Issuu on Google+

Economic Papers and Notes Volume 14 (1)

North American Review of Finance Volume 12, Number 4


Economic Papers and Notes Volume 14, Number 1

Articles Heterogeneous Panel Causality between FDI and Growth in COMESA Countries Arcade NDORICIMPA ………………………………….………………………………..1 - 13 Export-Led Growth in Europe: Where And What to Export? Paula Gracinda Santos, Ana Paula Ribeiro & Vitor Manuel Carvalho…14 – 27 Empirical Examination of Impact of Mobile Phone Usage on the Well-being of Fishermen in Rural Bangladesh Md. Sadik Pavela, Seikh Ruksana Burhan & Muktadir Mahfuj Jamee……………………………………………………………….28 – 57 The Observational Equivalence of Commons and Anti-Commons Bingyuan Hsiung…………………………………...…………………………..…………58 - 63

Copyright © 2014 by North American Academic Journals


Heterogeneous Panel Causality between FDI and Growth in COMESA Countries

Arcade NDORICIMPA1

Abstract

an

d

N

ot e

s

This study examines the causal links between FDI and economic growth in COMESA countries. Granger causality tests in heterogeneous panels recently developed by Dumitrescu and Hurlin (2012) are applied. The results show that Homogeneous Non Causality hypothesis is rejected for both causal directions. The results further suggest that FDI Granger cause economic growth only for Libya where the causal impact was found to be negative. For the rest of the countries, FDI does not affect growth. Concerning the reverse causality, the findings indicate that Growthdriven FDI hypothesis is only evident in Comoros. The findings imply that COMESA Countries should strive to improve their absorptive capacity for significant positive growth effects of FDI to be realized. JEL Classifications: F23, F36, F43, C23

Ec

on

om

ic

Pa

pe

rs

Keywords: COMESA, FDI-led Growth hypothesis, Growth-driven FDI hypothesis, Granger causality, Heterogeneous Panels

1

Department of Economics, University of Dar Es Salaam (Tanzania) Email: arcade_ndoricimpa@yahoo.com

1

North American Academic Journals, 2014


1. Introduction

s

Due to imbalances between domestic saving and investment, countries are competing in attracting foreign saving especially Foreign Direct Investment (FDI). COMESA (Common Market for Eastern and Southern Africa) countries have not been left out of the race and have been attracting a significant amount of FDI although unequally distributed across countries with Egypt, Libya and Sudan being the largest FDI recipients and, Comoros, Burundi and Eritrea being the least recipients of FDI in the region2. This unequal distribution is due to the differences in pull factors such as endowments in natural resources, skilled labor and infrastructure, differences in market size, differences in costs of labor but also the differences in the host government’s policy framework, business facilitation activities and business conditions (Metha, 2007, Anyanwu, 2012).

N

ot e

FDI net inflows to COMESA have been increasing decade by decade. For COMESA as a whole, the region attracted FDI amounting on average to USD 189.7 Million in the 1970s, USD 788.3 Million in the 1980s, USD 1477.7 Million in the 1990s and USD 11203.6 Million for 200020113.

rs

an

d

It should be noted that promoting FDI is one of strategies proposed by the New Partnership for Africa’s Development (NEPAD) for African countries to boost their economic growth and progress towards the Millennium Development Goals (MDGs) (Asiedu, 2004). The question one would ask is, “Does FDI promote economic growth in COMESA countries?”

on

om

ic

Pa

pe

Two hypotheses are found in the literature on FDI-growth nexus, namely, FDI-led growth hypothesis and Growth-driven FDI hypothesis (Klasra, 2011; Yalta, 2013). The proponents of the first hypothesis suggest that FDI promote economic growth in the host countries. They argue that FDI increases not only the volume of investments but also its efficiency. Moreover, FDI promotes economic growth in the host country through technology transfer, diffusion, and spillover effects (Nair-Reichert and Weinhold, 2001; Driffield and Jones, 2013). And as a report of UNCTAD (1999) emphasizes, FDI boosts economic growth of host countries because of its impact on financial resources and investment in the host country, through its ability to enhance the technological capabilities in the host economy; boost export competitiveness in the host country; generate employment and strengthen the skills base in the host country, etc.

Ec

However as the report of UNECA (2006) and Farkas (2012) point out, the impact of FDI on growth is dependent on the host countries’ absorptive capacity, determined by factors such as the level of technology used in domestic production in the host country, the extent to which the financial sector is developed, the host country’s human capital quality, the degree of openness etc. Thus, countries with low absorptive capacity might not enjoy the potential FDI growth effects. In addition, according to UNCTAD (2001), for Multinational Companies (MNCs) operating in the primary sector, FDI might not affect growth since in this case the linkages between MNCs and domestic firms are limited, as a consequence the benefits FDI is supposed to bring to the host country are not transmitted, hence limited spillover effects.

2 3

By analyzing data from United Nations Conference on Trade and Development (UNCTAD), online database Data from UNCTAD statistics, online database

Economic Papers and Notes, vol.14, no.1

2


Moreover, Akinlo (2004), Fortanier (2007) and Moura and Forte (2010) argue that FDI can also have a negative impact on economic growth of the host country. They give a number of reasons for their argument. First, they argue that FDI may negatively affect the host country’s economic growth if profits and dividends of MNCs are remitted back to the home country since this tends to negatively affect the host country’s capital account. Second, FDI may have a negative impact on growth if transnational corporations (TNCs) obtain substantial concessions from the host country. Thirdly, FDI may also have a negative impact on growth by crowding out domestic savings and investment if MNCs are operating in imperfectly competitive sectors.

s

Economic growth can also affect FDI especially market-seeking FDI where MNCs intend to capture the local market, which is known as growth-driven FDI hypothesis (Zhang, 2001, Agiomirgianakis et al., 2006).

N

ot e

A number of studies have investigated the links between FDI and economic growth using different methodologies for developed and developing countries. These studies can broadly be categorized into two groups, namely; country specific studies using time series data and panel data studies.

pe

rs

an

d

Among the studies which used time series data, one finds Balamurali and Bogahawatte (2004), Ayanwale (2007), Samad (2009), Chowdhury and Mavrotas (2006), Jayaraman and Choong (2006), Ozturk and Kalyoncu (2007), Shawa and Shen (2013), Ndeffo et al. (2013), etc. The findings from these studies are mixed and provide contradictory conclusions with some studies supporting “FDI-led growth hypothesis”, others supporting “growth-driven hypothesis” and still others failing to support either of the two hypotheses.

ic

Pa

Some studies have also used panel data to examine the link between FDI and economic growth. Among them, are Nair-Reichert and Weinhold (2001), Hansen and Rand (2004), Lyroudi et al. (2004), Turkcan et al. (2008), Tiwari and Mutascu (2011), Asrafuzzaman and Hasanuzzaman (2012), Iftikhar (2012), etc.

Ec

on

om

As Dumitrescu and Hurlin (2012) point out however, the major problem for most of panel data studies is the failure to account for heterogeneity among the cross section units of the panel, by assuming that countries share common characteristics (Emirmahmutoglu and Kose, 2011) hence the impact of a variable on another is the same for all the countries. This is misleading since countries do not have the same economic, political and even institutional structures; there is no way therefore one should expect variables to relate the same way in all the countries of a panel. Dumitrescu and Hurlin (2012) overcome this shortcoming noted in other panel data studies by proposing a methodology of panel causality testing accounting for heterogeneity among the cross-sections of the panel. This study therefore follows Dumitrescu and Hurlin (2012) and examines the causal links between FDI and economic growth for COMESA countries using panel causality test in heterogeneous panels. To the best of our knowledge, no study has used this methodology in examining the FDI-growth nexus in COMESA countries. The rest of the paper is organized as follows: Section 2 describes the methodology used. Section 3 presents the data and empirical results and section 4 concludes the study.

3

North American Academic Journals, 2014


2. Econometric Methodology This study applies heterogeneous panel causality tests advanced by Dumitrescu and Hurlin (2012) to examine the causal nexus between FDI and growth in COMESA Countries. According to Hurlin and Venet (2005), assuming homogeneous causality can lead to faulty conclusions, inferring a causal relationship in all the cross-sections of the panel when it is only present in a subset of them or, rejecting the presence of a causal relationship for all the cross-section units yet it is present in at least one of them.

d

N

ot e

s

Dumitrescu and Hurlin (2012) proposed a test whose structure is similar to the panel unit root test in heterogeneous panels suggested by Im, Pesaran and Shin (2003). Testing for causality in heterogeneous panels involves testing first the Homogeneous Non Causality (HNC) hypothesis. Under the null hypothesis of HNC, there is no causal link between two variables in all the crosssections of the panel. Unlike Holtz-Eakin, Newey and Rosen (1988) who assume Homogeneous Causality (HC) hypothesis under the alternative hypothesis, that is, causality exists for all the cross-section units, Dumitrescu and Hurlin (2012) consider that under the alternative hypothesis, there are some cross-section units, N1 of the panel, for which there is no causality and ( N − N1 ) cross-section units for which there exists causality between the variables.

an

Dumitrescu and Hurlin (2012) propose the following model to test for causality between two stationary variables x and y : K

rs

K

yi ,t = α i + ∑ λi ( k ) yi ,t − k + ∑ ϕi( k ) xi ,t − k + υi ,t k =1

(1)

pe

k =1

om

ic

Pa

where i = 1, 2,..., N . K is the lag order which is the same for all cross sections of the panel and the panel is balanced. The autoregressive parameters λi( k ) and the regression slope coefficients ϕi( k ) are constant over time but differ across the cross-sections of the panel and the model is a fixed coefficients model with fixed individual effects.

(2)

on

To test the Homogeneous Non Causality (HNC) Hypothesis, the null hypothesis is set as: H 0 : ϕi = 0, ∀i = 1, 2,..., N

Ec

where ϕi = (ϕi(1) , ϕi(2) ,..., ϕi( K ) ) Under the null hypothesis, there is no causality for all the cross-section units of the panel. However under the alternative hypothesis, there exists a subgroup of cross-sections units of the panel with dimension N1 for which the variable x does not Granger cause the variable y and another subgroup of dimension ( N − N1 ) for which the variable x Granger causes y , since at least one regression slope coefficient associated to xi ,t − k is statistically different from zero. H 1:

ϕi = 0, ∀i = 1,..., N1 ϕi ≠ 0, ∀i = N1 + 1, N1 + 2,..., N

(3)

If N1 = N , there is no causality for all the individuals of the panel, and we get the Homogeneous Non Causality (HNC) in this case. In the opposite case if N1 = 0, there is causality for all the

Economic Papers and Notes, vol.14, no.1

4


individuals of the sample; in this case a causality relationship exists for all the cross-section units of the panel (Homogeneous Causality Hypothesis). However, if N1 > 0 , the causality relationship is heterogeneous; the causality links are different according to the cross-section units of the panel. In order to test the null hypothesis of Homogeneous Non Causality, Dumitrescu and Hurlin (2012) suggest the use of the average of individual Wald statistics associated to the null hypothesis of no causality for each cross-section unit of the panel i = 1,..., N . The average Wald statistic proposed is as follows: N

∑W

i ,T

s

1 N

i =1

(4)

ot e

WNHNC ,T =

where Wi ,T is the individual Wald statistic for each cross-section unit i associated to the null

N

hypothesis H 0 : ϕi = 0 .

an

d

However, according to the authors, in small T samples like ours, the average Wald statistic does not follow the standard chi-square distribution. In addition to the average Wald statistic therefore, Dumitrescu and Hurlin (2012) propose the asymptotic standardized statistic ( z NHNC )

rs

HNC and the approximated standardized statistic ( z N ) which follow a normal distribution.

