and Hurd, 2014), however these financial support schemes need to be considered as tools for encouraging enrolment rather than tools for enhancing achievement. Financial support affects students from different economic backgrounds differently, but financial constraints affect low socioeconomic groups more than those of other groups in higher education (Callendar & Jackson, 2005). Students from low socioeconomic backgrounds were less inclined to incur long term debts and are more averse to debt, rather than appreciating the investment that scholarships would afford them in terms of academic mobility. It is evident that access to economic capital plays a significant role in students’ ability to engage and participate in tertiary education. MOTIVATION AND SELF-ESTEEM Motivation is crucial in understanding student behaviour and achievement. Students engage in different activities based on whether they are intrinsically or extrinsically motivated by that activity. The selfdetermination theory (SDT) describes behaviour as either intrinsically motivated, extrinsically motivated, or demotivated, and is used to, firstly, “Make the critical distinction between behaviours that are volitional and accompanied by the experience of freedom and autonomy…and those that are accompanied by the experience of pressure and control and are not representative of one self” (Ryan and Deci, 2000, p. 65). Secondly, it is used to understand student motivation and engagement in the learning process 176 | Research: The Platform for Innovation, Competitiveness and Growth

which affect the level of student academic achievement (Ryan and Deci, 2000). While the assumption of extrinsic motivation is that it is undesirable motivation and, hence, negative for student achievement, the SDT demonstrates that not all extrinsic motivation is undesirable. Extrinsic motivation regulates autonomous motivation (self-determined motivation) and controlled motivation (low self-determination) which, suggests that the more autonomous or self-determined the motivation, the more successful the outcome (Deci & Ryan, 2000). However, in higher educational institutions, culture (academic content, rules, and coursework) can conflict with student expectations resulting in low self-determination and low academic achievement. Contrary to popular belief, not all extrinsic motivation is undesirable; extrinsic motivation for students has its place and should not be used in isolation to advance the case for male students’ underachievement (Deci and Ryan, 2000). Students are also motivated to participate in the learning environment if they feel comfortable and are able to fit into the culture of the learning environment (O’Hara, 2007). As students access the tertiary institutions, they express optimism about their social and academic life, and enter the institution with a set of expectations as to what the learning environment can offer them (Nelson, 2002). Accessing tertiary/higher education represents a major development for students as they adjust to the academic challenges presented at the college level (Kreig, 2013). Students expect the institution to offer a variety of support services (Pike et al., 2006), and other experiences different from their prior educational settings. However, for some students, tertiary/higher education can be overwhelming and stressful as the collegiate experience fails to meet their expectations (Tinto, 1987). The inability to make a satisfactory transition into tertiary/higher education is associated with instances of early withdrawal from college and underachievement for some of these students (Lowe and Cook, 2003). Students come to college with unrealistic expectations which, when not met, could cause stress and anxiety as these expectations are incongruent with initial expectations of the learning environment (Kreig, 2013), but when students’ experiences and expectations of college are congruent, students’ academic outcomes are positive (Miller et al., 2006). Studies suggest that the difficulties of transitioning into tertiary/ higher education are compounded by the lack of cultural capital from students’ past learning experiences (Leese, 2010; Young, Glogowske & Lockyer, 2007). TEACHER-STUDENT INTERACTIONS Student achievement levels within the classroom are significantly enhanced through teacher interactions. This being the case, the teacher-student dynamics become central to understanding student academic performance (Lawrence, 2005). Teachers’ perceptions of their students affect academic achievement (Leese, 2010), as teachers have the ability to shape, not only the learning experience, but also the academic progress of the student (Figueroa, 2000). However, in cases of students’ failure, teachers generally disassociate themselves from any blame arguing that achievement, or lack thereof, should be the sole responsibility of the students (Lawrence, 2005). When teachers express enthusiasm about their work, are engaged and knowledgeable about content and pedagogy, they influence student learning (Kandiko and Mawer, 2013). Teachers use a relevant, engaging curriculum that helps improve the quality of the student learning experience. However, while it is possible to elevate students’ cultural capital to an equitable position to participate in higher education, teachers need to be aware of the subtle messages within the learning environment that make student learning difficult. Male students who feel they are supported by their teachers are more likely to perform and be successful in academics (Howard, 2002). However, male students’ success greatly reduces “when they felt that teachers are not concerned about them or their academic performance” (Noguera 2003: 449). The concern for students can be shown by simply paying attention to them and recognising their achievements, consequently, if students believe that the teachers respect their views, work with their skills level and offer continuous encouragement, they are more likely to be successful in the classroom (Kandiko and Mawer, 2013).

A Compilation of Working Papers by OECS Scholars

177

178 | Research: The Platform for Innovation, Competitiveness and Growth

CONCLUSION The lack of motivation, socioeconomic background, socialisation, teacher-student relationship, and student identity are all factors that challenge the level of male student performance. When these factors are negatively validated, they lead to student underachievement. Students who are not motivated to achieve display little engagement and little mastery of content. Such behaviours put them at risk of producing negative academic outcomes. The evidence of studies of underachievement in tertiary education suggests that students who display low motivation are usually unable to cope with the stress and academic challenges of college life. Based on the literature, this research takes the view that underachievement at the tertiary level is perpetuated by inequalities within social class, which results in different levels of educational outcome, and that material and cultural differences that become embedded dispositions incline disadvantaged learners within tertiary education to disengage from the learning environment. FINDINGS Young males in tertiary education are predestined to underachieve as a result of their limited composition of social capital. Their lack of investment in social capital facilitates their academic failure and limits their upward mobility. Additionally, the lack of financial information, lack of access to financial support, lack of institutional support, teacher practices and low expectations lower student engagement and participation in tertiary education. The underachievement of male students in tertiary education can be attributed to the fact that the majority of them do not consider the long-term value and benefits of acquiring a good education, but are more focused on the financial gain derived from other forms of employment. The desire for positive academic achievement is not considered a priority especially if the outcomes are not seen as enhancing their physical and social status. Underachieving male students perceive this to be as a result of their limited access to social capital. Underachieving male students perceive that their parent's social capital presents a level of advantage and disadvantage for them. Parental involvement or lack thereof in the young males’ education increases or decreases the potential for improved educational outcomes. If a parent does not have the social capital to assist the young adult, there is the possibility that that young adult will not perform as expected. The absence of parental support makes it difficult for these young men to develop appropriate learning skills as they are constantly exposed to negative views and opinions about college education; as a result, it is difficult for them to develop a positive attitude about their learning. These young men expressed that their parents were unsupportive of their participation in higher education, concluding that their parents have limited knowledge about that level of education and cannot really understand the benefits of such an education. This gives an indication that student social networks within the home do not work in their favor, and suggests that parental lack of involvement and not valuing education influences male students’ commitment to learning. Further, parents are frustrated with the young men for not leaving college completely to engage in full-time employment to help supplement income in the home. For most of these young men, their parents preferred to support them in trying to get gainful employment rather than supporting their academic pursuits. The negative parental perception about learning at the level of tertiary education plays a significant role in how the young men view education- they view education as not important and therefore, not much effort or interest is expended in their learning at that level. Moreover, the young men’s perceptions of underachievement are derived from the negative influences that surround them on a daily basis; these negative influences are manifested through behaviour of family and friends who believe that the young men are incapable of learning at the higher level of education and, therefore, do not foresee their academic success at that level. Unfortunately, for many of these young men, their parents are unaware of A Compilation of Working Papers by OECS Scholars

179

their potential, and have low expectations of them, and this impacts on how these young men negotiate and approach their learning. Parents who do not understand the importance of a college education, and are not familiar with tertiary level education, cannot offer encouragement or give academic support to these young men and this impacts negatively on students’ academic outcomes. They are less likely to understand the importance of higher education, and are less likely to value that level of education, which results in students either leaving college or performing poorly academically. While the blame is levelled on parents for not offering enough support to young male students, it must be understood that that parents’ inability to invest time and effort to encourage academic achievement stems from their stress as a result of economic hardships. Consequently, success is limited for students who have no emotional connection with their parents and have a negative home-based learning environment. Parents who fail to acknowledge their child’s potential, or abilities, typically reinforce the development of negative self-concepts in these students, which has an influence on their academic outcomes. Unfortunately, within society, many parents are not interested in ensuring educational advantage for these young men, which results in academic disengagement and failure. Nonetheless it must be noted that that academic underachievement is not only experienced by students from low socioeconomic backgrounds whose lack of access to social, cultural and economic capital denies them the opportunity for upward mobility, but is also experienced by students who are privileged by virtue of their social and cultural capital. In this case, parents pay less attention to these young men because their attention is focused more on the accumulation of additional economic capital so as to ensure they remain within the dominant social class. LACK OF FINANCIAL SUPPORT The lack of financial support affects how male students navigate the learning environment. Their academic performance is affected by the lack of financial support indicative of their socioeconomic status. Parental income, an indicator of resource availability is seen to determine the level of students’ participation in college. Male students are distracted from their learning because of their inability to afford important learning resources and other related educational support systems necessary for the improvement of their learning and achievement. Coming from single parent homes where resources are not readily available, makes regular attendance at college very challenging. While limited financial support causes male students to engage less with the learning process, underachieve or depart from learning at different levels within tertiary education, they are also faced with the added disadvantage of harboring negative perceptions about seeking financial assistance, as they believe that there is a stigma attached to asking for assistance. Two reasons are attributed to this: firstly, male students do not want others to know that they are not financially able and, secondly, the fear that they would be treated differently by their peers. In this instance, male students are acutely aware of their social class, and demonstrate frustration at the notion of being judged by their peers for belonging to a different socioeconomic class. There is a certain degree of tension and internal conflict experienced by male students once they have to disclose their economic status. Consequently, for underachieving male students, the refusal of financial assistance invariably affects their academic outcomes and engagement within the learning environment. While access to economic capital has been viewed as influencing student retention and academic achievement, underachieving male students can be seen as not appreciative of the economic capital at their disposal. This disparity highlights that financial assistance reinforces the idea of social capital inequality to which most low-income students do not want to be ascribed to, or associated with. Underachieving male students who are usually from lower socioeconomic backgrounds are debt averse and are therefore not willing to take up loans as they are not convinced that these loans would actually accrue to investment in human capital. The lack of access to financial resources hinders male student’s 180 | Research: The Platform for Innovation, Competitiveness and Growth

investment in their education, and the inability to purchase advantage and academic mobility exacerbates inequalities in tertiary/higher education which in turn leads to academic underachievement. MOTIVATION A contributing factor to male student underachievement is the lack of motivation. Within the learning environment, students are neither motivated by the college curriculum, their teachers, nor by the success of their peers. While male students are aware that the female students are performing much better than they are, their desire to improve their position power is nonexistent because according to them they have nothing to prove to anyone, unlike the female students. They are more concerned that what they are being taught in tertiary education is not challenging and that teacher strategies used to deliver content are not engaging enough. In this case the male students are not self-determined nor intrinsically motivated to want to achieve. While extensive research has shown that intrinsic motivation is more conducive to learning (Ryan and Deci, 2002), extrinsic, and not intrinsic motivation, is more conducive for these underachieving male students as those external motivators enhance their sense of self. External motivators are perceived to be more beneficial to these male students as they can stay in school and maintain their social relationships without having to actively participate in the learning environment. These relationships demand a significant amount of students’ time; as the more female relationships male students have, the greater their level of acceptance amongst their peers. Extrinsic motivation experienced by students is therefore viewed as counterproductive to their academic success as it encourages the development of social behaviour that distracts them from the achievement of their goals. Posing one of the greatest motivational challenges for underachieving male students is the curriculum which does not challenge male students’ creativity, it is highly theoretical in nature and provides limited opportunity for practical experience. The curriculum is rigid, inadequate, not flexible nor malleable and fails to meet student’s personal, social or academic needs. The rigid curriculum encourages inattentiveness, disorganization, and lack of interest. When students do not have clear goals, illustrated by the value they place on the curriculum, they demonstrate the inability to exert effort toward academic achievement resulting in student poor performance and underachievement. There is therefore, a curriculum mismatch within tertiary education that does not meet the needs of male students. TEACHER STUDENT RELATIONSHIPS Teachers’ attitude and stereotyping are viewed as factors that negatively influence male student academic achievement. Teachers unconsciously humiliate and embarrass male students through their actions toward them, and also in the way that they communicated to them, creating an uncomfortable learning atmosphere, which male students internalize to mean that the teachers are stereotyping them as learners. Teachers have no empathy or patience, and treat them as individuals who they cannot learn. Teachers do not try to understand male students, have low expectations of them and their ability to achieve causing tension and internal conflict. The limited teacher-student interactions create communication barriers caused by the social distance between teachers and students where students feel neglected and abandoned. Male students feel uncomfortable approaching teachers for academic assistance, making it easy for students to disengage within the learning environment. When students feel that their teachers undervalue their knowledge, they disengage from the learning process and ultimately underachieve. PRIOR LEARNING Students’ prior educational experiences are found to negatively influence students’ ability to participate in the college. The majority of the underachieving male students struggle to fit into the new learning environment as a consequence of having insufficient or no information to make informed decisions. A Compilation of Working Papers by OECS Scholars

181

The lack of adequate institutional guidance encompassing orientation programs, financial guidance and student support services, serves as a barrier to students’ upward mobility and affects their academic performance. Underachieving male students establish that orientation programs offering career guidance fail to give adequate information to allow them the opportunity to make informed decisions about choice of programmes and career development. Academic support and advice from teachers is inconsistent, leaving students poorly informed about self-management and career options, resulting in students’ feelings of insecurity about their learning potential. RECOMMENDATIONS From a teaching-learning perspective, underachieving male students have different social, economic and educational needs that challenge their ability to effectively engage in the learning environment. ● Teachers need to be more aware, attentive and responsive to the needs of this particular group of students. A greater element of trust would be needed between teachers and students in order to enhance the student-teacher relationship. ● Teachers need to establish positive teacher attitudes towards the limitations of underachieving students, value and acknowledge student contributions, show interest and be mindful of students’ wellbeing. ● Teachers need to revise the teaching strategies used in the learning environment regularly, utilizing a more student-centered approach that creates avenues for greater teacher-student trust. ● Teachers need to improve their communication and interpersonal skills through periodic training and staff professional development programmes in order to understand student behaviour and background, making it harder for teachers to misunderstand the actions of students; but recognize and take into account the common characteristics that male students have within the learning environment ● The learning institution needs to develop programmes that would encourage collaborative participation between parents, teachers and students in an effort to institute awareness in parents about the benefits of tertiary education and the importance of their support to the positive growth and development of male students. ● Teachers engage in periodic programme review of courses within the existing curriculum to ensure that practical components of the courses are being implemented and meet the standards of workbased programs. This would increase student motivation and interest in what they are learning and help them develop skills necessary for upward mobility. In order to support ease of transition into tertiary education and to assist student's decision-making processes, academic support services such as academic advisement be ongoing throughout students’ college lives, as this would give students the opportunity to utilize such services regularly, and in the process develop the skills and knowledge necessary to improve their academic outcomes. Institutional support services such as academic orientations and advisory programmes provide adequate information and guidance in relation to financial access and alternative avenues to access such information in order to assist student decision making processes. Financial awareness programmes need to be made available to students so that students understand the benefits of accessing finances to aid their educational investment. The availability of information regarding other financial assistance programmes would give students greater option choices, develop a level of self-worth and minimize the dropout or non-attendance of male students in tertiary education. 182 | Research: The Platform for Innovation, Competitiveness and Growth

Educational officials in an effort to increase student completion rates in programme courses should consider formulating plans and strategies to include remedial education programmes in the curriculum to positively motivate and influence male students’ participation in tertiary education. There is a need for greater collaboration between secondary and tertiary education in order that there is a transfer of knowledge and information about tertiary education expectations. This would create a smoother transition for students as they become better prepared and more knowledgeable about what to expect of tertiary education. CONCLUSION This research sought to determine factors that contribute to students’ academic failure in tertiary education. This was achieved using Bourdieu’s Social and Cultural capital concept (1989) as the theoretical framework to examine and conceptualise the experiences of underachieving male students. The findings of this research demonstrate that, within the context of tertiary education, the education system encourages the reproduction of social class and alienates those who possess limited social, cultural and economic capital. For underachieving male students, socioeconomic status, financial limitations, lack of motivation, student-teacher relationships, lack of institutional guidance and support have created obstacles that limit students’ ability for upward mobility. While it may seem that, at the level of tertiary education, students have developed the maturity to cope with academic pressures and take responsibility for their learning, this research has shown that male students need intervention strategies that cater to their individual needs, in particular, strategies directed at developing a student-centered curriculum catering to the diverse and creative interests of these young men. Young men are interested in learning and have the aspirations of achieving upward mobility but are challenged by the absence of adequate social, cultural and economic capital. Intervention methods, therefore, need to focus on how these forms of capital can be manipulated to function in favour of underachieving students.

REFERENCES Archer, L. (2007) Diversity, equality and higher education: a critical reflection on the ab/uses of equity discourse within widening participation. Teaching in Higher Education, 12(5-6), 635-653 Aries, E. and M. Seider (2005) ‘The Interactive Relationship between Class Identity and the College Experience: The Case of Lower Income Students.’ Qualitative Sociology 28: 419–43. Babour, R. S. (1998) Mixing qualitative methods: Quality assurance or qualitative quagmire? Qualitative Health Research, 8(3), 352-361. Bailey, B. and Brown, M. (1999) Schooling and Masculinity: Boys’ Perceptions of the School Experience,” Caribbean Journal of Education, Vol. 21, No.1 Bailey. B. (2000) Issues of Gender and Education in Jamaica: What about the Boys? UNESCO, Kingston, Jamaica. Bailey, W. Clement, B. and Aldrie Henry-Lee (2002) Gender, Contest and Conflict in the Caribbean: Lessons from Community-based Research, Kingston, Institute of Social and Economic Studies. Bailey, B. (2009) “Needed: A Paradigm Shift in Addressing Boys’ Underachievement in Education”, Paper presented at 18th Meeting of Council of Social and Human Development (COHSOD) Montego Bay, Jamaica June 3. A Compilation of Working Papers by OECS Scholars

183

Bailey, B. (2009) Gender and Political Economy in Caribbean Education Systems: An agenda of inclusion. Cambridge: Commonwealth education partnership Bailey. W. Branche, C. and Stuart S. (1998) Family and the quality of gender relations in the Caribbean. Mona, Jamaica. Institute of Social and Economic Research. Bourdieu, P. (1984). Distinction: A Social Critique of the Judgement of Taste. London, Routledge Bourdieu, P. (1986) The forms of capital. In Handbook of Theory for Research in the Sociology of Education, Richardson, J.G. (ed.) 241-58. Westport, CT: Greenwood Press. Bourdieu, P. (1989) “How Schools Help Reproduce Social Order.” Current Contexts 8: 16. Bourdieu, P. (1993) Language and Symbolic Power. Cambridge, MA: Harvard University Press. Bourdieu, P (1998) ‘Acts of Resistance: Against the New Myths of Our Time’ [translated by Richard Nice], Cambridge, Polity Press Bourner, T. and Race, P (1990) How to Win as a Part-time Student: A Study Skills Guide. (London: Inform Publishing Services Ltd) Cadinu, M. Maass, A. Rosabianca, A., and Kiesner, J. (2005) Why do women underperform under stereotype threat? Psychological Science, 16, 572-578. Carbonaro, W. (2005). ‘Tracking, students’ effort, and academic achievement, Sociology of Education, 78(1), 27-49. Chevannes, B. (1999) What we sow and what we reap – problems in the cultivation of male identity in Jamaica. Kingston, Jamaica: Grace, Kennedy Foundation. Coard, B. (2006) An Exploration of African Caribbean Boys Underachievement British School System. London: New Beacon Books. Coleman, J. (1990) Foundations of social theory. Cambridge, MA: Harvard University Press. Connell, R. W. 1996. “Teaching the Boys: New Research on Masculinity, and Gender Strategies for Schools.” Teachers College Record, vol. 98 no. 2, 206- 235 Davis, J. (2002) “Boys to Men: Masculine Diversity and Schooling”. Paper presented at the School Leadership Centre of Trinidad and Tobago. Deci, E. L. and Ryan, R. M. (2000) The “what” and “why” of goal pursuits: Human needs and the selfdetermination of behavior. Psychological Inquiry, 11, 227-268. Desforges, C. (2003) The impact of parental involvement, parental support and family education on pupil achievements and adjustment: A literature review. London: Department for Education and Skills. Dillon, M. (2010) Introduction to Sociological Theory: Theorists, Concepts, and their Applicability to the Twenty-First Century. Chichester: Wiley-Blackwell. Dumais, Susan A. (2002) “Cultural Capital, Gender, and School Success: The Role of Habitus.” Sociology of Education 75: 44-68. Eckert, C. and Imhof, M. (2012) Leseleistung = Schulform + Domain Identification + Situation Geschlecht? Vortrag auf dem 48. Kongress der DGfP in Bielefeld, 23–27 September 2012. Edmund-woods, M. (2011) A literature review exploring academic underachievement among males in 184 | Research: The Platform for Innovation, Competitiveness and Growth

Trinidad and Tobago. Evers, F. and Mancuso, M. (2006) Where are the boys? Gender imbalance in higher education. Higher Education Management and Policy, 18(2), 1–13. Howard, C. (2002) Telling their side of the story: African American students’ perceptions of culturally relevant teaching. The Urban Review, 33(2), 131-149. Hurtado, S. Laird, T. and Perorazio, T. (2003) The transition to college for low-income students: The impact of the Gates Millennium Scholars program. Center for the Study of Higher and Postsecondary Education, University of Michigan. Jacob, B. A. (2002) “Where the Boys Aren’t: Noncognitive Skills, Returns to School and the Gender Gap in Higher Education.” Economics of Education Review 21(6):589-98. Jha, J. and Kelleher, F. (2006) Jamaica: Alienation and high dropout rates. In J. Jha & & F. Kelleher (Eds.) Boys’ underachievement in education: An exploration in selected Commonwealth countries (pp. 82–95). London: Commonwealth Secretariat; Vancouver, Canada: Commonwealth of Learning. Jones, S. and Myhill, D. (2004) ‘Troublesome Boys’ and ‘Compliant Girls’: Gender Identity and Perceptions of Achievement and Underachievement - British Journal of Sociology of Education, Vol. 25, No. 5 (Nov., 2004), pp. 547-561. Joseph, S. Jackman, M. and Moore, Z. (2012) An Investigation into Male perspectives of their educational experiences in Trinidad and Tobago. Journal of Education and Development in the Caribbean. Vol. 14, NO. 1- 2012 Kafer, K. (2004) Girl Power: Why girls don’t need the Women’s Educational Equity Act. School ReformNews. Available at: http://www.heritage.org/Research/Education/wm563.cfm (Accessed 20 January 2012) Lawrence, J. (2005) Addressing diversity in higher education: Two models for facilitating student engagement and mastery. Paper presented at Higher Education and Research Development: Society of Australia Conference. http:// conference.herdsa.org.au/2005/pdf/refereed/paper_300.pdf Leese, M. (2010) Bridging the gap: Supporting student transitions into higher education. Journal of Further and Higher Education Lincoln, Y. S. and Guba, E. G. (1985) Naturalistic inquiry. Beverly Hills, CA: Sage. Lowe, H, and Cook, A. (2003) ‘Mind the Gap: are students prepared for higher education?’, Journal of Further and Higher Education, 27(1): 53–76. Martino, W. and Kehler, M. (2006) Male teachers and the “boy problem”: An issue of recuperative masculinity politics. McGill Journal of Education, 41(2), 1–19. McCoach, D. B. and Siegle, D. (2003) The structure and function of academic self concept in gifted and general education samples. Roeper Review, 25, 61 – 65. Miller, T.E. Kuh, G.D. Paine, D, and Associates (2006) Taking student expectations seriously: A guide for campus applications. NASPA: Student Affairs Administrators in Higher Education Miller, E. (1986) The marginalization of the black male: insights from the development of the teaching profession. Kingston, Jamaica: Institute of Social and Economic Research. Miller, E., 1991. Men at Risk – Jamaica Publishing house, Kingston Jamaica A Compilation of Working Papers by OECS Scholars

185

Miller, E. (1994) Marginalization of the Black Male: Insights from the Development of the Teaching Profession, 2nd Edition Mona, Jamaica: Canoe Press Noguera, P. A. (2003) The trouble with Black boys: The role and influence of environmental and cultural factors on the academic performance of African-American males. UrbanEducation, 38(4), 431-459. Organisation of Eastern Caribbean States (2012) Education Sector Strategy (OESS) 2012- 21 Retrieved from www. https://www.collegesinstitutes.ca/wp-content/uploads/2014/05/OECS-Educ-Sector-Strategy2012-2021-OESS-final-2012-05-18.pdf (Accessed December 14 2017) Parry, O. (2000) Male Underachievement in High School Education in Jamaica, Barbados and St. Vincent and the Grenadines. Canoe Press, UWI, Mona. Plummer, D. (2007) Is learning becoming taboo for Caribbean boys? Paper based on the research on Masculinities commissioned by Commonwealth Secretariat. Trinidad & Tobago: University of the West Indies. Reay, D. J. Davies, J. M. and Ball, S. J. (2001) Choice of degree or degrees of choice? Class, race and the higher education choice process, Sociology, 35(4), 855874. Simmons, A. B., Glenda D. M., Choong Geun Chung (2005), “Persistence Among First -Generation College Students in Indiana: The Impact of Precollege Preparation, College Experiences, and Financial Aid,” IPAS Research Report #05-01, April. Smith, E. (2005) Analysing underachievement in schools. London: Continuum International Publication Group. Swain, C. M. (2006) An Inside Look at Education and Poverty. Academic Questions, 19(2), 47-53. Saint Lucia Ministry of Education and Culture. (2010). Educational statistical digest. Retrieved from http:// www.stats.gov.lc (Accessed June 10 2015,) Swain, C. M. (2006) An Inside Look at Education and Poverty. Academic Questions, 19(2), 47-53. Tinto, V. (1987) Leaving College: Rethinking the Causes and Cures of Student Attrition. Chicago: University of Chicago Press. Vryonides, M. (2007) Social and cultural capital in educational research: Issues of operationalisation and measurement. British Educational Research Journal 33, no. 6: 867–85. Wray, J. Aspland, J. and Barrett, D (2014) “Choosing to Stay: Looking at retention from a different perspective”, Studies in Higher Education. 39(9), 1700-1714.

186 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Dr. Maria Mbinda-Lashley

Senior Lecturer Sir Arthur Lewis Community College

Dr Maria Mbinda- Lashley is a Senior Lecturer at the Sir Arthur Lewis Community College, Saint Lucia and is dedicated, diligent and committed to the teaching profession. Her teaching, management and research areas are in the fields of Business, Entrepreneurship, Communication, Gender and Development. Dr Mbinda- Lashley joined the Business department at the Sir Arthur Lewis Community College in 1993 and has had over 26 years of teaching experience, working with young adults in an array of educational disciplines and levels including the Associate degree and Under Graduate degree level, Continuing education, Teacher education and Early childhood education. In addition, she also has regional affiliations, having worked with the regional examination body, Caribbean Examination Council (CXC) in the capacity of Examiner and Assistant Chief Examiner for CAPE Communication Studies and Entrepreneurship. She is also a co- author of publications in Business Communication and Entrepreneurship with the Commonwealth of Learning (COL). Most recently her focus has been on advocating for the engendered conversation on Male Academic Upliftment in the Post –Secondary and Tertiary levels of education, with a view to identifying relevant issues, trends, solutions and intervention methods in relation to Male student academic performance. Dr. Mbinda -Lashley holds a Doctorate of Social Science (DSocSci) from the University of Leicester- England in the research area of Male Academic Underachievement.

A Compilation of Working Papers by OECS Scholars

187

11 The Effects of Tactile Learning Strategies on Attitudes of Form 4 CCSLC Mathematics Students Pascalina Stanislas-Inglis and Nitha Mauricette-Phillip

188 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT Attitude to a subject can determine how well a student performs in the particular subject area. A change in teaching strategy may be helpful in improving attitudes toward a subject area and by extension the students’ performance. This study implements tactile learning strategies in the teaching of geometry to a group of Form 4 CCSLC Mathematics students of School X, in an effort to improve attitudes and performance in Mathematics. This Form 4 class is a group of 18 students all from the Vieux-Fort district with ages from 15 years to 17 years. Data collected during the period of the research study was generated from the use of pre-tests and post-tests, attitude scales checklist, exit slips completed by the students and an observation checklist used by the researcher. Some of the data collected also include thoughts shared verbally between the researcher and the students. Analysis of the data revealed that the students responded positively to the tactile learning strategies implemented during the lessons and overall showed an improvement in attitude toward the subject.

INTRODUCTION TO PROBLEM Students in the Form 4 CCSLC Mathematics class of the School X, exhibit a lack of interest toward the subject. This is demonstrated in poor attendance rates or 18.46% for the last school year. Students’ lack of interest was also evident in their responses during class time and in preparation for class. Very often the students would make comments belittling their mathematical ability such as “I can’t do this”, “Why do I need to do this”, “I will never again need to use this in my life”, “It’s a waste of time, I just can’t do this” and “I will never get this”. The simple task of acquiring a calculator to be used during class proved difficult as they just did not see the need for one. Additionally, students at School X generally perform poorly in Mathematics and in particular Geometry, and this is evident from the students’ class tests results as well as the school’s unsatisfactory performance in the CSEC examinations every year, with pass rates below 50%. Therefore, the researcher realised that school X and in particular the form 4 students have a negative attitude towards mathematics. This negative attitude is linked to low motivation and also poor performance. Singh, Granville & Dika state that “students who displayed school behaviour associated with low motivation (e.g., coming late to school, skipping classes, coming unprepared without books and homework) had a more negative attitude toward mathematics.” (as cited in Monteiro, Mata, & Peixoto, 2012, pg. 3). Therefore, the researcher hopes to investigate the effects of using Tactile Learning Strategies to help improve Attitudes and Performance of Form 4 CCSLC Mathematics Students of the School X. STATEMENT OF PROBLEM This action research investigates the impact of tactile learning strategies on the attitudes and performance of Four 4 CCSLC Mathematics students of the School X in the teaching of geometry. RESEARCH QUESTIONS 1. What are Form 4 CCSLC Mathematics students’ attitudes towards mathematics before applying tactile learning strategies? 2. How does the use of tactile learning strategies affect Form 4 CCSLC Mathematics students’ attitudes toward Mathematics? RESEARCH VARIABLES Independent Variables Tactile learning strategies or kinaesthetic strategies – refers to the use of manipulatives and hands-on activities to enhance learning by doing, touching and creating.

A Compilation of Working Papers by OECS Scholars

189

Dependent Variables Student attitude- this refers to the scores obtained by students after completion of the attitude scale checklist. Student performance- this refers to students’ achievement scores on subtopic and end of unit test. QUALITATIVE INSTRUMENTS

Exit Slip, Observation Checklist

QUANTITATIVE INSTRUMENTS

Attitude scale checklist, students’ grades.

SIGNIFICANCE OF THE RESEARCH The finding of this research should provide school x with the data needed to help effect changes at the school, which would improve the general learning climate at the school. Additionally, the findings of this research are very significant to not only school X, but also other low achieving schools on the island of Saint Lucia. The findings can also provide the policy makers at the ministry of education with empirical data to help them make informed decisions in the management of the schools nationally. This research is founded on the constructivist theory of learning that all can learn if the conditions are tailored to the learner and that there is a need for every student to leave the secondary school system with a working knowledge of Mathematics as put forward in the Education Sector Development Plan, 2015-2020. It is believed that with the right conditions there exist the possibility to impact positive change in the lives of these students. The results of this research study will seek to influence how teachers at School X teach Mathematics and eventually change the overall performance of students at the school, in spite of the students having gained entry with low Mathematics grades.