N

z NHNC =

i =1

→ N (0,1)

N

∑ Var (W

i ,T

i =1

)

( z NHNC ) is defined as:

(5)

ic

N

−1

Pa

−1 N [WNHNC ,T − N ∑ E (Wi ,T )]

pe

For small T samples, the Asymptotic standardized statistic

on

Wald statistic Wi ,T .

om

where E (Wi ,T ) and Var (Wi ,T ) respectively denote the mean and the variance of the individual

N

Ec

Furthermore, for small T samples, Dumitrescu and Hurlin (2012) makes the following HNC approximations to define the approximated standardized statistic ( z N ) :  i ,T ) =K * (T - 2 K -1) N −1 ∑ E (Wi ,T ) ≈ E (W (T - 2 K - 3) i =1 N

 i ,T ) = 2 K * N −1 ∑ Var (Wi ,T ) ≈ Var (W i =1

(6)

(T - 2 K -1) 2 *(T - K - 3) (T - 2 K - 3) 2 *(T - 2 K - 5) HNC

The Approximated standardized statistic ( z N ) is hence defined as:

5

North American Academic Journals, 2014

(7)


HNC z N =

 N [WNHNC ,T − E (W i ,T )] → N (0,1)  i ,T ) Var (W

(8)

HNC If the value of ( z NHNC ) or ( z N ) is superior to the normal critical value for a given level of significance, the Homogeneous Non-Causality (HNC) hypothesis is rejected.

ot e

s

If the hypothesis fails to be rejected, the implication is that the variable x does not cause y in all the N cross-sections of the panel, the non-causality is homogeneous and therefore the testing procedure goes no further than this. But if the Homogeneous Non Causality hypothesis is rejected, it means that causality exists at least in one cross-section unit of the panel and the next step is to test in which of the cross-sections causality exists and in which of these cross sections, causality does not exist. This is accomplished using the usual Granger causality test for each cross-section unit of the panel. However, since our sample size is small (T = 32), we use bootstrap causality tests4 for each member of the panel to this effect.

N

3. Data and Empirical Results

on

om

ic

Pa

pe

rs

an

d

The study uses annual data on growth rate of Real GDP (GDPGR) and the ratio of foreign direct investment, FDI as a percentage of GDP (FDIR) for 17 COMESA countries, Burundi, Comoros, Democratic Republic of Congo, Egypt, Ethiopia, Kenya, Libya, Madagascar, Malawi, Mauritius, Rwanda, Seychelles, Sudan, Swaziland, Uganda, Zambia and Zimbabwe for the period 19802011. Data on Growth rate of Real GDP have been accessed from the World Development Indicators (WDI, 2012) of the World Bank and data on FDI from UNCTAD (United Nations Conference on Trade and Development) online database. We examine the order of integration of the variables prior to any detailed analysis. Three panel unit root tests are employed, namely, Harris-Tzavalis (HT, 1999) test, Im, Pesaran and Shin (IPS, 2003) test and the Cross-section Augmented Dickey-Fuller (CADF) test of Pesaran (PESCADF, 2005). The first two tests assume cross-sectional independence of individual time series in the panel while PESCADF test allows for cross-sectional dependence. Moreover, Harris-Tzavalis (1999) test assumes homogeneous autoregressive coefficients between the cross-sections while IPS (2003) and PESCADF (2005) tests allow for heterogeneity between the cross-section units. Panel unit root tests results are presented in the next subsection.

Ec

3.1 Panel Unit Root Tests Results Panel unit root tests results are presented in Table 1. They indicate that HT (1999), IPS (2003) and PESCADF (2005) tests fail to reject the null hypothesis of panel unit root for the ratio of FDI as a percentage of GDP (FDIR). However, when differenced once, the same tests reject the presence of panel unit root at 1% level. For growth of real GDP (GDPGR), all the tests used, HT (1999), IPS (2003) and PESCADF (2005) robustly reject the null hypothesis of panel unit root at 1% level. This implies that in the estimation of the panel VAR for causality testing, GDPGR is considered in level while FDIR needs to be differenced once to be stationary.

4

According to Hacker and Hatemi-J (2010), if the sample size is small, the Wald statistic of no Granger causality does no longer follow the assumed asymptotic distribution. This creates an inferential bias leading to an over-rejection of the null hypothesis of no causality. The solution proposed by Hacker and Hatemi-J (2010) is to use Bootstrap causality tests which give precise critical values in that case.

Economic Papers and Notes, vol.14, no.1

6


Table 1: Panel Unit Root Tests Results IPS-test HT-test PESCADF-test W[t-bar] [Rho-Z ] Z[t-bar] Constant C & T Constant C & T Constant C&T FDIR 4.401 -0.292 2.870 2.467 2.333 1.313 (1.000) (0.385) (0.998) (0.993) (0.990) (0.905) -10.91*** -8.93*** -36.57*** -20.36*** -4.411*** -3.745*** ∆ FDIR (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) GDPGR -10.91*** -11.43*** -28.54*** -17.78*** -6.186 *** -5.761*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Notes: [1] C & T indicates a model with constant and trend; [2] Out of 7 lags, AIC chooses 2 lags when the deterministic term included is constant and 3 lags when it is constant and trend. [3] Between the brackets (.) are the p-values; [4] *** indicates rejection of the null hypothesis of panel unit root at 1% level respectively.

N

ot e

s

Series

3.2. Granger Causality tests results in Heterogeneous Panels

pe

rs

an

d

As Dumitrescu and Hurlin (2012) indicate, the first step consists of testing whether there is no causal relationship for all the cross-section units (countries), that is, Homogeneous Non Causality (HNC) hypothesis. The results of Homogeneous Non Causality Hypothesis test between FDI and economic growth for a panel of 17 COMESA countries are presented in table 2. The results are based on three test statistics, namely the average Wald statistic, WHNC ; the

on

om

ic

Pa

asymptotic standardized statistic, z HNC and the approximated standardized statistic based on finite sample moments, z HNC . The results suggest that Homogeneous Non Causality hypothesis from FDI to Economic growth is robustly rejected for 2 and 3 lags at 1% level for both the asymptotic standardized statistic, z HNC and the approximated standardized statistic, z HNC . However, regarding the reverse causality from economic growth to FDI, Homogeneous Non Causality Hypothesis (HNC) is not robustly rejected, it is rejected only when we consider the asymptotic standardized statistic, z HNC for 2 and 3 lags.

Table 2: Homogeneous Non Causality Hypothesis Tests Results FDI ⇒ Economic Growth Economic Growth ⇒ FDI K=1 K=2 K=3 K=1 K=2 K=3 0.9607 3.9103 6.4904 0.6803 1.5613 3.3977 WHNC -0.1146 7.8762♣ 17.6257♣ -0.9321 -1.8088♥ 2.0083♠ z HNC (0.9088) (0.0000) (0.0000) (0.3513) (0.0705) (0.0446) z HNC -0.2966 3.0521♣ 4.4407♣ -1.0124 -1.0597 0.1725 (0.7668) (0.0023) (0.0000) (0.3113) (0.2893) (0.8631) Notes: [1] Between the parentheses are the p-values related to the test statistics; [2] ⇒ indicates the direction of causality being tested; [3] All the computations are done using Matlab software. A Matlab code written by Dumitrescu and Hurlin (2012) available at www.runmycode.org/CompanionSite/Site51 was used, we are very grateful to the authors for making the code available. [4] ♥, ♠ and ♣ indicate that the null hypothesis is rejected at 10%, 5% and 1% respectively; [5] K is the number of lags considered.

Ec

Test Stat

7

North American Academic Journals, 2014


Rejecting Homogeneous Non Causality (HNC) Hypothesis for both directions of causality implies that there exists a causal link between FDI and economic growth in both directions for at least one country of our panel. The remaining task is to investigate in which country (crosssection unit) the causal links exist and in which unit they do not exist. This is accomplished using Bootstrap causality tests since the Wald statistic for Granger causality in small samples like ours (T=32) can be biased, leading to an over-rejection of the null hypothesis of no causality (Hacker and Hatemi-J, 2010). Bootstrap causality tests results for causality running from FDI to economic growth are reported in table 3.

ot e

s

The results indicate that the null hypothesis of no Granger causality can only be rejected for Libya at 1% level of significance. However, the sign of the sum of the estimated causal parameters is negative, which implies a negative causal impact of FDI to economic growth in Libya, suggesting that FDI is harmful to growth in that country.

rs

an

d

N

For the rest of COMESA countries, the results suggest that the null hypothesis that FDI do not Granger cause economic growth, could not be rejected even at 10 % level. Although the findings show that the causal impact of FDI is positive for Comoros, DRC, Egypt, Malawi, Mauritius, Seychelles, Sudan, Uganda and Zambia, it is however not significant. Failure to support FDI-led growth hypothesis in this group of countries is probably due to the fact that the absorptive capacity in most of African countries is low. Moreover, FDI flowing to COMESA region is mainly resource-seeking which does not create many linkages with domestic firms, limiting hence spillover effects.

pe

Table 3: Results of Bootstrap Causality Tests from FDI to Economic Growth

Pa

H0: FDI Does Not Granger Cause Economic Growth Optimal lag Wald Statistic Bootstrap Critical values 1% 5% 10% Burundi 1 0.157[-] 8.665 4.603 3.027 Comoros 3 2.743 [+] 14.439 8.949 6.896 DRC 1 0.132 [+] 8.021 4.203 2.878 Egypt 4 2.138 [+] 20.438 13.079 9.901 Ethiopia 4 0.198[-] 22.145 13.217 10.251 Kenya 1 0.811[-] 7.711 4.187 2.891 Libya 4 28.250***[-] 24.269 15.034 11.216 Madagascar 4 9.086 [-] 20.434 12.579 9.659 Malawi 1 1.392 [+] 8.272 4.267 2.890 Mauritius 1 0.038 [+] 7.786 4.286 2.911 Rwanda 1 0.013 [-] 7.752 4.346 2.981 Seychelles 1 0.055 [+] 7.776 4.221 2.914 Sudan 1 0.831 [+] 8.151 4.469 2.964 Swaziland 1 0.186 [-] 7.923 4.451 3.036 Uganda 1 0.008 [+] 7.564 4.134 2.909 Zambia 1 0.001 [+] 7.850 4.169 2.854 Zimbabwe 1 0.343 [-] 7.897 4.263 2.910 Notes: (1) Bootstrap causality tests are conducted using a GAUSS code written by Hacker and Hatemi-J (2010) available in the statistical software components archive; (2) Optimal lag was

Ec

on

om

ic

Country

Economic Papers and Notes, vol.14, no.1

8


selected by Akaike Information Criterion (AIC) for a maximum of 5 lags; (3) Between brackets [.] is the sign of the sum of the estimated causal parameters; (4) Bootstrap Critical Values are based on 10000 simulations; (5) *** indicates that the null hypothesis of no causality is rejected at 1% level.

d

N

ot e

s

Table 4 presents the results of bootstrap causality tests from economic growth to FDI. The findings indicate that the null hypothesis of no Granger causality can only be rejected for Comoros at 5% level. For the rest of COMESA countries, the null hypothesis of no Granger causality from economic growth to FDI could not be rejected even at 10% level. The findings hence support the growth-driven FDI hypothesis only for Comoros. Implying that economic growth in Comoros could be a pulling factor for Multinational companies to invest in that country. For the rest of COMESA countries, failure to support the growth-driven FDI hypothesis is not surprising since most of the countries in the region are oil-rich and mineral-rich countries and FDI flowing in those countries is mainly resource-seeking FDI which naturally does not depend on the host country’s growth performance since MCNs investing in those countries do not have intention to capture the local markets.

an

Table 4: Results of Bootstrap Causality Tests from Economic Growth to FDI

Ec

on

om

ic

Pa

pe

rs

H0: Economic Growth Does Not Granger Cause FDI Country Optimal lag Wald Statistic Bootstrap Critical values 1% 5% 10% Burundi 1 0.219 8.214 4.689 3.162 Comoros 3 10.303** 15.830 9.658 7.257 DRC 1 0.024 7.696 4.333 2.982 Egypt 4 2.672 18.652 11.830 9.240 Ethiopia 4 5.577 21.202 12.912 9.853 Kenya 1 2.440 7.897 4.300 2.976 Libya 4 0.754 18.637 12.151 9.380 Madagascar 4 8.688 19.859 11.688 8.957 Malawi 1 0.001 7.846 4.106 2.850 Mauritius 1 0.473 8.000 4.289 2.884 Rwanda 1 0.367 8.179 4.389 2.945 Seychelles 1 0.407 7.494 4.261 2.894 Sudan 1 0.012 8.484 4.484 6.967 Swaziland 1 0.909 8.847 4.671 3.167 Uganda 1 0.046 8.154 4.276 2.898 Zambia 1 1.687 7.756 4.146 2.835 Zimbabwe 1 1.226 8.410 4.357 2.889 Notes: (1) Bootstrap causality tests are conducted using a GAUSS code written by Hacker and Hatemi-J (2010) available in the statistical software components archive; (2) Optimal lag was selected by Akaike Information Criterion (AIC) for a maximum of 5 lags; (3) Bootstrap Critical Values are based on 10000 simulations; (4) ** indicates that the null hypothesis of no causality is rejected at 5% level.

9

North American Academic Journals, 2014


4. Conclusion

N

ot e

s

The objective of this study was to examine the causal nexus between FDI and economic growth in COMESA countries. The study applied Granger causality tests in heterogeneous panels advanced by Dumitrescu and Hurlin (2012), to test for Homogeneous Non Causality hypothesis. The findings indicate that Homogeneous Non Causality hypothesis for both directions of causality was rejected although not robustly for the causality running from economic growth to FDI. Having rejected the Homogeneous Non Causality hypothesis, the study proceeded to check in which countries, the causal link between FDI and economic growth exists and in which it does not. To this effect, bootstrap causality tests were applied since our sample size was small (T = 32) and the findings indicate that FDI Granger cause economic growth only for Libya. However, the causal impact was found to be negative, suggesting that FDI in that country is harmful to growth. For the rest of the countries, no causal link could be found in that direction. Concerning the reverse causality, the results support growth-driven FDI hypothesis only in Comoros suggesting that economic growth is not a major determinant of FDI flowing to the COMESA region.

rs

an

d

The findings imply that COMESA Countries should strive to improve their absorptive capacity by improving things like production technologies, financial sector, human capital, etc. for positive growth effects of FDI to take place.

pe

References

Pa

[1] Agiomirgianakis G.M., Asteriou D. and Papathoma K., “The Determinants of Foreign Direct Investment: A Panel Data Study for the OECD Countries,” Working Papers 03/06, 2003, Department of economics, City University London

om

ic

[2] Akinlo A.E., “Foreign Direct Investment and Growth in Nigeria: An Empirical Investigation,” Journal of Policy Modeling, vol. 26, no.5, July 2004, pp. 627-639

on

[3] Anyanwu J.C., “Why Does Foreign Direct Investment Go Where it goes? New Evidence from African Countries,” Annals of Economics and Finance: vol. 13, no. 2, 2012, pp. 433-470

Ec

[4] Asiedu E., “Policy Reform and Foreign Direct Investment in Africa: Absolute Progress but Relative decline,” Development Policy Review, vol. 22, no.1, January 2004, pp. 41-48 [5] Asrafuzzaman and Hasanuzzaman S., “Does Foreign Direct Investment (FDI) Matter for Growth? A Longitudinal Disaggregated Panel Data Analysis,” International Journal of Applied research in Business Administration and Economics, vol. 1, no 4, September 2012, pp. 39-49 [6] Ayanwale A.B., “FDI and Economic Growth: Evidence from Nigeria,” AERC Research Paper 165, African Economic Research Consortium, Nairobi, April 2007 [7] Balamurali N. and Bogahawatte C., “Foreign Direct Investment and Economic Growth in Sri Lanka,” Sri Lankan Journal of Agricultural Economics, vol.6, no.1, 2004, pp. 37-50

Economic Papers and Notes, vol.14, no.1

10


[8] Chowdhury A. and Mavrotas G., “FDI and Growth: A Causal Relationship,” UNU-WIDER, Research Paper No. 2005/25 [9] Driffield N. and Jones C., “Impact of FDI, ODA and Migrant Remittances on Economic Growth in Developing Countries: A Systems Approach,” European Journal of Development Research, vol. 25, 2013, pp. 173–196 [10] Dumitrescu E.-I. and Hurlin C., “Testing for Granger non-Causality in Heterogeneous Panels,” Economic Modelling, Elsevier, vol. 29, no.4, 2012, pp. 1450-1460

s

[11] Emirmahmutoglu F. and Kose N., “Testing for Granger causality in heterogeneous mixed panels,” Economic Modelling, vol.28, no.3, May 2011, pp. 870–876

N

ot e

[12] Fortanier F., “Foreign direct investment and host country economic growth: Does the investor’s country of origin play a role?,” Transnational Corporations, vol. 16, no.2, August 2007, pp. 41-76

an

d

[13] Farkas B., “Absorptive Capacities and the Impact of FDI on Economic Growth,” Discussion papers no.1202, 2012, DIW Berlin, German Institute for Economic Research

pe

rs

[14] Hacker S. and Hatemi-J A., “A Bootstrap Test for Causality with Endogenous Lag Length Choice: Theory and Application in finance,” CESIS Electronic Working Paper Series, Paper No. 223, April 2010

ic

Pa

[15] Hacker S. and Hatemi-J A., “HHcte: GAUSS Module to Apply a Bootstrap Test for Causality with Endogenous Lag Order,” Statistical Software Components G00012, 2010, Boston College Department of Economics, Available at: http://ideas.repec.org/s/boc/bocode.html.

om

[16] Hansen H. and Rand J., “On the Causal Links between FDI and Growth in Developing Countries,” Discussion Papers 04-30, 2004, Institute of Economics, University of Copenhagen

on

[17] Holtz-Eakin D., Newey W. and Rosen H., “Estimating Vector Autoregressions with Panel Data,” Econometrica, vol. 56, November 1988, pp. 1371-1396.

Ec

[18] Hurlin C. and Venet B., “Testing for Granger Causality in Heterogeneous Panel Data Models,” Revue Economique, vol. 56, 2005, pp. 1-11 [19] Hurlin C. and Venet B., “Financial Development and Growth: A Re-Examination Using a Panel Granger Casuality Test,” Working Papers 2008, halshs-00319995, HAL [20] Iftikhar A., “Foreign Direct Investment and Economic Growth in Selected SAARC Countries: A Causality Investigation Using Heterogeneous Panel Analysis,” Interdisciplinary Journal of Contemporary Research in Business, vol. 4, no 3, July 2012, p121 [21] Im K.S., Pesaran S.H. and Shin Y., “Testing for Unit Roots in Heterogeneous Panels,” Journal of Econometrics, vol.115, no.1, July 2003, pp. 53-74

11

North American Academic Journals, 2014


[22] Jayaraman T.K. and Chee-Keong C., “Foreign direct investment in the South Pacific Island Countries: a case study of Fiji,” World Review of Entrepreneurship, Management and Sustainable Development, Inderscience Enterprises Ltd, vol. 2, no.4, January 2006, pp. 309-322. [23] Klarsa M.A., “Foreign Direct Investment, Trade Openness and Economic Growth in Pakistan and Turkey: An investigation Using Bounds Test,” Qual Quant, vol. 45, no.1, 2011, pp. 223-231 [24] Kutan A.M. and Vukšić G., “Foreign Direct Investment and Export Performance: Empirical Evidence,” Comparative Economic Studies, 2007, vol.49, no.3, September 2007, pp.430-445

ot e

s

[25] Lyroudi K., Papanastasiou J. and Vamvakidis A. (2004), “Foreign Direct Investment and Economic Growth in Transition Economies,” South Eastern Europe Journal of Economics, vol.2, no. 1, 2004, pp. 97-110

N

[26] Mehta S., “Inwards FDI to South Asia: Identifying the Driver,” SAARC Journal of Human Resource Development, vol.3, no.1, 2007, pp. 46-64

rs

an

d

[27] Moura R. and Forte R., “The Effects of Foreign Direct Investment on the Host Country Economic Growth - Theory and Empirical Evidence,” FEP Working Papers, no. 390, November 2010

Pa

pe

[28] Nair-Reichert U. and Weinhold D., “Causality Tests for Cross-Country Panels: A New look at FDI and Economic growth in Developing countries,” Oxford Bulletin of Economics and Statistics, vol. 63, no.2, May 2001 pp. 153-171

om

ic

[29] Ndeffo L.N., Kamdem D. and Nanfosso R.T., “Foreign Direct Investments and Economic Growth in Sub-Saharan African Countries: A Comparative Analysis between Landlocked Countries and Countries Having Access to the Sea,” International Journal of Economics and Finance, vol. 5, no.6, June 2013, p157

Ec

on

[30] Ozturk I. and Kalyoncu H., “Foreign Direct Investment and Growth: An Empiricial Investigation Based on Cross-Country Comparison,” Economia Internazionale, vol. 60, no.1, February 2007, pp.75-82 [31] Ponce A.F., “Openness and Foreign Direct Investment: The Role of Free Trade Agreements in Latin America,” Munich Personal RePEc Archive Paper No. 4187, 2006 [32] Pradhan R.P.P., “The FDI- Led- Growth Hypothesis in ASEAN- 5 Countries: Evidence from Cointegrated Panel Analysis,” International Journal of Business and Management, vol.4, no.12, December 2009, p153 [33] Reza A. and Mojtaba G., “Openness, Economic Growth and FDI: Evidence from Iran, ” Middle-East Journal of Scientific Research, vol.10, no.2, 2011, pp.168-173 [34] Samad A., “Does FDI Cause Economic Growth? Evidence from South-East Asia and Latin America,” Woodbury School of Business, Working Paper 1-09, 2009

Economic Papers and Notes, vol.14, no.1

12


[35] Sarbapriya R., “Impact of Foreign Direct Investment on Economic Growth in India: A Co integration Analysis,” Advances in Information Technology and Management, vol.2, no.1, 2012, pp.187-201 [36] Shawa M.J. and Shen Y., “Causality Relationship between Foreign Direct Investment, GDP Growth and Export for Tanzania,” International Journal of Economics and Finance, vol.5, no.9, August 2013, p13 [37] Tiwari A.K. and Mutascu A., “Economic Growth and FDI in Asia: A Panel-Data Approach,” Economic Analysis & Policy, vol. 41, no.2, September 2011, pp.173-187

ot e

s

[38] Turkcan B., Duman A. and Hakan I.Y., “How Does FDI and Economic Growth affect Each Other? The OECD Case,” 3rd International Conference on Emerging Economic Issues in a Globalizing World, 2008, pp.21-40

d

N

[39] United Nations Conference on Trade and Development, “Foreign Direct Investment and the challenge of Development,” World Investment Report, 1999

an

[40] United Nations Conference on Trade and Development, “Promoting Linkages,” World Investment Report, 2001

pe

rs

[41] United Nations Conference on Trade and Development, “Transnational Corporations and Export Competitiveness,” World Investment Report, 2002

ic

Pa

[42] United Nations Economic Commission for Africa, “Economic Report on Africa: Accelerating Africa’s Development through Diversification,” ECA Publications and Conference Management Section (PCMS), 2007

om

[43] Zhang K.H., “Does Foreign Direct Investment Promote Economic Growth? Evidence from East Asia and Latin America,” Contemporary Economic Policy, vol.19, no.2, April 2001, pp.175-185,

Ec

on

[44] Yalta A.Y., “Revisiting the FDI-led growth Hypothesis: The case of China,” Economic Modelling, vol.31, 2013, pp. 335-343

13

North American Academic Journals, 2014


Empirical Examination of Impact of Mobile Phone Usage on the Well-being of Fishermen in Rural Bangladesh Md. Sadik Pavela,1, Seikh Ruksana Burhanb, Muktadir Mahfuj Jameec a

b

Research Associate, Bangladesh Foreign Trade Institute (BFTI), Dhaka-1215, Bangladesh. Email: srburhansust@gmail.com

MBA Student, BRAC University, Dhaka-1212, Bangladesh. Email: jamee2005@gmail.com Correspondence Author

an d

Abstract

N

ot

1

es

c

Department of Economics, Shahjalal University of Science & Technology, Sylhet-3114, Bangladesh. Email: sadikpavel@gmail.com

om ic

Pa

pe

rs

Information technology via mobile phone could bring new possibilities to develop fisheries communities’ life standard. To evaluate contribution of mobile phones, this study examined 205 household from fisheries community in Sylhet region of Bangladesh, where uses of mobile phones is extensive. After examining findings indicate that the income of the fishermen who use mobile phone is slightly higher than the fishermen who are not the user. Key findings suggested that to provide communicate business related information through mobile phone technology ensured better rural livelihood. Mobile phone also reduces cost and risk of doing business, increase savings, help to fishermen to empower.