LITERATURE REVIEW Kotelnikov (2018) defines attitude as the established ways of responding to people and situations that we have learned, based on the beliefs and values and assumptions that we hold. It stands to reason then, the student's attitudes to Mathematics is a response to previous situations and beliefs and values about Mathematics having been developed over time. Subki (2012) further highlights that our attitudes are influenced by many things as we develop, including our parents, and interactions with other people. The attitudes of students toward Mathematics is by extension a response to how they have interacted with the subject in the past, taking into account the role of the teacher, other students and the environment within which the subject is taught. Monteiro, V., Mata, L., M., & Peixoto, (2012), posit that negative attitudes are the result of frequent and repeated failures or problems when dealing with mathematical tasks and these negative attitudes may become relatively permanent. The authors further explained that when children first go to school, they usually have positive attitudes towards mathematics. However, as they advance their attitudes become less positive and frequently become negative at secondary school. (Monteiro, V., Mata, L., M., & Peixoto, 2012). Additionally, in a research with secondary students, it was observed that in a study with secondary school students with better academic performance have more positive attitudes regarding math than those with poorer academic performance (Mato, M & De la Torre, E., 2010). Consequently, the research clearly indicates a clear link between attitude and achievement or performance in mathematics. McKinney (2016) describes performance as how well the student has prepared for and performed in class and in addition considers how well the student has assimilated the material presented to them. 190 | Research: The Platform for Innovation, Competitiveness and Growth

A positive attitude towards self, the subject matter can produce enhanced proficiency in the subject area (Brown, 2000). Marzano (1992), has proven that students’ attitudes on learning determine their ability and willingness to learn and that if negative attitudes are not altered the students desire to learn will be stunted, effectively crippling their performance. It is noteworthy that a positive attitude towards mathematics increases a student’s motivation to learn. Fredericks (2005) shows that motivating students can be accomplished by including various activities within the lesson including: allowing for the students to demonstrate their creativity and variety, offering differentiated instruction to cater for the needs of all students and allowing the students to share their accomplishments. Research shows that active learning is a process wherein students are actively engaged in building understanding of facts, ideas, and skills through the completion of instructor directed tasks and activities. In essence active learning is any sort of activity which encourages students to get involved in the learning process. (Promoting Active Teaching and learning, 2014). For teachers, the main concern is ensuring that all students within the classroom acquire as much knowledge as possible when teaching and learning is taking place. Kitson (2012) explains that there are persons who can be described as kinaesthetic or tactile learners, where these individuals learn by touch rather than by the usual visual or auditory modes, thereby these persons and everyone can benefit from, and enjoy physical hands-on activity. Tactile learning strategies used in the teaching and learning of Mathematics include hands-on activities and use of manipulatives. Therefore, tactile learning strategies promote active learning. Tactile learning strategies encourage the students to engage with the material provided during the lesson while actively participating and allows them to interact and collaborate with each other. In lessons where active learning is being explored the students must be allowed to demonstrate a process in place of merely listening and memorizing (Promoting Active Learning, 2000). Grunert (1997) has shown that students learn more when they participate in the process of learning which is in opposition to the traditional styles of teaching where students are expected to sit and listen to information being presented to them. Some benefits coming out of the application of Tactile Learning Strategies include, the facilitation of independent, critical and creative thinking, effective collaboration among peers and increased student investment, motivation and performance (Promoting Active Learning, 2000). Candler (2009) shares that tactile or active learning is much more than hands-on activities and instruction. For active learning to be effective both the hands and the minds of the students are to be engaged. Malouf (2008) recommends the use of engaging teaching methods and an appealing teaching style. Students will reflect the attitude of the instructor, as a result it may be necessary to justify to yourself as the instructor the benefits of the topic you are about to teach and use a teaching style that is likeable and comfortable. Sakar (2009) identifies teaching strategies as the structure, methods, techniques, procedures and processes that a teacher used during instruction and learning activities as the teacher-guided instructional tasks or assignments for students. Active hands-on teaching strategies are designed with the purpose of allowing students to step away from the traditional way of doing things and to become active participants in their own learning (Sakar, 2009). Tactile learning strategies can include: ● students taking breaks during lessons and moving around ● students writing down their own notes ● students standing or moving while reciting information or learning new material ● incorporating multimedia resources (computer, video camera, OHP transparencies, photography camera, etc.) into programs (teacher presentations and student presentations) Provide lots of tactile-kinesthetic activities in the class (Salazar, 2010) A Compilation of Working Papers by OECS Scholars

191

Shaughnessy (1994) posits from his research that use of hands-on models in the Mathematics classroom led teachers to reports that there were dramatic improvements in the student skills, demonstrated both in their ability to work with other students and in their ability to take control of their own learning. Salazar (2010) further posits that kinaesthetic learning strategies help with a plethora of common classroom problems that get in the way of learning, such as: (a) discipline, (b) student focus and interest, (c) motivation, (d) expression, and (e) persistence. The NCTM encourages the use of manipulatives in teaching a wide variety of topics in Mathematics including recognizing geometric shapes and understanding relationships among geometric shapes making use of measurement. Both learning theory and educational research recommends the use of manipulatives within the Mathematics classroom (Research on the Benefits of Manipulatives). Marilyn Burns considers manipulatives as an essential tool for the teaching of Mathematics to students of all levels. Through investigations she has discovered that the use of manipulatives allows almost all learners to access Mathematics content while allowing opportunities to students who are able to catch on quickly to concepts being taught. Marsh and Cooke (1996) along with Ruzic and O’Connell (2001) allude that the use of manipulatives is particularly suited to the teaching of low achievers. Okebukola (2002) and Ekwueme (2007) together affirm that Mathematics and science are practical activities and are best learnt through manipulation of various objects. They assert that the hands-on approach is a teaching method that leads students to attain knowledge by experience. Haury and Rillero (2015) posit that the “hands-on learning approach involves the child in a total learning experience which enhances the child’s ability to think critically”. Ekwueme (2015) recommends the use of tactile learning or hands-on approach to teaching Mathematics and science as a means to improve the students’ academic achievement and level of understanding of various abstract concepts, that can be made clear to the students through the use of manipulatives.

METHODOLOGY PARTICIPANTS The participants involved in this research paper are all students for the Form 4 CCSLC Mathematics class of School X. This group of students comprised Eighteen (18) students participated in this research; 4 females and 14 males. The participants all live in the Vieux-Fort and Laborie district. SETTING The research will be carried out during regular Mathematics lessons scheduled for the Form 4 students. This involves a total of 8 periods out of a 48-period cycle without any interruptions to the regular school day. RESEARCH DESIGN The research design selected for this study was action research. This design was selected because of the nature of the action research and the setting within which the research will be taking place. Action research allows for the careful examination of educational practices and for the elaboration of research in cycles. DESCRIPTION OF INTERVENTION This action research was elaborated over a five to six weeks period. During the research period the students will make use of various resources such as geoboards, popsicle sticks, coloured paper and measuring instruments during the implementation of the lesson to facilitate learning. The participants will use these materials in various hands-on activities set out during the lesson. 192 | Research: The Platform for Innovation, Competitiveness and Growth

INSTRUMENTS For the purpose of this study the researcher used an array of instruments to facilitate the collection of data. The researcher used the participants’ report books to collect their grades (Appendix N) in Mathematics during the previous school year and term. The participants were asked to complete an attitude rating scale to establish a baseline for attitudes toward the subject before the initialization of the application of the tactile learning strategies in the lessons. The rating scale (Appendix H) to be completed was prepared by Brooskstien (2011) of the Kaput Center for Research and Innovation in Stem Education and was designed to determine students’ attitudes toward Mathematics in the classroom. Permission was sought and granted for the use of this checklist in this action research. The rating scale used in this study has been modified, as some of the questions in the original checklist were not important to this research study. The use of this instrument will seek to provide answers to what are Form 4 CCSLC Mathematics students’ attitudes towards Mathematics before and after applying tactile learning strategies? Towards the end of each session where the strategies were applied the students were asked to complete an exit slip and also interacted and communicated verbally their perceptions to the researcher. The exit slip (Appendix I), was designed by the researcher to determine how the participants felt after experiencing the lesson and to determine if learning objectives for the class were being fulfilled. Bafile (2004), encourages that exit slips can be used to determine what students learn each day. The application of the instrument in this manner will set out to determine how does the use of tactile learning strategies affect Form 4 CCSLC Mathematics students’ attitudes toward Mathematics? During the time that the participants were engaged in the exercises the researcher made use of an observation checklist (Appendix K) to assess participation and how students interacted with the material. Kay Burke (1994) describes an observation checklist as ‘‘a strategy to monitor specific skills, behaviours, or dispositions of individual students or all the students in the class.’’ It is recommended that the observation checklist is used as an indicator to indicate if a student accomplished a stated objective. At the end of the research period students were all given an attitude scale checklist to establish students’ attitudes after the application of the tactile learning strategies. PROCEDURE FOR ADMINISTERING INSTRUMENTS The participants completed the exit slips at the end of the lessons where the interventions were applied. The attitude scale checklist was completed in confidence by the participants, before the start of the research period and after the last intervention was applied. It was necessary to ensure that participant responses were not influenced by any outside sources. The observation checklist was used during the lessons where the interventions were being applied. ETHICAL ISSUES For the purpose of this study the researcher informed the principal of School X that the study would be undertaken at the school and the nature of the participants’ involvement in the study. Due to the ages of the participants, that they all are minors, consent forms were sent out to parents to give authorization for their children to participate in the study. Parents were informed of the study and the nature of the study that their child/ward would be involved in. In order to use the attitude scale checklist to determine students’ attitudes, it was necessary to seek permission from the author. Authorization was sought for and received in writing via email. APPLICATION OF INTERVENTION The interventions selected for each lesson was dependent upon the objectives for each class and the availability of each of the resources and materials to be used in each case. The evaluations for the lessons were often dependent upon the completion of the activity or the completion of a worksheet designed for each class.

A Compilation of Working Papers by OECS Scholars

193

In the elaboration of the topic of Triangles (Appendix C), one of the objectives outlined was for students to identify the different types of triangles. Students through the use of videos and a projector explored the various types of triangles and their properties. Students were then required to build the various triangles that they had come to learn of using coloured popsicle sticks, glue and paper. The students were placed in groups as this was to be carried out within the group. Within the group they would build and name each triangle. The teacher would ensure that each group had the resource required to carry out the task and that there was collaboration among group members.

FIGURE 1: CONSTRUCTION OF TRIANGLES WITH POPSICLE STICKS

FIGURE 2: COLLABORATIVE GROUP WORK WITH MANIPULATIVES

FIGURE 3: GROUP WORKING ON TRIANGLE CONSTRUCTION

194 | Research: The Platform for Innovation, Competitiveness and Growth

Another implementation involved a class on circles (appendix d), where students would be required to correctly identify the parts of the circle. The lesson began with a review of what the students already knew. A diagram of a circle and its parts clearly highlighted was posted on the board and each child was given a sticky note upon which they were given one minute to write the part of the circle that they knew and place it correctly on the diagram. Students were encouraged to ignore where their classmates had placed their notes but be mindful of only their positioning. Upon the completion of the task, parts of the circle were correctly labelled and explained after which each child was given a paper plate and markers. The participants would use the plate as their circle and correctly place and label each part of the circle. Collaboration among their peers was encouraged but within limits, each child had to work on their own plate. FIGURE 4: RESULT OF STUDENTS WORKING WITH MANIPULATIVES FOR THE CIRCLE LESSON

FIGURE 6: EVIDENCE OF USE OF TECHNOLOGY IN LESSON

FIGURE 5: FOLDABLE CREATED BY STUDENT DESCRIBING THE PARTS OF A CIRCLE

FIGURE 7: CHART USED AS PART OF INTRODUCTION TO LESSON

During the elaboration of the class on the perimeter of squares (Appendix B), rectangles and triangles, the participants used geoboards and rubber bands. Due to the limited number of geoboards the students were placed in groups to complete the activity. Building on the knowledge that the perimeter of the shape is the distance around the boundary of the shape. The students were taught how to make a shape (triangle, square, rectangle), with a specific perimeter and given the lengths of the sides using a geoboard. A short period of competition was elaborated where each group was asked to make a shape using the geoboard and rubber bands with specific dimensions. The group able to complete the task correctly first won. During the execution the members of the group collaborated to ensure that the triangles met the specifications required. A Compilation of Working Papers by OECS Scholars

195

For composite shapes (Appendix F), the researcher again used the geoboard and rubber bands during the elaboration of the lesson. One of the objectives for this specific lesson was for students to identify composite shapes and the shapes within a composite shape. Students were placed in groups and asked to create composite shapes given a series of regular shapes. For example, a composite shape involving 1 rectangle and 1 square. The first phase involved drawing the diagrams on paper with assistance from all the group members. During the second phase students were to build their composite shapes using the geoboard and rubber bands. The researcher needed to ensure that all members of the group participated and as a result the number of shapes that needed to be created was enough that each group member had an opportunity to create their own with consultation from other group members. The strategy was also applied during the elaboration of a lesson with objectives of identifying base and height of various triangles (Appendix G). In this lesson the use of geoboards was employed. Students were asked to create triangles given specific dimensions. In each case the students were required to identify the type of triangle they were building and to identify the base and the height in each case. It allowed for students to identify the base of the triangle given the height or the height of the triangle having identified the base of the triangle. The researcher ensured that every student got an opportunity to create a triangle and identify the base or the height of the triangle. The students would also be required to identify the measurement of the base or height of the triangle. FIGURE 8: STUDENTS INTERACTING WITH TECHNOLOGY DURING THE MATH LESSON

196 | Research: The Platform for Innovation, Competitiveness and Growth

Tactile learning was also used during the elaboration of a lesson on areas of regular shapes. The focus for the lesson was on areas of squares and rectangles and the use of area in square units was part of the lesson. After the introduction of square units (Appendix F), students were placed in groups and given a geoboard and rubber band.

FIGURE 9: SAMPLE OF STUDENT WORK USING SQUARE UNITS TO CALCULATE AREA OF A SHAPE

Students were asked to create a shape with specific dimensions, after which students were asked to determine the area of the shape in terms of square units. Students were given a list of shapes and dimensions to determine the area using the geoboard. After some time, students were asked to try to identify the relationship between the sides of their shape and the area that they had identified. In every application of the intervention the students were engaged and remained on task. Students collaborated within and outside their groups at times when they were consulted by other groups. In each case it could be noted that students were engaging with the knowledge being acquired and evaluation of the tasks completed would confirm this.

A Compilation of Working Papers by OECS Scholars

197

DATA ANALYSIS AND INTERPRETATION DATA ANALYSIS A variety of instruments were used to collect data for this study and the table below shows the instruments based on each research question. Both qualitative and quantitative methods were used to analyse the data of the study.

TABLE 1: TABLE OF INSTRUMENT SPECIFICATION Research Question

Proposed Instrument

What are Form 4 CCSLC Mathematics students’ attitudes towards Mathematics before applying tactile learning strategies?

Pre-Attitudinal scale

How does the use of tactile learning strategies affect Form 4 CCSLC Mathematics students’ attitudes toward Mathematics?

Observation Checklist

Post Attitudinal scale

Exit Slips

This section of the research seeks to provide answers to the research questions through investigation of results obtained. Research Question 1: What are Form 4 CCSLC Mathematics students’ attitudes towards Mathematics before applying tactile learning strategies? To identify a baseline for the research process, each participant was given an attitude scale checklist to complete before the initialization of the application of intervention. Moving forward each participant will be referred to by a number. The questions from the attitude scale have been grouped into 3 sections, each dealing with a specific aspect of attitude.

TABLE 2: TABLE SHOWING DISTRIBUTION OF QUESTIONS FROM THE CHECKLIST Attitude

Question Number

Total Questions

Attitude 1: Indicator of positivity toward Mathematics and school

1, 2, 7, 13, 14, 22, 23

Attitude 2: Indicator of effect of working collaboratively

4, 12, 15,16, 18, 19, 21, 24

Attitude 3: Indicator of working privately 3, 5, 20

198 | Research: The Platform for Innovation, Competitiveness and Growth

The Table below shows the results obtained from the initial application of the checklist. The scores for each child for each attribute was recorded and a composite score was calculated based on the values of the responses. This composite score was calculated as ∑_i^1 m_i /w_i, where m is the maximum score and w is the weighting applied to that question. For the purpose of this study each item in the attitude group has an equal effect on the attitude. The minimum score to be obtained is 0- Strongly Disagree and the maximum score is 4-Strongly Agree. The mean was then analysed based on the following criteria: 0 – 0.9 very poor attitude 1 - 1.9 poor attitude 2 – 2.9 positive attitude 3- 4- very positive attitude TABLE 3: TABLE SHOWING RESULTS OF INITIAL CHECKLIST Attitude 1

Attitude 2

Attitude 3

Student 1

1.22

1.33

Student 2

2.33

1.33

Student 3

2.86

2.33

Student 4

3.29

2.78

1.33

Student 5

2.43

1.67

Student 6

3.29

2.56

Student 7

2.71

2.44

1.67

Student 8

2.29

1.56

2.67

Student 9

2.29

2.22

Student 10

2.43

1.44

Student 11

1.71

2.33

Student 12

2.57

1.67

3.67

Student 13

1.71

2.89

1.67

Student 14

2.71

2.33

3.67

TABLE 4: ANALYSIS OF INITIAL RESULTS N

Minimum

Maximum

Mean

Std. Variance Deviation

Skewness

Statistic

Std. Error

Attitude1

1.71

3.29

2.5204

.50307

.253

-.110

.597

Attitude2

1.00

2.89

2.0317

.59505

.354

-.293

.597

Attitude3

1.33

3.67

2.1190

.76914

.592

1.156

.597

Valid N (listwise)

A Compilation of Working Papers by OECS Scholars

199

Analysis of the initial data shows that students’ positivity towards learning, Mathematics and school to be relatively the same for each participant, given the low calculated values for the standard deviation (0.5) and the variance (0.25). The mean of the composite score for attitude 1 and attitude 2 are very close to the maximum recorded score and also speaks to the similarity in the opinions of the participants in this research. Consideration was also given to comments made by students upon entry to Mathematics lessons and informal interactions with the researcher. Some of these comments included: “Miss, I can’t do that” John. “I’m never ever going to need what we are doing there now, so I don’t see why I have to do it.” Trudy. “Must we have class today.” Jonathan. “Math again.” Mark. These statements speak to the negative desire to participate and to attend the stipulated Mathematics class. The students of the Form 4 CCSLC Class of School X, before the application of intervention recorded attitudes ranging from poor attitudes to Mathematics to positive attitudes in Mathematics. Most of the students resulted falling within this range with a few having very positive attitudes. The results indicate when it comes to working together with other students the attitude is the same, with most of the values falling with the range of poor attitudes. The results also indicate that the students showed a preference toward working privately. The results from the attitude scale indicate no evidence of students having very poor attitudes toward Mathematics, although they had indicated this through their comments. Research Question 2: How effective is the use of tactile learning strategies towards improving Form 4 CCSLC Mathematics students’ attitudes toward Mathematics? The use of the tactile learning strategies during the lesson had a positive impact on the students. The researcher noted that some most students before the application of the interventions were not responsive to teacher questions, however after a few interventions the students began to display more interest in class and became more willing to answer questions posed by the teacher. The students displayed confidence in their responses. The researcher also noted a change in the comments made by students which include: “That’s so easy.” James. “That’s what I did not know before.” Jennifer. “Miss, you should have more classes like these.” Jack. The researcher also noted an increase in the frequency of attendance of the participants. The researcher used a classroom walkthrough checklist while observing the students during the hands-on activities. Through use of the observation checklist during the activities the researcher noted that in each group students encouraged one another, they worked collaboratively to complete the task given and there was sharing of knowledge within the groups. During the whole class engagement, students would participate in the discussions and the level of work produced by either the whole group or the individual demonstrated levels of application and analysing.

200 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 5: RESULTS OF POST APPLICATION OF CHECKLIST N

Minimum

Maximum

Sum

Mean

Std. Deviation

Variance

Statistic

Std. Error

Statistic

Attitude 1

1.57

3.43

29.29

2.4405

.14760

.51131

.261

Attitude 2

1.22

3.00

24.89

2.0741

.15372

.53252

.284

Attitude 3

1.00

3.00

21.00

1.7500

.16979

.58818

.346

Valid N (listwise)

Table 5 demonstrates the changes in attitude of the participants at the end of the intervention.

TABLE 6: CHANGE IN ATTITUDES AFTER THE INTERVENTION Min

Attitude 1

Attitude 2

Attitude 3

Max

Mean

Std. Deviation

Before

1.79

3.29

2.52

0.5

After

1.57

3.43

2.44

0.51

Change

0.22

-0.14

0.12

-0.01

Before

2.89

2.03

0.6

After

1.22

2.07

0.53

Change

-0.22

-0.11

-0.04

0.07

Before

1.33

1.67

2.12

0.77

After

1.75

0.59

Change

0.33

-1.33

0.37

0.18

Table 6 shows a positive change in the component scores for attitude 1 indicating an improvement for students in positivity towards learning and Mathematics. An improvement in the maximum component score indicates that the students are agreeing more with the statements that belong to this group. An improvement in the component scores for attitude 2 implies that the students are more comfortable with working together in groups and sharing their ideas of Mathematics in the classroom. A reduction in the standard deviation for this section is indicative of similar scores for the participants. The lower values for attitude 3 after the application of the intervention is proof that the students are more comfortable working as a group instead of on their own, although the increase in the maximum component score is evidence that there are students who still prefer to work alone.

A Compilation of Working Papers by OECS Scholars

201

TABLE 7: CORRELATION WITH CHANGES IN ATTITUDES USING SPEARMAN’S RHO Correlations ChangeAtt1

ChangeAtt2

ChangeAtt1

Correlation Coefficient

1.000

ChangeAtt2

Correlation Coefficient

.822**

1.000

ChangeAtt3

Correlation Coefficient

.043

.216

ChangeAtt3

1.000

**. Correlation is significant at the 0.01 level (2-tailed).

TABLE 8: DISTRIBUTION OF INDICATORS FROM OBSERVATION CHECKLIST Indicators of Positivity Toward Mathematics

Indicator of Student Collaboration.

● Asking and Responding to Questions.

● Students have defined responsibilities.

● Listening & Note Taking.

● Students encourage one another.

● Participating in the Discussion. ● Participating in guided Practice.

● Collaboratively producing a product. ● Collaboratively problem-solving. ● Participating in discussion.

Indicator of Student working Privately ● Independently producing product. ● Independently solving a problem. ● Independent practice. ● Presenting. ● Writing Activities. ● Researching information. ● Silent reading.

● Presenting.

The researcher noted from observing the students when placed in groups that most of the time all of these conditions were satisfied. However, the students showed some difficulty working independently. After the application of intervention various changes were noted in attitudes. There was a slight increase in the attitude towards working collaboratively, which indicated that the students developed some appreciation for working with each other. There was a noted decrease in the attitude to working independently which means that the students now felt more comfortable working with others as opposed to preferring to work alone. The results also showed a positive change in the attitudes of students toward Mathematics with a new maximum score in this area of 3.43, an indicator of very good attitude. 202 | Research: The Platform for Innovation, Competitiveness and Growth

REFLECTIONS Coming out of this action research project are four major findings. The first being that the use of hands-on strategies in the Mathematics classroom can positively influence students’ attitude toward learning, Mathematics and school. Secondly, use of hands-on resources and activities in the Mathematics classroom in a collaborative setting positively influences students’ attitudes to group work and Mathematics. Thirdly, regardless of the positive effect of the use of tactile teaching strategies in Mathematics, a few students still preferred to work alone rather than in groups. Fourthly, an improvement in Mathematics students’ interest and desire can be attributed to the use of hands-on teaching strategies in Mathematics lessons.

RECOMMENDATIONS Based on the results of this action research project the researcher recommends the following: 1. The use of hands-on teaching strategies and activities to be utilized as often as possible within the Mathematics classroom as this helps to improve students’ attitudes toward Mathematics. Given the nature of the students entering School X, that is, students performing very poorly in Mathematics, it is necessary to try to change the way that these students view the subject and one way to do this is through the use of tactile learning as demonstrated in this research project. 2. The acquisition of Mathematics resources for example, geoboards, beets and straws, fraction tiles, algebra tiles, etc., that can be used in and out of the Mathematics classroom. 3. Setting up a Mathematics room with hands-on resources to be used by students both in the classroom under the direction of the class teacher and by students in their spare time to encourage development of mathematical skill. 4. Proper design of lessons where hands-on activities are to be used, these lessons should cater for both the development of the mind and hands of the students involved. Use of hands-on activities to impart knowledge must be deliberate to avoid students enjoying the activity without gaining any knowledge. The teacher must ensure that the design and use of the strategy must foster complete development of the student. 5. Implementation of tactile learning in the lower forms of School X to remedy or prevent the development of negative attitudes to Mathematics. 6. Further investigation to determine the effect of tactile learning on the performance of students in the Mathematics class. During the course of this research project one test was administered and the results were favourable, however the time dedicated to this action research was not enough to determine satisfactorily the effect of the tactile learning strategy on performance.

CONCLUSION In conclusion this research shows that the use of tactile learning strategies has a positive impact on form 4 CCSLC Mathematics students of School X. Gennerman (2006), describes an intervention as a specific skill building strategy implemented and monitored to improve a targeted skill and to achieve adequate progress in a specific area. Intervention is applied usually through a change in the instruction method. In this regard the use of tactile learning strategies within the Mathematics classroom, is one such intervention that can be applied to encourage learning. Accommodation as detailed by Gennerman (2006), allows for a change in how a student accesses information, participates in school activities and demonstrates learning. Accommodation is not an advantage, rather it is providing what is necessary to allow the student to learn what is being taught more effectively. Imagine not being able to see clearly, one would not be able to read, however with the accommodation that prescription glasses provide it makes the words more visible and allows them to be A Compilation of Working Papers by OECS Scholars

203

read without any advantage to the individual wearing the glasses or to the one who does not need them. Students harbouring negative attitudes toward Mathematics, can present as a disability as it makes learning Mathematics difficult. In the teaching of Mathematics, it may be necessary to provide glasses to the students to clarify their view of the subject or in the case of this research the use of manipulatives. The addition of tactile learning strategies with the use of manipulatives, to Mathematics lessons provides one such accommodation and has been proven through this research project to improve the attitudes toward learning Mathematics. DeGeorge (2004), has demonstrated through research that harnessing the power of manipulatives has proven invaluable in the teaching of Mathematics and that the students are better able to visualize math concepts and gain insights. Spectacular results can be achieved when learning is taken off the chalkboard and literally put into the hands of the learners themselves. Improved attitudes as resulting from this study, is one of the responses to the intervention of tactile learning strategies, which was administered throughout this research process.

REFERENCES Alice Krueger, J. S. (2002). Why Teach Math with Manipulatives. Arden Brookstein, S. H. (2011). Measuring Students Attitude in the Mathematics Classroom. Kaput Center for Research and Innovation in STEM Education. Bafile, C. (2004). “Let It Slip!” Daily Exit Slips Help Teachers Know What Students Really Learned. Barbara DeGeorge, A. M. (2004, December). Manipulatives: A Hands-on Approach to Math. Bobby Ojose, L. S. (2009). The Effect of Manipulative Materials on Mathematics Achievement of First Grade Students. The Mathematics Educator. Burns, M. (2007). About Teaching Mathematics, A k-8 resource (3rd Edition). CA: Math Solutions. Candler, L. (2009). Actively engage Students using Hands-on & Minds-on Instruction. Retrieved from Teachhun.com: http://www. teachhub.com/actively-engage-students-using-hands-minds-instruction Candler, L. (n.d.). Actively Engage Students Using Hands-on & Minds-on Instruction. Retrieved from www.teachhub.com Cecilia O. Ekwueme, E. E.-N. (2015, November 16). The Effect of Hands-on Approach on Student Academic Performance in Basic Science and Mathematics. Calabar, Nigeria. Claesessens, A. (2013). Hands-On Learning. The University of Chicago and Mimi Enge, Vanderbilt University. Clearinghouse. (2009). What Works. Dewey, J. (1934). Individual Psychology and Education. The Philosopher, 12. Fredericks, A. D. (2005). Top 10 Motivational tips for the classroom. Retrieved from Teacher Vision: https://www.teachervision. com/top-10-motivation-tips-classroom Gennerman, T. (2006). Accommodations, Modification and Intervention. Hagner, P. (2001, January 25). Interesting Practices and Best Systems in Faculty Engagement and Support. National Learning Infrastructure Initiative. Hien, T. T. (2009). Why is Action Research Suitable for Education? VNU Journal of Science, Foreign Languages.

204 | Research: The Platform for Innovation, Competitiveness and Growth

Hoke, D. M. (2008). Effects on Students Performance of Using Hands-on Activities to Teach Seventh Grade Students Measurement Concepts. John M. Malouff, S. E. (2008). Methods of Motivational Teaching. Jones, S. (2013, April). Instructional Strategies Motivate and Engage Students in Deeper Learning. Retrieved from www.sreb.org Kitson, T. (2012, January 1). Enhancing Teaching Using Tactile Objects. Retrieved from https://eic.rsc.org/feature/enhancingteaching-using-tactile-objects/2020133.article Korn, J. (2014). Teaching Conceptual Understanding of Mathematics via a Hands-on Approach. Kotelnikov, V. (2018). Attitude Motivation. Retrieved from http://www.1000ventures.com/business_guide/crosscuttings/ motivating_attitude.html Krueger, J. S. (2002). What We Know About Mathematics Teaching and Learning. L. Cohen, L. S. (2005). Major Education Philosophies. Lambdin, D. V. (2009). Benefits of Teaching Through Problem Solving. Lewin, K. (1946). Action research and Minority Problems. Journal of Social Issues 2, 34. Marzano, R. J. (1992). A Different Kind of Classroom. Mckinney, D. (2016). Student Performance Standards. Retrieved from http://www.ce.utexas.edu/prof/mckinney/ StudentPerformanceStandards.htm Meghan M, J. A. (2013). Kinesthetic and Tactile Learners. Ministry of Education Guyana. (2016, August 29). About Students’ Attitude on learning. Retrieved from Ministry of Education, Guyana: http://www.education.gov.gy/web/index.php/teachers/tips-for-teaching/item/2192-about-student-attitudes-on-learning Moore, D. S. (2014). Why Teach Mathematics with Manipulative? Mueller, J. (2016). Authentic Assessment Toolbox. North Central College. National Research Council. (2001). Helping Children Learn Mathematics. NCSM. (2013). Hands-on Learning: Proven to Increase Student Outcomes. National Council of Supervisors of Mathematics. Noel, J. (1993, April). Practical Reasoning: Constructivist Theory and Practice in Teacher Education. O’Neill, S. (2009, June). Effective Teaching. Australia: Department of Education and Training. Retrieved from https://www. education.wa.edu.au/documents/43634987/44524721/Effective+Teaching.pdf/5dcc8207-6057-3361-ade8-cf85e5a2c1ab Plourde, J. D. (2008, December). The Effect of Inquiry Base, Hands-on Math Instruction Utilized in Combination with Web Based, Computer Assisted Math Instruction on 4th-Grade Students’ Outcomes. Nebraska, Omaha. Promoting Active Learning. (2000). Retrieved from Stanford Teaching Commons: https://teachingcommons.stanford.edu/ resources/learning-resources/promoting-active-learning Promoting Active Teaching and learning. (2014). Retrieved from James Cook University: https://www.jcu.edu.au/__data/assets/ pdf_file/0006/227868/Promoting-Active-Teaching-and-Learning.pdf Research on the Benefits of Manipulatives. (n.d.).