Keywords: Empowerment, Fishermen, Market Information, Mobile Phone Service,

on

Bangladesh.

Ec

JEL Classification: I31, Q13, Q22, D83.

Economic Papers and Notes, vol.14, no.1

14


1. Introduction In Bangladesh, millions of people directly and indirectly depend on agriculture and fisheries businesses. River irrigated Bangladesh has numerous rivers, haors, beels, ponds and those are favorable to fisheries sector. According to Economic Review 2013, in 2011-12 agriculture and fisheries together contributed 19.41 percent of total GDP, production of fisheries during this economic year was 3.222 million metric ton. But small scale fishing communities are the disregarded communities in Bangladesh in terms of socioeconomic condition, most of them intending to change their profession. Main reasons of struggling of small scale fishing communities are ancient pattern of cultivation, inadequate marketing experience, poor communication facilities and middleman.

rs

an d

N

ot

es

Currently, at the time of technology revolution, it is possible to alleviate poverty as well as improve social development of fisheries community. Bangladesh, as a developing country practicing to be enriched with information and communication technology. Mobile phones are believed to represent one of the most rapidly adopted information technologies in the history of technological innovation. Mobile phones are more easy and affordable than computers, moreover require fewer infrastructures, do not require the user to have much technological knowledge or even to be able to read or write and are easy to carry from place to place. Due to the low cost of labor, the use of mobile phone user increasing day by day and according to Bangladesh Telecommunication and Regulatory Commission (BRTC), at the end of December 2013 there are 113.784 million mobile phone subscriber.

om ic

Pa

pe

Market information reduced market margin and improved the market transparency, competiveness, efficiency (Nikolov, G. and Hughes, D., 2000). Efficient market information provision ensure better negotiation, increase welfare (Shepherd, A. W., 1997). Availability of mobile technology enables new mode of cooperation, using mobile phones in culturally enhancing and ecologically oriented ways improve the working and living condition of fishers community (Sreekumar, T. T., 2011). Human capital and income per capita have a positive effect on technology adoption, mobile community feature improve the quality of rural life through sharing information, experience (Comin, D., and Hobijn, B., 2004; Lu, N., and Swatman, P. M.C., 2009).

Ec

on

This study investigate the impact of mobile phone usage on economic and social conditions of special target group of Sylhet region, Bangladesh. The targeted group includes the fishermen engaged in different locations of Sylhet and Hobiganj districts and are living mainly on fishing. The rationale behind choosing the special group is firstly for the nature of their product, raw fish, which is a very short lasting good and thus market turnovers have to be made quickly; and secondly for the nature of fish market, where price fluctuations are evident along with wide scale existence of middle men. This study also determine the channel through which a modern communication system like mobile phone can be used in the above mentioned production and marketing structure. Data for this paper were collected in first half of 2011 from a surveys in Sylhet and Hobiganj. The sample consist of 205 sardine fishing units, in which 121 data from Sylhet region and 84 data from Hobiganj region. For data collection purpose, randomly choose two villages of Sylhet region Mirzargaon and Uzanmiragaon from the local government village list. About 95% people of these villages are fishermen. In Sylhet region, 77 data obtained from Mirzargaon and 44 data obtained from Uzanmiragaon randomly. Following the same procedure, 84 data randomly obtained from two villages of Hobiganj; Chorgaon and Umednagar. Data obtained from both the fishermen who are mobile phone users and the

15

North American Academic Journals, 2014


fishermen who are not. The respondents were the head of the household. The sample characteristics are shown in Table-1 and Table-2.

N (%) or mean (±S.D.) 205

ot

es

77 (37.6 %) 44 (21.5 %) 41 (20.0 %) 43 (21.0 %) 152 (74.1 %) 205 (100 %) 36.18 (±13.67) 63 (30.7 %)

N

117 (57.1 %) 1 (0.5 %) 87 (42.4 %) 7.23 (±3.22)

an d

Table 1: Sample Characteristics of the Fishermen Variable Sample size Study Area Mirzargaon Ujanmiragaon Chorgaon Umednagar Mobile Users Gender (Male) Age (years) Primary Education Marital status Married Widowed/widower Single Number of person per household

om ic

Pa

pe

rs

The questionnaire is also included with qualitative questions about their lifestyle that how they manage fishing, or they receive any microcredit or have to pay any usury, about their productivity, do they face any harassment or do mobile phone use can reduce their search cost and improve markets or they receive any health assistance through initializing mobile phone etc.

Ec

on

Table 2: Professional Characteristics of the Fishermen Professional Related Question Frequency Ownership of fishing zone Self 162 Other 43 How do you manage fishing? Self 82 Other 123 Type of Jal Normal 179 Current 26 Have you receive any microcredit? Yes 66 No 139 Do you pay any usury? Yes 58 No 151

Economic Papers and Notes, vol.14, no.1

Percentage 70.02% 20.97% 40% 60% 87.31% 12.68% 32.19% 67.80% 26.34% 73.65%

16


2. Methods Information through mobile phones reduce price dispersion, increase fishermen’s profit and consumer’s welfare and due to profit self-sustaining mobile phone companies provides information, at the same time fishermen willing to pay to providers because of their increased earnings (Jensen, R., 2007). With four period difference in difference (DID) data analysis among three region of South Indian Fisheries sector, Jensen, R. (2007) also found that Introduction of mobile phone leads to a significant increase in arbitrage, 30-40% of fisherman sell outside their local market from an initial situation of autarky.

ot

es

Islam, M. A. and Grönlund, A. (2007) assessed agricultural market information E-service in Bangladesh using reality gap, their result suggested although deficiencies of adoption but combination with call centers and locally human resources mobile phone technology is the most important factors of success. Mobile phone service was effective in ‘push’ group while it was ineffective in ‘pull’ group. Since effectiveness of rural e-service depends on design and delivery service so that to keep the service sustainable, social business must be planned with rational financial model (Islam, M. A. and Grönlund, A., 2010).

pe

rs

an d

N

Adoption of technology is important, almost all respondent expressed use of mobile phones makes their daily lives easier due to mobility characteristics, help to overcome barriers of time, location and improves productivity (Islam, M. S. and Grölund, A., 2011). After evaluating the role of telecommunication within the context of Bangladesh rural development in general and poverty reduction Bayes, A. (2001) concludes that mobile phone has positive effect on women empowerment and relatively poor mobile phone owner tends to raise their social status, pave the way for change in the social equilibrium.

om ic

Pa

Akter, J. C. (2008) empirically analyzed impact of cell phones on grain markets in Niger with the detailed information on mobile phone coverage from 2001 to 2006 and also collected information on grain market operation from 2005 to 2007. The dataset includes monthly grain price data from 1996-2006, after analysis Akter, J. C (2008) found that introduction of mobile phones reduced price dispersion across grain market and improved consumer as well as trader welfare in Niger.

Ec

on

Mobile phones requires basic literacy so that accessible to large segment of the population and advantages of mobile phone technology can be used for the purpose of health, education, commerce and Governance (Rashid A. T. and Elder, L., 2009). Direct impact are less and greater use of mobile phone for social purposes and emergencies, rather than dedicated economic activity; mobile phone’s impact on small enterprises, farmers, and the selfemployed is not clear-cut (Souter et al., 2005; Donner, 2004 and 2008). Gordon, J. (2007) used three critical situation (a) Chinese SARS outbreak (2003), (b) the south-east Asian tsunami (December 2004) and (c) the London bombings (July 2005) as case studied and explored the influence of mobile phones. After the analysis, author concludes that mobile phone technology amongst the public sphere helped ameliorate the effects of those situation by being providing information via voice or sms. For development income is the one condition and in this regard, access to market information is important factor. Market information ‘must be for the farmers’ not just ‘be there’ in such way that they can understand and able to utilize (Islam, M. S., 2011). Mobile phone information technology will be better choice to provide information to the fisheries community.

17

North American Academic Journals, 2014


Major impediment for raising productivity among smallholders are access, affordability, efficiency of information and use of mobile phone reduce information asymmetries, enable producer to arbitrage over price, coordinate sales, helped to increase incomes and sales volumes (Muriithi, A.G. et al. 2009).

ot

es

A socio-economic questionnaire was administrated in each household. The questionnaire included quantitative semi-open and closed (multi-choice) questions. The questionnaire was structured to collect detailed information on household income, general socio-economic background (number of family members, age, and education), productive assets, and livelihood strategies (on-farm and non-farm activities). Income referred to the household’s income earned in cash plus payment in kind that could be valued at market prices. The cash earning components of income included crop and vegetable sales, petty trade, remittance and fish sales. Also, it was hypothesized that wealth would also influence the household income structure.

N

2.1 Econometric Tool

rs

an d

A major part of the paper relies on comparative study of different socio-economic aspects between these two groups of fishermen. We also have developed a probit model to show if there is any economic as well as person specific determinants providing incentives to use mobile phone.

Pa

pe

A probit model (Bliss C. I., 1935) works well when the probabilistic dependent variable can take limited values. Probit model serves an appropriate framework for statistical analysis, avoids likelihood direct evaluation and thus avoids calculating choice probabilities associated problems (Daykin, A. R. and Moffatt, P. G., 2002; McCulloch, R. and Rossi, P. E., 1994).

Ec

on

om ic

A probit model is a popular specification for an ordinal or binary response model that employs a probit link function. This model most often uses standard maximum likelihood procedure, an estimation technique suitable for optimization. Suppose, the binary response variable Y can have only two possible outcomes: users of mobile phone = 1 and non-users = 0. We also have a vector of determinants X, which are assumed to influence the outcome Y. Specially, we assume that the model takes form Pr (Y=1|X) = Φ (X´β),

Where, Pr denotes probability and Φ is the Cumulative Distribution Function (CDF) of the standard normal distribution. The parameters β are typically estimated by maximum likelihood. It is also possible to motivate the probit model as a latent variable model. Suppose there exists an auxiliary random variable Y* such that: Y* = X´β +ε, Where ε ~ N (0, 1). Then Y can be viewed as an indicator for whether this latent variable is positive: Y = 1 {y*>0} = {1

if Y* > 0 i.e. – ε < X´β, 0 otherwise.}

Economic Papers and Notes, vol.14, no.1

18


The binary response variable predicted by the model is the decision to use mobile phone. Research objective lies in searching a probable impact of mobile phone usage on income profile of fishermen households. The X vector thus include three major independent variables representing impact of mobile phone use on individual income: the change in fishing income (Fishych), the change in other income (Otherych) and the capability of selling all (almost all) of the last couple of catches (Sell all).

an d

N

ot

es

Fishing income is actually defined as fishing firm earnings in this study. It can also be viewed as fishing firm ownerâ&#x20AC;&#x2122;s income. Other income means the income earned by activities other than fishing which includes agricultural income, remittances, wage or salary, business profit, interest income etc. The change in the above two types of income is captured either by the difference between incomes earned before & after starting the use of mobile phone; or by the difference incomes earned current year & the previous year. We also used a dummy variable showing reception of microcredit. Because of the absence of any well-defined and generalized indicator of household wealth status, we use the above dummy to identify if a household has the opportunity of easing its financial constraint. Age, education and marital status are also included as independent variables to show the impact of individual characteristics in determining the decision for using mobile phone.