A Compilation of Working Papers by OECS Scholars

205

Rudhumbu, N. (2014). Motivational Strategies in the Teaching of Primary School Mathematics in Zimbabwe. International Journal of Education Learning and Development. Sakar, A. (2009). Active Teaching Strategies and Learning Activities. Jones and Bartlett Publishers. Salazar, J. B. (2010). Mathematics in Motion: a Handbook of Kinesthetic Teaching Strategies. Salazar, J. B. (2010). Mathematics in Motion: A Handbook of kinesthetic Teaching Strategies. Retrieved from https://www. hindawi.com/journals/cdr/2012/876028/ Shaughnessy, J. M. (1994, November). Promoting Student Mathematics Learning Through a Hands-on and Visual Math Program. Portland, Oregon. Torre, M. M. (2010). Evaluaciones de las Actitudes hacia Las Matemáticas y el Rendimiento Académico. PNA. V. Monteiro, L. M. (2012). Attitudes Toward Mathematics: effects of Individual, Motivational and Social Support Factors. Retrieved from https://www.hindawi.com/journals/cdr/2012/876028/ Vogt, K. J. (2006). The Effect of Hands-on Strategies on Student Understanding and Motivation in Science. e-journal for student teachers and new teachers. W. Car, S. K. (1986). Becoming Critical: Education, Knowledge and Action Research. London: Palmer Press. Wenglinsky, H. (2002). How schools matter: The link between teacher classroom practices and student academic performance.

206 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX TABLE A: ATTITUDE SCALE CHECKLIST

A Compilation of Working Papers by OECS Scholars

207

TABLE B: ATTITUDE SCALE RATING SCALE MODIFIED

DARTMOUTH ATTITUDE SURVEY

CIRCLE THE APPROPRIATE RESPONSES BASED ON THE KEY BELOW

0 STRONGLY DISAGREE

1 DISAGREE

2 NEUTRAL/UNDECIDED

3 AGREE

4 STRONGLY AGREE

1. I think mathematics is important in life.

2. In primary school, my math teachers listened carefully to what I had to say.

3. I learn more about mathematics working on my own.

4. I do not speak in public.

5. I prefer working alone rather than in groups when doing mathematics.

6. I get anxious in school.

7. In primary school, I learned more from talking to my friends than from listening to my teacher.

8. I like my own space outside school the majority of the time.

9. I enjoy being part of large groups outside of school.

10. I do not participate in many group activities outside of school.

11. I do not like school.

12. I like math.

13. I feel confident in my abilities to solve math problems.

14. In the past, I have not enjoyed math class.

15. I receive good grades on math tests and quizzes.

16. When I see a math problem, I am nervous.

17. I am not eager to participate in discussions that involve mathematics.

18. I enjoy working in groups better than alone in math class.

19. I like to go to the board and share my answers with peers in math class.

20. I enjoy hearing the thoughts and ideas of my peers in math class.

21. Mathematics interests me.

22. I sometimes feel nervous talking out-loud in front of my classmates.

208 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE C: EXIT SLIP

What did you learn today?

What did you like about class today?

What did you not like about the class today?

A Compilation of Working Papers by OECS Scholars

209

TABLE D: OBSERVATION CHECKLIST Teacher/Grade/Subject

LEGEND CLASSROOM WALKTHROUGH CHECKLIST

Date/Start time/End time

+ - Evidence NE - NO Evidence NA - Not Available

Focus on LEARNERS & RELEVANCE QUALITY EVIDENCE

OBSERVATIONS

Student Engagement

Follow-up FOCUS

□ Authentically on Task □ Passive/Compliant □ Disengaged/Disruptive Whole Class □ Asking & responding to questions □ Listening & note taking □ Participating in discussion □ Participating in guided practice Small Group or Paired □ Students have defined responsibilities □ Students encourage one another □ Collaboratively problem-solving □ Participating in discussion □ Presenting Individual □ Independently producing a product □ Independently solving a problem □ Presenting □ Silent reading □ Writing activities □ Researching information Level(s) o Student Work □ Remembering □ Understanding □ Applying □ Evaluating □ Creating Strengths Areas of Need

210 | Research: The Platform for Innovation, Competitiveness and Growth

GUIDING QUESTIONS

About the Authors

Pascalina Stanislas-Inglis Teacher Corinth Secondary School

Nitha Mauricette-Philip

Lecturer Sir Arthur Lewis Community College

Mrs. Pascalina Stanislas-Inglis is a dedicated teacher currently assigned to the Corinth Secondary School where she teaches Mathematics to form ones, form fours and form fives. Mrs. Inglis holds a Bachelor’s degree in Computer Science and Mathematics from the University of Santa Clara, Cuba. She also holds a post graduate degree in the teaching of Secondary School Mathematics from the University of the West Indies. She is fluent both in oral and written Spanish and also fluent in her country’s native creole language. She has also completed various training programs including teaching with technology, instructional programs with CDELTA, differentiated instruction, child centered learning and various youth leadership programs. In addition, she has been aiding students with special needs during the national common entrance examination for the past four (4) years as an Assessment Service Provider (ASP).

Nitha Mauricette-Philip is an experienced Mathematics lecturer at The Sir Arthur Lewis Community College; attached to the division of Teacher Education and Educational Administration (DTEEA). She is currently the Subject head for mathematics and Head of Department for Practical teaching at DTEEA. She is also the Chair of the Regional Joint Board of Studies for Mathematics for Joint Board of Teacher Education (JBTE). She holds a Masters Degree in the Teaching of Mathematics, from Lesley University, Cambridge MA, USA. A Bachelor’s Degree in Secondary Mathematics, Mona Campus, UWI, Jamaica. She is trained in project management, e-learning, blended learning, Moodle, project-based learning, k- 12 Education and course writing.

A Compilation of Working Papers by OECS Scholars

211

12 Measuring Vulnerability: A Multidimensional Vulnerability Index for the Caribbean Justin Ram, J. Jason Cotton, Raquel Frederick, Wayne Elliott

212 | Research: The Platform for Innovation, Competitiveness and Growth

EXECUTIVE SUMMARY This paper updates and revises the Caribbean Development Bank’s (CDB) Vulnerability Index previously estimated by Crowards (2000) and Hartman (2011) and widens the scope to include social vulnerability and climate change components. This paper seeks to quantify and gain deeper understanding of the relative vulnerabilities of CDB’s Borrowing Member Countries (BMCs). It aims to answer the questions:

(A) How vulnerable are Caribbean small states?

(B) Where are the vulnerabilities concentrated?

Among the main findings are that BMCs, on average, can be considered middle˗to˗high vulnerability countries with an average vulnerability index score of 0.54 for 2017, slightly above the score of 0.52 for 2016. The vulnerabilities are concentrated in the areas of dependence on a few major export products and trading partners, high levels of energy imports and related products, social challenges, such as crime, and exposure to natural hazards and climate change. Although resource˗rich countries had lower vulnerability scores, the paper notes that these economies have unique challenges related to sharp boom and bust cycles and higher levels of inequality, among others.

INTRODUCTION The challenges among small states are well known and include: trade openness; dependence on energy; limited diversification; susceptibility to natural disasters and climate change; restricted access to external capital; and weaknesses in institutional capacity. Recent research1 by the Caribbean Development Bank (CDB) elaborated on these challenges for small states in the Caribbean region and articulated a vision for regional economic transformation. Although these challenges are not new, in recent years small states are taking a collective stand to highlight their development challenges in international fora, draw attention to their concerns, and spur support for their need of increased concessionary international development assistance. Small states seek increased access and eligibility to concessional development finance to address their vulnerabilities and development challenges. Central to this goal is the quantification of the vulnerabilities of small states. As early as 1994, the Barbados Programme of Action2 urged that vulnerability indices integrating ecological fragility and economic vulnerability be developed to ensure that small states can access supplementary resources.3 The resurgence of global interest in the quantification of vulnerability indices follows the United Nations (UN) Third International Conference on Small Island Developing States (SIDS) in 2014. This conference called for a SIDS Accelerated Modalities of Action (SAMOA) Pathway to continued eligibility for concessional aid, given the vulnerability to climate change and natural disasters, and the gains made towards the increased recognition of the special needs of least developed countries in the 2015 UN Climate Change Conference. Previous efforts were made by the CDB to estimate the vulnerability of small states by Crowards (2000) and Hartman (2011). The focus was the estimation of an Economic Vulnerability Index (EVI) that was used to support evidence˗based policy formulation, planning and decision-making, but also to guide CDB’s development financing architecture and, particularly, for its concessional resources. This research updates CDB’s EVI and considers expanding the scope of the index beyond the typical economic measures to consider social vulnerability and susceptibility to natural hazards, and to include vulnerability to climate change. This provides a more holistic perspective on vulnerability while maintaining strong strategic alignment with the development priorities in countries of the Caribbean. This A Compilation of Working Papers by OECS Scholars

213

paper seeks to quantify and gain deeper understanding of the relative vulnerabilities of CDB’s Borrowing Member Countries (BMCs) (see Appendix 1). It aims to answer the questions:

(a) how vulnerable are Caribbean small states?

(b) where are the vulnerabilities concentrated? And

The structure of the paper is as follows. Section 2 reviews vulnerability metrics; Section 3 highlights the methodology for estimating the Vulnerability Index; Section 4 discusses the results of the paper; and Section 5 concludes and provides some initial policy recommendations. It is anticipated that additional CDB research and policy papers on the topics of vulnerability and resilience will provide additional policy recommendations for Caribbean policymakers.

REVIEW OF VULNERABILITY METRICS DEFINITIONS The concept of vulnerability is complex but critical for development, and is highly policy˗relevant. The term is multidimensional in nature, which contributes to the challenges in defining and measuring it. In 2011, the Economic Commission for Latin America and the Caribbean [ECLAC (2011)] noted that the concept of vulnerability has several dimensions.4 These dimensions begin with vulnerability as an internal or intrinsic risk factor (which is universally accepted) and can be broadened to a multidimensional approach which includes the physical, economic, social, environmental and institutional characteristics of the grouping being assessed. It is defined in this paper as the exposure to sharp external shocks, either fiscal, trade or climate-related, and can be distinguished from the term fragility, which is a consequence of the tenuous institutional or societal mechanisms within a country to mediate internal pressures, causing it to either implode or face the stresses of conflict and economic collapse. The term economic vulnerability refers to the inherent, permanent or quasi-permanent features of a country, which render that country exposed to economic forces outside its control. Deriving an index of economic vulnerability is regarded as a challenging, but worthwhile, exercise. The index attempts to combine what are perceived to be the root causes of economic vulnerability into an aggregate composite index. In this paper, the focus of economic vulnerability is on the structural characteristics of small states that make them more vulnerable to external shocks than their larger country counterparts. These structural characteristics are independent of a country’s political will or policy-induced factors and therefore do not result from recent policy choices of the government. The key building blocks of the economic vulnerability argument within small states are linked to structural factors that are associated with:

(a) remoteness from global markets;

(b) lack of diversification;

(d) susceptibility to natural disasters;

(e) small internal markets and lack of economies of scale; and

(f) dependence on non-renewable sources of energy.

214 | Research: The Platform for Innovation, Competitiveness and Growth

It should be noted that susceptibility to natural disasters was assessed as an important component as it can exacerbate the downside effects of economic shocks.

MEASURING VULNERABILITY Considerable work has been done on the estimation of economic vulnerability metrics. However, there remains tremendous scope to develop policy consensus on the choice of vulnerability metrics, the operationalisation of such metrics, and the agreement on weighting and aggregation. Much of the variation depends on the purpose and objectives of the index. This is a key component in guiding the design, as it distinguishes the selection of variables and estimation methods in one vulnerability index compared with another. Additional considerations include the availability of data and theoretical underpinnings. Several development agencies have also undertaken the design of vulnerability metrics including UN Committee for Development Policy (UNCDP); UN Office of the High Representative for Least Developed Countries, Landlocked Developed Countries and Small Island Developing States (UNOHRLLS); ECLAC; UN Development Programme (UNDP); UN Department of Economic and Social Affairs (UNDESA); the Commonwealth Secretariat; and CDB. The following paragraphs will briefly review these metrics and provide information pertaining to their indicators, methodology, strengths and limitations. UNCDP flags three criteria to identify the least developed countries, benchmarks that are useful in measuring vulnerability in upper-middle and middle-income small states. The criteria include: (a) gross national income per capita; (b) the Human Assets Index, with four health and education indicators; and (c) EVI, which computes the structural vulnerability to economic and environmental shocks and incorporates two sub-indexes. Firstly, the exposure index factors in population, remoteness, merchandise export concentration, share of agriculture, forestry and fishing in gross domestic product (GDP), and the share of population in low-elevated coastal zones. Secondly, the shock index quantifies the instability of goods and services, victims of natural disasters, and instability of agricultural production. However, concerns about data availability weaken the applicability of the metrics and the rationale for the relative weights is unclear. UNOHRLLS adapts EVI to create a weighted index to capture the interactions and interdependence between the selected indicators. The steps in the construction of the index include a normalisation methodology to the data. The composite index is estimated as the simple arithmetic average of the exposure and shock indices. The strengths of this approach include: the choice of indicators facilitates comparability (128 countries); it is relatively simple; it advocates the use of EVI with other indicators; and includes an approximation procedure for missing data. The limitations of the approach are as follows: vulnerability is confined mainly to economic factors; the rationale for the weights is unclear; it excludes service exports; and resilience is not explicitly covered by EVI. ECLAC prepared a study on the vulnerability and resilience of Caribbean SIDS in 2011. The study highlights various indicators and indices of vulnerability and resilience and data requirements, methodological issues and disadvantages in deriving the economic vulnerability, environmental vulnerability and social vulnerability indices. Notably, the study highlighted that the development of appropriate vulnerability indices for the Caribbean will be severely compromised unless the data paucity challenge is addressed in a holistic manner. The study highlights that strategies for building resilience, particularly to natural environmental impacts across the Caribbean Region, are not homogenous although there are crossingcutting issues, such as gender equity, that would have universal application. Additionally, economies of scale can be realised through a regional effort to mainstream vulnerability reduction and resilience building into development planning. The Commonwealth Secretariat started its work in the area of vulnerability and resilience in 2004. The methodology involved in estimating EVI is based on the seminal work of Briguglio (1992). The basic

A Compilation of Working Papers by OECS Scholars

215

criteria adopted to construct the Briguglio Vulnerability Index and that underlies the selection of the indicators in the current study are: ● simplicity: the index should not be too complicated to construct; ● ease of comprehension: the overall composite index must have an intuitive meaning; and ● suitability of international comparison: the index should lend itself to international comparisons. The Commonwealth’s EVI has four equally˗weighted components of: (a) Trade Dependence Index; (b) Export Concentration Index; (c) Dependence on Strategic Imports Index; and (d) Disaster Proneness Index. These components are complemented with a proposed resilience index that is grouped into three equally-weighted components of the: Macroeconomic Stability Index; the Market Flexibility Index; and Political, Social and Enviro˗Governance Index. For both the vulnerability and resilience indices, the total score is the simple average of the different components; and vulnerability is the risk of being hurt by an external economic shock minus resilience. This approach provides an aggregate score upon which aid can purportedly be apportioned. The main limitations of the approach relate to determining the appropriate proxy indicators and the lack of timely, reliable and consistent data on small states; and considering the effects of climate change in the disaster reduction index. The Commonwealth Secretariat has gone a step further in exploring the development of a vulnerability and resilience framework that is country focused. UNDESA has been implementing a project to strengthen the capacity of SIDS to mitigate risks and reduce vulnerability. The project promotes the Vulnerability and Resilience Country Profile (VRCP) developed by UNDESA as a tool for the self-assessment of progress on the SAMOA Pathway and the Sustainable Development Goals. VRCP consists of an assessment of a country’s vulnerabilities and its capacity to cope with these vulnerabilities. The VRCP methodology is based on a systematic and participatory process that: ● builds on a baseline study that is prepared by national experts and assembles relevant disaggregated data on the thematic areas in the SAMOA Pathway; ● uses an inclusive process based on multi-stakeholder and multi-disciplinary consultations; ● provides a numerical score on a scale of 1 to 5, to assess the extent of vulnerabilities and resilience; and ● presents the scores graphically within a low-to-high range showing the vulnerability and associated resilience of each identified thematic area. The VRCP methodology involves five steps:

Step 1: selecting priority themes and major issues for each theme;

Step 2: selecting criteria for determining vulnerability and resilience of each theme;

Step 3: selecting indicators for each criterion;

Step 4: assessing and rating; and

Step 5: justifying and mapping.

It provides SIDS with a pictorial presentation of the vulnerability resilience nexus using existing information and data that can aid decision-making and serve as a practical tool for policymakers. The methodology is guided by expert and inter˗agency coordination, is country-focused, and brings together data from several sources. However, it is resource and time-intensive and must be country-driven. 216 | Research: The Platform for Innovation, Competitiveness and Growth

CDB also estimates an EVI. The design and methodological approach was guided by the work of Briguglio (1992, 1997,) and was initially computed by Crowards (2000). The CDB EVI consisted of the following six sub-indices5 and 11 proxy indicators: ● peripherality and accessibility, measured by freight and insurance costs for imports as a percentage of total imports, and provides an indication of remoteness from major economic trading partners; ● dependence upon imported energy, measured by imports, net of exports of energy (largely in the form of oil), as a percentage of total energy consumption; ● export concentration, measured as the percentage of total export receipts and accounted for by the major export and the top three exports, includes both export of goods and services and is combined with information on the openness of the economy measures as total export earnings as a percentage of GDP; ● convergence of export destination, measured in terms of the percentage of total export receipts, accounted for by the single most-important destination and the top three most-important destinations. This includes the exports of goods and services and is combined with information on the openness of the economy, that is measured as total export earnings as a percentage of GDP; ● reliance upon external finance, measured by a combination of two variables, i.e. overseas development assistance as a proportion of annual gross fixed capital formation and foreign direct investment as a proportion of annual gross fixed capital formation; and ● susceptibility to natural disasters, measured as the cumulative number of persons affected and deaths caused by natural disasters between 1950 and 1998, each as a proportion of the total population. As mentioned earlier, the EVI has a role in the allocation of CDB’s concessional financial resources. In CDB, the Special Development Fund (SDF) is the single largest source of concessionary resources. The distribution of these concessional resources is a two-stage process. Currently, access to SDF is based solely on per capita income. Only then, are concessional resources allocated using a number of metrics, including the vulnerability index score. As a result, the vulnerability index score is one of several criteria that is used to determine the size of the allocation of each country that has access to SDF. Therefore, revising and updating the Vulnerability Index is necessary as a component of CDB’s financial resource allocation framework.

NEW METHODOLOGY The CDB Vulnerability Index combines what are perceived to be the root causes of vulnerability into an aggregate composite index. It provides a static view of the vulnerability of a country at a point in time, relative to other Caribbean small states. It quantifies the extent of the exposure6 of the country to exogenous shocks and is updated using data for the years 2016 and 2017. This paper updates and revises the CDB Vulnerability Index, previously estimated by Crowards (2000) and Hartman (2011), and widens the coverage of the vulnerability analysis to include social vulnerability and a climate change component that considers not only historic natural hazard events but also predicts how the environment is likely to cope with future events. The term Multidimensional Vulnerability Index (MVI) was considered a more appropriate title for the revised index rather than EVI. The rationale for introducing the social vulnerability and climate change components of the Vulnerability Index and the proxy indicators is further discussed below.

A Compilation of Working Papers by OECS Scholars

217

MODIFICATIONS TO THE VULNERABILITY INDEX Social vulnerability, like susceptibility to natural hazards, is not a cause of vulnerability but was assessed as an important component of the Vulnerability Index because it exacerbates the downside risks of economic shocks. Social vulnerability can be defined as the inability of human units (individuals, households or families) to cope with, and recover from, stresses and shocks; to adopt and exploit changes in physical, social and economic environments; and to maintain and enhance future generations. Among the factors that may influence social vulnerability are: access to resources (knowledge, technology); political power and representation; physical disabilities; and beliefs and custom. The selection of the proxy indicators for the social vulnerability sub-index was guided by the social vulnerability index developed by St. Bernard (2007). St. Bernard (2007) proposed social vulnerability indicators including: (a) education (proportion of population with tertiary education, adult literacy rate); (b) health (life expectancy at birth); (c) security, social order and governance (murders per 100,000 population); (d) resource allocation (proportion of children and working age persons belonging to the poorest quintiles); and (e) communications architecture (computer literacy rate). The strength of this approach is that it embraced criterion of simplicity in the choice of indicators, and adopted a pioneering and strong methodological and a theoretical framework. The limitations include the paucity of social data and that the results were not subject to empirical testing. The rationale for including climate change in the susceptibility to natural hazards sub-index is because the factors contributing to environmental vulnerability are varied and, in some cases, interconnected. These include, but are not limited to: ● natural disasters: increased intensity and frequency; ● climate change: sea level rise, coastal erosion; ● oceans and seas: exploitation of marine resources; ● biodiversity: deforestation and desertification and invasion of alien species; ● water: over exploitation of surface ground and coastal water and saline intrusion; and ● waste: insufficient waste treatment. For consistency with the definition of vulnerability outlined earlier, the paper considered appropriate exogenous variables to be included in the Index (such as climate change and natural disasters), rather than endogenous variables such as resource degradation (biodiversity and waste). Additional details related to the susceptibility to natural hazards sub-index are provided in Appendix 2. Additional adjustments made to the methodology utilised in the previous CDB work of Crowards (2000) and Hartman (2011) include the: ● Peripherality and Energy Sub-Index, that was renamed the Strategic Imports Sub-Index (SISI), and the proxy indicator freight and insurance7, that was removed due to concerns regarding the veracity of its data sources and its relevance given the narrowed focus of SISI. A new proxy indicator, food, was included in the SISI measured as food imports as a percentage of total merchandise imports. As a result, the two parameters in the SISI are energy and food (as a percentage of total merchandise imports). The energy proxy indicator, which previously represented mainly crude oil, was widened to include fuels, natural and manufactured gas and lubricants and related materials, to better reflect the country’s dependence on energy and its by˗products. Strategic imports refer to essential products, that tend to be price and income inelastic, therefore the demand for such products does not decrease enough to compensate for income decreases; and 218 | Research: The Platform for Innovation, Competitiveness and Growth

● Official Development Assistance (ODA) proxy indicator that was replaced with Remittances in the External Finance Sub-Index. ODA was excluded because, while concessional flows have increased globally (including after the global economic crisis), Caribbean small states have become increasingly less successful in accessing international development assistance. Multilateral flows have approximately the same trend as bilateral flows. The BMCs received approximately 0.8% of global ODA in 2017 (of which 80% was directly to Haiti). Moreover, many Caribbean countries are classified as middle and upper-middle-income, so they do not qualify for aid, or are a low priority. SUB-INDICES, PROXY INDICATORS AND DATA SOURCES The updated and revised Vulnerability Index comprises three dimensions of vulnerability (economic, social and environmental) and six sub-indices (export concentration; concentration of export destination; dependence on strategic imports; reliance upon external finance; social vulnerability; and susceptibility to natural hazards and climate change). Within the 6 sub-indices there are 15 proxy indicators. The following paragraphs provide a summary on the sub-indices, proxy indicators, data sources and weighting used in the Vulnerability Index. ● Export Concentration relates to the dependence of the country on a few major exports (goods and services). The rationale for including export concentration is intuitive, the greater the dependence on a few major exports (goods and services) the more vulnerable that economy will be to shocks in the demand and supply of those exports. The extent of export concentration is measured by two proxy indicators, the percentage of total exports represented by the top three export categories, including tourism, and a measure of economic openness taken as the total exports of goods and services as a percentage of GDP (see Figure 1). The main data sources for the export concentration subindex were the UN Conference on Trade and Development (UNCTAD) for categories of merchandise exports, according to the Standard International Trade Classification (SITC) 2-digit codes, and total merchandise exports. The UNCTAD database was used to source the service exports data (see Appendix 3), with travel utilised as a proxy for tourism exports. ● Concentration of Export Destination occurs when a large proportion of a country’s exports are supplied to a limited number of trading partners. In this instance, the economy will be vulnerable to changing patterns of trade, economic performance and changing preferences in major trading partners. Two proxy indicators measure the extent of the concentration of export destination: (a) the proportion of total exports of goods and services converging on the top three export destinations; and (b) the proportion of total tourists from the top three source market countries. The main data sources for the concentration of export destination sub-index were the International Monetary Fund’s (IMF) Direction of Trade Statistics, Yearbook 2016. Tourism arrivals data available from the Caribbean Tourism Organisation were used to estimate the direction of trade for tourism, in the absence of a breakdown of tourism expenditure by country of origin. ● Dependence on Strategic Imports relates to the dependence of the country on critical imports, which can have direct and indirect effects on domestic production and consumption. The two proxy indicators used in this sub-index were the dependence on imported energy and food. The greater the dependence on imported energy the more susceptible the economy will be to fluctuations in international market prices of energy. Imports of energy are taken as net imports, since many fossil fuel producers will be involved in some importation, either for domestic use or for re˗exports, and are expressed as a percentage of domestic energy consumption. Further, countries that are dependent on food imports are also vulnerable to the vagaries in price and supply of international markets for their food. Ideally, the dependence on food imports should compare with the level of domestic food consumption; however, given challenges with estimating the latter, the measure of food imports as a percentage of total imports will proxy this dependence in this study. The main data sources for the dependence on strategic imports sub-index were the UN Energy Statistics Yearbook 20148 and UNCTAD for imports of goods and services and total imports.

A Compilation of Working Papers by OECS Scholars

219

● Reliance Upon External Finance relates to the dependence on external financial flows to support existing levels of consumption and investment. Investment is productive capital and is an essential ingredient in achieving a level of income that is sustainable. However, allocating resources towards investment requires forgoing some current consumption for the sake of greater consumption in the future. This allocation often is not the case in small and open economies that have relatively low levels of savings and investment. Additionally, the small size of the economy might impede the development of financial markets. Where limited financial markets restrict opportunities for reallocating resources, as with small economies, funds may be derived from external sources. These external sources include foreign direct investment (FDI), borrowing from the international private sector, and remittances. The reliance upon the external finance sub-index is measured by two proxy indicators: (a) the ratio of annual flows of FDI to GDP; and (b) the ratio of annual remittances to GDP. ● Social Vulnerability, the extent of which is measured by three proxy indicators: (a) the number of murders per 100,000 population; (b) the rate of unemployed persons in the labour force; and (c) the rate of persons living in poverty. The main data sources for the social vulnerability sub-index were the United States of America Overseas Security Advisory Centre and UN Office on Drugs and Crime (murders per 100,000, and official country statistics for poverty and unemployment). The main data sources for these proxy indicators were data on FDI (inward flows) extracted from the UNCTAD database and remittance receipts as a percent of GDP sourced from UNCTAD and the World Bank (WB). ● Natural Hazards and Climate Change assesses a country’s vulnerability to environmental factors outside of its control. Natural disasters and climate change can have catastrophic impacts, which can encompass damage to infrastructure, loss of life, injury, ill health and environmental damage. Crowards (2000) utilises the number of people affected by disasters and the number of deaths attributable to disasters as the proxy indicators in the Vulnerability Index. Notably, there are concerns about the Emergency Events Database (EM-DAT), which is the primary source of information related to the number of persons affected by natural disasters. These concerns surround the missing data for the years 1950-1998, that in part is due to the classification criteria9 for recording events in EM-DAT. Also, Crowards (2000) did not include a proxy indicator for the macroeconomic impacts of natural disasters. The rationale was mainly that the impacts on macroeconomic variables, such as income, trade and debt, are difficult to isolate from the plethora of non˗disaster related influences on the macro-economy and that the time scale of available data is limited for some countries. Following a review of the available climate vulnerability data sources, the Dara Climate Vulnerability Monitor (DCVM) was selected for use in the Vulnerability Index. DCVM was developed in 2010 and estimates the human and economic impacts of climate change and the carbon economy for 184 countries in 2010 and 2030, across 34 indicators in four impact areas (environmental disasters, habitat change, health impact, and industry stress). DCVM was updated in 2012. It uses current peer-reviewed scientific research, in-country field research, and critical input from two separate external advisory bodies. The main strength of DCVM is that it includes both the natural disaster and climate change aspects of environmental vulnerability—not only based on the past but also the near future.

220 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 1: MULTIDIMENSIONAL VULNERABILITY INDEX

The proxy indicators in the Vulnerability Index are as follows: Vulnerability = average (E3, O); (D3, T3); (Fd, Fl); (R, If); (C, U, P); (NDa, NDd; CE, CP)

C 3 = the proportion of total exports of goods and services represented by the top three export categories. O

= total exports of goods and services10 as a percentage of GDP.

D 3 = the proportion of total exports of goods converging on the top three export destinations. T 3

= the proportion of total tourists from the top three source countries.

F d

= Food imports as a percentage of total imports.

F I

= Fuel imports as a percentage of total imports.

= the ratio of annual remittances to GDP.

= the ratio of the annual flow of FDI to GDP.

= the number of intentional homicides per 100,000 population.

= the rate of unemployed persons in the labour force.

= the rate of persons living in poverty.

NDa = the number of persons affected by natural disasters, as a proportion of total population. NDd = the number of deaths resulting from natural disasters, as a proportion of total population. CE

= the economic losses or gains of climate change (Dara Climate Index).

= the economic losses or gains of carbon (Dara Climate Index). A Compilation of Working Papers by OECS Scholars

221

CONSTRUCTING AND AGGREGATING The Vulnerability Index is a composite index calculated from various sub˗indices. The composite index is estimated as the simple arithmetic average of the sub˗indices. There are four steps in constructing the index:

Step 1: determining the causes of vulnerability;

Step 2: selecting and compiling proxy indicators;

Step 3: applying normalisation methodology to data; and

Step 4: computing sub-indices and aggregating the index.

The index is created for each country by averaging across the 6 sub-indices representing 15 proxy indicators. The grouping of proxy indicators within each of the sub-indices represents the structural characteristics that can reasonably be used as a proxy for the particular measure of vulnerability. The aggregation of separate sub-indices into a single index introduces issues related to the direct comparability of these parameters, and the potential need for the scaling of the parameters to facilitate more direct comparison. Crowards (2000) examined five different methods of scaling11 (see Appendix 4). The preferred method of scaling was to apply a fixed exponent of 0.3 to the entries in each series prior to normalisation. This is the degree of exponential scaling (i.e. raising each entry to the power of 0.3) that results in medians, that are, on average, midway between the series minimum and maximum. This produces series that are relatively evenly distributed, avoiding a situation where the majority of entries are bunched together at the lower end of the scale, making the difference between them almost indiscernible. The trade-off in any such transformation is the loss of information regarding the degree to which the more extreme values exceed the rest of the values in each series. The loss of information is considered justified given that: ● a country should not be considered highly vulnerable based primarily on an extreme value in a single series, as could be the case when employing untransformed data; and ● transforming the data into relatively evenly-distributed series enables differences between the majority of countries to be reflected in the final index, which is not the case for some of the other methods of scaling. WEIGHTING After scaling the variables and applying a normalisation transformation, equal weights were applied to each of the sub-indices. A number of studies have recommended and embraced the assignment of equal weighting [Briguglio (1995); Crowards and Coulter (1998); Commonwealth Secretariat (2014); and Bernard (2007)]. The Human Development Index also assigns equal weights to all three of its dimension indices (long and healthy life, knowledge and a decent standard of living). SENSITIVITY ANALYSIS The study conducted a sensitivity analysis on the proxy indicators in the Vulnerability Index to gauge the robustness of the results. A correlation matrix was prepared and used to assess the extent of the relationship between these proxy indicators. In instances where there was evidence of correlation, the 222 | Research: The Platform for Innovation, Competitiveness and Growth

proxy indicator was either deleted from the index or combined with another proxy indicator. Further, the selection of weights for proxy indicators is another important consideration in the quantification12 of the Vulnerability Index. As a result, alternative weighting scenarios were evaluated to assess how the results of the study are affected. LIMITATIONS Critics argue that: the vulnerability index has too much subjectivity; access to high-quality data is a challenge; justification for weighting should be strengthened; and none of the variables can be tested for their relationship with economic vulnerability. The research paper aimed to address these limitations in several ways. With respect to the subjectivity in the selection of proxy indicators13, there is a growing body of literature on this topic. Some of this literature was identified in Section 2 (Review of Vulnerability Metrics) and is aligned with the work of Crowards (2000). More broadly, the selection of proxy indicators was guided by criteria identified by Briguglio (1997), including relevance, simplicity, transparency and reproducibility. Further, the paper conducted sensitivity analysis of the proxy indicators. Care was taken to use proxy indicators with a wide coverage of countries in the event that the study needed to be expanded. All of the data utilised in the study was sourced from well-established international, regional and country sources, such as the central banks. The MVI in this paper focused specifically on the BMCs to better understand the inherent characteristics that make them vulnerable, and how this has evolved over time. However, there is scope for expanding this study, particularly in the coverage of countries to facilitate more direct comparison with earlier CDB studies14. It should be noted that it is well established in the literature that vulnerability scores tend to be higher for smaller countries than larger countries. For example, Cordina (2008) revealed that seven out of the eight vulnerability indices reviewed had statistically-significant positive correlation coefficients between country size and vulnerability scores, implying that, in general, the indices tend to agree that small countries are more economically vulnerable than larger ones. Also, Crowards (2000), indicated that SIDS are more vulnerable than other larger country groupings due to their inherent dependence on energy imports, high level of export concentration, and exposure to environmental hazards such as climate change, among other reasons. Notwithstanding, the Commonwealth Secretariat (2018) noted that there remains a need to build international consensus on defining and measuring economic vulnerability. In this regard, they have proposed the development of a new universal economic vulnerability index to focus on the economic, environmental and socio-political causes of vulnerability. An obvious challenge of widening the sphere of the Vulnerability Index (to include the social and natural hazard and climate change dimensions) is that it introduces further complexity in the construction of the index. This complexity is related to the selection of appropriate proxy indicators given the paucity of social and climate change data. Indicators are selected based on availability and reliability, as well as suitability. For example, it is difficult to source appropriate proxy indicators to track susceptibility to natural disasters and climate change, and social vulnerability. Crowards (2000) noted that there are significant hurdles to deriving a suitable measure to natural disasters. The main difficulty being that assessing the vulnerability to natural disaster and climate change requires predicting the likelihood of events occurring in the near future and their degree of impact, which may not be done correctly through analysing historic frequencies of events and the estimates of their magnitudes. Relatedly, it can be argued that while social factors15 influence economic vulnerability, identifying suitable proxy indicators is challenging given the data limitations and that some of these social factors may also be policy induced. The Turks and Caicos Islands and the Virgin Islands were excluded from the computation of the index. A Compilation of Working Papers by OECS Scholars

223

The Virgin Islands does not compile trade data and, therefore, no data was available for the Export Destination and Export Concentration sub-indices. Trade data by SITC classification and disaggregated by the main trading partner was not available for the Turks and Caicos Islands, therefore no data was available to compile the Export Destination and Export Concentration sub˗indices. Data gaps in other Caribbean states were overcome in other instances by extrapolating trends from historical data, and in others by leveraging existing data from similar countries to create estimates. The Vulnerability Index provides a static view of vulnerability at a point in time. As a result, the relative ranking would be sensitive to the year the Index was calculated. Concerns have been raised about the potential for wide variations in the index scores, particularly in years where there are economic or natural disaster shocks. Whilst the utility of measuring vulnerability in a country at a point in time is clear, the possibility of large changes in the Index can be problematic for some policy applications. As a result, a dual approach for calculating the Index is proposed that involves a point-in-time measure and a moving average (at least three years) measure. The moving average will reduce the potential for large changes in the index and can be particularly useful to smooth the effects on the data when there are outlier years that may be beneficial in some policy applications of the Index. While the moving average approach was not calculated in this paper, it will be included in subsequent revisions of the Index.

RESULTS The study provides estimates of the MVI for BMCs (with the exception of the Turks and Caicos Islands and Virgin Islands) for the years 2016 and 2017. The results show that the vulnerability of BMCs is concentrated in the areas of: ● the extent of trade openness with other countries; ● dependence on a few major exports and trading partners; ● dependence on the imports of energy and related products; ● social challenges such as crime; and ● exposure to natural hazards and climate change (number of persons affected) (see Appendix 5). The results of the estimation also support the view that, in spite of the high human development status of BMCs, most are within the medium-high classification of vulnerability to external shocks. To determine this, the study utilised the Common Vulnerability Scoring System16 (CVSS) as a classification criterion for the Vulnerability Index for each country. This classification criteria were adjusted to align the lower threshold for vulnerability scores (0.332) with that utilised by the Commonwealth Secretariat 17. The vulnerability scoring system utilised in the study is: 0 to 0.33 as low vulnerability; 0.34 to 0.49 as mediumlow vulnerability; 0.50 to 0.69 medium-high vulnerability; and 0.70 to 1.00 as high vulnerability. Based on the number of countries that had scores in the medium vulnerability category, two tiers were added to the classification system: (a) medium (low) vulnerability comprising index scores of 0.34 to 0.49; and (b) medium (high) vulnerability comprising index scores of 0.50 to 0.69. The classification of BMCs in the vulnerability scoring system is illustrated in Table 1. Tourism-based economies appear to be more vulnerable than commodity˗based economies (see Table 1). In part, the low vulnerability classification of commodity˗based economies is linked to them being net exporters18 of crude oil (e.g. Trinidad and Tobago and Suriname) and their geographic location that, in the past, resulted in the lower likelihood of being impacted by the Region’s more severe natural disasters (particularly hurricanes). 224 | Research: The Platform for Innovation, Competitiveness and Growth

A Compilation of Working Papers by OECS Scholars

225

TABLE 1: MULTIDIMENSIONAL VULNERABILITY INDEX SCORING SYSTEM 2017 Country19

High Vulnerability (0.70 to 1.00)

Med-High Vulnerability (0.50 to 0.69)

Med-Low Vulnerability (0.34 to 0.49)

2016 Low

High Vulnerability (0.70 to 1.00)

Med-High Vulnerability (0.50 to 0.69)

ANG

0.54

0.52

ANT

0.54

0.50

BAH

0.57

0.52 0.48

BAR BZE

0.47

0.48

0.49

DOM

0.54

GRE

0.58

0.52

GUY

0.56

0.58

HAI JAM

Low Vulnerability (0 to 0.33)

0.59

0.60

CAY

Med-Low Vulnerability (0.34 to 0.49)

0.48

0.69

0.71

0.61

MON

0.60 0.42

0.43

SKN

0.55

0.52

SLU

0.63

SVG

0.52

0.54

SUR

0.47

0.34

0.43 0.31

Source: Author’s calculations

Dependence on the imports of crude oil and refined crude oil products is a major source of vulnerability in BMCs. The dependence of BMCs on strategic imports, and particularly energy, is clearly visible in the following scatter plot (Figure 2) that illustrates on a Cartesian plane the cost of domestic electricity tariffs per kilowatt hour with the vulnerability of the country to the strategic imports index, and particularly of energy and its by-products. With the exception of Trinidad and Tobago, Suriname and Belize to a lesser extent, all of the other BMCs have very high energy costs due to their dependence on imported fuels.

226 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 2: ENERGY IMPORTS SUB-INDEX AND DOMESTIC ELECTRICITY COSTS

Source: Author’s calculations

Being rich in natural resources has allowed some countries to accumulate large financial assets abroad, and enabled them to invest in schools, hospitals, and roads to promote growth and diversification. However, resource-based economies have challenges. The reality is that these resources will eventually be depleted and, in the absence of a more diversified economic base, the country would have to implement painful corrective measures to fiscal and monetary policies to achieve macroeconomic stability and growth. Resource-based economies also tend to be associated with higher levels of inequality. Research by Berry (2006), Buccellato and Mickiewicz (2008), El-Katiri et al. (2011), and Freije (2006), supports this point. In part, this may be due to the nature of resource-based economies and particularly the petroleum industry that is capital intensive and employs a very small, but highly paid, proportion of the labour force, which tends to increase income inequality within the society. The average vulnerability index score of BMCs was 0.54 in 2017, and 0.52 in 2016 (medium-high vulnerability, see Figure 3). The evolution of the vulnerability score over time can provide important signals about the changing productive and export structure of a country, its competitiveness in international markets, the level of diversification in its energy mix, the extent of the connect with its diaspora, challenges related to social cohesion, and susceptibility to the impacts of natural disasters and climate change. Among the factors that contributed to the increase in vulnerability score in 2017 are: the rise in the number of murders; increased dependence on energy-related imports; the impact of natural hazards, particularly hurricanes on the number of persons affected in countries; higher remittances (related in part to countries affected by Hurricanes Irma and Maria); and also, economic shocks and higher FDI, that may also be related to natural hazard and economic factors in BMCs. The Vulnerability Index in this paper focused specifically on BMCs to better understand the inherent characteristics that make them vulnerable, and how this has evolved over time. The Vulnerability Index can be expanded to include other classifications of countries (such as small states, non-small states and large states). A Compilation of Working Papers by OECS Scholars

227

FIGURE 3: MULTIDIMENSIONAL VULNERABILITY INDEX—BMCS 2016 AND 2017

Source: Author’s calculations

A detailed review of the MVI sub-indices provides insights into what were the factors that contributed to the changes in the index score in each country between 2016 and 2017. For example, in Haiti the increase in the index score from 0.68 in 2016, to 0.71 in 2017, was mainly due to higher inflows of FDI that increased to 4.4% of GDP in 2017, compared with 1.4% of GDP in 2016, and the increased prevalence of crime (murders per 100,000). In Dominica, the MVI index score rose from 0.49 in 2016, to 0.54 in 2017. The increase can be attributed to the country’s susceptibility to natural hazards and, in particular, the devastating impact of Hurricane Maria in 2017; the increased prevalence of criminal activity; and increased dependence on remittances. The increases in these sub-indices in Dominica were able to more than offset declines in other sub-indices such as export concentration. In Trinidad and Tobago, the MVI index score increased from 0.30 in 2016, to 0.34 in 2017. The increased prevalence of crime (murders per 100,000), and higher dependence on remittances and FDI contributed to the rise in the Index and more than offset a slight decline in the dependence on food imports. SENSITIVITY ANALYSIS A correlation matrix (see Table 2) was utilised to assess the extent of relationship between the proxy indicators that were included in the Vulnerability Index. In instances where there was evidence of correlation, the proxy indicator was either deleted from the Index or combined with another proxy indicator, as outlined below: ● The Export Concentration parameter: C1 (exports of goods and services—top category) and C33 (exports of goods and services—top three categories) variables were correlated, and the C1 proxy indicator was removed. ● The Export Destination parameter: the proportion of total exports of goods converging on the top one export destination (D1) was correlated with the proportion of goods converging on the top three export destinations (D3), therefore the D1 proxy indicator was removed. Similarly, the variables capturing the proportion of tourists from the top source location (T 1) and the top three source locations (T3) were correlated, therefore the T1 indicator was removed. Table 2 highlights the results of the correlation matrix after the adjustments highlighted. 228 | Research: The Platform for Innovation, Competitiveness and Growth

The selection of weights for proxy indicators is an important consideration in the quantification of the Vulnerability Index. As a result, alternative weighting scenarios were evaluated to assess how they affect the results of the study. For simplicity, the paper utilised three weighting scenarios (see Table 3). The first weighting scenario (which represents the results presented in the paper) and is the same in principle as that utilised by Crowards (2000), is equal weighting by sub-index. It assumes that each of the six broad dimensions of vulnerability are equal influencers of a country’s overall vulnerability. Therefore, in forming the composite EVI, each parameter was given an equal weight of 16.66%. The second weighting scenario (equal weights by proxy indicator) is similar to the first in that it also utilises the concept of equal weighting, but the unit of emphasis changes to the proxy indicator as compared to the broad dimension of vulnerability. In this weighting scenario, the broad dimension of vulnerability with more proxy indicators (such as natural hazards and climate change and social vulnerability) would receive a larger overall weight than sub-indices with a lower number of proxy indicators. This resulted in the natural hazards and climate change and social vulnerability dimensions of vulnerability receiving weights of 26.7 and 20, respectively, with the other dimensions of vulnerability (export concentration, export destination, strategic imports, and external finance) receiving weights of 13.3, respectively. TABLE 2: CORRELATION MATRIX (2017)

C3 O D

F(F)

F(R)

NDa

NDd

1.00

-0.02

0.21

0.43

0.06

0.18

-0.26

0.14

-0.19

0.54

0.11

0.12

0.21

-0.03

0.02 -0.38 -0.17

0.11

-0.52 -0.23

-0.19

-0.16

-0.48 -0.13

0.20

0.38

0.06

1.00 1.00

0.13

0.01

-0.04

0.29

0.00

-0.31

-0.01

-0.18

0.40

0.33

1.00 -0.13

0.06

0.28

-0.12 -0.11

-0.15

-0.36

0.04

-0.12 -0.14

1.00

-0.04 -0.60

0.69

0.15

0.41

0.78

0.66

0.13

-0.05

0.12

1.00

0.16

0.35

0.00

0.50

-0.10

0.02

0.41

-0.14

0.13

1.00

-0.44 -0.39

-0.30

-0.69

-0.06

0.24

-0.38

0.09

1.00

0.24

0.46

0.54

0.60

0.33

-0.06

1.00

-0.22

0.06

-0.32 -0.38 -0.24

1.00

0.42

0.31

0.57

-0.12

1.00

0.39

0.00

0.44

1.00

0.55

-0.13

0.16

1.00

0.09

0.46

1.00

0.06

Fl F(F) F(R)

C U P NDa ND

CE CP

0.35

1.00

Source: Author’s calculations

The third weighting scenario also assumes equal weighting but eliminates the social vulnerability dimension. This enables the research to assess how the removal of this sub˗index would affect the overall index score. As a result, the weight for the five remaining dimensions changed to 20, respectively. The weight assigned to each proxy indicator and the six dimensions of vulnerability in the three weighting scenarios are displayed in Table 3. While in each of the scenarios the economic dimension of the Index is weighted the most, the weight ranges from 53.2% to 80%, the weight of the social dimension ranges from 17% to 20%, and the weight of the environmental dimension ranges from 17% to 26.8%. A Compilation of Working Papers by OECS Scholars

229

TABLE 3: ALTERNATIVE WEIGHTING SCENARIOS (Scenario 1) Base Scenario—equal weights by sub-index Category Parameter

Variable

Economic (66%)

Social (17%)

Environmental (17%)

Export Concentration

Export Destination

Strategic Imports

External Finance

Social Vulnerability

Climate and Natural Hazards

16.67%

8.3%

F(F)

F(R)

8.3% 8.3% 8.3% 8.3% 8.3% 8.3%

5.6%

NDa

NDd

5.6% 4.2% 4.2% 4.2%

(Scenario 2) Alternative Scenario—equal weights by proxy indicator Category Parameter

Variable

Economic (53.2%)

Social (20%)

Environmental (26.8%)

Export Concentration

Export Destination

Strategic Imports

External Finance

Social Vulnerability

Climate and Natural Hazards

13.3%

20.0%

26.7%

6.7%

F(F)

F(R)

6.7% 6.7% 6.7% 6.7% 6.7% 6.7%

6.7%

NDa

NDd

6.7% 6.7% 6.7% 6.7%

(Scenario 3) Eliminate Social Vulnerability sub-index Category Parameter

Variable

Economic (80%)

Social (0%)

Environmental (20%)

Export Concentration

Export Destination

Strategic Imports

External Finance

Social Vulnerability

Climate and Natural Hazards

20%

F(F)

F(R)

NDa

NDd

10%

Source: Author’s calculations

The results of the weighting sensitivity test generally validated the results of the Vulnerability Index (see Table 4). Notably, there were not significant changes in the Vulnerability Index score in the alternative weighting scenarios in BMCs. Haiti and Saint Lucia remained the first and second ranked countries in terms of vulnerability, whilst Trinidad and Tobago remained the least-vulnerable country across all three weighting scenarios. Small differences in the index score are likely due to the variances in weights across scenarios and data limitations with the climate and natural hazards metrics. This provides further validation for the Scenario 1 methodology, as it enables better smoothing across missing data.

230 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 4: ILLUSTRATION OF ALTERNATIVE WEIGHTING SCENARIOS ON THE MULTIDIMENSIONAL VULNERABILITY INDEX (2017)

Weighting Scenario 1

Weighting Scenario 2

Weighting Scenario 3

Index Score

Rank

Index Score

Rank

Index Score

Rank

ANG

0.54

0.53

0.57

ANT

0.54

0.53

BAH

0.57

0.53

0.58

BAR

0.48

0.43

0.49

BZE

0.59

0.55

0.57

CAY

0.48

0.45

0.56

DOM

0.54

0.56

0.52

GRE

0.58

0.56

GUY

0.56

0.57

0.55

HAI

0.71

0.69

0.73

JAM

0.61

0.56

0.59

MON

0.42

0.45

0.40

SKN

0.55

0.53

SLU

0.63

0.59

0.62

SVG

0.52

0.51

0.46

SUR

0.47

0.46

0.45

0.34

0.31

AVERAGE

0.54

Country

0.52

0.53

Source: Author’s calculations

VULNERABILITY BY CATEGORY Table 5 gives a snapshot of economic, social and environmental vulnerabilities of each BMC. These disaggregated tables provide a baseline and highlights the resilience building priorities for each BMC. For example, building resilience in Anguilla should place the highest priority on reducing economic vulnerability. Meanwhile, in Saint Vincent and the Grenadines and Grenada, the highest priority should be on building social and environmental resilience, respectively. These disaggregated tables offer a useful guide for policy discussion in country about how to build resilience at the national level and how to allocate scarce development resources taking into account vulnerabilities.

A Compilation of Working Papers by OECS Scholars

231

TABLE 5: ECONOMIC, SOCIAL AND ENVIRONMENTAL VULNERABILITY SUB-INDICES

Economic Country

2016

Social

2017

2016

Environmental 2017

2016

2017

Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank ANG

0.55

0.58

0.42

0.39

0.00

0.55

ANT

0.52

0.55

0.50

0.60

0.41

0.42

BAH

0.57

0.65

0.51

0.52

0.32

0.30

BAR

0.58

0.43

0.45

0.11

0.10

BZE

0.63

0.65

0.71

0.32

0.30

CAY

0.61

0.60

0.05

0.07

0.45

0.40

DOM

0.45

0.48

0.54

0.62

0.56

0.66

GRE

0.51

0.60

0.69

0.41

0.39

GUY

0.56

0.55

0.64

0.61

0.59

0.56

HAI

0.73

0.77

0.64

0.60

0.59

JAM

0.67

0.69

0.70

0.68

0.22

0.20

MON

0.42

0.41

0.62

0.26

0.24

SKN

0.49

0.53

0.61

0.00

0.42

SLU

0.63

0.69

0.79

0.68

0.46

0.34

SVG

0.53

0.50

0.73

0.78

0.41

0.33

SUR

0.39

0.47

0.60

0.56

0.39

0.35

0.34

0.38

0.46

0.51

0.00

AVERAGE 0.54

0.57

0.32

0.36

Source: Author’s calculations *Key

High Vulnerability

Med-High Vulnerability

Med-Low Vulnerability

Low Vulnerability

DISCUSSION AND CONCLUSION This Paper updates and revises CDB’s Vulnerability Index previously estimated by Cowards (2000) and Hartman (2011), and widens the scope to include social vulnerability and climate change. The Vulnerability Index is important as it is a component of CDB’s financial resource allocation framework and has the potential to assist in the more effective allocation of scarce concessional resources. The Vulnerability Index also provides a solid basis to develop more effective strategies to build resilience and foster the economic growth that is needed. This work is also timely, as it contributes to the international and regional dialogue about increased access to development finance for small states to help them address their vulnerabilities.

232 | Research: The Platform for Innovation, Competitiveness and Growth

The paper seeks to quantify and gain deeper understanding of the relative vulnerabilities of BMCs and proposes a framework to inform the preparation of a resilience index. The paper also seeks to:

(a) answer the question of how vulnerable BMCs are;

(b) identifies where these vulnerabilities are concentrated; and

Among the main findings are that BMCs, on average, can be considered middle-to-high vulnerability countries with an average vulnerability index score of 0.54 for 2017, slightly above the score of 0.51 for 2016. The vulnerabilities are concentrated in the areas of dependence on a few major export products and trading partners, dependence on the imports of energy and related products, social challenges such as crime, and exposure to natural hazards and climate change. These results are not surprising since there is a need within the Region to improve competitiveness and strengthen small and micro enterprises to increase non-traditional exports. Further, fossil fuels account for 95% of the Region’s energy needs. Significant volatility in energy prices has contributed, in large part, to the competitiveness challenges that many Caribbean industries face, and one of the major development challenges now threatening the development agenda within the Caribbean Region is the thorny issue of increasing crime and violence. In the natural environment, 2017 was one of the most destructive hurricane seasons for the Region with two major hurricanes—Irma and Maria—causing over USD100 billion in damage and loss across several Caribbean countries. These were a stark reminder that the Region, ranked as the second highest in terms of climate vulnerability, faces a future characterised by more intense and destructive meteorological systems and, possibly, more frequent and intense natural hazards. These findings give credence to the emphasis of CDB on the greater use of renewable energy and energy efficiency, disaster risk management, and resilience building. The findings corroborate with Crowards (2000), and the identified areas of vulnerability align with those identified by the IMF study on strengthening growth and boosting resilience prepared by Alleyne et. al. (2017). Although resource-rich countries had lower vulnerability scores, the paper notes that these economies have unique challenges related to sharp boom and bust cycles and higher levels of inequality, among others. The paper, therefore, makes the following policy recommendations: ● There is need for sustained and deeper effort to reduce the dependence of Caribbean economies on hydrocarbons and intensify the utilisation of more sustainable fuel sources. This can involve a deeper investigation into the impact on government revenues, private sector interest and the social outcomes of higher dependence on renewable energy sources compared with hydrocarbon resources, and strategies for transitioning to renewable energy sources. ● There may be the need for a regional approach to address vulnerability and, in particular, to close the financing gap caused by natural disasters given the increased intensity and the insufficient finances currently available. The regional approach can include a disaster contingency fund, which will seek to provide immediate liquidity to the affected countries and will complement other ex˗post financing, including domestic and external credit, budget reallocation, donor assistance and relief, and parametric insurance. ● Regional sovereigns can also explore the potential of state contingent debt instruments, e.g. GDPlinked bonds or disaster-linked bonds. These instruments can assist in building resilience in the face of shocks by providing an option for a moratorium20 on the payment of interest and principal in the event of a natural hazard of a predetermined magnitude. Sovereigns in the Caribbean may need capacity building to design and negotiate these instruments as they are new to the Region.

A Compilation of Working Papers by OECS Scholars

233

● There is the potential for the MVI to be used by BMC’s to assist in determining and or justifying their development priorities within an evidence-based framework. For CDB, the MVI can play a greater role in its policy advisory and financing frameworks. The possibilities include providing information on the relative vulnerability profile of BMCs over time, serving as a gauge or assessing the effectiveness of the policies implemented by countries. Additionally, this information can be utilised to design new and innovative financial instruments, and source targeted funding (economic, social, and environmental) to assist with building resilience. ● The MVI shows that although many of the BMC’s are classified as middle-income countries, economic security of these countries is highly vulnerable due to small size, economic and social structures, the high annual probability of many individuals being affected by natural disasters and, to a lesser extent, individuals losing their lives as a result of natural disasters. The individual country vulnerability scores for 2016 and 2017 (the change from year-to-year), show that when an event occurs vulnerability can change significantly from one period to the next, highlighting the economic insecurity of the Region. This analysis shows that ex-ante resilience building should be an urgent requirement in all BMCs.

NOTES 1

A Policy Blueprint for Caribbean Economies.

A UN policy document in 1994 on sustainable development and vulnerability in SIDS.

While lower-income small states are eligible for concessional funds to address vulnerability, upper-middle and middle-income small states—that equally require development finance to mitigate their vulnerability—are not entitled to such resources. Multilateral development banks consider these countries to have less need for concessional finance because their higher per capita income level theoretically allows them to mobilise domestic and international capital. 3

Birkmann (2005) highlights the various dimensions of vulnerability, that begins with vulnerability as an internal risk factor (intrinsic vulnerability), which can gradually be widened to vulnerability as the likelihood to experience harm (human-centred). Vulnerability could also be conceived as a dualistic approach of susceptibility and coping capacity; it can be further widened as a multiple structure that considers susceptibility, coping capacity and exposure, adaptive capacity and, ultimately, vulnerability can be considered in a multidimensional context encompassing, physical, social, economic, environmental and institutional features. 4

The Peripherality and Energy Import Dependence sub-indices were combined into a single sub-index. This partly reflects their focus on imports, but also serves to implicitly reduce the emphasis placed on each sub˗index due to data limitations related to the proxy indicators used for the relative cost of importation. 5

The Vulnerability Index focuses on a country’s exposure to exogenous shocks, and not the probability or intensity of risk from those shocks. 6

The peripherality parameter was proxied by freight and insurance costs for imports as a percentage of total import costs. However, Crowards (2000) notes that the accuracy of the data and the conflicting data sets that are available cast some doubt on the legitimacy of deriving an index based on such figures (freight and insurance costs). 7

UN Energy Statistics Yearbook (last published in 2014) provides oil import and export estimates for BMCs in metric tonnes. Net oil is converted to gallon units for comparison with domestic oil consumption data. Estimates are calculated after 2014. 8

EM-DAT includes information on natural disasters during the period 1900 to present, that satisfy the following criteria: (a) 10 or more dead; (b) 100 or more affected; (c) the declaration of a state of emergency; and (d) a call for international assistance. 9

234 | Research: The Platform for Innovation, Competitiveness and Growth

Includes tourism and financial services.

Crowards (2000) examined the following methods for scaling: (a) normalisation; (b) condensed-decile normalisation; (c) the Borda Rule; (d) fixed exponential scaling; and (e) variable exponential scaling. 11

EVI is a composite index, i.e. a weighted mean of multiple proxy indicators. The process of weighting variables based on perceived importance introduces subjectivity to the analysis, therefore, these weights were the focus of initial sensitivity testing. 12

Principal Component Analysis (PCA) is a frequently utilised technique in the selection of proxy Indicators. However, it is a requirement of the PCA method that the constituent variables be positively correlated, so that variations in each series contribute to the cumulative variation in the overall index. Given that a correlation matrix was utilised in the range of approaches to guide the selection of proxy indicators, it significantly reduced the extent of correlation in the remaining proxy indicators and, as such, the utility of the PCA method. 13

Crowards (2000) included 136 countries in the study, however, sufficient data was available for 95 countries for the calculation of EVI. 14

Cutter, Boruff and Shirley’s (2003) identified the following factors that influence social vulnerability: lack of access to resources (including knowledge and technology); limited access to political power and representation; social capital (including social networks and connections); beliefs and customs; building stock and age; frail and physically-limited individuals; and type and density of infrastructure and lifelines. 15

CVSS provides a way to capture the principal characteristics of a vulnerability and produce a numerical score reflecting its severity. The numerical score can then be translated into a qualitative representation (such as low, medium, high, and critical) to help organisations properly assess and prioritise their vulnerability management processes. 16

Commonwealth Secretariat (2014) noted that the threshold between high and low vulnerability scores was set at 0.332. Crowards (2000) identified vulnerability as high when the index score exceeded 0.54 (the 60th percentile). 17

Suriname recently improved its crude oil refining capacity.

Cf. Appendix 1

The trigger for the moratorium option usually is linked to indicators such as GDP, the magnitude of the natural hazard, and commodity prices. 20

Climate change impacts on the natural environment can be seen as chains of causation with first, second and third-order effects. The first order effects of climate change indicate rising temperatures and shifting precipitation patterns. The second order effects include rising sea levels and increasing temperatures. These rising sea levels can lead to flooding, erosion damage and altered ecosystem distribution, which we may refer to as the third order effect of climate change. Climate change impacts increase the intensity and frequency of the occurrence of natural hazards on the environment. 21

A Compilation of Working Papers by OECS Scholars

235

REFERENCES Acevedo, S. (2016). “Gone with the Wind: Estimating Hurricane Climate Change Costs in the Caribbean, International Monetary Fund, Working Paper No. 16/199. Adepoju, A. A., Salau, A. S. and Obayelu, A. E. (2007). “The Effects of External Debt Management on Sustainable Economic Growth and Development: Lessons from Nigeria”, Munich Personal RePEC Achieve (MPRA), Paper No. 2147. Adger, W. N. (2006). Vulnerability. Global Environmental Change, vol. 16, No. 3, pp. 268-281. Adrianto, L., and Matsuda, Y. (2004). “Study on Assessing Economic Vulnerabilities of Small Island Regions”. Environment, Development and Sustainability, September 2004; vol. 6, Issue 3, pp 317-336. Alleyne, T., Ötker, I., Ramakrishnan, U., and Srinivasan, K. (2018). “Unleashing growth and strengthening resilience in the Caribbean, International Monetary Fund, Washington D.C, ISBN: 9781484315194. Atkinson, G., Dubourg, R., Hamilton, K., Munasinghe, M., Pearce, D. and Young, C. (1997). “Measuring Sustainable Development, Macroeconomics and the Environment, Edward-Elgar Publisher, Cheltenham, UK. Atkins, J. Mazzi, S. and Easter, C. (1998). “A Study on the Vulnerability of Developing and Island States: A Composite Index”, Commonwealth Secretariat, London. Auguste, S., and Cornejo, M. (2015). “Vulnerability of Small Island Economies, the case of the Caribbean”, Universidad Torcuato di Tella, Buenos Aires. Berry, A. 2006. “Employment and Income distribution experiences of minerals exporters and of countries achieving growth acceleration”. Human Sciences Research Council, Toronto. Binger, Al. (2002). “Vulnerability and Small Island States”, UNDP Policy Journal, Volume 1, 2002. Briguglio, L. (1992). Preliminary Study on the Construction of an Index for Ranking Countries according to their Economic Vulnerability, UNCTAD/LDC/Misc.4, Geneva. Briguglio, L. (1993). ‘The Economic Vulnerabilities of Small Island Developing States”, study by CARICOM for the Regional Technical Meeting of the Global Conference on the Sustainable Development of Small Island Developing States, Port-of-Spain, Trinidad and Tobago, July, 1993. Briguglio, L. (1995). ‘Small Island States and their Economic Vulnerabilities’, World Development, vol. 23, pp 1615–1632. Briguglio, L. (1997). “Alternative Economic Vulnerability Indices for Developing Countries”. Report prepared for the United Nations Department of Economic and Social Affairs, December 1997. Briguglio, L. (2000). “An Economic Vulnerability Index and Small Island Developing States, Recent Literatures. Working Paper, Kagoshima University Pacific Islands Studies Center, Kagoshima, November 29, 2000. Briguglio, L. (2004). “Economic vulnerability and resilience: concept and measurements", in Briguglio, L and EJ Kisanga (Eds.), Vulnerability and Resilience of Small States, Commonwealth Secretariat and the University of Malta, Formatek, Malta, 43-53. Bruckner, M. (2012). “Climate Change vulnerability and the identification of least developed countries”, United Nations, Department of Economic and Social Affairs, CDP Background Paper No. 15, ST/ESA/2012/CDP/15. Buccellato, T and T. Mickiewicz. (2008). “Oil and Gas; a blessing for few, Hydrocarbons and within-region inequality in Russia”. Centre for the study of Economic and Social Change in Europe, Economics working paper No. 80.