3. Socio-economic Characteristics

om ic

Pa

pe

rs

In this paper, demographic characteristics such as age, sex and education as well as socioeconomic characteristics such as household assets and livelihood activities are assessed. These characteristics provide an overview on the background of the respondents, which in turn provides an overview about the suitability of the study population. Without necessarily being the source of poverty, it has been pointed out that having a particular characteristic may be associated with poverty. For example, most households that depend on agriculture, livestock and fishing keeping are more likely to be poor.

Ec

on

We have also collected data about their profession that how they manage fishing. Also, data has been collected about their economic activities like have they received microcredit or do they pay any usury. Here in the following table we show their professional activities. NonUsers of mobile phone number is 53 out of 205, which is 25.85% and number respondent using mobile more than 1 year is 92 out of 205, which is 60.53%. These characteristics are described in table 1, 2 and 3. Table 3: Information (Qualitative) about Mobile Phone of the User Subject Number of User Percentage Length of using mobile phone <6month 18 11.8% 7 to 12 months 42 27.6% >1 year 92 60.5% Reason of using mobile phone Family 20 13.2% Business 93 51.2% Other 39 25.7%

19

North American Academic Journals, 2014


The level of literacy rate is 30.7 % which is less than rural Bangladesh illiteracy rate 50.6% (BBS, 2013). Mobile phones are easy to afford and do not require the user to have much technological knowledge or even to be able to read or write, so this group with primary or no education can operate this. Table 4: Income Status For Mobile Users (N = 152) Variable

84 (41.0 %)

N (%) 75 (36.6 %) 13 (6.3 %)

es

N (%) 87 (42.4 %) 34 (16.6 %)

ot

Household monthly income Lower Income (0-10000) Average Income (1000015000) Higher Income (>15000) For Non-Mobile Users (N = 53) Variable Income [Mean (±S.D.)] (Bangladeshi taka)

115 (56.1 %)

N

Income [Mean (Bangladeshi taka)

Income Before use Income after mobile use mobile (±S.D.)] 13068.39 (± 11840.35) 20854.72 (±22868.66)

Present Year 4134.08 (±9743.71)

rs

an d

Last Year 4666.97 (±11390.37)

Ec

on

om ic

Pa

pe

In order to understand the socio-economic status of the households, a number of household assets and livelihood activities were assessed through multiple-response questions. As all respondents are fishermen, so their main earnings are from this occupation. At the same time, they are also involved in some agricultural productions. Also, some respondents are partly involved in business. As we collected the whole information of their family, so the other source of income with fishing by the other members of the family are also involved. The other member either may be migrated or jobholder or may be involved in business like storekeeper. So, in our data, all of the income earnings through these income sectors are also included. The Income group are separated in three groups with their monthly income. The lower income group with the income from 0 to 10000 taka, average income group with the income from 10000 to 15000 taka and higher income group with the income higher than 15000 taka. For mobile phone user, the average income before mobile phone use was 13068.39 taka where after the use of mobile phone this average income raised to 20854.72 taka. For the non-user of mobile phone we collected the data about income for the previous year. The estimated average income of the last year was 4666.97 taka where in present period this income has decreased to 4134.08 taka. The estimated income of the mobile phone user and non-user are given by the following Table-4.

4. Impacts of Mobile Phone Use on Fishing Community’s Business The study has sought to identify the impacts of the mobile phone in fishing community life. Impact refers to the difference that access to the mobile phone has meant to the individuals in

Economic Papers and Notes, vol.14, no.1

20


the study areas. Assessments of impacts are based on the self-reported advantages of mobile phone access that the interviewees have indicated. Access to the mobile phone has above all meant the creation of more opportunities and choices, but it has also provided help in managing uncertainty. Moreover, existing business relations have been strengthened. We have asked several questions related mobile phone use in their business and they answered (Table- 5). The advantages that the users feel the mobile phone has given them in business transactions is above all relate to the reduced access time to information. Reduced communication expenses are also important to many.

N

ot

es

Table 5: Business related advantages of mobile phone users (N=152) Mobile phone has helped in: Agreed Easy to use 125 (82.2%) Easy to access market information 128 (84.2%) Reducing search cost and improve market knowledge 110 (72.4%) Reducing risk 119 (78.3%)

Pa

pe

rs

an d

Most of the respondents say that they call mostly for their business purpose. Before introducing mobile phone they had to accept the middlemanâ&#x20AC;&#x2122;s price offer for fishes because they had no other way to know the market price for fishes in the bigger markets. Mobile phone gave them the opportunity to verify the market price of fishes. Now, before sell fishes to middlemen, they do verify the market price in the nearby markets and only agree to sell when they get a good price. Now they feel much more confident as they have gained bargaining power with the middle men who mostly deprive them from their profit.

Ec

on

om ic

Table 6: Business information of mobile phone users (N=152) Business Related Agreed Strongly Neither Questions Agreed Agreed nor N (%) Disagreed N N (%) (%) After using mobile 102 (67.1%) 5 (3.28%) 31 (20.39%) phone income has increased After using mobile 78 (51.32%) 6 (3.95%) 33 (21.71%) phone savings have increased After using mobile 82 (53.95%) 17 (11.18%) 37 (24.34%) phone expenditure has increased

Disagreed N (%)

Strongly Disagreed N (%)

14 (9.21%)

0 (0%)

35 (23.03%) 0 (0%)

14 (9.21%)

2 (1.32%)

After using mobile phone, 67.1 percent fishermen have agreed that their income has increased, 3.28 percent have strongly agreed. Again, about the increase in savings, after mobile phone use 51.32 percent have agreed, 3.95 percent have strongly agreed. Finally, about the increase in expenditure. 53.95 percent agreed that their expenditure has increased,

21

North American Academic Journals, 2014


11.18 percent have strongly agreed. 24.34 percent are indifferent whether 9.21 percent have disagreed and 1.32 percent strongly disagreed about the increase in price (table- 6). They are agreed mostly that the impact of mobile phone on rural market is that the rural suppliers could more easily get market information, they could more easily get price information and they find out that the market is expanding. Also they find out that the introduce of mobile phone strengthening their relationships with business partners, motivating himself in taking new initiatives and creating new economic or income generating opportunities.

ot

es

Findings of table- 7 suggest that contribution of mobile phones were enabling rural households in Sylhet region to overcome vulnerabilities related to social exclusion .The phones were also reducing travel times and monetary costs; decreases physical risks; and increases the outcomes of those necessary journeys. Furthermore, increased temporal accessibility enables people to manage several activities regardless of their physical location.

om ic

5. Empowerment

Pa

pe

rs

an d

N

Table 7: Qualitative information about mobile phone user’s (N=152) business Livelihood and development aspects Agreed N (%) Does mobile phone can reduce risk? 119 (78.29%) Do you get the assistance of health service initializing mobile 35 (23.03%) technologies? Do mobile application and practices can increase the benefit to 68 (44.74%) women? Do you face any harassment by others? 37 (24.34%) Do your productivity rise? 103 (67.76%)

on

Empowerment is the reduction of dependency, owners as well as users have experienced a variety of changes after access to the mobile phone. In rural Bangladesh, people have very little scope for choice in work or social relation but remain confined to the village and its limited income earning opportunities.

Ec

Economic empowerment refers not only to increases in income but also to having control over resource and resource management, decision making power, involvement in and control over economic transactions. Mobile phone, besides financial gain, could also facilitate the economic empowerment of women. Mobile phone has created an income generation opportunity for rural women. It has also provided scope for interacting with a wider crosssection of people. Obviously, mobile phone as a business venture provides an opportunity for financial gain for the users. Almost most of the fishermen 70.38% (Summation of “agreed” and “strongly agreed”, table- 6) have said their income have increased through mobile phone. In the majority of cases the income of the fishermen has increased with the length of the mobile phone owning period. The greater the length of ownership, the higher has been the increase in income. So, apparently, as an income opportunity, the mobile phone has been a success for the fishermen.

6.1 Probit Regression

Economic Papers and Notes, vol.14, no.1

22


Probit regression analysis attained in table-8 suggested mobile phone has a significant impact on social and economic condition.

Number of observation LR chi2 (7) Prob > chi2 Log likelihood

fish

an d

(1)*, **& *** denote 10%, 5% and 1%level of significance respectively, (2) Standard Error is repeated in parenthesis. (3) Otherych=Income from other source; Fishych = Income from fishing; Credit= Micro-credit; Sellall = Sell all

om ic

Note:

rs

Sellall

pe

Credit

0.0024*** (0.0007) 0.06 (0.23) 0.97*** (0.29) 205 59.28 0.0000 -88.560767

Pa

Fishych

N

ot

es

Table 8: Probit Regression: Dependent Variable- Probability of using mobile phone Variables Coefficient Constant -0.24 (0.47) Age -0.01335 (0.0085) Education 0.89** (0.29) Marital Status 0.29 (0.23) Otherych 0.0018** (0.0008)

Ec

on

Probit regression suggested that an increase in Age decreases the predicted probability of mobile phone use by 0.013. However, it can be easily seen that age has no significant influence on probability of mobile phone use. The coefficient of “Education” shows that an increase in education increases the probability of using mobile phone by 0.89. This result was significant at 5% level. Another coefficient of “marital status” was 0.29, which means that there is a positive impact of marriage in the predicted probability of income. One of the most important coefficient “income from other sources” (Otherych) was 0.00175, the result was significant at 5% level. This means that increase in income from other source than fishing cause increase in total income. The coefficient of “income from fishing” (Fishych) was 0.0024. This means the increase income from fishing increases the total income. This result was significant at 1% level. The coefficient of “microcredit” (Credit) was 0.06. This means that an increase in receiving microcredit causes an increase in total income. Coefficient of “selling all fish” (Sell all) was 0.97. This means that the increase in selling all fish causes an increase in total income. This result was significant at 1% level.

23

North American Academic Journals, 2014


The constant term is -0.24 which describes that predicted probability of income of the fishermen through mobile phone is extremely low if all of the predictors (Age, education, marital status, otherych, fishych, credit and sell all) are evaluated at zero. The Likelihood Ratio (LR) Chi-Square (χ2) was 59.28 assuming that the model converged with all the parameters. Here, the value of log-likelihood is -88.56, which is negative, indicating better fit of this model. Prob > χ2 - If Prob > χ2 tends to zero then there is no heteroscedasticity problem. Our probability of χ2 value is 0.0000 that rules out existence of heteroscedasticity problem. 6.2 Marginal Effects after Probit

Note:

ot N

(1) (+) dy/dx is for discrete change of dummy variable from 0 to 1 (2) *, **& *** denote 10%, 5% and 1%level of significance respectively. (3) Otherych=Income from other source; Fishych = Income from fishing; Credit= Micro-credit; Sellall = Sell all

on

fish

om ic

Pa

pe

rs

an d

Table 9: Marginal Effects after Probit Regression Variables dy/dx Age -0.004 (0.002) + Education 0.21*** (0.06) + Marital Status 0.08 (0.07) Otherych 0.0005** (0.0002) Fishych 0.0007*** (0.0002) + Credit 0.017 (0.07) + Sell all 0.33** (0.11)

es

Marginal effects after probit is taken to find out the variation in the probability of increasing mobile phone use of the respondents. The marginal effects are calculated in Table- 9.