236 | Research: The Platform for Innovation, Competitiveness and Growth

Chambers, Robert and Gordon R. Conway (1992). “Sustainable Rural Livelihoods: Practical Concepts for the 21st Century”, Institute of Development Studies Discussion Paper, Issue 296. Chander, R. (1999). “Measurement of the Vulnerability of Small States”, Report prepared for the Commonwealth Secretariat, London. Commonwealth Secretariat (2018). “Background Paper: Building Consensus on Defining and Measuring Vulnerability”, Presented at the Commonwealth Secretariat Ministers Meeting, Bali, Indonesia, October 10, 2018. Commonwealth Secretariat (2018). “Background Paper: Defining and Measuring Economic Vulnerability”, Presented at the Commonwealth Secretariat Ministers Meeting, Bali, Indonesia, October 10, 2018. Cordina, G. (2004). “Economic vulnerability and economic growth: some results from a neo-classical growth modelling approach”, Journal of Economic Development, Volume 29, No. 2, pp 21-40. Crowards, T., and Coulter, W. (1998). “Economic Vulnerability in the Developing World with Special Reference to the Caribbean”, Annual Review Seminar, Research Department, Central Bank of Barbados. Crowards, T. (1999), “An Economic Vulnerability Index for Developing Countries with Special Reference to the Caribbean (alternative methodologies and provisional results)”, A Summary Draft, Social and Economic Research Unit, Economics Department, Caribbean Development Bank, Barbados. Crowards, T. (2000). “An Index of Inherent Economic Vulnerability for Developing Countries”, Staff Working Paper No. 4/00, Social and Economic Research Unit, Economics Department, Caribbean Development Bank, Barbados. Cutter, Susan L., Bryan J. Boruff and W. Lynn Shirley (2003). “Social Vulnerability to Environmental Hazards”. Social Science Quarterly vol. 84(2): pp 242-61. Easter, C. (1998). ‘Small states and development: a composite index of vulnerability’, Small States: Economic Review and Basic Statistics, vol. 24-46. ECLAC (2011). “Study on the vulnerability and Resilience of Caribbean Small Island Developing States”, Economic Commission of Latin America and the Caribbean, publication no. 2011 LC/ CAR ECLAC/ L. 354. El-Katiri, L., B. Fattouh and P. Segal 2011. “Anatomy of an oil-based welfare state: Rent distribution in Kuwait”, LSE Centre for the study of Global governance. No. 13. Freije, S. 2006. “Income distribution and redistribution in an oil-rich economy: the case of Venezuela”. www.cid. harvard.edu/events/papers/0604caf/Freije.pdf. Füssel, H.M. (2010). “How inequitable is the global distribution of responsibility, capability, and vulnerability to climate change: a comprehensive indicator-based assessment”, Global Environmental Change, vol. 20, no. 4, pp. 597-611. Guillaumount, P. (2008). “An Economic Vulnerability Index: Its Design and Use for International Development Policy.” United Nations University World Institute for Development Economics Research, Research Paper No. 2008/99. Guillaumont, P. (2011). “The concept of structural economic vulnerability and its relevance for the identification of the Least Developed Countries and other purposes”, CDP Background paper no. 12, ST/ESA/2011/CDP/12. Nijkamp, P. and Vreeker, R. (2000). “Sustainability assessment of development scenarios: methodology and application to Thailand”, Ecological Economics vol. 33, pp 7-27. Pantin, D. (1994). “The Economics of Sustainable Development in Small Caribbean Islands”, University of the West Indies, St. Augustine, Trinidad and Tobago, Center for Environment and Development.

A Compilation of Working Papers by OECS Scholars

237

Ram, J., Frederick, R., Ramrattan, D., Hope, K. and Elliott, W. (2018). “A Blueprint for Caribbean Economies”, Caribbean Development Bank, Working Paper No: CDB/WP/18/01. Turvey, R. (2007). ‘Vulnerability assessment of developing countries: the case of small island developing states’, Development Policy Review, vol. 25: no. 2 pp 243-264. Wang, CC (2013). “Reconsidering the Economic Vulnerability Index of the United Nations”, Canadian Journal of Development Studies, vol 34(4), pp 553-68. Wells, J. (1996). “Composite Vulnerability Index: A Preliminary Report”, Commonwealth Secretariat, London. Wells, J. (1997). “Composite Vulnerability Index: A Revised Report, Commonwealth Secretariat, London. St. Bernard, G. (2004). “Toward the Construction of a Social Vulnerability Index: Some Theoretical and Methodological Considerations”, Journal of Social and Economic Studies, vol. 53. no. 2, pp. 1-29. Gonzales, A (2000). “Policy Implications of Smallness as a Factor in the Lome FTAA and WTO Negotiations, Final Report, Caribbean Regional Negotiating Machinery Technical Paper (RNM/IDB Project No. ATN/JF/SF-6158-RG), Kingston, Jamaica. IPCC (2014). “Climate Change, IPCC Fifth Assessment Report Working Group II (Impacts, Adaptation and Vulnerability), available at https://ar5-syr.ipcc.ch/ipcc/ipcc/resources/pdf/IPCC_SynthesisReport.pdf. Commonwealth Secretariat (2014). “Building the Resilience of Small States: A Revised Framework”, ISBN (e-book) 978-1-84859-918-5. Hartman, M. (2011). “Revised Economic Vulnerability Index, 2008”, Presented at the Caribbean Development Bank President’s Discussion Series, December 15, 2011, Barbados.

238 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX APPENDIX 1: CARIBBEAN DEVELOPMENT BANK BORROWING COUNTRIES Country

Acronym in Tables

Anguilla

ANG

Antigua and Barbuda

ANT

Bahamas, The

BAH

Barbados

BAR

Belize

BZE

Cayman Islands

CAY

Dominica

DOM

Grenada

GRE

Guyana

GUY

Haiti

HAI

Jamaica

JAM

Montserrat

MON

St. Kitts and Nevis

SKN

Saint Lucia

SLU

St. Vincent and the Grenadines

SVG

Suriname

SUR

Trinidad and Tobago

Turks and Caicos Islands*

TCI*

Virgin Islands*

VI*

(*Excluded from the computation of the Index).

APPENDIX 2: SUSCEPTIBILITY TO NATURAL HAZARDS (SUB-INDEX)

This research widened the scope of the susceptibility to natural hazards sub-index beyond natural disasters to include climate change21. Higher frequency and intensity of weather events, including floods, storms, landslides, hurricanes, droughts (13 per year since 1990, EM-DAT) are linked to a changing climate in the Caribbean and negative future scenarios [The Intergovernmental Panel on Climate Change (2014)]. These weather events are also leading to depressed economic growth and increased costs for public budgets. The events of the last two hurricane seasons certainly support this view. Sea level rise is the main threat, as most Caribbean communities and infrastructures are located in the coastal areas and coastal infrastructures are not resilient. Acevedo (2016) noted that between 1950 and 2014, hurricanes in the Caribbean resulted in an average annual cost of 1.6% of GDP and, based on reported damage, the cost increases to 2.5% of GDP

A Compilation of Working Papers by OECS Scholars

239

when missing data on damage was included, and to 5.7% of GDP utilising all available information on hurricanes in the Caribbean. The cost of disaster damage in the Caribbean (2.5% of GDP) is six times larger than the average country in the world (0.4% of GDP). These disasters have reversed years of gains made in infrastructure, economic and social advancement. Building resilience against these disasters is imperative for the continued success of the Region. Several climate change indicators/databases were reviewed and a criteria utilised to guide the selection of an appropriate indicator for the study. Included in the criteria was the selection of an indicator that fulfilled the conditions identified earlier of simplicity, ease of comprehension, and suitability for international comparison. In addition, the climate indicator should quantify the vulnerability of BMCs with respect to the effects of climate change and human pressures on the quality of the environment and susceptibility to disasters. The indicator should also predict how the environment is likely to cope with future events. DCVM was selected for use in the Vulnerability Index, among the other climate vulnerability data sources. The other climate vulnerability data sources considered included the Global Climate Risk Index. This index is produced annually by the think tank Germanwatch and analyses the extent to which countries have been affected by the impacts of weather-related loss events, such as storms, flood, heat waves, etc. The index has an extensive database spanning the period 1997 to 2016, with coverage of more than 180 countries. The variables used in the estimation of the index are fatalities (annual average), fatalities per 100,000 inhabitants (annual average), losses in millions of United States dollars (purchasing power parity), and losses per unit of GDP in percentage. The main weakness of this index is that it does not include a climate change component nor take into account important aspects of environmental vulnerability, such as sea level rise and warmer seas. Moreover, the data only reflects the direct impacts (direct losses and fatalities) of extreme weather events, whereas the indirect impacts (e.g. droughts and food security) are not considered. The main drawback of the Secretariat of the Applied Geoscience Commission methodology is that it is highly data-intensive, deriving information from 50 indicators covering areas related to weather and climate, geology, geography, resources and services and human populations. Other climate risk data sources include the Notre Dame Global Adaptation Index developed by the Climate Change Adaptation Programme out of the University of Notre Dame’s Environmental Change Initiative. The index summarises a country’s vulnerability to climate change and other global challenges, in combination with its readiness to improve resilience. The index utilises two decades of data across 45 indicators (36 vulnerability indicators; 9 readiness indicators) to rank 181 countries annually based upon their vulnerability and readiness to adapt. APPENDIX 3: DATA SOURCES FOR THE VULNERABILITY INDEX

Strategic Imports: UNCTAD Statistics Data Centre. Export Concentration: UNCTAD Statistics Data Centre. Available online at http://unctadstat.unctad. org/EN/. Export Destination: IMF Direction of Trade Statistics, Yearbook 2017; Caribbean Tourism Organisation, Tourism Statistical Tables, 2017 (tourism arrivals). External Finance: WB Development Indicators (FDI); UNCTAD Statistics Data Centre (Remittances). Natural Disasters and Climate Change: EM-DAT: The OFDA/CRED International Disaster Database www.emdat.be, Université Catholique de Louvain, Brussels, Belgium; DCVM (economic gains or losses from climate change and carbon. Social Vulnerability: United States Overseas Security Advisory Centre; UNODC (murders per 100,000); Official country statistics for poverty and unemployment rates. 240 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX 4: CONVERTING VARIABLES TO A COMMON SCALE

The Crowards (2000) study considered five different methods of scaling and combining the data. These are: ● Normalisation: each variable is converted to a scale between zero (applied to the lowest value of the series) and one (applied to the highest value in the series) and the minimum value for a series is subtracted from each of its entries in turn; the result of which is divided by the difference between the maximum and the minimum of the series. This is a standard transformation procedure, maintaining relative proportions within the series, but suffers from the influence of singularly high (or low) values causing the rest of the series to be bunched at the lower (or upper) end of the series. ● Decile-condensed Normalisation: in order to reduce the impact of extreme values, normalisations are carried out, but with the top decile of entries in a series attributed a value of one, and the bottom decile attributed a value of zero. This reduces the bunching effect of extreme values, but leads to a loss of information about differences between entries within the top and bottom deciles. ● Borda Rule: a ranking of countries is determined for each component variable and an aggregate score calculated for each country as the sum of its ranks, rather than normalised values across variables. This ranking removes the influence of extreme values, but ignores the extent of differences between the entries in each series. ● Fixed Exponential Scaling: prior to normalising, each series is transformed by applying a fixed exponent. In the present study, each entry is raised to the power of 0.3. This reduces, but does not remove, the impact of extreme values in positively skewed distributions, which is appropriate in this study. ● Variable Exponential Scaling: before normalising, each series is transformed by applying an exponent that meets a prior objective, such as minimising the skewness of the series or achieving median value that is midway between the minimum and maximum of the series. This has the advantage of producing relatively evenly distributed series, but involves treating each of the series differently and applying radical transformations to some of the series. Other transformation methods, such as applying logarithms or inverting the data, were proposed but rejected since they served to increase the skewness in some variables.

A Compilation of Working Papers by OECS Scholars

241

APPENDIX 5: MULTI-DIMENSIONAL VULNERABILITY INDEX (PROXY INDICATORS 2017)

BAH

ANT

ANG

0.48

0.57

0.54

Sub-Weight

0.65

0.50

0.76

0.92

0.61

8.33%

0.92

1.00

0.92

0.72

0.93

0.94

8.33%

0.77

1.00

0.77

0.50

0.93

0.00

0.63

8.33%

0.61

0.00

0.97

0.78

0.75

0.99

0.83

0.72

8.33%

0.46

0.64

0.68

0.00

0.59

0.60

0.48

0.41

0.29

8.33%

0.49

0.77

0.67

0.54

0.52

0.63

0.72

0.76

0.71

0.63

8.33%

0.22

0.20

0.22

0.23

0.19

1.00

0.20

0.22

0.24

0.19

0.31

8.33%

0.57

0.86

1.00

0.67

0.53

0.61

0.00

0.59

0.47

0.33

0.43

0.51

8.33%

1.00

0.82

0.99

0.23

0.43

0.26

0.48

0.00

0.81

0.33

0.72

0.57

0.54

5.56%

0.89

0.22

0.26

0.59

0.69

0.60

1.00

0.68

0.20

0.46

0.47

0.52

0.67

0.40

5.56%

0.72

0.70

0.63

0.79

0.46

1.00

0.79

0.81

0.70

0.00

0.84

0.55

0.33

0.54

0.21

5.56%

0.04

0.15

0.16

0.07

1.00

0.10

0.25

0.39

0.08

0.03

0.16

0.04

0.22

4.17%

NDa

0.00

0.29

0.67

0.95

0.92

0.81

0.71

1.00

0.72

0.28

0.13

0.37

0.58

0.88

4.17%

NDd

0.28

0.00

0.77

0.09

0.14

0.18

1.00

0.20

0.23

0.46

0.03

0.22

0.19

4.17%

0.35

0.00

0.29

0.40

0.17

0.03

0.25

0.35

0.42

1.00

0.10

0.19

0.45

0.85

4.17%

Climate and Natural Hazards

BAR 0.59 0.37 0.88 0.56

0.27

1.00

0.77

0.25

0.38

0.45

0.84

0.90

0.00

0.57

Social Vulnerability

BZE 0.48 0.21 0.85 0.59

0.92

0.51

0.76

0.28

0.47

0.78

0.46

0.54

0.18

External Finance

CAY 0.54 0.70 0.79

0.99

1.00

0.52

0.28

0.22

0.60

0.32

0.19

0.62

Strategic Imports

DOM 0.58 0.64 0.68

0.78

0.34

0.71

1.00

0.26

0.23

0.79

0.54

Export Destination

GRE 0.56 0.84 0.73

0.10

0.72

0.43

0.64

0.00

0.32

0.56

Export Concentration

GUY 0.71 0.62 0.78

0.65

0.60

0.65

0.00

0.12

0.50

Overall VI

HAI 0.61 0.00

0.86

0.90

0.25

0.57

0.00

0.26

17%

JAM 0.42 0.35

0.87

0.25

0.71

0.32

0.58

17%

MON 0.55 1.00

0.77

0.64

0.38

0.52

17%

SKN 0.63

0.56

0.95

0.53

0.64

17%

SLU 0.52

0.67

0.82

0.62

17%

SVG 0.47

0.59

0.85

17%

SUR 0.34

0.59

Weight

0.54

0.24

AVG

A score close to 1 indicates high vulnerability with respect to that variable.

242 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX 6: MULTI-DIMENSIONAL VULNERABILITY INDEX AND RANK 2017

JAM

HAI

GUY

GRE

DOM

CAY

BZE

BAR

BAH

ANT

ANG

0.55

0.42

0.61

0.71

0.56

0.58

0.54

0.48

0.59

0.48

0.57

0.54

0.93

0.60

0.39

0.68

0.76

0.72

0.78

0.54

0.65

0.82

0.71

0.74

0.93

0.78

0.25

0.75

0.69

0.22

0.89

0.96

0.43

0.58

0.38

0.98

0.77

0.62

0.96

0.42

0.68

0.28

0.64

0.72

0.49

0.64

0.74

0.61

0.65

0.61

0.26

0.61

0.66

0.62

0.56

0.46

0.11

0.43

0.35

0.33

0.41

0.54

0.60

0.44

0.38

0.40

0.50

0.39

0.34

0.28

0.31

0.41

0.56

0.78

0.68

0.61

0.62

0.68

0.64

0.61

0.69

0.62

0.07

0.71

0.45

0.52

0.60

0.39

0.35

0.33

0.34

0.42

0.24

0.20

0.59

0.56

0.39

0.66

0.40

0.23

0.10

0.30

0.42

0.55

Rank

MON 8

0.63

0.66

0.68

Country

SKN 2

0.52

0.81

External Social Natural Hazards Finance Vulnerability and Climate Change

SLU 12

0.47

Strategic Imports

SVG

0.00

Index Export Export Concentration Destination

SUR

0.51

0.36

0.22

0.57

0.16

0.38

0.45

0.55

0.71

0.63

0.34

0.72

0.54

TT AVG

243

A Compilation of Working Papers by OECS Scholars

Overall Vulnerability Rank

APPENDIX 7: MULTI-DIMENSIONAL VULNERABILITY RANK 2017

SKN

MON

JAM

HAI

GUY

GRE

DOM

CAY

BZE

BAR

BAH

ANT

ANG

Country

SLU 12

External Social Natural Hazards Finance Vulnerability and Climate Change

SVG

Strategic Imports

SUR

Overall Export Export Vulnerability Concentration Destination Rank

244 | Research: The Platform for Innovation, Competitiveness and Growth

APPENDIX 8: MULTI-DIMENSIONAL VULNERABILITY INDEX (PROXY INDICATORS 2016)

BAR

BAH

ANT

ANG

0.59

0.47

0.52

0.50

0.52

Sub-Weight

0.38

0.69

0.52

0.77

1.00

0.98

8.33%

0.86

0.95

1.00

0.94

0.73

0.96

0.92

8.33%

0.48

0.34

0.94

0.79

0.63

0.60

0.00

0.28

8.33%

0.30

0.34

0.00

0.93

0.78

0.76

0.97

0.84

0.74

8.33%

1.00

0.48

0.64

0.69

0.00

0.59

0.63

0.46

0.41

0.30

8.33%

0.75

0.49

0.76

0.68

0.65

0.43

0.66

0.70

0.78

0.64

0.62

8.33%

0.43

0.80

1.00

0.56

0.35

0.42

0.25

0.48

0.31

0.13

0.24

0.35

8.33%

0.16

0.14

0.13

0.04

0.05

0.16

0.12

1.00

0.06

0.11

0.12

0.09

0.20

8.33%

0.78

1.00

0.82

0.91

0.28

0.52

0.26

0.24

0.00

0.80

0.23

0.68

0.26

0.66

5.56%

0.85

0.89

0.21

0.25

0.64

0.65

0.62

1.00

0.68

0.14

0.48

0.52

0.51

0.69

0.39

5.56%

0.90

0.72

0.70

0.63

0.79

0.57

1.00

0.79

0.81

0.70

0.00

0.84

0.55

0.33

0.54

0.21

5.56%

0.00

0.30

0.70

1.00

0.12

1.00

0.16

0.29

0.31

0.14

0.07

0.21

0.00

4.17%

NDa

0.52

0.00

0.20

0.45

0.60

0.26

0.59

0.97

0.84

0.75

0.72

0.76

0.57

0.14

0.38

0.60

4.17%

NDd

0.28 0.35

0.00 0.00

0.77 0.29

0.09 0.40

0.09 0.17

0.14 0.03

0.18 0.25

1.00 0.35

0.20 0.42

0.23 1.00

0.46 0.10

0.03 0.19

0.22 0.45

0.19 0.85

4.17%

Climate and Natural Hazards

BZE 0.49 0.54 0.89 0.75

0.93

0.57

0.66

0.19

0.13

0.63

0.60

0.48

0.32

Social Vulnerability

CAY 0.48 0.55 0.82

1.00

0.43

0.28

0.30

0.18

0.32

0.15

0.62

External Finance

DOM 0.52 0.75 0.66

0.74

0.37

0.70

1.00

0.48

0.13

0.76

0.54

Strategic Imports

GRE 0.58 0.73 0.75

0.59

0.73

0.47

0.66

0.00

0.54

Export Destination

GUY 0.69 0.65

0.76

0.68

0.75

0.66

0.49

0.11

0.17

Export Concentration

HAI 0.60 0.00

0.88

0.62

0.51

0.56

0.00

0.38

Overall VI

JAM 0.43 0.31

0.90

0.30

0.05

0.36

0.60

17%

MON 0.52 0.86

0.79

0.41

0.53

17%

SKN 0.63

0.67

0.87

0.47

0.62

17%

SLU 0.54

0.61

0.80

0.57

17%

SVG 0.43

0.47

0.85

17%

SUR

0.31

0.62

17%

0.52

Weight

AVG

245

A Compilation of Working Papers by OECS Scholars

SKN

MON

JAM

HAI

GUY

GRE

DOM

CAY

BZE

BAR

BAH

ANT

ANG

Overall Vulnerability Rank

Export Concentration

Export Destination

Strategic Imports

External Finance

Social Vulnerability

APPENDIX 9: MULTI-DIMENSIONAL VULNERABILITY RANK 2016

SLU 6

Country

SVG 16

Natural Hazards and Climate Change

SUR 17

246 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

J. Jason Cotton

Country Economist Caribbean Development Bank

Mr. Cotton, a Trinidad and Tobago national, joined the Caribbean Development Bank (CDB) in 2015 as a country economist and is currently assigned the responsibility for monitoring Barbados and Jamaica. Previously, he was the country economist to Suriname. While at the CDB, he has functioned as the team leader for the CDB Policy Based Loans (2018 and 2019) to the Government of Barbados. He also participated in technical assistance (TA) missions to Antigua and Barbuda and the Turks and Caicos Islands. The TA missions assisted the Governments of Antigua and Barbuda and the Turks and Caicos Islands in the preparation of a Medium-Term Development Plan (2016-2020) and a concept paper for the establishment of a Sovereign Wealth Fund, respectively. Mr. Cotton is a graduate of the University of the West Indies, St. Augustine, Trinidad with a Master of Science and Bachelor of Science degrees in Economics. He has completed several professional development programmes from: Harvard University, Mc Gill University and the University of Wolverhampton in topics including: Public Policy, International Development and Results Based Management. His research publications focus on public finance, the external sector and Vulnerability and Resilience in Small Island Developing States. Prior to joining the CDB, Mr. Cotton functioned as a Senior Economist at the Central Bank of Trinidad and Tobago and part-time lecturer at the University of the West Indies, St. Augustine

A Compilation of Working Papers by OECS Scholars

247

To Have or Not to Have Private Health Insurance Coverage? A Review of the 2016 Saint Lucia Survey of Living Conditions and Household Budgets Saint Lucia Mr. Janai Leonce and Dr. Marisa Jacob-Leonce

248 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT A logistic model was used to determine the contributing factors to a Saint Lucian’s likelihood of having health insurance coverage. Income, union status, educational attainment and place of employment were statistically significant, positive contributing factors while age, location, religion and proxies of health awareness proved insignificant. More specifically, individuals with post- secondary, tertiary education, married, non-self-employed or earning more than $5,000 monthly had the highest likelihood of having health coverage. Stylized facts however show that less than 15% of the populace earn more than $5,000 monthly and that only 14% have either post or tertiary levels of educational attainment. Our results, which are based on the 2016 Survey of Living Conditions and Household Budgets provide increasing evidence for the need for health reform which increases access to health interventions by subsidizing or reducing cost. INTRODUCTION Article 25 of the United Nations’ 1948 Universal Declaration of Human Rights posit that health, wellbeing and medical care are inalienable human rights. The 2030 Sustainable Development Goals (SDG) reiterates this principle with SDG 3 speaking directly to the need for healthy lifestyles and well-being1. These core tenants of human decency mandate that societies facilitate access to quality health care. Bloom Canning and Jamison (2004) suggest that nations which facilitate appropriate health infrastructure reap the rewards of having a more productive and economically vibrant society. Cognizant of the importance of the medical well-being Saint Lucia is currently working with the World Bank and other relevant stakeholders to strengthen the islands’ health systems and consequently improve health outcomes. Reform of the health system is warranted as research shows that out-of-pocket health expenditures in Saint Lucia average 48.02 percent, which is 2.5 times higher than recommended as per the World Health Organization (WHO). Similarly estimates of total health expenditures, as a percentage of GDP are 2.53 percent, lower than the 6.0 percent4 target expressed by the international community. This relatively low health of health spending and high out of pocket expenditures renewed interest in and spurred a desire for comprehensive reform. The ensuing debate on the level of coverage of health reform and its subsequent financing has oscillated between a legally mandated nationwide insurance scheme and a social health system funded through increased taxation. A consequence of this debate was the bringing to the fore of results from the 2016 Survey of Living Conditions and Household Budgets which shows that only 18.2 percent of the population had health insurance. On the surface this sounds low particularly when juxtaposed against regional peers5 such as Barbados at 27.06 per cent and St Kitts and Nevis at 35.07 per cent. While the decision to have non-state health insurance is heavily dependent on the breadth and depth of easily accessible public health systems it does raise an interesting point of enquire as to why some persons choose to get as opposed to others given the prevailing socio-economic realities. In a climate of ‘low’ public health expenditures and ‘high’ out of pocket expenses why don’t more persons supplement these ‘deficiencies’ via non state health insurance and relatedly what caused those who do have private health insurance to get it? Intuitively the decision to obtain private health insurance is motivated by either an absence of a national health system, a desire to prevent financial distress and hardship on account of a medical condition and or simply a desire to recoup out of pocket health expenditures. Relatedly one’s risk aversion, the existence of a medical condition and financial standing are all factors which impact the decision to have private health insurance. A sound understanding of how, and in which groupings of persons the aforementioned and other contributing factors affect the decision to obtain health insurance is needed, particularly in an environment where the government is actively considering health reform options which may center on mandating broad based insurance coverage. This information can allow for policy makers to better target PSA campaigns and possibly inform risk-based compliance audits should be obtaining health insurance be legally mandated. The paper is also useful for regional counterparts who are thinking of embarking on similar reforms. A Compilation of Working Papers by OECS Scholars

249

LITERATURE REVIEW In both developed and developing countries a debate exists on both the role and efficacy of private health insurance (PHI’s) schemes. Colombo and Tapay 2014 note that within the OECD some see private health insurance as a way of improving health outcomes while others express concern on the possibility of PHI’s to lead to inequitable health outcomes and market failures. Within a developing country context Drechsler and Jutting 2007 point out that there is a significant need, given fiscal and economic weaknesses, for sustainable mechanism(s) to deliver better health interventions and that there is a role for private health insurance to do so. Notwithstanding this debate Besley, Hall and Preston 1999 and others, point out the predominance of public health spending in most societies and that private health insurance is an important part of any societies’ health system. The perceived importance of private health insurance has led some health researchers see Gruber and Simon 2007 to ask whether expansions in public health insurance schemes have led to “crowding-out” i.e. reduced private insurance coverage. Gruber and Simon 2007 suggest that crowding out exists within the US context and argue that this is a counterintuitive effect of efforts to provide public insurance coverage to otherwise uninsured persons. The literature of private health insurance is mixed and a distinction is made between private health insurance and national legally mandated or in some cases voluntary health insurance see Preker, Scheffler and Basset 2007. Preker et al 2007 notes that developing countries often finance health care via a 20:80 mix of prepaid to out of pocket expenditures while in developed countries this ratio is 80:20. The optimal mix is suggested to be closer to 80:20 and that most developing countries become stuck at a ratio of 40:60 which closely mirrors Saint Lucia’s experience. This optimal mix and the proposed interventions of the government necessitate through research on better understanding consumer behavior re prepaid/private/quasi private insurance. Testing the determining factors to obtain private health insurance is a key aim of this paper. King and Mossialos 2005 assessed a similar aim in the British context where the National Health Service (NHS) is a well renowned, compulsory and comprehensive publicly financed system. Nonetheless, King and Mossialos, using household surveys and found that education, income, sex, waiting times and political preference were among some of the statistically significant determinants of obtain PHI. Further to these variables the importance of risk aversion and the perceived loss due to illness are important considerations that are intuitively necessary to modelling but empirically difficult to assess. Bound, Schoenbaum, Stinebricker and Waidmann (1998) who looked at the interplay between health status and labour force behavior in the US also sought to determine factors which influence a person’s health decisions. In that piece proxies for age, income, work ethic and preferences were included as factors which influence health related decisions. In addition to self-evident socio –economic proxies which influence health decisions risk preference is also an economic condition that is central to the decision making of persons particularly as it relates to health insurance in a voluntary context. Kimball, Sahm and Shaprio (2008) sought to proffer guidance on imputing risk tolerance from health-related survey responses. The note that a sound basis for imputing risk is that survey questions are designed in a manner which provide each respondent with objective trade off based questions. METHODOLOGY The decision to have health insurance is a dichotomous one as an individual can either have or not have private health insurance. Given that Saint Lucia is in the process of reviewing compulsory mandates to obtain health insurance the objective of our paper is timely, as we seek to assess which, if any, socioeconomic traits alter the probability or likelihood of an individual making this choice i.e. obtaining health insurance. To answer or satisfy this objective this paper uses logit regressions which are of the form;

250 | Research: The Platform for Innovation, Competitiveness and Growth

Equation 1

Where S is the set of all observations j, so that yi is not equal to 0, F(z) = ez/(1+ez), and wi could be optional weights (Stata 13 manual pg. 1087). Logit regressions are maximum likelihood models with a dichotomous dependent variable, often codified as 0 or 1 with 1 representing ‘success’. In our paper success is defined as having health insurance coverage and consequently codified as 1 while not having private health insurance coverage is codified as 0. The decision to purchase health insurance is thought to be motivated by several thematic areas namely; risk aversion8, financial standing (wealth and risk), prevailing health status and awareness (level of health and knowledge) and socio-economic status. In this paper (see table 1) socio-economic status encapsulates one’s age, level of educational attainment, household status i.e. head of the household or not, employment status, marital status and gender. While financial standing is proxied by income earned and the nature of and type of job. Prevailing health status is interpreted to mean the extent to which an individual has either a communicable or non-communicable disease. Most of the aforementioned are directly codified via a dummy variable from the survey except age which is continuous, risk aversion and health status. Risk aversion is more challenging as it attempts to capture the extent which persons mitigate against uncertainty. Ideally trade off based questions9 can be codified into measures of risk tolerance but in our case the survey did not ask such questions. Risk tolerance is therefore proxied by their vulnerability score which is the quintile of their level of consumption. We contend that the extent to which someone has a non-communicable disease or has visited the hospitals on island may have increased their sensitization to the cost associated with health care and consequently the need for health insurance coverage. In keeping with O’Donnell et al (2008) we ensure that the survey design of the 2016 SLCHB such as sampling weights and units are an appropriate factor into the analysis by using the ‘svyset’ command in Stata 15. TABLE 1: THEMATIC AREAS AND PROXIES

Risk Aversion

Financial Standing

Prevailing Health Socioeconomic Status Status

Vulnerability (Score)

Employment Status (Employed/Unemployed)

Presence of disease

Age

Type of Occupation (Temp/ Permanent)

Hospitalization

Gender

Nature of Occupation (Self Employed or not)

Level of Education Marital Status Head of Household Religon Urban/Rural

A Compilation of Working Papers by OECS Scholars

251

STYLIZED FACTS The 2016 SLC-HBS survey was based on a stratified, two-stage probability design. The survey instrument was administered by the Central Statistical Office (CSO) to a randomly selected sample of 1,493 households, which represented 2.7 percent of the population of Saint Lucia. The survey consisted of 4,568 participants of which 51.8 percent were female and 48.8 percent were male. In the survey 801 persons indicated that they were covered by health insurance (18.0%) while 3,767 (82%) persons indicated that they were not covered. Of those 801 persons covered by health insurance 493 were above the age of 18 and below 65 the remainder were primarily children and dependents of the 493 persons. Consequently, the descriptive statistics below centers mainly on the 493 i.e. working age population. Of our target sample 56.5 percent of those covered were women, suggesting a disproportionate 10 rate of coverage given that women made up 50 per cent of the target sample. Table 3 (in the appendix) which shows the descriptive statistics11 of the underlying data12 show that those who are covered by health insurance are more likely to have higher levels of academic achievement, hail from urban areas and have higher average incomes. The average income of individuals with health insurance was $3,113.0 while those without was 38.2 percent lower at $1,925.00. As shown in table 2 below a deeper review of health coverage by income showed that for every category of worker with the exception of self-employed with employee persons with coverage had higher average incomes.

TABLE 2: AVERAGE INCOME BY STATUS IN EMPLOYMENT

Status in Employment

Yes

Total

Difference (Have/Don’t Have Insurance)

Central Govt Employee

$ 3,557.91

$ 1,894.87

$2,584.68

$ 1,663.04

Employee of Statutory Board

$ 3,309.53

$ 1,860.60

$2,321.59

$1,448.83

Private Employee

$ 2,630.71

$ 1,528.96

$1,783.21

$1,101.75

Apprentice

$ 800.00

$ 750.00

$762.50

$50.00

Self-employed with Enployee

$ 8,403.12

$ 3,763.12

$4,505.52

$4,640.00

Self-Employed without Employee

$ 1,860.68

$ 2,289.32

$2,261.05

$(428.64)

Unpaid family worker

$ 400.00

$ 1,143.33

$1,037.14

$(743.33)

$ 8,000.00

$8,000.00

$(8,000.00)

Member of production cooperative Other

$ 3,084.40

$ 1,894.87

$1,363.06

$2,438.57

Average

$ 3,080.09

$1,873.64

$2,147.06

$1,206.45

252 | Research: The Platform for Innovation, Competitiveness and Growth

An occupational classification (see table 3 in the appendix) of those with and without health insurance coverage showed that 45.2 percent of those covered were managers, professions and technicians which is in contrast to a similar comparison of those without health insurance where the equivalent ratio (i.e. managers and professionals) was 15.6 percent. The cohorts (those with and without coverage) also differ by the highest level of educational attainment. Approximately 50.0 percent of those with insurance coverage had post-secondary or tertiary education compared with 85.7 percent of persons without health insurance having attained below secondary schooling education. In addition to education and income, the union status13 and relative positions in the household also appear to differ between the cohorts. Of the 814 categories of union status, 34.3 percent of those with health insurance coverage were married, compared to 17.5 percent of persons without coverage being married. If common law relationships are lumped with married persons, the statistic rises to approximately half of the persons with insurance being married or in common-law relationships despite the fact that that cohort (married or common law) represented 8.4 percent of the whole sample15. Whether someone was self-employed or not was also a key differentiator between the cohorts as only 8.9 percent of those with coverage were self-employed. Unlike union status and education, a first look at the data suggest that the relative standing in the home i.e. head or spouse of head of household and our proxy for disease and knowledge of health awareness were similar between the cohorts. Of those who had health insurance 42.0 per cent16 were household heads which is similar to the per cent17 of persons who did not have insurance i.e. 42.0%. Similarly, roughly the same ratios existed for our visited St Jude or Victoria Hospital and our had diabetes and or hypertension between persons who had or did not have insurance coverage. RESULTS18 Two variants of logistic regressions using equation 1 (see table 4 in the appendix) were used. This was required given that our union status and gender variable were collinear. We subsequently estimated models which varied whether they were included or not. Our model without gender showed that education, income and being married positively impacted your odds of having health insurance coverage while changes in location between urban and rural, having a non-communicable disease or religion did not have a statistical bearing on health insurance coverage. The model which included gender but omitted union status showed similar results with the exception that age, occupational type and gender were now statistically significant. With age and gender19 being positive contributors and occupation type negative. Table 5 shows the categorical breakdown of our variables including education and occupations (this model also had two specifications). Income is significant at all typical levels of significance and is robust to various specifications. Each dollar increase in income raises the odds of having health coverage by 1.0. Figure 1supports this finding as it shows that the probability of having insurance coverage increases to 80.0 percent if someone earns $8,000 monthly from 30.0 percent if earnings are $2,000 monthly. The 50.0 percent probability of having insurance is achieved in instances where monthly income is approximately $5,000. The issue of education having post-secondary and or tertiary (university) education boosted the odds of having coverage by a factor of 3.35 and 4.8 both of which represented the highest factors in our study. Put differently we found that the probability of someone with postsecondary education or tertiary education having health coverage was 38.2 and 47.1 percent compared with 15.5 per cent with no education. Figure 2 visually charts the progression or change in probability by educational attainment and income. In figure 2 we can see that the interaction between education and income is a positive one since for the same income referenced in figure 1 having different types of education can bolster or reduce the estimated predictions. In figure 1 we saw that an individual earning $2,000 monthly had a 30.0 percent change of having coverage. Figure 2 shows if that same individual has post-secondary education then their likelihood will increase to approximately 50%.

A Compilation of Working Papers by OECS Scholars

253

FIGURE 1: MODEL PREDICTED PROBABILITY OF HAVING HEALTH COVERAGE (INCOME) (18-65 YEARS OLD)

Further to one’s income and education, their standing in the household, i.e. whether they were the head of the household or not was observed to have a statistically significant positive impact on their odds of having health coverage. Being the head of the household raised the odds of having health coverage by a factor of 1.87 suggesting that the title of head brought with it some level of responsibility and as such a need to insure against health calamities. Notwithstanding this the average probability of head having coverage was only 19.3 percent suggesting that although being a head boosted your health coverage likelihood many heads did not have insurance. In addition to one’s position in the home their occupational classification also had a statistically significant impact on the likelihood of having health coverage. Model 220 in table 5 show that being a craft worker or a skilled agricultural worker lowered 21 the likelihood of having health coverage by a factor of –0.74 and -1.35 respectively (see figure 5 in the appendix). Compounding the findings of table 5 we see in table 6 (in the appendix) that being selfemployed significantly lowered the odds of having health coverage. In keeping with the findings above that income is a factor in determining health coverage we found evidence that someone’s union status also significantly altered their odds of having health insurance. Table 522 shows that when compared to a base of being married not being in a union, being in a common law relationship or in a visiting relationship lowered the odds of having coverage (see figure 3).

254 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 2: EFFECT OF EDUCATION AND INCOME ON HEALTH COVERAGE

In our tables 4 and 5 specification age and gender provide insignificance while table 6 which introduces employer type and number of live births show that age and gender are positively associated with having coverage. We found no evidence that living in an urban or rural area, or having a non-communicable, or visiting a hospital influenced health coverage. This finding coupled with the stylized facts and aforementioned suggest that the driving factor in the Saint Lucia context to having health coverage was income and factors which may impact income such as education and marital status. Marriage and education, were two socioeconomic factors shown to boost the odds of having insurance and they are both intuitively positively correlated with income. In our sample average incomes of married persons (individually) were higher than that any other union status at $2,219. This higher individual income, means that a married household potentially has joint income of just under $4,500 which is significantly higher than the average income of a single person ($1,754). In our sample only 6.023 percent of persons were married. Educational attainment, particularly post-secondary and tertiary education, were significant predictors of insurance coverage but only 7.3 and 7.2 percent of people have these levels of educational attainment. Earlier our results showed that one’s likelihood of having a 50.0 percent chance of having health coverage coincided with a monthly income of $5,000. That level of income however is outside the reach of most of the populace as only 15.0 percent of the population earn more than $5,000 monthly. Further to this percentage the mean income in the sample was $2,147 with a median of $1,400. Our a priori initially was that the low levels of health insurance coverage cited by various reports and media may have been driven by cultural and other factors such as asymmetry of information related to the true cost of health care. This belief led to the creation of the “visit to the hospital” and the “have a noncommunicable disease” variable(s) as we believed that persons who had made contact with the health system would be better informed about the cost of healthcare and therefore take steps such as getting insurance to mitigate said cost. This and other such variables were insignificant. Taking all our findings together they paint a different picture and suggest that far from cultural attitudes it may be the case that the low levels of health insurance coverage is due to the fact that too many Saint Lucians simply don’t earn enough to afford health coverage coupled with not being part of institutional arrangements which can bolster their incomes such as marriage and or having higher forms of educational attainment. A Compilation of Working Papers by OECS Scholars

255

POLICY RECOMMENDATIONS Based on the aforementioned results the demographic of persons most likely to not have health insurance coverage are self-employed, non-married persons with less than post-secondary education while those with the highest likelihood would be married, non-self-employed, females who have tertiary (university) education. This information should prove valuable to policy makers who may be considering advertising and other campaigns to sensitize persons to the benefits of any health insurance scheme. Furthermore, a review of the contributing factors to having health insurance coverage show that income is a key determining factor. Figure 3 shows that approximately 65.0 percent of Saint Lucians earn less than $2,200 monthly. This finding has significant implications for policy makers. The relatively low absolute earnings suggest that persons simply may not be able to afford health insurance. A typical health insurance policy may cost an individual $100-$200 monthly. This works out to almost 5.0 percent at the lower end and 10.0 per cent at the upper end of one’s monthly income. The results suggest that this percentage serves as a barrier to getting coverage. In the short term the incomes of the populace are difficult to change and as such highlights the need for efforts to cross subsidize health care.

FIGURE 3: HISTOGRAM OF MONTHLY INCOME

the two groups was similar to their prevalence in the study i.e. one third of survey respondents were rural while a third of those who had health coverage hailed from a rural area. Furthermore, there was not statistically difference24 in earnings between the two areas. This is an important finding for policy makers who may have intuitively thought that rural areas had higher incidences of no health coverage and consequently may have been inclined to focus on that demographic rather than the urban one. This is not the case. The finding that self-employed people are less likely to have health coverage with an average rate of less than 10.0 per cent (see figure 6) suggests that the group has to be a key demographic which policy makers focus on. Self-employed persons with employees had one of the highest average incomes of approximately $9,000 monthly (see table 2) but still had low health insurance take up rates.

NOTES 1

https://sustainabledevelopment.un.org/sdg3

PENMNADU

https://www.paho.org/salud-en-las-americas-2017/?p=4211

https://www.who.int/health_financing/en/how_much_should_dp_03_2.pdf

Who don’t have national health insurance schemes

Ministry of Health Barbados Paper on Proposal for Reform of the Health Care System in Barbados, 2018

Estimates from St Kitts-Nevis Financial Services Regulatory Commission, 2018

https://halshs.archives-ouvertes.fr/halshs-01935846/document

Which of the statements comes closest to the amount of financial risk that you and your (spouse/partner) are willing to take when you save or make investments? 9

take substantial financial risks expecting to earn substantial returns

take above average financial risks expecting to earn above average returns

take average financial risks expecting to earn average returns

not willing to take any financial risks

Women comprise 51% of the sample but 56% of those with insurance.

Non survey adjusted.

For persons between the ages of 18 and 65 which is a sample of 493 persons, lower than the 801 sample size where there are no age restrictions. 12

Married or unmarried.

Crowards (2000) included 136 countries in the study, however, sufficient data was available for 95 countries for the calculation of EVI. 14

Due to missing values the number of overlapping respondents was 1,764.

204/493 A Compilation of Working Papers by OECS Scholars

257

947/2255

All regressions are done for those between the ages of 18 and 65.

Codified as 1 if female.

With gender but not union status.

When compared to a base of a central government employee.

In table 4 we had used a dummy variable for 1=married or common law which showed a positive relationship. In table 5 we use the full categorical breakdown for union status. 22

105/1766 (missing values lowered the sample size).

Mean incomes in rural areas were $2,361 and $2,032 in urban ones. T-test confirmed that this difference was not statistically significant. 24

REFERENCES (n.d.). Goal 3; Sustainable Development Knowledge Platform. Progress of goal 3 in 2019 Retrieved from https://sustainabledevelopment.un.org/sdg3 Bloom, D. E., Canning, D., & Jamison, D. (2004, March). Health, Wealth, and Welfare. www.imf.org/external/pubs/ft/fandd/2004/03/pdf/bloom.pdf

Retrieved from https://

Saint Lucia. (n.d.). Retrieved from https://www.paho.org/salud-en-las-americas-2017/?p=4211 Colombo, F. and N. Tapay (2004), “Private Health Insurance in OECD Countries: The Benefits and Costs for Individuals and Health Systems”, OECD Health Working Papers, No. 15, OECD Publishing, Paris, Retrieved from https://doi.org/10.1787/527211067757. Drechsler, Denis & Jütting, Johannes. (2007, July). Different Countries, Different Needs: The Role of Private Health Insurance in Developing Countries. Journal of health politics, policy and law, Retrieved from https://www. researchgate.net/publication/6314155_Different_Countries_Different_Needs_The_Role_of_Private_Health_ Insurance_in_Developing_Countries Besley, John, Preston, Ian, & Timothy & Hall. (1999, January 1). The demand for private health insurance: do waiting lists matter? Retrieved from https://ideas.repec.org/a/eee/pubeco/v72y1999i2p155-181.html Gruber, J., & Simon, K. (2007, January). Crowd-out ten years later: have recent public insurance expansions crowded out private health insurance? Retrieved from https://www.nber.org/papers/w12858.pdf Preker, A. S., Scheffler, R. M., & Bassett, M. C. (2007). Private Voluntary Health Insurance in Development Friend or Foe? Retrieved from https://openknowledge.worldbank.org/bitstream/handle/10986/6641/382810Private0101O FFICIAL0USE0ONLY1.pdf?sequence=1 King, D., & Mossialos, E. (2005, February). The Determinants of Private Medical Insurance Prevalence in England, 1997–2000. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1361133/pdf/hesr_00349.pdf Bound, J., Schoenbaum, M., Stinebrickner, T., & Waidmann, T. (1998, November). The Dynamic Effects of Health on Labor Force Transitions of Older Workers. Retrieved from https://www.nber.org/papers/w6777.pdf

258 | Research: The Platform for Innovation, Competitiveness and Growth

Kimball, M. S., Sahm, C. R., & Shapiro, M. D. (2008, September 1). Imputing Risk Tolerance from Survey Responses. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2856097/ Doorslaer, Eddy & O’Donnell, Owen. (2008). Measurement and Explanation of Inequality in Health and Health Care in Low-Income Settings. Retrieved from https://www.researchgate.net/publication/23547680_Measurement_ and_Explanation_of_Inequality_in_Health_and_Health_Care_in_Low-Income_Settings Ssempala, Richard. (2018, February 26). Factors Influencing Demand for Health Insurance in Uganda. Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3124179 Abu Bakar, A., Regupathi, A., Aljunid, S. M., & Omar, M. A. (2012, November 27). Factors affecting demand for individual health insurance in Malaysia. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3507976/ Bhat, R., & Jain, N. (2007, January). A Study of Factors Affecting the Renewal of Health Insurance Policy. Retrieved from https://web.iima.ac.in/assets/snippets/workingpaperpdf/2007-01-02_rbhat.pdf Barrett, R. D. (2012, October 22). The role of Regional Health Insurance in Caricom on the path to Universal Health Coverage. Retrieved from https://www.paho.org/hq/dmdocuments/2012/Barrett-RHIM-Bar-2012.pdf Holt, C. A., & Laury, S. K. (2002, December). Risk Aversion and Incentive Effects. Retrieved from http://community. middlebury.edu/~jcarpent/EC499/Holt and Laury (2002) AER.pdf

A Compilation of Working Papers by OECS Scholars

259

APPENDIX TABLE 1: DESCRIPTIVE STATISTICS OF VARIABLES (18-65 YEARS OLD)

260 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 2: LOGISTIC REGRESSION RESULTS (NON CATEGORICAL) (18-65 YEARS OLD)

Age Head Household Married or Common Law

(1)

(2)

With Married (No Gender)

With Gender (No Married)

0.0111

0.0191**

(1.03)

(2.53)

-0.189

-0.0838

(-0.80)

(-0.59)

0.442* (1.95)

Education Urban=1 Has Diabetes or hypertension Visited St Jude or Victoria Self employed Monthly income Occupation Type Religion

0.402***

0.485***

(4.03)

(6.91)

0.166

0.0276

(0.75)

(0.16)

0.381

0.316

(1.22)

(1.16)

-0.134

0.0349

(-0.58)

(0.21)

-0.695*

-1.045***

(-1.81)

(-4.07)

0.000420***

0.0000150

(4.50)

(0.57)

-0.0795

-0.155***

(-1.37)

(-3.87)

-0.0134

-0.0167

(-0.50)

(-0.89)

Gender (Female=1)

0.256* (1.94)

_cons N

-3.361***

-2.844***

(-4.22)

(-4.79)

796

1788

t statistics in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01

A Compilation of Working Papers by OECS Scholars

261

TABLE 3: LOGISTIC REGRESSION RESULTS (CATEGORICAL) (18-65 YEARS OLD) (1)

(2)

With Union Status (No Gender)

With Gender (No Union Status)

0.0135

0.00752

(1.12)

(0.89)

-0.255

0.631***

(-0.60)

(2.96)

0.0472

-0.307

(0.10)

(-0.97)

0.221

-0.121

(0.55)

(-0.50)

0.0847

-0.273

(0.24)

(-1.09)

0.321

-0.0492

(0.51)

(-0.10)

spouse/partner of child of head/ spouse/partner

2.067*

2.027***

(1.95)

(2.93)

grandchild of head/spouse/partner

-0.109

-0.963**

(-0.14)

(-1.98)

-0.0772

-0.185

(-0.15)

(-0.50)

-0.197

-0.391

(-0.26)

(-0.71)

Age

Head (base= parents of head) spouse of head (husband/wife) partner of head child of head and spouse/partner child of head only child of spouse/partner only

other relative of head/spouse/partner other non-relative Union Status (base=never had partner) common-law

-1.137*** (-2.72)

visiting

-1.004** (-2.28)

no longer living with husband

-1.769 (-1.53)

no longer living with common law partner

-1.840 (-1.45)

262 | Research: The Platform for Innovation, Competitiveness and Growth

not stated

(1)

(2)

With Union Status (No Gender)

With Gender (No Union Status)

-1.941*** (-3.59)

not in union

-1.371*** (-2.94)

Education (base=none) pre-primary (infant) or primary lower / junior secondary (forms 1-3) / senior primary upper secondary (forms 4 & 5) post secondary, non-tertiary (diploma or associate degree) tertiary (university) other URBAN (base= rural) Has diabetes or hypertension Visited St Jude or Victoria Income

Occupation (base=managers) Professionals Technicians and associate professionals Clerical support workers

-1.388

-0.618

(-1.17)

(-0.92)

-1.487

-0.730

(-1.21)

(-1.07)

-0.554

0.462

(-0.47)

(0.68)

0.343

1.211*

(0.29)

(1.72)

0.620

1.576**

(0.50)

(2.22)

-0.452

0.703

(-0.33)

(0.89)

0.228

-0.0578

(0.95)

(-0.32)

0.294

0.270

(0.97)

(1.06)

-0.170

0.123

(-0.71)

(0.71)

0.000364***

0.00000667

(3.61)

(0.26)

(.)

0.349

0.163

(0.66)

(0.47)

0.835

-0.00472

(1.56)

(-0.01)

0.700

0.213

(1.38)

(0.58)

A Compilation of Working Papers by OECS Scholars

263

(1)

(2)

With Union Status (No Gender)

With Gender (No Union Status)

0.113

-0.443

(0.22)

(-1.36)

-1.352**

(.)

(-2.23)

-0.0689

-0.748*

(-0.09)

(-1.90)

Plant and machine operators, and assemblers

1.831***

0.0169

(2.67)

(0.04)

Elementary occupations

-0.0743

-0.797*

(-0.11)

(-1.95)

Service and sales workers Skilled agricultural, forestry and fishery workers Craft and related trades workers

Gender (female=1)

0.201 (1.30)

Constant Observations

-1.185

-1.676*

(-0.82)

(-1.92)

799

1784

t statistics in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01

264 | Research: The Platform for Innovation, Competitiveness and Growth

TABLE 4: LOGISTIC RESULTS EMPLOYMENT TYPE AND NUMBER OF LIVE BIRTHS (1)

(2)

With Union Status (No Gender)

Without Union Status

0.0178**

0.0166

(2.07)

(1.62)

(.)

0.252*

0.0659

(1.77)

(0.28)

(.)

0.615***

0.461*

(2.93)

(1.71)

-0.412

-0.950**

(-1.31)

(-2.01)

-0.124

-0.296

(-0.52)

(-1.11)

-0.287

-0.383

(-1.16)

(-1.29)

0.0388

-0.234

(0.08)

(-0.36)

spouse/partner of child of head/ spouse/partner

2.043***

2.969***

(3.21)

(3.99)

grandchild of head/spouse/partner

-1.039**

-2.311***

(-2.21)

(-3.17)

newhealth_insurance age last birthday Male Female head spouse of head (husband/wife) partner of head child of head and spouse/partner child of head only child of spouse/partner only

parents of head/spouse/partner

0 (.)

other relative of head/spouse/partner domestic employee other non-relative none

-0.307

-0.405

(-0.88)

(-0.94)

(.)

-0.591

-0.783

(-1.08)

(-1.09)

(.)

A Compilation of Working Papers by OECS Scholars

265

(1)

(2)

With Union Status (No Gender)

Without Union Status

-0.601

0.0762

(-0.87)

(0.07)

-0.594

0.00839

(-0.86)

(0.01)

0.649

1.338

(0.93)

(1.31)

post secondary, non-tertiary (diploma or associate degree)

1.620**

2.544**

(2.27)

(2.45)

tertiary (university)

1.995***

2.603**

(2.84)

(2.52)

0.984

2.013*

(1.21)

(1.76)

(.)

0.0410

-0.0233

(0.24)

(-0.11)

0.208

0.117

(0.79)

(0.40)

0.111

0.176

(0.70)

(0.95)

0.0000189

0.0000136

(0.67)

(0.68)

(.)

-0.252

0.0997

(-0.61)

(0.23)

-0.121

0.116

(-0.68)

(0.54)

0.235

1.017

(0.17)

(0.63)

-1.205**

-1.229**

(-2.49)

(-2.36)

-1.061***

-1.109***

(-3.43)

(-3.12)

pre-primary (infant) or primary lower / junior secondary (forms 1-3) / senior primary upper secondary (forms 4 & 5)

other RURAL URBAN diabities_hypertension stjude_victoria monthly_income central govt employee employee of statutory board private employee apprentice self-employed with employee self-employed without employee

266 | Research: The Platform for Innovation, Competitiveness and Growth

unpaid family worker member of production cooperative other

(1)

(2)

With Union Status (No Gender)

Without Union Status

(.)

0.162

0.00799

(0.27)

(0.01)

1=0 to 2, 2=3 to 5kids, 3=more than 5 kids=1

0 (.)

1=0 to 2, 2=3 to 5kids, 3=more than 5 kids=2 Constant Observations

-0.478 (-1.51) -2.435***

-2.638**

(-3.01)

(-2.18)

1844

1402

FIGURE A: EFFECT OF UNION STATUS AND INCOME ON HEALTH COVERAGE

A Compilation of Working Papers by OECS Scholars

267

FIGURE B: EFFECT OF OCCUPATIONAL STATUS AND INCOME ON HEALTH COVERAGE

FIGURE C: EFFECT OF EMPLOYER TYPE AND INCOME ON HEALTH COVERAGE

268 | Research: The Platform for Innovation, Competitiveness and Growth

About the Authors

Janai Leonce

Chief Economist, Department of Finance Government of Saint Lucia

Dr. Marisa Jacob-Leonce

Assistant Professor Spartan Health Sciences University

Janai Leonce, is currently the Chief Economist at the Research and Policy Unit in the Department of Finance (Saint Lucia). Mr. Leonce holds a master’s degree from Durham University in Finance and Investment and a bachelors in Economics and Management from the UWI (St. Augustin). Prior to working for the Government of Saint Lucia, Mr. Leonce worked at the Eastern Caribbean Central Bank as an Economist and also lectured at the Sir Arthur Lewis Community College (Saint Lucia). Mr. Leonce is the author of several working papers covering topics on unemployment and gender wage discrimination. He has also written several articles in local newspapers on issues relating to productivity on behalf of the National Competitiveness and Productivity Council (NCPC) and Sir William Arthur Lewis.

Dr. Marisa Jacob-Leonce is an Assistant Professor in the Department of Clinical Sciences, at Spartan Health Sciences University, School of Medicine, where she previously graduated from in 2011. After, graduating from Spartan Health Sciences University, she completed Anatomic and Clinical pathology residency training in 2015, and Surgical Pathology fellowship training in 2016, at the University of Iowa Hospital and Clinics in Iowa City, Iowa. Dr. Jacob-Leonce, then went on to pursue subspecialty training in Forensic Pathology at the Hennepin County Medical Examiner’s Office, where she also served as the Deputy Medical Examiner for: Hennepin, Dakota and Scott counties, in Minnesota from July 2016 to June 2017. Dr. Jacob-Leonce then returned to Saint Lucia in July 2017, where she has performed private consultancy work to include: An Assessment of the National Blood Bank Service in Saint Lucia, for the Ministry of Health and Wellness and served as a consultant Forensic Pathologist for the Ministry of Justice from June 2018 to April 2019. She is currently a founding director of JL Consulting Inc. and volunteers part-time at the St. Jude Hospital Pathology Department. A Compilation of Working Papers by OECS Scholars

269

14 The Impact of Motivation on Implementing Innovation in Organization Kurt Augustin

270 | Research: The Platform for Innovation, Competitiveness and Growth

ABSTRACT The world is continuously evolving through development and in order for businesses to stay competitive in the business arena innovation must be considered. Innovation provides an organization with the ability to deal with rapid change by methods of problem solving, ideation and development. Many workplaces have a very difficult time implementing innovation therefore methods have to be deduced to ensure proper implementation. One of the concepts which is assumed to provide generous support to innovation is motivation. Motivation is used to capture productivity and satisfaction of employees therefore it can be seen as an easy methodology to drive innovation. Majority of motivational theories categorizes an individual as being intrinsically or extrinsically motivated therefore this implies that an employee can have the zeal to be creative by having the ideal behaviour to fulfil organizational goals. Creativity is known to bring ideas to a company and this is one of the most important attributes of innovation. After highlighting theoretical background on innovation, motivation, job performance and employee satisfaction this paper will seek to identify the impact of motivation on the creation of an innovative workplace.

CHAPTER ONE - INTRODUCTION INTRODUCTION In order to remain competitive in the business arena, a need exists to keep up with the constant demands for the rapid changes in products and services, which will keep consumers satisfied in an increasingly unpredictable business environment (Aleksandra, Ristovska, & Gramatnikovski, 2015). To ensure that an organization remains competitive, innovation is a key strategy used to produce new products, services and business models with the intention of delivering value to consumers. In addition, to give way to success, an innovative organization should facilitate a work environment, which makes allowances for failure, creativity, experimentation and developing new ideas (Intel, 2014). On the contrary, implementing an innovative organization involves an extensive transformation of the business models, operations and most importantly culture. Capgemini (2017) states that approximately 62% of issues faced, in relation to innovation transformation are based on culture, therefore the most challenging aspect for implementing a successful innovative organization is the employees as they are the root of the organizational culture. Since it has been established that the employees of an organization is the driving force of innovation, it is important to focus on the strategies used to manage an innovative organization. Koudelková & Milichovský (2015) state that motivation is the primary stimulus in the implementation of an innovative organization, as motivation would make a positive contribution to the working environment. In order to fully grasp the concept of motivation, several theories were created with different ideologies on attaining the goal of satisfying the needs of an individual. The theories of motivation are categorised according to the factors, which influence individuals to be motivated (content) and how individuals are motivated by those factors (process) (Glasberg & Ouerghemi, 2011). Furthermore, all theories of motivation are based on two broad categories, which are intrinsic and extrinsic motivation. To briefly define the two categories, intrinsic motivation is the result of an individual achieving full satisfaction when completing a particular task whereas extrinsic motivation is where the individual gains satisfaction from receiving a reward for completing a task (Richard & Edward, 2000).

CONTEXT AND RATIONALE- MOTIVATION FOSTERING INNOVATION Suhag, Solangi, Larik, Lakho & Tagar. (2017) concluded that after several studies innovation has a direct positive impact on the performance of the organization as a whole especially as there is a constant change of products and services. As a result, the efficiency, flexibility and productivity of the employee would dictate the level of innovation which any organization can acquire. Motivation achieves employee satisfaction which consequently drives organizational development and performance. Therefore, motivation can be seen as the main component in the implementation of innovation. A Compilation of Working Papers by OECS Scholars

271

In an ever-changing world with advancements in culture and technology the fate of a business relies on a product or service which can satisfy consumers. Consumer satisfaction can be acquired through innovation which foster continuous development and evolution of products and services (Augustin, 2018). Hence organizations require a motivated force to achieve revolutionized innovative products and services. In order to motivate an employee, there are many contributing factors which must be considered. This research will explore each of these factors which include; salary, non-monetary incentives, teamwork, leadership, organizational culture & structure, working environment, personal life, training & development and job design. Each of the factors will be analyzed to identify their effects on job performance and employee satisfaction. After the analysis the relationship between motivation and innovation will be investigated and an implementation plan will be developed to foster innovation through motivation in organizations. Organizations will benefit by developing products and services through innovation and satisfying their employees. For the purpose of this research a conceptual model (see figure 1.1) was developed to establish the relationships which will be analyzed in this study. FIGURE 1: CONCEPTUAL MODEL FOR IDENTIFYING THE RELATIONSHIP BETWEEN INNOVATION IN THE WORKPLACE BY THE USE OF MOTIVATION

272 | Research: The Platform for Innovation, Competitiveness and Growth

AIM AND OBJECTIVES The main aim of this research is to reveal that motivation is a major contributing factor to innovative development and implementation in an organization. Motivation theories will be examined and reference will be made to see how it contributes to creating an innovative working environment. Furthermore, job performance and employee satisfaction will be examined to identify relationships with innovation and motivation. Furthermore, an organization will have an insight on the most influential motivating factors which contributes to the implantation and sustenance of innovation in the workplace. Recommendations will also be made on how an organization can implement innovation with the use of motivation. The following question was asked to fulfill the objectives of the research: 1. Does motivation have a positive effect on job performance and employee satisfaction in the workplace? 2. Does motivation contribute to workplace innovation? 3. What motivational factor contributes most to innovation? 4. What is the best way to implement motivation through innovation? Objective execution will give clear answers to the proposed research question which was highlighted previously (Augustin, 2018). The objectives are as follows: 1. Determine if a relationship exists between motivation, job performance and employee satisfaction. 2. Determine the factor of motivation which is most influential to employee creativity 3. Explain the relationship between motivation and innovation in the workplace. 4. Ascertain whether motivation has a positive effect on the development of innovative workers. Offer insight to an organization on how to achieve innovation in the workplace through employee motivation. CONTENT The project is broken down into five chapters and is as followings; Chapter One- This chapter which introduces the project, provide the aim and objectives and the rationale for this project, Chapter Two- Is a literature review of the main concepts studies which are motivation, innovation, job performance and employee satisfaction, Chapter Three take a look at the methodology used for this research, Chapter Four make comparison through analysis of all concepts examined in this project, Chapter Five concludes the project and makes recommendations for implementation of innovation in an organization.