Ec

All else held constant, education increases the probability of income by 21.28% and this was significant at 1% level. As the literate fishermen who are more educated can operate mobile phone effectively than the illiterate fishermen and are more informed about new inventions. They know that the use of mobile phone can minimize their cost by proving various market information including ups and downs in prices, pick demand etc. Relying on the middlemen, instead do not provide the mark; they information about price and market demand. So, the use of mobile phone among the fishermen who are educated is higher than the fishermen who are not educated. Also, the income from other source rather than fishing, when all else held constant, increases the total mobile usage by 0.05 %, with a significance of 5% level .That is, the change in income from other sources like agriculture, remittances, businesses, wage or salaries and interest earnings were influenced by the change in attitude towards mobile use. Again, earnings from fishing increases the total income by 0.07 % and this result is significant at 1% level. That is the income earnings from fishing change the mobile usage

Economic Papers and Notes, vol.14, no.1

24


positively. The fishermen who were using mobile phone can sell their fishes with better prices. So, the total income of the fishermen who were mobile phone user greater than the fishermen who were not mobile phone user. We have to admit however, that the real impacts on the probability to use mobile phone by both income variables are quite low. Finally, selling all fish remaining increases mobile phone use by 33.14% and this result was significant at 1% level. So, the fishermen who are mobile phone user can sell all the fishes whereas the fishermen who were not the user of mobile phone cannot sell all the fishes very fast compared to fishermen who were the user of mobile phone. As fish is very perishable good, it becomes a vital incentive for the fishermen to use mobile phone.

ot

es

On the other hand, the impact of marital status and credit has no any significant impact on the change in total income. The status â&#x20AC;&#x153;Marriedâ&#x20AC;? for person increases the probability of increasing income by 8.14% which was not significant. Receiving credit has a positive probability and increases the probability of increasing total income by 1.67% which was not significant. That the receiving microcredit may enhance their wealth but the wealth status is not much more different of both the fishermen who are mobile phone user or not.

an d

N

Finally, the probability of increase in total income is negatively related to the age. As age increases, the probability of increasing income decreases at 0.37% rate, which was not significant.

rs

7. Conclusion

Ec

on

om ic

Pa

pe

The econometric estimation shows a clear incentive for using mobile phones from economic point of view. Though the returns to fishing income and other income are dismal, they show strong significance. One possible reason behind a weak influence of these two income variables is the small sample size for which distinguishing mobile phone users according to their extent of usage could not be made. However, it is clearly depicted that mobile phone users get the benefit when the sale of their stock is concerned. Such benefits are easily foreseen by the literate persons. It could be argued that income could be related to many other determinants other than mobile usage. But we have a strong relationship between education and mobile phone use, it is natural that smart people would use mobile phone more and can reap the benefits of using them easily. Hence use of mobile phone ease communicational constraints for marginal people like fishermen and provide scopes to trickle down their costs of marketing. Moreover, people enjoy scopes of increasing their income from other sources too. Mobile phone usage provides rural households with faster and easier mode of communication, by which they can improve their ability to access to livelihood assets and overcome their vulnerabilities. The mobile phones contribute to improve socio economic status through a number of ways. First, by expanding and strengthening social networks; increase peopleâ&#x20AC;&#x2122;s ability to deal with emergencies and to work together thereby reducing costs and increasing productivity. Secondly, mobile phones enable rural people to cut down travel costs, minimize physical risks, and expand efficiency of activities. Thirdly, using mobile phones information rural traders and farmers will be able to secure better markets and prices, save time and money, and promptly communicate business related information. However, mobile phones have contributions to improve incomes of rural households but not much significantly.

25

North American Academic Journals, 2014


ot

es

With the incremental impact of mobile phone usage on total income of a marginal group of people, the fishermen, we support all policies supporting enhanced mobile phone usage in our country, probably in lower costs. Government taxes in this sector can be relaxed more. In excess to this government can provide more incentives to mobile phone operators to work sincerely for marginal people and to provide necessary training along with motivational programs.

N

8. References

an d

[1] Akter, J.C. (2008), “Does digital divide or provide? The impact of cell phones on grain markets in Niger”, Center for global development, Economics Department, Fletcher School of Law and Diplomacy, Tufts University.

pe

rs

[2] Bangladesh Bureau of Statistics (BBS), (2013), “Literacy Assessment Survey (LAS) 2011”, Statistics and Informatics division, Ministry of Planning, Government of Bangladesh.

Pa

[3] Bangladesh Economic Review 2013, (2013), Finance Division, Ministry of Finance, Government of Bangladesh.

om ic

[4] Bayes, A. (2001), “Infrastructure and rural development: insights from a Grameen Bank village phone initiative in Bangladesh”, Agricultural Economics, Vol. 25, pp. 261-272.

on

[5] Bliss, C. I. (1935), “The calculation of the dosage-mortality curve”, Applied Biology, Vol. 22, pp. 134-167.

Ec

[6] Comin, D. and Hobijn, B. (2004), “Cross-country technology adoption: making the theories face the facts”, Journal of Monetary Economics Vol. 51, pp. 39-83. [7] Daykin, A. R. and Moffatt, P. G. (2002), “Analyzing ordered responses: A review of the ordered Probit Model”, Understanding Statistics, Vol. 1, No. 3, pp. 157-166. [8] Donner, J. (2004), “Microentrepreneurs and Mobiles: An Exploration of the Uses of Mobile Phones by Small Business Owners in Rwanda”, Information Technologies and International Development, Vol. 2 No. 1, pp. 1-21. [9] Donner, J. (2008), “Research Approaches to Mobile Phone Use in the Developing World: A Review of Literature”, The Information Society, Vol. 24, pp. 140-159. [10] Gordon, J. (2007), “The Mobile Phone and the Public Sphere: Mobile Phone Usage in Three Critical Situations”, Convergence, Vol. 13 No. 3, pp. 307-319.

Economic Papers and Notes, vol.14, no.1

26


[11] Islam, M. S. and Grölund, A. (2007), “Agriculture market information E-Service in Bangladesh: A stakeholder-oriented case analysis”, LNCS series, Vol. 4656/2007, pp. 167178. [12] Islam, M. S. and Grölund, A. (2010), “An agricultural market information service (AMIS) in Bangladesh: evaluating a mobile phone based e-service in a rural context”, Information Development, Vol. 26 No. 4, pp. 289-302. [13] Islam, M. S. (2011), “Creating opportunity by connecting the unconnected mobile phone based agriculture market information service for farmers in Bangladesh”, Örebro University, Örebro Studies Informatics 4.

es

[14] Islam, M. S. and Grölund, A. (2011), “Factors influencing the adoption of mobile phones among the farmers in Bangladesh: Theories and Practices”, International Journal on Advances in ICT for emerging regions, Vol. 4 No.1, pp. 4-14.

N

ot

[15] Jensen, R. (2007), “The digital provide: Information (Technology), market performance, and welfare in the South Indian fisheries sector”, The quarterly journal of Economics, Vol. 122 No. 3, pp. 879-924.

an d

[16] Lu, N. and Swatman, P.M.C. (2009), “The MobiCert project: integrating Australian organic primary producers into the grocery supply chain”, Journal of Manufacturing Technology Management, Vol. 20 No.6, pp. 887-950.

pe

rs

[17] McCulloch, R. and Rossi, P. E. (1994), “An exact likelihood analysis of the multinomial probit model”, Journal of Econometrics, Vol. 64, pp. 207-240.

Pa

[18] Mobile phone subscriber in Bangladesh, (2013), Government of Bangladesh, Bangladesh telecommunication regulatory commission (BTRC), Dhaka. Available at: http://goo.gl/YgiLrA (accessed February 8, 2014).

on

om ic

[19] Muriithi, A. G. and Bett, E., et al. (2009), “Information technology for agriculture and rural development in Africa: Experiences from Kenya”, Conference on International research on food security, Natural resource management and rural development, University of Hamburg.

Ec

[20] Nikolov, G. and Hughes, D. (2000), “Market information services in economies in transition: The case of Bulgaria”, Journal of Euromarketing, Vol. 8 No. 3, pp. 53-76. [21] Rashid, A. T. and Elder, L. (2009), “Mobile phones and development: An analysis of IDRC-supported projects”, The Electronic journal on information system in developing countries, Vol. 36 No. 2, pp. 1-16. [22] Shepherd, A.W. (1997), “Market information services: Theory and practices”, Agricultural services bulletin, No. 125, FAO, Rome. [23] Souter, D., Garforth, C., Jain, R., Mascarenhas, O., McKerney, K. and Scott, N. (2005), “The Economic Impact of Telecommunications on Rural Livelihoods and Poverty Reduction: A Study of Rural Communities in India (Gujarat), Mozambique, and Tanzania”, Commonwealth Telecommunications Organisation for UK Department for International Development. [24] Sreekumar, T. T. (2011), “Mobile phones and the cultural ecology of fishing in Kerala, India”, The information Society: An international Journal, Vol. 27 No. 3, pp. 172-180.

27

North American Academic Journals, 2014


EXPORT-LED GROWTH IN EUROPE: WHERE AND WHAT TO EXPORT? Paula Gracinda Santos 1, Ana Paula Ribeiro 2 and Vitor Manuel Carvalho 3

Abstract From the late 70s onwards, the literature has produced numerous studies, mostly for developing countries, relating exports and economic growth. Since several European Union

es

(EU) countries face strong recessions in the sequence of the economic crisis and the related fiscal consolidation measures, exports emerge as a meaningful source of growth for

N

ot

developed countries with rather stagnant domestic markets.

In this context, we assess if and how the product and the destination structures of exports

an

d

shape the growth dynamics for the EU countries. Using panel data estimation to 23 of the 27 EU members over the period 1995-2010, we find that economic growth is foster through

rs

export specialization in high value-added products, such as manufactures and high-

pe

technology. Moreover, we find evidence that higher growth is fostered by export

Pa

diversification across partners while enlarging the portfolio of partners, mainly to less developed and more distant countries, has negative impacts on European growth.

ic

Unambiguously, relative concentration of exports should be directed towards higher growth

on om

countries.

Keywords: Economic growth; Product structure of exports; Exports’ destination; European Union; Panel data.

Ec

JEL classification codes: C23; F10; O40; O52.

1

Faculdade de Economia, University of Porto. Faculdade de Economia, University of Porto, and CEF.UP – Center for Economics and Finance at University of Porto. CEF.UP is financially supported by FCT (Fundação para a Ciência e a Tecnologia). Corresponding author: please email to aribeiro@fep.up.pt or address to Rua Dr Roberto Frias, 4200-464, Porto, Portugal. 3 Faculdade de Economia, University of Porto, and CEF.UP. 2

Economic Papers and Notes, vol.14, no.1

28


INTRODUCTION From the late 70s onwards, the literature has produced several studies relating exports and economic growth, although with an unclear unique-direction causal relationship. For the European Union (EU) countries, facing strong domestic recessions in the sequence of the economic and financial crisis and the related public debt correction measures, exports’ growth emerges as a meaningful source of economic growth. Moreover, given that exports are a potential source of growth, a more refined analysis is in order: with a view to maximize the effects on growth, criteria on what and where to export may not be negligible. In this study, we aim at assessing if and how product and destination structure of exports shape the output

es

growth dynamics for the EU countries.

ot

Among the relevant literature, most studies focus on the Export-led growth (ELG) hypothesis,

N

motivating and testing to what extent an increase in the volume of exports contributes to

d

higher economic growth in the country of origin. In parallel, but to a rather small extent, there

an

is some research that focuses, alternatively, on the product structure of exports or on the destination of exports as determinants of economic growth or/and exports’ growth. In this

rs

context, our research contributes to the literature because it tests, simultaneously, how the

pe

product and destination structure of exports influences the economic growth in the country of

Pa

origin. Additionally, and given the current environment constraints on the European growth prospects, our study relies on panel data estimation for the EU countries, whereas most of the

ic

collected literature on exports as a growth device applies to developing countries. Finally, we

on om

make an attempt to suggest export-supporting policy guidelines on where to and what should the European countries export.