A Compilation of Working Papers by OECS Scholars

273

CHAPTER TWO - LITERATURE REVIEW MOTIVATION Motivation can be defined as the zeal which influences a person to commit to the desire of satisfying a basic need or want (Pardee, 1990). In other words, motivation takes place when someone makes a pronouncement to perform efficiently at an organization with a target to satisfy different desires which normally come in the form of incentives. Job performance is positively influenced by motivation since many employees attain a sense of satisfaction (Ganta, 2014). INTRINSIC AND EXTRINSIC MOTIVATION Theories and models of motivation have been developed around two types’ motivational factors which are intrinsic and extrinsic. Richard, R. & Edward, D. (2000) states that intrinsic motivation relies on behavioural attributes of an employee which leads to satisfaction such as enjoyment, growth or passion. For instance, an employee can gain satisfaction just by the challenge or the fun which accompanies the job rather than monetary incentives and perks. Furthermore, a passionate employee might continue to work on a piece of software even after working house without even considering compensation. Social demands such as comfort and sustenance of human survival are essentially the most important to an employee therefore a majority of persons are inclined to extrinsic motivation since it provides income which sustains social needs of any individual. Richard et al. (2000) states that extrinsic motivation occurs when employees are more satisfied with the rewards rather than the joy of the task. Since the outcome of doing a task is more important than enjoying the task, extrinsically motivated employees would enjoy the rewards/incentives way more. For instance, an employee will complete a building in a timely fashion to get paid and not really because he/she enjoys designing and architecture. To compare intrinsic motivation is governed by interest whereas extrinsic is governed by result or rewards (see figure 2.1).

FIGURE 2.1: INTRINSIC AND EXTRINSIC MOTIVATION COMPARISON

274 | Research: The Platform for Innovation, Competitiveness and Growth

MOTIVATIONAL THEORIES Many theories have been developed over the years to identify the contributing factors to motivation and they were led by influences such as needs, reinforcement, performance and expectations. Motivational theories are categorized into the content and process theories each of which identifies particular factors influencing motivation. Glasberg & Ouerghemi (2011) mentioned that the content theories focused on attributes influencing human behaviour whereas process theories concentrate on the methods used to drive different outcomes of human behaviour.

CONTENT THEORIES MASLOW’S NEEDS HIERARCHY Maslow’s theory as the name suggests is based on acquiring a certain need in order to influence motivation and it by far one of the most popular motivational theories (Saylor, 2018). It is based on a bottom-up pyramid hierarchical structure (see figure 2.2) - the needs at the bottom of the pyramid (biological and physiological needs) have to be satisfied first before moving up the other needs. Huitt (2007) stated that the needs can be categorized as growth needs; biological and physiological, safety, belongingness and love and esteem or deficiency needs; cognitive, aesthetic and self-actualization where each of the deficiency needs must be satisfied at the lower level before moving up. After the deficiency needs are satisfied the growth needs would be the focus since it gears towards one’s self.

FIGURE 2.2: MASLOW’S NEED HIERARCHY, SAYLOR, 2018

Source: https://saylordotorg.github.io/text_organizational-behaviour-v1.1/s09-theories-of-motivation.html

A Compilation of Working Papers by OECS Scholars

275

ALDERFERS’S ERG THEORY The ERG theory is an amendment to Maslow’s theory. (Saylor, 2018) stated that three classes exist within the theory and they are existence, relatedness and growth (see figure 2.3). Furthermore, unlike Maslow’s theory the needs are not in a bottom-up structure but anyone can be at any level on the pyramid and take an upwards or downwards turn according to influences on the needs. The influence on the needs can either be satisfaction or frustration which can be simulated by emotions to a progressive or regressive manner respectively.

FIGURE 2.3: ALDERFER’S ERG STRUCTURE AS IT RELATED TO MASLOW’S THEORY

MCCLELLAND’S ACHIEVEMENT THEORY Mcclelland’s theory is an expansion of Maslow’s theory but take into consideration three motivators (see figure 2.4); (i) achievement- the need to exhibit superiority or proficiency, (ii) affiliation- the need for feel love, belongingness and relatedness and (iii) power- the need to have complete influence of one’s tasks or other task, essential for accomplishing each need (Huitt, 2007). Saylor (2018) also states that theory is aimed at satisfying needs which are already in existence instead of development and creations of new needs.

276 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 2.4: MCCLELLAND’S ACHIEVEMENT THEORY. WHAT MAKES A GOOD LEADER

Source: http://www.whatmakesagoodleader.com/employee-motivation.html

HERZBERG’S TWO-FACTOR THEORY Herzberg theory is also known as the motivation-hygiene as a result of having two factors which influences motivation, the first factor consists of motivators which foster satisfaction and the second factor is de-motivators (hygiene) which is gear towards dissatisfaction. Tan & Waheed (2011) states that there are five motivators which enable satisfaction and they are achievement, recognition, work itself, responsibility, advancement and five hygiene elements contributing to dissatisfaction and they are company policy and administration, supervision, salary, interpersonal relationships, working conditions. Figure 2.5 illustrated that if the motivating factors are high the employees would be extremely satisfied on the other hand if the hygiene factors are low then employees would be extremely dissatisfied.

FIGURE 2.5: HYGIENE FACTOR SATISFACTION GRAPH

Source:https://www.safaribooksonline.com/library/view/the-little-book/9780273785262 A Compilation of Working Papers by OECS Scholars

277

PROCESS THEORIES SKINNER’S REINFORCEMENT THEORY This theory uses positive and negative stimulus to reinforce or reprimand behaviour in individuals thus an employee’s behaviour is related to the consequences of such behaviour (Maleka, 2015). Furthermore, it is assumed that the behaviour would be repeated if there is a positive consequence whereas for negative consequences it is assumed that there is not repetition. Figure 2.6 shows that positive reinforcement and positive behaviour would stimulate motivation.

FIGURE 2.6: REINFORCEMENT THEORY OF MOTIVATION, MAHLOMOLA STEVENS MALEKA

VROOM’S EXPECTANCY THEORY Motivation is built of expectation in Vroom’s theory and assumes that an individual will behave in a particular manner according to valence which when acted on will lead to rewards based upon instrumentality (Eerde & Thierry, 1996). A mathematical equation was deduced from this theory and is as follows Motivation= Valance X Expectancy X Instrumentality. Consequently, if a factor is not present/ calculates to zero it means that there is no motivation present.

278 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 2.7: EXPECTANCY THEORY EQUATION EXPLAINED, JIM MCGRATH & BOB BATES

Source: https://www.safaribooksonline.com/library/view/the-little-book/9780273785262/html/chapter-033. html

ADAMS’S EQUITY THEORY Huseman, Hatfield & Miles (1987) stated the equity theory stems from exchange and social comparison theories and focuses on the management of the relationship of individuals with other entities. Furthermore, the theory is captured using four objectives; (i) Evaluation of relationship is based on an assessment of the ratio of the inputs to the relationship and the output from the relationship, (ii) Inequity exists if the input/out ratio is not balanced, (iii) higher inequity levels bring greater distress and (iv) greater distress make it more difficult to work toward restoration of equity. Figure 2.8 shows how an employee can analyse equity based on the input and outputs- if the input weighs more than the outputs then an employee will be considered demotivated.

FIGURE 2.8: EQUITY THEORY DIAGRAM. BUSINESS BALLS

Source: https://www.businessballs.com/improve-workplace-performance/adams-equity-theory-on-job-motivation-4045 A Compilation of Working Papers by OECS Scholars

279

LOCKE’S GOAL SETTING THEORY Locke & Latham (1991) mentioned that all living things are governed by goal directed actions therefore in an organization this theory applies since employees are directed by organizational goals. Furthermore, the theory highlights that not all goals are considered purposeful to human beings therefore in the working environment purpose must be given to tasks in order for employees to have to desire to fulfill such goals. Furthermore, setting goals influences performance in an organization since it encourages persistence and strategic development. INNOVATION Innovation was described by Fagerberg (2003) as the commercialization of an idea. Henceforth, innovation happens whenever profit is made in a product that consumers see beneficial. According to Hernandez (2010), innovation and invention are linked; invention is considered as the building of new products while innovation involves generating unique ideas to ensure profit is made with the invention. TYPES OF INNOVATION Innovation can be achieved in one of two ways: Radical Innovation involves new ideas which has never been implemented while Incremental Innovation involves ideas which already exist. The innovation matrix uses the Business Model and Technology model to bring rise to radically new and close to existing innovation. This is displayed in figure 2.9 below.

FIGURE 2.9: INNOVATION MATRIX, BY MANUEL LORENZO HERNÁNDEZ

280 | Research: The Platform for Innovation, Competitiveness and Growth

Incremental Innovation has a lower risk since it involves alteration of existing ideas, products and services. Therefore, it is the most commonly used form of innovation among businesses (Roderkerken, 2011). A classic example of this is the Android Operating System which was originally developed by Google but is being adjusted by numerous mobile companies to suit their own products. Radical innovation, on the other hand, is more challenging for companies and if it fails can cause risks to company’s profit. However, if it is successful, it can be used as a major competitive edge for the business (Roderkerken, 2011). This is because radical invention implements new products and services. This is evident in Apple’s first smartphone invention which contributed to their major success as a mobile company. Technology innovation involves experts with immense knowledge and skill. It can be considered slightly radical because it involves enhancements in technology which causes a significant difference in performance. Patents and other preservation methods of intellectual property is necessary since changes can be unique to the organization or individual. Business model innovation, however, places emphasis on the current operations of the business and creates ways to improve existing business models. The dynamics of competition and marketing are the components required to achieve such innovation (Roderkerken, 201).

INNOVATION STRATEGIES Open or Closed Innovation are the two different approaches used by many organizations and is highlighted in table 1 below:

TABLE 1: PRINCIPLE OF OPEN AND CLOSE INNOVATION COMPARED, JOÃO P. C. MARQUES Closed innovation

Open innovation

1. All the smart people work in our organization.

1. Not all smart people work in our organization.

2. To profit from R&D we have to discover, develop and supply everything ourselves.

2. External R&D can create value for our organization.

3. Only if we discover it will we manage to get it to market first.

3. Internal R&D is needed to grasp that value.

4. If our organization is the first to commercialize an innovation, we will beat our rivals.

4. We have to be involved in basic research to benefit from it, but the discovery does not have to be ours.

5. If we create the most and best ideas in our industr, we will win.

5. If we make better use of external and internal ideas and unify the knowledge created, we will win.

6. If we have fullcontrol over the innovation process, our rivals will not be able to profit from our innovative ideas.

6. We should optimize the results of our organization, combining the sale or licensing of our innovation with the purchase of external innovation processes whenever they are more efficient and economic.

A Compilation of Working Papers by OECS Scholars

281

In Close innovation, all development is done internally and remains within the organization. Consequently, a great investment is made by the business. All concepts and ideas are unique to the business and therefore creates an advantage over competitors. In contrast, Open innovation allows companies to combine both internal and external resources to develop a successful and valuable product (Marques, 2014). Open innovation is more economical and less tedious to a company and as such can be a bigger investment for a business. Sharing of knowledge by organization essentially creates a win-win business environment. INNOVATION PROCESS MODEL To implement innovation in an organization there are many models that can be considered. The classic innovative management model comprises a group of activities which must be done step by step to implement innovation. Acklin (2010) mentioned that the class model takes the form of a funnel since the top most part would require more effort. Figure 2.10 shows that the funnel is made up of idea generation, research and development/ concept, product and market test and implementation.

FIGURE 2.10: CLASSIC INNOVATION MANAGEMENT MODEL-FUNNEL APPROACH, CLAUDIA ACKLIN

JOB PERFORMANCE One of the world’s self-made billionaires Bill Gates quotes “In business, the idea of measuring what you are doing, picking the measurements that count like customer satisfaction and performance... you thrive on that” (Brandon, 2016). Performance entails measuring the results base on the effort put into completing a task. In the context of an individual performance is an amalgamated property of various behaviours over a specific period of time (Weiner, Schmitt & Highhouse, 2012). It is important the performance focuses on behaviours rather than outcomes since employees may find the easiest way to attain the outcome which can be disadvantageous to organizations. Cook (2008) defined job performance as behaviours by employees which can be observed while completing tasks associated with their job which was essential for the attainment of organizational goals. 282 | Research: The Platform for Innovation, Competitiveness and Growth

Job performance is vital for an organization because it directs the productivity of a workplace. Borman and Schmit (1997) mentioned that performance relies on an evaluation aspect of behaviour which is responsible for measuring job performance or performance ratings. In another definition, Bin Shmailan (2016) states that that generally job performance relies on actions and/or behaviours that is pertinent for the success of company goals and there are three factors applicable; (i) job performance should be behaviour oriented rather than results (ii) behaviour has to be aligned with the organization’s goals (iii) job performance exist within multiple dimensions. DIMENSION OF JOB PERFORMANCE Job performance is a construct that cannot be measured directly therefore it is made up of multiple dimensions which comprise of measurable indicators (Koopmans, Bernaards, Hildebrandt, Schaufeli, De Vet & J Van Der Beek, 2011). Figure 2.11 below summarizes the dimension of performance as well as the measurable indicators. The dimensions are as follows:

FIGURE 2.11: HEURISTIC FRAMEWORK OF INDIVIDUAL WORK PERFORMANCE, KOOPMANS, LINDA & BERNAARDS, CLAIRE & HILDEBRANDT, VINCENT & SCHAUFELI, WILMAR & DE VET, HENRICA & J VAN DER BEEK, ALLARD, 2011

A Compilation of Working Papers by OECS Scholars

283

TASK PERFORMANCE This dimension of job performance is aligned with an employee’s proficiency to perform a task relating to the job. Task performance can also be categorized as in-role performance, job-specific task proficiency and technical proficiency and it comprises the quality, quantity and job knowledge towards work (Koopmans et. al, 2011). Campbell, Gasser & Oswald (1996) implied that task performance is multidimensional since there are sub dimensions under each of Campbell’s eight dimensions. The dimensions comprise of the following: 1. Job-specific task proficiency: The ability for an employee to fulfill the core technical requirements. 2. Non-job-specific task proficiency: The ability to perform general tasks which are not specific to the job but is an organizational requirement. 3. Written and oral communications: The ability to converse both verbally or nonverbally to any audience. 4. Demonstrating effort: The commitment which an employee has toward a task and the desire to work feverishly towards accomplishment. 5. Maintaining personal discipline: Avoiding disorder behaviour such as drug abuse, breaking the rules and not being present. 6. Facilitating team and peer performance: The ability to work efficiently and cordially in a group or unit. 7. Supervision: The ability to lead subordinates through direct interaction. 8. Management and administration: The ability to perform non-supervisory tasks which directs the goals of the company. To illustrate the multidimensional aspect of Campbell’s framework, the 8th dimension- management and administration will be examined. Krausert (2009) states that there are sub dimensions stemming from management and administration such as: (i) planning and organizing, (ii) guidance, direction, and motivation for subordinates as well as providing feedback, (iii) training and development of subordinates, (iv) effective communication among employees. CONTEXTUAL PERFORMANCE Krausert (2009) revolves around behaviours which enhances the function of an organization at its current state and proactive behaviours which are aimed at developing the procedures and processes of the organization. Labels for this dimension include extra-role performance and organizational citizenship behaviour just to name a few. All components of contextual performance focus primarily on behaviour and go beyond just completing the organizational goals. For instance, employees can be open to additional tasks, take initiative to complete tasks not issued to them and coaching persons who work with them. Koopmans et. al. (2011) stated that six of Campbell’s eight dimensions can be considered contextual performance and they include: written and oral communications, demonstrating effort, maintaining personal discipline, facilitating peer and team performance, supervision and leadership, and management and administration. ADAPTIVE PERFORMANCE In organizations there are constant changes and normally employees have difficulty in co-mingling with these changes therefore adaptive performance would be the solution to this issue. Adaptive performance occurs when an individual has a set of skills or behaviour which allows consistency with performance when there is an unexpected organizational change (Calarco, 2016). Adaptive performance can either be proactive; foresee future changes thus performing actions to adjust to these future changes, or reactive; deal with unexpected changes by behavioural modifications. For instance, in a proactive work 284 | Research: The Platform for Innovation, Competitiveness and Growth

environment a budget can be put in place for the next five years for new equipment since the current equipment has a lifespan of five years. Jundt, Shoss & Huang (2014) states that the changes can take place internally with modifications in the organizational structure, technology and job requirements therefore employees need to take up new roles, acquire relevant skills and change their overall behaviour to adapt. Furthermore, external changes such as economic, cultural and climate would require adaptation as well. COUNTERPRODUCTIVE PERFORMANCE Counterproductive performance essentially relies on behaviour which is detrimental to the fate of the company. For example, tardiness, engaging in behaviour not relating to the task, drug/substance abuse and theft (Koopmans et. al., 2011). Majority of performance frameworks highlight counterproductive dimensions such as unruliness, destructive/hazardous behaviours, downtime behaviours and absenteeism. Krings & Bollmann (2011) states that counterproductive performance stems from intentional harm done by an individual, in other words it is done purposely and it is different from an accident which leads to harm. Counterproductive performance can have consequences which includes low job satisfaction by victims which eventually influence a high turnover rate at the organization as well as mental and physically related health issues (e.g. exhaustion and depression). In order to curb the effects of counterproductive performance measures such as organization justice, organization rules and working conditions and effective leadership.

EMPLOYEE/JOB SATISFACTION A satisfied employee is typically very happy and content with all aspects of their organizations. Aziri (2011) mentioned that true job satisfaction is an amalgamation of environmental, psychological and physiological attributes both internal and external to the organization. Satisfaction is typically a feeling generated by an individual when their needs are beyond expectation. Job satisfaction is also captured when expectations are met and are combined with real rewards. Organizations can face many consequences if an employee is not satisfied with the work hence issues such as absenteeism, increased incidents and lack of loyalty can arise. Mishra (2013) states that a high level of job satisfaction implies that employees are mentally and motion stable thus they will be committed and effective with the task at hand. Furthermore, job satisfaction is also a good indication that the organization is performing efficiently. Rane (2011) mentioned that there are certain traits and organization will have if the level of employee satisfaction is high and this includes; (i) Opportunity for Grow-employee can get promoted and grow within the organization, (ii) Exceptional Compensationemployees are satisfied with their salary and other benefits such as perks, vacation and affiliations, (iii) Boss is a Mentor- The leaders in the company uplift their subordinates to grow and become successful, (iv) Solid Company Structure- The organization is very well organized and structured which makes processes very streamlined. Job satisfaction is derived from the attitudes of the employees and this attitude is associated with three areas which are job specific factors, character/personality of the employee and external/internal organizational relationships (Mishra, 2013). Moreover, in order to attain complete satisfaction all three attributes must be considered in the analysis. FACTORS OF JOB SATISFACTION Job satisfaction factors is categorized into three groups; personal-focuses on personality and characteristics of the employee, job specific- attributes of the job which can influence satisfaction and organizational- general organization processes and structure influencing employees (Mishra, 2013).

A Compilation of Working Papers by OECS Scholars

285

286 | Research: The Platform for Innovation, Competitiveness and Growth

Personal Factors ● Gender- Women are generally more satisfied with their job than men since they would rather a friendly work environment over exceptional pay whereas a man would be more satisfied with pay since he would be considered his family’s provider. ● Age- A young person would be a little less satisfied with his or her job since their career has just begun compared to an older person who is already settled in the working world. ● Education- More educated people are less satisfied since they would always be looking forward to promotions and changes in their job rather than a less educated person who are sometimes more content. ● Time of Job- Individuals are more satisfied when in the initial phase of the job.

Job Specific Factors ● Type of Work- Greater satisfaction is attained when jobs comprise less routine work and more exciting unpredictable tasks. ● Skills required- If an employee has all the skills necessary to complete a task required by the job they will be more satisfied since they will be successful at completing the task. ● Occupational Status- In the public realm your occupation shows the intelligence, education and financial status of an employee therefore a high-level occupation will indeed promote satisfaction. ● Responsibility- More responsibility given to employees would result in job satisfaction.

Organizational Factors ● Wages- The most important factor in satisfaction is the salary- the higher the salary the more satisfied an employee will be since income provides a means of comfort for an individual. ● Work Conditions- A comfortable working environment is essential for employee satisfaction. For example, office workers would be more comfortable in an air-conditioned area rather than ceiling fans. ● Benefits- This adds more value to the wages therefore benefits such as insurance, pension, relocation packages and allowances will contribute to job satisfaction. ● Security- The more stable and secure a job is the more satisfied an employee will be since they desire steady jobs and income. ● Opportunity for Promotion- Growth is very important to employees therefore advancement in career is really important for satisfaction. The factors of job satisfaction will determine if an employee is satisfied or dissatisfied with a job. Aziri (2011) implies that satisfaction and dissatisfaction will influence the efficiency of an employee (see figure 2.12).

A Compilation of Working Papers by OECS Scholars

287

FIGURE 2.12. DETERMINANTS OF SATISFACTION AND DISSATISFACTION, BRIKEND AZIRI

MEASURING JOB SATISFACTION Job satisfaction is usually measured using research methods such as questionnaires and the most popular measurement techniques are the Minnesota satisfaction questionnaire and job description index (JDI) (Aziri, 2011). Minnesota satisfaction questionnaires can be facilitated in both individuals and in groups and does take gender into consideration. Furthermore, it has twenty work features examined on five different levels which are very satisfied, satisfied, neither satisfied nor dissatisfied, dissatisfied and very dissatisfied. On the other hand, the JDI is a questionnaire with five work features; pay, promotions and promotion opportunities, coworkers, supervision, and the work itself and is answered with either yes, no or indecisive (Mishra, 2013).

288 | Research: The Platform for Innovation, Competitiveness and Growth

CHAPTER THREE - RESEARCH APPROACH The research will be conducted using an inductive approach as a result of the research objective would be investigated using existing theories and research. QPlace (2018) stated that an inductive approach involves exploration of the research topics by discovering and evaluation methods which will eventually lead to a conclusion through precise observation. The inductive research approach does not need an initial theory to conduct research therefore this methodology provides flexibility for the purpose of this research. The initial stage of the research would be to make observations then analyze patterns which derive a tentative hypothesis leading to a theory (see figure 3.1). Furthermore, this research method aims in the development of current theories and practices.

FIGURE 3.1. DEDUCTIVE VS INDUCTIVE RESEARCH METHODS, OTHMAN AYMAN & TAWFIK AHMED

Source: https://www.researchgate.net/publication/243457733_Towards_lean_construction_using_quality_management_as_a_tool_to_minimise_waste_in_the_Egyptian_construction_industry Copyright 2013 Department of Construction Economics and Management University of Cape Town

RESEARCH DESIGN The design of this research will be an amalgamation of causal and exploratory methods which will be used to identify the impact of motivation on creating an innovative organization and the most suitable motivators which can be used to develop innovation. Exploratory research framework will provide the avenue to make ideas and insight discoveries which is essential for finding out the impact of motivation on innovation. Furthermore, casual research will give the ability to identify relationships between different variables. Dowling (2014) mentioned that the casual framework is best suited for cause and effect with an experimental approach for data collection. Table 3.1 shows a comparison between the three main design methods.

A Compilation of Working Papers by OECS Scholars

289

TABLE 3.1. RESEARCH DESIGN COMPARISON, NARESH MALHOTRA Exploratory Objective:

Characteristics:

Methods:

Descriptive

Causal

Discovery of ideas and insights

Describe market characteristics Determine cause and effect or functions relationships

Flexible, versatile

Marked by the prior formulation Manipulation of independent of specific hypotheses variables, measure the effect on dependent variables, Preplanned and structured control mediating variables design

Often the front end of total research design Expert surveys, Pilot surveys, Case studies Secondary data: Qualitative analysis Qualitative research

Secondary data: Quantitative analysis, Surveys, Panels, Observation and other data

Experiments

The two research methods were chosen because of the advantages which they pose. Exploratory design has the following advantages: (i) increased understanding allows the researcher to gain more knowledge on topic which allows better conclusion to be made (ii) Flexible data source allows the utilization of secondary resources such as case studies and discussions to formulate the study (ii) Better findingsince the findings of this method are very extensive it would allow the research to come up with better recommendations and conclusions. On the other hand, the casual design has the following advantages: (i) identifies the reasoning of many processes and impacts of changes on existing theories (ii) provides insights into experimental research. DATA COLLECTION METHOD Data and information will be gathered using qualitative methods on academic areas by the collection of secondary data which consist of existing theories, case studies, statistics from reputable organizations and accredited books and journals. Dowling (2014) mentioned that secondary has prime advantages which makes it really effective in research and those advantages are; (i) gathering data is very time efficient, cheap and effective, (ii) many sources of data are available and (iii) ability to cover a large geographical area.

290 | Research: The Platform for Innovation, Competitiveness and Growth

QUALITATIVE RESEARCH FIGURE 3.2. QUALITATIVE RESEARCH BREAKDOWN, SUSAN DOWLING

Qualitative research allows for a collection of data to be open to interpretation. For the purpose of this research document will be analyzed as well as case studies. Figure 3.2 shows the major advantages of using such an approach. Motivational and innovation theories and other sources of data will be explored as part of this research. RESEARCH PROCESS The process will first start by identifying the theories which will be analyzed to conclude the research. For the case of this research, motivation, innovation, job performance and employee satisfaction theories will be examined. The literature found will be used to provide a clear insight on attain the objectives of the research. The following steps highlights how this research was conducted: Step One Research was conducted on all theories and literature necessary to satisfy the aim and objectives for the research. Information and data will be collected which will be analyzed in line with the construct of the research. Step Two Specific literature was chosen from the research and a structure was formulated to provide solutions to the aim of the research. Ethical issues were addressed by ensuring all literature is referenced. In text citations will be used to ensure that the work is not plagiarized. A Compilation of Working Papers by OECS Scholars

291

Step Three The results will be discussed and comparisons will be made among theories which will generate recommendations and a conclusion for this research.

CHAPTER FOUR - ANALYSIS & DISCUSSION

EFFECTS OF MOTIVATION ON JOB PERFORMANCE

Halisch & Kuhl (1987) emphasizes that motivation and job performance are related since motivation impacts goal-oriented tasks which relies entirely on job performance. Since performance (P) is the rational difference between the quality of work (Q) and the working time (t) then motivational factors would have a positive on each factor (Halisch & Kuhl, 1987). For instance, if an employee is motivated about attaining a set goal, they would want to submit a high quality deliverable in a time manner. Hence it is evident that a majority of motivational factors positively influences the level of performance among employees at the workplace. Hoskins (2014) stated that in order to develop a high performing workplace the office environment is key in attain such goal. It is essential that employees have a choice of office space or a very comfortable working environment to perform very effectively. For example, Facebook allows its employees to create or choose their own work space design specifications to ensure full comfort and productivity at their offices (Hoskins, 2014). Furthermore, Nuru, Islam, Dip & Hossain (2017) concluded in a study that more than 70% of employees believe that both intrinsic and extrinsic motivational factors would improve or increase their overall productivity and performance at accomplishing tasks. Extrinsic motivational factors such as compensation packages, monetary incentives and salary were identified as the highest (over 80%) influencer for increased performance from employees. Other motivational factor such as job enrichments, growth opportunities and work relationships contributing to performance were seen as having a positive relationship among more than 60% of employees (Nuru, Islam, Dip & Hossain, 2017). This reveals that more than half of employees are motivated to perform by both extrinsic and intrinsic motivational factors. Vroom’s Expectancy Theory further supports job performance since it assumes that once an employee is aware that a reward is available for performing a set task, they will be motivated to do it to the best of their ability (Vroom, 1964). For instance, an employee would perform beyond exceptional if they are aware that there is an employee of the quarter reward. Furthermore, performance is the main focus of Vroom’s Theory since it’s based on Effort leading to Performance and Performance leading to Rewards. Trskova (2016) emphasizes in Figure 4.1 that motivation of employees starts with the satisfaction of personal employee goals which then contribute to the success of company goals. Furthermore, the factors of motivation combined with successful stimuli for performance will generate efficiency among employees (Trskova, 2016). In other words, Trskova’s concept uses both intrinsic and extrinsic motivational factors along with employee qualities such as skills and determination to successfully attain an efficiently performing workforce.

292 | Research: The Platform for Innovation, Competitiveness and Growth

FIGURE 4.1. RELATIONSHIP BETWEEN MOTIVATION AND EMPLOYEE PERFORMANCE & EFFICIENCY, KRISTÍNA TRŠKOVÁ

EFFECTS OF MOTIVATION ON JOB SATISFACTION The better working environment not only consists of a productive workforce but a satisfied employee. Job satisfaction encompasses a positive feeling toward a job and includes many factors which contributes to the fulfillment of such satisfaction such as salary, working conditions and leadership. Essentially job satisfaction is related motivation since motivational factors are usually the same. Roos & Eeden (2006) states that since motivational factors focus on the fulfilling the needs of employees as well as improving productivity in the organization, employees will attain great satisfaction once they are motivated. Osakwe (2014) concluded in a study that most employees agree that both intrinsic and extrinsic factors of motivation contribute to the overall job satisfaction. Training and development, promotions and good salaries where the top factors which employees would desire in order to be satisfied. The Society of Human Resource Management (SHRM) conducted a survey in 2017 in the US which identified the greatest contributors to employee job satisfaction to be employee treated with respect by superiors and overall compensation which accounted a level of importance of 65% and 61% respectively (SHRM, 2017). Smerek & Peterson (2006) established in a study that a combination of personality, job characteristics and motivational factors directly contributes to job satisfaction in the working environment (see figure 4.2). Both intrinsic and extrinsic factors contribute to job satisfaction when accompanied with other supporting entities. This concludes and every single factor of motivation would contribute to the satisfaction of any employee in the workplace.