The paper proceeds as follows. In section 1, we provide a review of the ELG hypothesis

Ec

(briefly compared with the, alternative, “growth-driven exports” hypothesis) and an exhaustive review on how the product structure and destination shape economic growth and exports’ growth. Section 2 presents data, methodology and the analysis of the estimation results. In section 3, we tentatively produce a note on export policy, given the current structure of European exports in terms of product and destination. Finally, we present the final remarks in section 4.

29

North American Academic Journals, 2014


1. EXPORT-LED

GROWTH

A LITERATURE OVERVIEW ON THE ROLE OF PRODUCT

STRUCTURE AND DESTINATION OF EXPORTS

The link between exports and economic growth has been, for a long time now, an important and attractive area of research, widely explored in the literature (e.g., Michaely, 1977, Balassa, 1978; Feder, 1983; Awokuse, 2008). Although the findings are not unanimous, a substantial amount of literature supports the export-led growth (ELG) hypothesis, both on theoretical and empirical grounds. A first argument for the ELG is that “openness” enlarges market dimension, and an increase

es

in production and sales arises as a result of higher demand pressure (Ramos, 2001; McCann,

ot

2007; Hesse, 2008; Andraz and Rodrigues, 2010; Soukiasis and Antunes, 2011). Moreover, an expansion in exports may also promote specialization, particularly in the production of

N

tradable goods, promoting a better reallocation of resources from (relatively) inefficient non-

d

tradable sectors to higher productivity export-oriented sectors, while enabling comparative

an

advantages; thus, as exports enlarge, domestic production rises through productivity growth

rs

(Awokuse, 2008; Andraz and Rodrigues, 2010; Soukiasis and Antunes, 2011; Lorde, 2011).

pe

Additionally, export effort involves facing stronger competitiveness which favors the exploitation of economies of scale and contributes to an acceleration of technical progress and

Pa

a greater integration of production processes (Ramos, 2001; Awokuse, 2008; Andraz and

ic

Rodrigues, 2010). International trade is found to favor "spillover-effects" from technology

on om

and knowledge transfers (e.g., Coe and Helpman, 1995, Keller 2004, Kali et al., 2007, Soukiasis and Antunes, 2011). Finally, export growth relaxes the external financial constraint of the country: it increases the potential demand of the economy and, consequently, increases the ability to save more and to capital accumulation; at the same time, it enables the country

Ec

with larger capability to import intermediate capital goods. Both effects contribute to growth (Ramos, 2001; Awokuse, 2008). Since we find significant theoretical and empirical support for exports to work as an engine of growth, a more refined analysis on the nature of this relationship requires a further review on detailed aspects of exports. The product structure and the destination of exports are often presented in the literature as non-neutral characteristics in driving economic growth.

Economic Papers and Notes, vol.14, no.1

30


1.1. Product structure of exports A country cannot simply increase its exports to ensure economic growth since the composition and concentration of the exported goods are found to be also relevant factors (McCann, 2007 and Hausmann et al., 2007, among others). The decision on what to export depends on production costs, specific costs of the product at the destination in question, market structure and consumer preferences and income (Amador and Opromolla, 2008); additionally, the pattern of product specialization is not independent of the level of development of the origin country (Spilimbergo, 2000).

es

The development, production and consumption of new goods (usually embedded with

ot

growth-delivering technology) are more likely to occur, first, in more advanced countries, arising only later in less developed countries (Stokey, 1991). On the demand-side, this is

N

explained, for example, through the theory of product life-cycle according to which the

d

demand for certain types of consumption goods is higher in countries with higher income

an

(Vernon, 1966). On the supply-side, Grossman and Helpman (1991) argue that advanced

rs

economies are endowed with technological advantages, particularly when it comes to R&D.

pe

Thus, the more developed regions, rich in skilled labor and superior technology, producing (and thus exporting) more sophisticated goods, the greater is the potential for transmission of

Pa

knowledge and skills and therefore to higher economic growth (Spilimbergo, 2000).

ic

Moreover, the export of more sophisticated goods also leads to more efficient management

on om

practices while stimulating innovation and technological advance (McCann, 2007). In this sense, it seems relevant to analyze exports taking into account their technological component. One of the recent studies on this issue, Guaresma and Wรถrz (2005), tests the

Ec

hypothesis that exports of high-tech industries have a greater potential for positive externalities and higher productivity (in terms of improved efficiency and economies of scale). They found evidence that there is a difference when considering exports disaggregated according to technological intensity: while technology-intensive exports have a significant positive effect on economic growth, exports of products with low technological intensity exhibit a negative effect on economic growth. The same study concludes that the better performance of high-tech exports is due to the difference in productivity relative to that in the domestic sector arising from greater openness to foreign trade and from exposure to international competition. However, conclusions are different for developed countries compared to developing countries. For the former the results are not significant, only accruing

31

North American Academic Journals, 2014


positive growth effects to developing countries: marginal increases in capital, labor or exports, the rate of economic growth is greater the lower the level of development is (Balassaeffect as in McCann, 2007). Some authors also disaggregate exports into commodities/natural resources and industrial and processed products. According to Herzer et al. (2004), there is evidence of a positive impact of manufacturing exports on economic growth, while exports of primary products exhibit negative impact on economic growth. Such findings can be interpreted as stemming from the effects of increased productivity associated with the industrial sector compared to those appending on primary goods (Herzer et al., 2004). Countries that export goods with high

es

levels of productivity benefit from faster economic growth (Hausmann et al., 2007). It is also

ot

argued that, based on endogenous growth theory, the diversification of exports towards export

N

more technology-advanced products, at the expense of "commodities", can contribute to

d

positive externalities in other sectors (Herzer and Nowak-Lehmann, 2006). Greenaway et al.

an

(1999) also test the impact of exports on GDP growth by disaggregating them into fuel, food, metals, other commodities, textiles and other manufactured goods. In contrast, they conclude

rs

that exports of fuels, metals and textiles to reveal important engine of economic growth, given

pe

the relative weight of the textile sector in developing countries and because metals and fuels

Pa

represent inputs of great importance to most developed countries. Ziramba (2011), by decomposing exports into merchandise exports, net gold exports, export of services and

ic

income receipts, finds that real merchandise exports lead growth and that there is evidence of

on om

reverse causality in the case of service exports and income receipts. For net gold exports there is no causal relationship in either direction. Another important issue for this analysis relates to product concentration in the exports

Ec

portfolio. It will then be useful to know if it reveals more advantageous to specialize in certain products for export or whether it is more productive to diversify and invest in a wider variety of goods. On the one hand, concentration of exports in certain products may allow economies of scale and enable the firms to move along the learning curve (Bebczuk and Berrettoni, 2006). For instance, the decrease in transport costs can lead to a reduction in the number of products produced domestically, thus promoting specialization (Dornbusch et al., 1977). Hausmann and Rodrik (2003) emphasize that investors face significant uncertainty about the costs in the production of new goods: if they are successful, the gains will be for the whole society

Economic Papers and Notes, vol.14, no.1

32


(information spillovers) but, in case of failure, losses will accrue to the private sector (investor). Thus, investment possibilities are withdrawn. On the other hand, a diversification strategy ensures the stability of profits, leading the company to invest in some sectors related to its current portfolio (Bebczuk and Berrettoni, 2006), and also contributing to the stabilization of export earnings in the long run (Ghosh and Ostry, 1994). Moreover, it is also argued that diversification is an endogenous process that moves along with economic development: under certain assumptions, the Engel effects imply that higher income levels demand for greater economic diversity of consumption goods, forcing, consequently, producers to invest in a wider range of sectors (Acemoglu and

es

Zilibotti, 1997). Additionally, one of the reasons, most frequently mentioned in the defense of

ot

diversification of exports, points to the knowledge transfer of new production techniques,

N

management or marketing to new industries (Hesse, 2008). The instability of exports is

d

another factor that contributes to the diversification of exports. Diversification will prove

an

beneficial for less developed economies as "commodities" are too volatile to price changes; countries dependent on these products might, thus, suffer negative consequences due to

rs

excessive oscillation once the elasticity of demand is too small (Hesse, 2008). A final

pe

argument used by apologists of export diversification that countries must export goods for

Pa

which world demand is increasing and that, regardless of a country producing more primary goods or manufactured, is the compatibility with the global demand that will determine the

on om

ic

growth of its exports (Alexander and Warwick, 2007). But export diversification prescription can differ in the context of more or less developed economies: the more developed countries tend to diversify their exports through innovate and invest in new technologies and not just by exporting a larger volume (Hummels and Klenow,

Ec

2005), while developing countries tend to imitate and to export the products where they have a greater advantage, namely those related to natural resource abundance and/or low cost of manpower (Hesse, 2008). Among the empirical literature, Al-Marhubi (2000) and Lederman and Maloney (2003) conclude for the positive impact of diversification of exports on economic growth. Some studies have, though, different conclusions when considering developed countries or countries at delayed phases of development (Hesse, 2008; Imbs and Wacziarg, 2000; McCann, 2007; Bonaglia and Fukasanu, 2003). Hesse (2008) and Imbs and Wacziarg (2000) conclude that specialization is beneficial for countries in more advanced stages of development while

33

North American Academic Journals, 2014


diversification is a best strategy for developing countries. In this sense, McCann (2007) and Bonaglia and Fukasaku (2003) also conclude that diversification is more important for developing countries and, thus, defend the these countries should be encouraged to diversify their exports to technologically more advanced sectors as to contribute to their economic growth. Note that this technological advance is recommended to occur in sectors where the country is already exporting before (â&#x20AC;&#x153;product proximityâ&#x20AC;?), notably with regard to countries with abundant natural resources for example, forestry and mineral sectors have been proof of that, recording a significant development in terms of technologies used (Bonaglia and Fukasaku, 2003).

es

Despite the diversification of exports being pointed out by many authors as a determinant of

ot

economic growth (Bonaglia and Fukasaku, 2003; Herzer and Nowak-Lehmann, 2006;

N

McCann, 2007; Hesse, 2008), there are some studies that find evidence that expertise in some

d

sectors also may prove beneficial to economic growth, as is the case of specialization in the

an

electronics sector (Amable, 2000) or in sectors with higher growth rates, generally more technologically advanced (Laursen, 2000; Guaresma and WĂśrz, 2005). Peneder (2002) also

rs

concludes that specialization in services represents a burden for future growth, while more

pe

technologically intensive exports have positive effects on economic growth. In this sense,

Pa

Hausmann et al. (2007) argue that countries specializing in goods that richer countries export

ic

exhibit faster growth than those specializing in the production of other goods.

on om

1.2. Destination of exports

The point that we want to explore next is the specific role of export destination on economic growth, an issue that the literature started to cover only recently and that still remains barely

Ec

explored. Internationalization is of strategic importance since, for instance, the expansion into new markets is among the main decisions in the life of a company. This option is often related with cultural or social links with former colonies, the need for more trading partners (which are also, usually, former colonies) or the proximity to (large) external markets (BaliamouneLutz, 2011). Basically, the decision to enter a new market proves to be as important as the decision to create a new company (Amador and Opromolla, 2008). In the literature we find a broad set of conflicting arguments in favor of destination diversification or for destination concentration. Moreover, and in particular, the literature also focuses on the optimal characteristics trade partners should exhibit.

Economic Papers and Notes, vol.14, no.1

34


A first set of arguments for destination diversification is related to the mechanism of technology and knowledge spillovers emerging from trade. To the extent that a country produces knowledge through research or experience, some countries generate more knowledge than others. In this sense, for the same export volume, the larger the number of trading partners, the greater the possibility of positive externalities resulting, namely, in terms of technology and exposure to new/different ideas (De Loecker, 2007). The adoption of new technologies helps to increase productivity and contributes to higher economic growth (Coe and Helpman, 1995). Moreover, countries that export to a wider range of markets benefit not only because they face

es

an enlarged and more diversified market to sell their products, but also because firms come

ot

across with new consumerâ&#x20AC;&#x2122;s tastes, government regulations and other business environments

N

(Lederman and Maloney, 2003).

d

Also, an increasing number of trading partners, resulting from the expansion of potential

an

markets, attracts local and foreign investment which is shown to play an important role in

rs

technology diffusion and innovation and, consequently, in economic growth (Grossman and

pe

Helpman, 1991).