A Compilation of Working Papers by OECS Scholars

293

FIGURE 4.2: MODEL FOR ASSESSING JOB SATISFACTION, RYAN E. SMEREK & MARVIN PETERSON

Popular theories of motivation both support the concept of job satisfaction in various ways by establishing the sources of motivation which in turns satisfies an employee. Maslow’s Theory identifies the needs of an employee in a hierarchical structure and from a bottom-up approach in the hierarchy once the employee is content with attainment of such needs they would be entirely satisfied with their job (Maksic, 2016). In other words, Maslow implies that once the employee is satisfied at each phase of the hierarchy then they would be very content. Herzberg Hygiene Theory also promotes job satisfaction by striking a balance between satisfiers and dis-satisfiers (Attrams, 2013). Furthermore, the satisfiers or motivators must surpass the dis-satisfiers in order for an employee to attain complete job satisfaction (Maksic, 2016). Pardee (1990) states that McClelland’s Theory also supports job satisfaction by allowing an employee to be motivated by a need that they are passionate about. Satisfaction is attainable when an employee uses behaviours such as determination to conquer a certain need/goal which is very rewarding. FACTORS OF MOTIVATION FOSTERING INNOVATION Due to the rapid evolution of business around the world innovation is by far the best means of surviving in a competitive environment. Employees can therefore assist in the development and performance of an organization by disbursing creative ideas to enhance products. The creativity in employees fosters innovation but we often need to think of how to enforce innovation in employees. Ngan (2015) states that motivational conditional stimulates an employee to be innovative by considering intrinsic and extrinsic motivational factors. In order to create an innovative employee, there needs to be focus on the concept of ideation. Motivational factors promote high performing employees in ideation which directly contributes 294 | Research: The Platform for Innovation, Competitiveness and Growth

to innovation (Bergendahl, Magnusson & Björk, 2015). For the purpose of this research the following motivational factors will be examined to identify their contribution to innovation in an organization: COMPENSATION & BENEFITS Hunt (2012) stated that a majority of employees seek reward for being innovative in an organization. Thus, with an ideal compensation model designed to eradicate inconsistencies that may cause envy or unfair treatment among employees, innovation can surely be simulated by compensation. For example, a monetary or symbolic rewards program can be created as an incentive for employees to be more innovative in the organization. Furthermore, employee ideation which fosters innovation can be developed by sharing the profits (bonuses) of the company with employees who came up with the idea for a successful product or service (Ngan 2015). This strategy would create a competitive environment for stimulating innovation since employees would feel valued as a part of the company which ensures their full commitment. Pay per performance is another compensation strategy which influences innovation Employees can also be influenced to innovate by monetary perks which are in addition to the base salary. This reward builds and attracts top innovative talent and can be given as reimbursements for investments in individual education, food and clothing vouchers, entertainment allowances and medical and family support (Bodell, 2014). The perk reward not only influences innovation but also increases the long-term performance of the employee. NON- MONETARY INCENTIVES Money alone cannot address the challenges that organizations face with fostering innovation in employees. Andersen, Murph & Börsch (2016) advised that organizations need to identify other forms of reward by taking a look at behavioural and psychological attributes of an employee. Deloitte Germany (2014) conducted a survey after initiating an innovation contest with employees and 51% of employees agreed that they would desire opportunities to publish or showcase an existing idea. Therefore, it innovation can be influences in the following form of non-monetary incentives: Acknowledge Performance When an employee goes above and beyond to achieve a given task, they not only desire proper compensation but also some form of acknowledgment. Pasmantier (2010) stated that non-financial rewards can positively contribute to the sustenance of employee participation in innovation. This would steal a feeling of belongingness which brings a desire to perform even more. For instance, if an employee is a high performer a project can be assigned specifically to that employee for implementation. Furthermore, employees can be given promotions or even a lunch with the CEO after exceptional performance. All these rewards can contribute extensively to innovation. Recognize Achievement When an employee is recognized for achieving a milestone it can provide a positive psychosocial benefit which can give the employee the zeal to do more thus promoting innovation (Callagher & Smith, 2017). This form of recognition focuses on satisfying the employee intrinsically and comes in many forms such as an increase in salary after completing training, giving days off after completing a very long project, posting an employee on company website or other media to showcase skills publicly and personal thank you notes just to name a few.

A Compilation of Working Papers by OECS Scholars

295

TEAMWORK & COLLABORATION The success of innovation relies on collaboration among employees due to the extreme pace of change therefore the quality of teamwork is definitely needed to win a project (Hoegl, 2008). Innovation requires team members to share information, accept criticism, meet very often and communicate effectively. This way team members are more likely to deal with the unpredictability of an innovative environment. Furthermore, Serinkan & Kızıloğlu (2015) concluded the teamwork significant for implementing innovation in the banking sector since a close neat team would provide the best solutions for competing in the banking through innovation. There is a need for well-designed collaboration to achieve innovative goals since innovations stem from high connected ideas which are dependent on all members of the team (Folkestad & Gonzalez, 2010). Consequently, team members involved in innovative projects would have to physically move to one location to ensure that every aspect of collaboration is covered since virtual or remote collaboration doesn’t have the impact that face to face interaction entails. The innovation process models comprise many different departments which all contribute to the success of an innovative product and service. It is mandatory that each department collaborate in an effort to achieve success in innovation. Open innovation requires an organization to bind forces with another separate organization to achieve innovation. Furthermore, if those companies do not collaborate effectively then the innovative product will most likely fail and both companies can surfer major losses. Consequently, the only form of protection from much failure is teamwork across both organizations (Kirschbaum, 2005). LEADERSHIP The fate of any organization relies on the valor of its leaders therefore good leadership is essential for any innovative organization. Steve Jobs legendarily highlighted that innovation really determines a leader from a follower (Llopis, 2014). An innovative leader normally has specific attributes which determine their success; they have a passion for what they do, they explore all options, they believe in collaboration and they are courageous enough to take risks to attain an innovation. For instance, Elon Musk of Space X and Tesla Motors almost went bankrupt but decided to use his own cash and invest in his vision while persuading stakeholders to believe in his products. Currently Elon Musk obtained a fortune by engaging his team and collaborating with stakeholders to keep his vision alive (Llopis, 2014). Lukowski (2017) mentioned that styles of leadership are very important for ideation and implementation of innovation. A combination of participative, charismatic and transformational leadership styles are ideal for innovation. Participative leadership decision making influences both the leader and subordinates which fosters the stimulation of creativity thus contributing to the growth of new ideas (Lukowski, 2017). In addition, charismatic leadership allows a leader to have exceptional character which normally attracts employees to want to be a part of a project. As a result, employees would be more engaged and willing to collaborate, which is essential for innovation. Transformational leadership would promote change in an organization by inspiring employees to idealize and implement change in the organization. An innovative leader is essential for the success of any innovative organization. ORGANIZATIONAL CULTURE In order to develop and implement innovation an organization would have to comprise certain characteristics which defines its culture. This culture has to adapt to continuous change and development therefore the organization environment has to cater for ideation, research and development, collaboration, continuous learning, incentives based on ideations and a democratic culture. Cancialosi (2017) identified three elements which would foster an organization’s innovative culture: ● Environment- A suitable environment would create an open door for ideation thus promoting collaborative efforts and challenge employees to be resilient and determined. Furthermore, there must be a level of freedom to experiment where failure is accepted and employees are given a chance to recuperate from such failure.

296 | Research: The Platform for Innovation, Competitiveness and Growth

● Talent – Continuous training and development is essential for innovation since the rapid change would require rapid growth of knowledge therefore employees should be able to absorb content quickly. It is also essential for employees to be placed in different teams which would enhance all their skills through collaboration. ● Process-Specific processes should be formulated to identify and generate ideas which can be developed through innovation. With a reasonable process then there might be a lack in focus when implementing an innovative project. Organizations can have programs or plans to develop innovation but without the proper culture to support then most often failure is the result of that development. JOB DESIGN To attain the true capability of a creative employee it is essential that their job is designed to suit such expectation. One of the challenges faced by organizations is mobilizing employees to attain innovative potential. Employees are constantly faced with the task of troubleshooting issues, finding solutions and the implementation of innovative creativity to master the task. De Spiegelaere, Van Gyes & Hootegem (2012) stated the main focal point for employee innovativeness is the characteristics of the job. Essentially job characteristics are defined by the job resources and job demand. Job resources take a look at the attributes of work which are essential for achieving goals focusing on the reduction of cost and stimulation of learning and development among employees. Job demands are the attributes of a job that has to be sustained which relies on the physical and psychological skills of an employee (De Spiegelaere, Van Gyes & Hootegem, 2012). Werleman (2016) states that employees engaged in innovative working environments are normally exposed to high intrinsic motivation and this motivation gives an experience of enriched job design. An enriched job design typically gives an employee an insight of how meaningful the work contributes to their overall satisfaction. Furthermore, when an employee’s psychological needs are satisfied, they are more likely to be very active and more involved in any changes and developments which are common to innovative organizations. TRAINING & DEVELOPMENT Employees drive the success of every company therefore specific talents are needed to foster growth and development. Consequently, for an innovative organization, specific talents such as new thinking, ideation, problem solving and radical/incremental implantation is key in ensuring the organization’s success. Innovation & Business Skills Australia (2009) identified a framework for employees to acquire the skills necessary for innovation and its call innovation@work skills. These skills give employees the ability to design new technologies and policies as well as adapting to a changing environment. Furthermore, it assists giving the employee the ability to generate new ideas and find ways to implement and make it better. The framework encompasses six essential skills which are needed for innovation; (i) Interpretidentification of opportunities or the need and research on the possibilities. (ii) Generate- the process of creative thinking of ideas and choosing the best one. (iii) Collaborate- working with other entities to generate and obtain feedback on ideas. (iv) Reflect- consideration of thoughts and feedback to improve the quality of the development. (v) Represent- the final presentation of the idea which is a product from all the previous steps. (vi) Evaluate- complete requirements for design/ide and finding the best practices to commence testing from the practicability of the idea. Studies show that more than 70% of employees believe that training and development is of great importance. Furthermore, in that same study, 89% of employees believe that training will foster creativity development and allow them to be more flexible at problem solving and 95% think that it will improve job performance (Edralin, 2011). Innovation can thus be stimulated by good human resource management which can align the proper training for employees. For instance, the ideal training methodology for A Compilation of Working Papers by OECS Scholars

297

innovation would be hands-on training on the job and team training which encompasses team building and ideation (Edralin, 2011). In an innovative environment training and development is critical for the survival of the organization. WORK ENVIRONMENT The ideal working environment for an innovative organization is one which fosters creativity among employees. There are many factors which generate creativity at a working place and the comfort and layout of the work environment is most important. Amabile, Conti, Coon, Lazenby & Herron (1996) concluded that rather than focusing on the personality characteristics of employees to foster creativity the working environment can be altered to generate and implement creative ideas to sustain an innovative organization. Walter (2012) mentioned that the greatest deterrent for fostering creativity in an innovative workplace are the lack of freedom, high pressure environment and the fear of taking risks. On the other hand, other factors such as adequate resources, a group work and team spirited environment, encouraged supervision add value to creativity. Furthermore, Walter (2012) amended a model for assessing the influences of harnessing creativity by adding a cultural aspect and the physical working environment (see figure 4.3). The culture of any society an organization is established in is essential since many persons from different ethnic backgrounds may have a slight difference in the factors of motivation which simulates innovation. Another important factor which drives creativity is the physical work environment such as offices, building face and general comfort of the work space.

FIGURE 4.3. MODEL OF UNDERLYING ASSESSMENT OF PERCEPTION OF THE WORK ENVIRONMENT FOR CREATIVITY, CHRISTIAN WALTER

298 | Research: The Platform for Innovation, Competitiveness and Growth

INNOVATIVE WORK SPACES Wagner & Watch (2017) mentioned that innovative spaces have revolutionized from being stylish to helping employees develop to achieve ambition and creativity. Innovation transforms space into a more open concept where different professions can collaborate and still have their departmental space as well as providing comfort and technology to harness ideation. Furthermore, Wagner & Watch (2017) highlighted some key findings about innovative spaces (see figure 4.4) and they are as follows: Openness – This translates to a physical office design which is very flexible and can be responsive to configuration which works well for anyone. Collaboration – The office should allow effortless collaboration among all employees therefore is needed to foster face to face communication among all levels of employees. Technology – Technology must be available which enables seamless research, manipulation of data and information as well as collaboration.

FIGURE 4.4. INNOVATIVE OFFICE SPACE, JULIE WAGNER & DAN WATCH

A Compilation of Working Papers by OECS Scholars

299

CONCEPTUAL MODEL FOR IMPLEMENTING INNOVATION BY MOTIVATION This research focuses on solely how innovation can be conquered using motivational techniques. The findings found from the use of empirical evidence from journals, websites and other literature it was deduced that a positive relationship exists between motivation and innovation. Furthermore, since job performance and employee satisfaction are two contributing factors to innovation, evidence shows the both entities are sub components of motivation. Additionally, eight (8) motivational factors were noted to have a very high influence on innovation. A conceptual model was created to show the true relationship among motivation and innovation (see figure 4.5). This model can be used to implement innovation in a workplace since it highlights all the attributes required to simulate innovation through motivating employees.

FIGURE 4.5. CONCEPTUAL MODEL FOR IMPLEMENTING INNOVATION IN THE WORKPLACE BY THE USE OF MOTIVATION

Source: Kurt Augustin, 2018

300 | Research: The Platform for Innovation, Competitiveness and Growth

CHAPTER FIVE - CONCLUSIONS & RECOMMENDATIONS

CONCLUSION The relationship between motivation and innovation is sufficiently covered in various researches and literature since many theories and surveys were available for analysis. A majority of the existing research supported the relationship therefore it was relatively simple to accomplish our aims and objectives. In order to examine the relationship between motivation and innovation two additional concepts (job performance & employee satisfaction) were analyzed since it was assumed that both of these concepts positively contributed to motivation and innovation. To assess the relationship a conceptual model was created (see figure 4.5) by choosing eight factors of motivation, identify their relationships with job performance and employee satisfaction and establishing the connection with innovation. The findings revealed the following: Effects of Motivation on Job Performance In the analysis of job performance literature shows that a positive relationship exists with motivation. Since motivation institutes factors which satisfy the needs of individuals, employees are inclined to perform to sustain those needs. Research has also shown that almost all employees are influenced by intrinsic motivation which normally includes factors such as salary and compensation. Furthermore, a majority of employees are motivated extrinsically; work relationships & work environment. In addition, Vroom’s motivation theory also concluded that performance leads to rewards and incentives. Effects of Motivation on Employee Satisfaction Employee satisfaction is positively influenced by motivation since the factors of motivation normally gratifies the need of an employee therefore satisfaction is generated. Studies have shown that more than 60% of employees get satisfaction from salary and good leadership which are all factors of motivation. Furthermore, other literature illustrated that the concatenation of personality, job design and motivational factor equate to job satisfaction. In further support, Herzberg’s and McClelland’s motivational theories are solely based on satisfaction. Factors of Motivation Contributing to innovation A positive relationship was also established between motivational factors and innovation since a majority of the factors foster creativity which is mandatory for the implementation of innovation. Furthermore, since it was established that job performance is related to motivation then normally a high performing individual will be motivated to create ideas which will contribute to innovation. The eight motivational factors all have a positive influence on innovation and they all are equally weighted in fostering innovation. All of the relationships between all concepts show positivity therefore it can be concluded that motivation has a positive impact on organizational innovation. Since change in this world is constant, innovation will soon be mandatory to all businesses therefore business should take the opportunity to implement innovation through motivation. RECOMMENDATIONS In order for an organization to successful implement innovation three recommendation will be made which stems from the findings of this research: Step One: Analyze Motivational Factors The organization should analyze each motivational factor mentioned in the model conceptualized in A Compilation of Working Papers by OECS Scholars

301

figure 4.5. The organization should ensure that every factor is implemented or evolved to achieve full motivation. Until these factors have not been fulfilled by the organization entirely they cannot move to the next step of implementation. Step Two: Performance and Satisfaction Check After completing the first step the organization should confirm that there is an increase in job performance and employee satisfaction. This is critical because this signifies that the organization is moving towards an innovative culture. If the checks are successful, then they move to the next step. Otherwise, the organization should revisit step one to check for any inconsistencies. Step Three: Innovation Process Implementation According to the nature of the organization a suitable innovative process framework should be implemented and then the organization can begin to move to an innovative direction.

REFERENCES

Acklin, C. (2010). Design Driven Innovation Process Model. Design Management Journal, 5, 50 - 60. doi:10.1111/j.1948-7177.2010.00013.x. Aleksandra S., Ristovska, N.& Gramatnikovski, S. (2015). Innovative Business Models a Factor for Competitive Advantage of the Companies. UTMS Journal of Economics, 6(1), 135–146. Retrieved from Amabile, T., Conti, R., Coon, H., Lazenby, J. & Herron, M. (1996). Assessing the Work Environment for Creativity. The Academy of Management Journal, 39(5), 1154-1184. Retrieved from http://people.wku.edu/richard.miller/ amabile.pdf. Andersen, N., Murph, T & Börsch, A. (2016). Nothing for Money: A Behavioural Perspective on Innovation and Motivation. Retrieved from https://www2.deloitte.com/insights/us/en/deloitte-review/issue-18/cultivating-innovationat-work.html Attrams, R. (2013). Motivation and Employee Satisfaction: Perceptions of Workers in Public and Private Health Care Facilities. Retrieved from http://ugspace.ug.edu.gh/ Augustin (2018). Research Proposal for MBA Project. Retrieved from https://ev.turnitinuk.com/app/carta/ en_us/?u=9537993&student_user=1&o=91774355&lang=en_us Aziri, B. (2011). Job Satisfaction: A Literature Review. Retrieved from http://mrp.ase.ro/no34/f7.pdf Bergendahl, M., Magnusson, M., & Björk, J. (2015). Ideation High Performers: A Study of Motivational Factors. Creativity Research Journal, 27(4), 361–368. https://doi.org/10.1080/10400419.2015.1088266 Bin Shmailan, A. (2016). The relationship between job satisfaction, job performance and employee engagement: An explorative study. Business Management and Economics, 4(1), 1-8. Retrieved from http://dx.doi.org/10.15739/ IBME.16.001 Bodell, L. (2014). Non-Traditional Ways to Reward Innovative Employees. Retrieved from https://www.tlnt. com/11-non-traditional-ways-to-reward-innovative-employees/ Borman, W. C., & Motowidlo, S. J. (1993). Expanding the criterion domain to include elements of contextual performance. San Francisco: Jossey-Bass

302 | Research: The Platform for Innovation, Competitiveness and Growth

Brandon, J. (2016). 25 Quotes from Bill Gates on How to Succeed. Retrieved from https://www.inc.com/johnbrandon/25-of-the-best-bill-gates-quotes-on-success.html Bushiri, C. (2014). The Impact of Working Environment on Employees’ Performance: The Case of Institute of Finance Management in Dar Es Salaam Region. Retrieved from http://repository.out.ac.tz/608/1/MHRM-DISSERTATION.pdf Calarco, H. (2016). Measuring the Relationship between Adaptive Performance and Job Satisfaction. Retrieved from https://jewlscholar.mtsu.edu/bitstream/handle/mtsu/5006/Calarco_mtsu_0170N_10615. pdf?sequence=1&isAllowed=y Callagher, L. & Smith, P. (2017). Innovation Awards: Reward, Recognition, and Ritual. International Journal of Innovation Management, 21(5), 1-19. doi: 10.1142/S1363919617400060 Campbell, J. P., Gasser, M. B., & Oswald, F. L. (1996). Individual Differences and Behaviour in Organizations. San Francisco: Jossey-Bass. Cancialosi, C. (2017). Why Culture Is the Heart of Organizational Innovation. Retrieved from https://www.forbes. com/sites/chriscancialosi/2017/02/07/why-culture-is-the-heart-of-organizational-innovation/#6baadef13f4d Capgemini. (2017). The Digital Culture Challenge: Closing the Employee-Leadership Gap. Retrieved from https:// www.capgemini.com/consulting/wp-content/uploads/sites/30/2017/07/dti_digitalculture_report.pdf Cook, A. (2008). Job Satisfaction and Job Performance: Is The Relationship Spurious? Retrieved from https://core.ac.uk/download/pdf/4277203.pdf De Spiegelaere, S., Van Gyes, G. & Hootegem, G. (2012). Job Design and Innovative Work Behaviour: One Size Does Not Fit All Types of Employees. Journal of Entrepreneurship, Management and Innovation, 8(4), 5-20. doi: 10.7341/2012841 Deloitte Germany (2014). Deloitte Germany Innovation Contest. Retrieved from http://dupress.com Dowling, S. (2014). An Investigation into The Benefits and Barriers of Shopper Marketing and the Direct Impact it has on the Consumer Buying Decision Process at Point Of Purchase Sales in the FMCG Sector in Generation X Females. Retrieved from https://esource.dbs.ie/bitstream/handle/10788/2182/ba_dowling_s_2014.pdf Edralin, D. (2011). Training and development practices of large Philippines companies. Asia Pacific Business Review, 17(2), 225-239. doi:10.1080/13602381.2011.533501. Eerde, W. & Thierry, H. (1996). Vroom’s Expectancy Models and Work-Related Criteria: A Meta-Analysis. Retrieved from https://pdfs.semanticscholar.org/e769/32a547cfb064ff641b17e5427184ea71fd60.pdf Fagerberg, J. (2003). Innovation: A Guide to the Literature. Retrieved from https://smartech.gatech.edu/bitstream/ handle/1853/43180/JanFagerberg_1.pdf Folkestad, J. & Gonzalez, R. (2010). Teamwork for Innovation: A Content Analysis of the Highly Read and Highly Cited Literature on Innovation. Advances in Developing Human, 12(1), 115-136. doi: 10.1177/1523422310365486 Ganta, V. C. (2014). Motivation in the Workplace to Improve the Employee Performance. International Journal of Engineering Technology, Management and Applied Sciences, 2(6), 221-230. Retrieved from http://www.ijetmas. com/admin/resources/project/paper/f201411201416479373.pdf Glasberg, R. & Ouerghemi, K. (2011). Innovation in Human Resources. International Conference on Economics, Business and Management, 22, 7-11. Retrieved from http://www.ipedr.com/vol22/2-ICEBM2011-M00004.pdf Halisch, F. & Kuhl, J. (1987). Motivation, Intention, and Volition (1st Ed.). Germany: Springer-Verlag.

A Compilation of Working Papers by OECS Scholars

303

Hernández, M. L. (2010). Basic Concepts of Innovation and Innovation Management. Retrieved from https://www. etsisi.upm.es/sites/default/files/Avisos/ModuloII.pdf Hoegl, M. (2008). Teamwork and innovation. Retrieved from https://www.pmi.org/learning/library/ teamwork-innovative-projects-7108 Hoskins, D. (2014). Employees Perform Better When They Can Control Their Space. Retrieved from https://hbr. org/2014/01/employees-perform-better-when-they-can-control-their-space Huitt, W. (2007). Maslow’s hierarchy of needs. Retrieved from http://www.edpsycinteractive.org/topics/regsys/ maslow.html Hunt, M. (2012). Use Compensation to Inspire Innovation. Retrieved from https://www.shrm.org/resourcesandtools/ hrtopics/compensation/pages/usecompensationtoinspireinnovation.aspx Huseman, R., Hatfield, J. & Miles, E. (1987). A New Perspective on Equity Theory: The Equity Sensitivity Construct. Retrieved from https://pdfs.semanticscholar.org/1d65/68e33f2ffcccf76d9c5b0a81657389d675cb.pdf Innovation & Business Skills Australia. (2009). Developing Innovation Skills. Retrieved from https://oce.uqam.ca/ wp-content/uploads/2015/01/1309_developing_innovation_skills.pdf Intel. (2014). Workplace Transformation. Retrieved from https://www.intel.com/content/dam/www/public/us/en/ documents/white-papers/workplace-transformation-vision-paper.pdf Jundt, D., Shoss, M. & Huang, J. (2014). Individual adaptive performance in organizations: A review. Journal of Organizational Behaviour, 36, 53-71.doi: 10.1002/job.1955 Kirschbaum, R. (2005). Open Innovation in Practice. Research-Technology Management, 48. 24-28. 10.1080/08956308.2005.11657321. Koopmans, L., Bernaards, C., Hildebrandt, V., Schaufeli, W., De Vet, H. & J van der Beek, A. (2011). Conceptual Frameworks of Individual Work Performance. Journal of occupational and environmental medicine / American College of Occupational and Environmental Medicine, 53(8) 856-866. doi: 10.1097/JOM.0b013e318226a763. Koudelková, P. & Milichovský, F. (2015). Successful Innovation by Motivation. Business: Theory and Practice, 16(3), 223–230. doi:10.3846/btp.2015.472 Krausert, A. (2009). Performance Management for Different Employee Groups. Berlin: Springer Verlag. Krings, F. & Bollmann, G. (2011). Managing Counterproductive Work Behaviours. Retrieved from https://www. researchgate.net/publication/266796582_Managing_counterproductive_work_behaviours Llopis, G. (2014). 5 Ways Leaders Enable Innovation in Their Teams. Retrieved from https://www.forbes.com/sites/ glennllopis/2014/04/07/5-ways-leaders-enable-innovation-in-their-teams/#2567d7758c4c Locke, E. & Latham, G. (1991). A Theory of Goal Setting & Task Performance. The Academy of Management Review, 16(2), 212-247. doi:10.2307/258875. Lukowski, W. (2017). The Impact of Leadership Styles on Innovation Management. Marketing of Scientific and Research Organizations, 24(2), 105-136. doi: 10.14611/minib.24.06.2017.12 Maksic, F. (2016). Employee Motivation and Satisfaction: The Case of Clinical Centre University of Sarajevo. Retrieved from http://www.cek.ef.uni-lj.si/magister/maksic2075-B.pdf Maleka, S. (2015). Dissertation -Critical Assessment for the Perfomance Management System. Retrieved from https://www.researchgate.net/publication/272692748_Dissertation_-Critical_Assessment_for_the_Perfomance_

304 | Research: The Platform for Innovation, Competitiveness and Growth

Management_System_2 Mark G. Resheske, M. (2001). A Descriptive Study of Job Satisfaction and Its Relationship with Group Cohesion. Retrieved from http://www2.uwstout.edu/content/lib/thesis/2001/2001resheskem.pdf Marques, J. (2014). Closed versus Open Innovation: Evolution or Combination? International Journal of Business and Management, 9(3), 196-203. doi:10.5539/ijbm.v9n3p196 McGrath, J & Bates, B. (2015). The Little Book of Big Management Theories. Retrieved from https://www. safaribooksonline.com/library/view/the-little-book/9780273785262/html/chapter-033.html Mishra, P. (2013). Job Satisfaction. Journal of Humanities and Social Science, 14(5), 45-54. Retrieved from http:// www.iosrjournals.org/iosr-jhss/papers/Vol14-issue5/F01454554.pdf Ngan, P. (2015). Organizational Innovativeness: Motivation in An Employee’s Innovative Work Behaviour. Scientific Bulletin – Economic Sciences, 14, 86-97. Retrieved from http://economic.upit.ro/RePEc/pdf/2015_3_10.pdf Nuru, N, Islam, M., Dip, TM., & Hossain, AA. (2017). Impact of Motivation on Employee Performances: A Case Study of Karmasangsthan Bank Limited, Bangladesh. Arabian Journal of Business and Management Review, 7(1), 1-8. doi:10.4172/2223-5833.1000293 Osakwe, R. (2014). Factors Affecting Motivation and Job Satisfaction of Academic Staff of Universities in SouthSouth Geopolitical Zone of Nigeria. International Education Studies, 7(7), 43-51. doi:10.5539/ies.v7n7p43 Pardee, R. (1990). A Literature Review of Selected Theories Dealing with Job Satisfaction and Motivation. Retrieved from https://files.eric.ed.gov/fulltext/ED316767.pdf Pasmantier, J. (2010). Incentives for Innovation. Retrieved from https://hr.toolbox.com/blogs/jpasmantier/ incentives-for-innovation-022511 QPlace, (2018). The Inductive Method. Retrieved from https://www.qplace.com/wp-content/uploads/2012/05/ Inductive-Method.pdf Rane, D. (2011). Employee Job Satisfaction: An Essence of Organization. Retrieved from https://pdfs. semanticscholar.org/defa/659c3154666b03c3c7d7399c0b794ce7edef.pdf Richard, R. & Edward, D. (2000). Intrinsic and Extrinsic Motivations: Classic Definitions and New Directions. Contemporary Educational Psychology, 25, 54–67. doi:10.1006/ceps.1999.1020 Roderkerken, J. (2011). Employee- Driven Innovation; The manager‟s guide to EDI. Retrieved from https://www. theseus.fi/bitstream/handle/10024/32972/Roderkerken_Jules.pdf?sequence=1 Roos, W. & Eeden, R. (2006). The Relationship between Employee Motivation, Job Satisfaction and Corporate Culture. SA Journal of Industrial Psychology, 34(1), 54-63. Retrieved from http://www.scielo.org.za/pdf/sajip/ v34n1/06.pdf Saylor. (2018). Organizational Behaviour. Retrieved from https://www.saylor.org/site/textbooks/Organizational%20 Behaviour.pdf Serinkan, C. & Kızıloğlu, M. (2015). Innovation Management and Teamwork: An Investigation in Turkish Banking Sector. Journal of Management Policies and Practices, 3(1), 94-102. doi: 10.15640/jmpp.v3n1a11 SHRM (2017). Employee Job Satisfaction and Engagement: The Doors of Opportunity Are Open. Retrieved from https://www.shrm.org/hr-today/trends-and-forecasting/research-and-surveys/pages/2017-job-satisfaction-andengagement-doors-of-opportunity-are-open.aspx

A Compilation of Working Papers by OECS Scholars

305

Smerek, R. & Peterson, M. (2006). Examining Herzberg’s Theory: Improving Job Satisfaction among Non-Academic Employees at a University. Research in Higher Education, 48(2), 229-250. doi: 10.1007/s11162-006-9042-3 Suhag, A., Solangi, S., Larik, R., Lakho, M. & Tagar, A. (2017). The Relationship of Innovation with Organizational Performance. International Journal of Research – Granthaalayah, 5(2), 292-306. https://doi.org/10.5281/ zenodo.345736. Tan, T. & Waheed, A. (2011). Herzberg’s Motivation-Hygiene Theory and Job Satisfaction in the Malaysian Retail Sector: The Mediating Effect of Love of Money. Retrieved from https://mpra.ub.uni-muenchen.de/30419/ Trskova, K. (2016). Motivation, Performance and Efficiency. Retrieved from https://www.researchgate.net/ publication/291765574_motivation_performance_and_efficiency Vroom, V. (1964). The motivation to work. New York: John Wiley Wagner, J. & Watch, D. (2017). Innovation Spaces: The New Design of Work. Retrieved from https://www.brookings. edu/wp-content/uploads/2017/04/cs_20170404_innovation_spaces_pdf.pdf Walter, C. (2012). Work Environment Barriers Prohibiting Creativity. Procedia - Social and Behavioural Sciences, 40, 642-648. doi: 10.1016/j.sbspro.2012.03.243 Weiner, I., Schmitt, N. & Highhouse, S. (2012). Handbook of Psychology, Vol. 12: Industrial and Organizational Psychology (2nd ed). New Jersey, US: Wiley. Werleman, A. (2016). The Effect of Enriched Job Design on Innovative Work Behaviour. Retrieved from http:// arno.uvt.nl/show.cgi?fid=142234

306 | Research: The Platform for Innovation, Competitiveness and Growth

About the Author

Kurt Augustin

Business Consultant KPA Technologies

Kurt Augustin is a Business and IT Consultant who amalgamates his experience in innovation, marketing, project management and technology to develop a very impactful and successful product or service. Kurt Augustin is the founder of a technology company (KPA Technologies) which aims to service the entire Caribbean region with solutions such as IT Consultancy, Website and Application Design, Internet Security, Business Development through Technology and Project Management. Mr. Augustin holds a Bachelors in Information Technology from the University of the West Indies St Augustine. He also obtained one of the world’s best internet security certifications (CompTIA Certified Advance Security Practitioner) which allows him to critically analyze and develop policies which motivates the security and success of any organization. He holds an MBA with specialization in Innovation and Project Management from a reputable Scottish University where he was able to conceptualize a theory which illustrated the relation between motivation and innovation.

A Compilation of Working Papers by OECS Scholars

307

308 | Research: The Platform for Innovation, Competitiveness and Growth