Pa

At the same time, the greater the diversity of trading partners, the stronger is the need for permanent development of innovations as to remain in a given market. Since fierce

ic

competition requires a continuous search for productivity gains, it impinges positively on

on om

economic growth (Kali et al., 2007).

Furthermore, the diversification of trading partners reveals positive because it minimizes the risk of relying on a small number of export markets and, thus, reduces the export-dependency

Ec

in case of idiosyncratic shocks (Baliamoune-Lutz, 2011). However there are also arguments in favor of export concentration on a smaller number of countries. Concentration can help minimize the costs associated with insufficient commercial infrastructure such as ports, airports, diplomatic posts, among others (Kali et al., 2007). Frankel et al. (1995), for example, point transportation costs as one of the main reasons for the emergence of specific trading blocks. Thus, when the infrastructure related to trade is not well developed, the concentration of trade destinies can help reducing transport costs (Kali et al., 2007).

35

North American Academic Journals, 2014


Besides the number of trading partners, the type of countries towards which exports are oriented to is also an important determinant for the role of exports in promoting economic growth. In this regard, the most obvious channel operates through external demand growth: the higher the average growth rate of the trading partners, the higher is their demand growth for imports, which directly contributes to a higher net exports growth of the country of origin (Arora and Vamvakidis, 2005). Moreover, and since countries at different stages of development demand, on the one hand, for different products and, on the other hand, influence differently the country of origin through technological spillovers, the choice of where to export is not innocuous (Coe and

es

Helpman, 1995). Additionally, and in this context, the choice of the menu of trading partners

ot

is rather limited: stating that "The G-7 countries accounted for about 84 percent of global

N

spending on R&D in 1995", Keller (2004: 752) argues that knowledge is concentrated in a

d

few countries.

an

Vacek (2010) finds that exports to more developed regions, which are pushing the world

rs

technological frontier forward, lead to higher productivity gains for the country of origin. On

pe

the one hand, exports to customers in more advanced countries requires a greater degree of attention to product quality and/or deliverance time, meaning that companies continually seek

Pa

to improve their performance by introducing innovations - improved methods of packaging

ic

and transport, adaptations to attract foreign consumers, product innovations, among others -

on om

(Kali et al., 2007). On the other hand, the most advanced countries are endowed with a greater learning potential, more sophisticated production techniques, marketing and management strategies, and better design of inputs (Vacek, 2010). In this sense, establishing trade relations with countries in a more advanced stage of development favors the exporting country as it has

Ec

access to a greater amount of knowledge (Damijan et al., 2004) and may also benefit from the expertise of their buyers (Clerides et al., 1998). Conversely, exports to less developed markets may lead to declining productivity, as an environment with fewer requirements for product quality and delivery timings would make exporters to become less efficient (Vacek, 2010). However, the above results related to technology and other efficiencies spillovers are also sensible to the degree of development of the export-origin country. According to Kali et al. (2007), the marginal benefit of an additional trading partner is different for poor or rich economies. If, on the one hand, new technologies increase the productivity of older technologies, the effect of an additional trading partner on growth should be lower for a poor

Economic Papers and Notes, vol.14, no.1

36


economy since it holds a smaller stock of knowledge to implement technological updates. On the other hand, the fact that the stock of knowledge is lower in poor countries, could imply that the contribution of an additional trading partner in terms of new knowledge (with impact on growth) is greater for developing economies. While these effects operate in opposite directions, both suggest asymmetric growth gains from trade accruing to rich and poor countries. A final note is in order: Amador and Opromolla (2008) found that destination and product diversification of exports are both determinants of growth. Their study relies on micro-data and an analysis is made for the dynamics of export structure of companies located in Portugal,

es

during the period 1996-2005. The authors conclude that multi-product and multi-destination

ot

firms are crucial in explaining the level and growth rates of Portuguese exports; in particular,

N

firms exporting four or more products and operating in four or more different markets are

d

responsible for about two thirds of total exports. The authors also find evidence that growth in

an

new markets is achieved mostly through the, simultaneous, introduction of new products in

rs

the firm’s export portfolio.

pe

Using a panel of more than 100 countries across four decades, Arora and Vamvakidis (2005) show that trading partners’ growth has a strong effect on domestic growth. Trading partners’

Pa

relative income levels are also positively correlated with growth, suggesting that the richer

ic

trading partners are, the stronger is conditional convergence. A general implication of the

on om

results is that countries benefit from trading with fast-growing and relatively more developed countries. Also, Baliamoune-Lutz (2011) concludes that where a country exports matters for the exporting country’s growth and development. Performing Arellano-Bond GMM estimations using panel data over the period 1995-2008 to explore the growth effects of

Ec

Africa’s trade with China, she finds that there is no empirical evidence that exports to China enhance growth unconditionally but export concentration enhance the growth effects of exporting to China, implying that countries which export one major commodity to China benefit more (in terms of growth) than do countries that have more diversified exports. 2. THE ROLE OF THE STRUCTURE OF EXPORTS TO THE ECONOMIC GROWTH OF THE EU 2.1. Data and methodology In this section, we estimate a simple export-augmented Solow-decomposition growth model in order to investigate the relationship between exports (including diversification of products

37

North American Academic Journals, 2014


and destinations) and real income per capita growth in the European Union. This framework is of rather widespread use in the literature (e.g., Feder, 1983, Guaresma and WĂśrz, 2005, Hausmann et al., 2007, and Dreger and Herzer, 2012). In particular, we estimate a panel growth regression using data for 23 EU countries from 1995 to 2010, following the standard panel-data specification in the literature: 4 đ?&#x2018;&#x2026;đ?&#x2018;&#x2019;đ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;?đ?&#x2018;&#x2019;đ?&#x2018;&#x; đ?&#x2018;?đ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D; đ??şđ??ˇđ?&#x2018;&#x192; đ?&#x2018;&#x201D;đ?&#x2018;&#x;đ?&#x2018;&#x153;đ?&#x2018;¤đ?&#x2018;Ąâ&#x201E;&#x17D;đ?&#x2018;&#x2013;đ?&#x2018;Ą = đ?&#x2018;Şđ?&#x2018;&#x2013;đ?&#x2018;Ą + đ?&#x153;ˇđ?&#x2018;żđ?&#x2018;&#x2013;đ?&#x2018;Ą + đ?&#x2019;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;Ą , đ?&#x2018;&#x201C;đ?&#x2018;&#x153;đ?&#x2018;&#x; đ?&#x2018;&#x2013; = 1, â&#x20AC;Ś , 23 đ?&#x2018;&#x17D;đ?&#x2018;&#x203A;đ?&#x2018;&#x2018; đ?&#x2018;Ą = 1995, â&#x20AC;Ś , 2010

(2.1)

The dependent variable is the average real per capita GDP growth rate; đ?&#x2018;Ş is the matrix of

es

constant terms (including potential cross-section and time effects); đ?&#x153;ˇ is the matrix of

ot

parameters to be estimated; and đ?&#x2019;&#x2013; is the vector of error terms. đ?&#x2018;ż is the matrix of independent

N

variables that includes variables of standard use in growth regressions:

d

â&#x20AC;˘ Population growth is measured by the growth rate of population, as percentage change on

consists of outlays on additions to the fixed assets of the

rs

â&#x20AC;˘ Gross capital formation

an

previous year;

pe

economy plus net changes in the level of inventories (measured as percentage of GDP);

Pa

â&#x20AC;˘ Inflation as measured by the consumer price. Inflation is often included as a macroeconomic stability control variable, usually impinging significantly and negatively

ic

on output growth (e.g., Arora and Vamvakidis, 2004, 2005). Additionally, we kept this

on om

variable because inflation differentials are a measure of price-competitiveness for most of the countries in the sample, i.e., the members of the European and Monetary Union. In addition, đ?&#x2018;ż includes refined indicators of exports motivated by the mechanisms explored

Ec

above:

â&#x20AC;˘ Number of partners is the total number of countries to where a country exports. â&#x20AC;˘ Partnerâ&#x20AC;&#x2122;s growth is a constructed index capturing a weighted average growth rate of the main trading partners of each country in our sample (i). Based on total exports by destination, we first calculate the share of exports for each country in the total exports of the origin country. Then, we select N representative partners (those receiving more than

4

Due to data restrictions, we have considered only 23 out of the 27 EU members: Austria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden and the United Kingdom.

Economic Papers and Notes, vol.14, no.1

38


1% of total exports from the origin country). After that, we calculate the relative weight of each trading partner on total exports for the N representative partners (đ?&#x2018;¤đ?&#x2018;&#x2014; ). The index is defined as:

đ?&#x2018;

đ?&#x2018;&#x192;đ?&#x2018;&#x17D;đ?&#x2018;&#x;đ?&#x2018;Ąđ?&#x2018;&#x203A;đ?&#x2018;&#x2019;đ?&#x2018;&#x;đ?&#x2018;  â&#x20AC;˛ đ?&#x2018;&#x201D;đ?&#x2018;&#x;đ?&#x2018;&#x153;đ?&#x2018;¤đ?&#x2018;Ąâ&#x201E;&#x17D; đ?&#x2018;&#x2013; = ďż˝ đ?&#x2018;¤đ?&#x2018;&#x2014; . đ?&#x2018;&#x;đ?&#x2018;&#x2019;đ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;?đ?&#x2018;&#x2019;đ?&#x2018;&#x; đ?&#x2018;?đ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D; đ??şđ??ˇđ?&#x2018;&#x192; đ?&#x2018;&#x201D;đ?&#x2018;&#x;đ?&#x2018;&#x153;đ?&#x2018;¤đ?&#x2018;Ąâ&#x201E;&#x17D;đ?&#x2018;&#x2014; đ?&#x2018;&#x2014;

đ?&#x2018;&#x2013;, đ?&#x2018;&#x2014; = 1, â&#x20AC;Ś , đ?&#x2018; ,

and computed values are presented in Annex A. Arora and Vamvakidis (2004, 2005) also consider the real per capita GDP growth rate of trading partners but as a simple average.

es

â&#x20AC;˘ HHI-destination measures the exportsâ&#x20AC;&#x2122; concentration among the trading partners as in Kali

ot

et al. (2007), where a low number indicates low concentration. It consists of a Herfindahl-

N

Hirschmann concentration index for exports from country i to partner j, constructed as

d

follows: đ?&#x2018;

an

đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;â&#x2020;&#x2019;đ?&#x2018;&#x2014; đ??ťđ??ťđ??ź â&#x2C6;&#x2019; đ?&#x2018;&#x2018;đ?&#x2018;&#x2019;đ?&#x2018; đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x203A;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;đ?&#x2018;&#x203A;đ?&#x2018;&#x2013; = ďż˝ ďż˝ đ?&#x2018; ďż˝ â&#x2C6;&#x2018;đ?&#x2018;&#x2014; đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;â&#x2020;&#x2019;đ?&#x2018;&#x2014;

2

rs

đ?&#x2018;&#x2014;

pe

where đ?&#x2018; and đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;â&#x2020;&#x2019;đ?&#x2018;&#x2014; denote the total number of trading partners and the total value of exports

Pa

between countries đ?&#x2018;&#x2013; and đ?&#x2018;&#x2014;, respectively. It should be noted that even though the HHI-

destination index described above is a function of the number of trading partners, these two

ic

variables are not necessarily related and, a priori, there should be no multicolinearity

on om

problem for the regression analysis.

â&#x20AC;˘ HHI-product refers to the product market concentration index; it is also a Herfindahl-

Ec

Hirschmann index, taken directly from the Unctadstat database and defined as: đ??ťđ??ťđ??ź â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;&#x;đ?&#x2018;&#x153;đ?&#x2018;&#x2018;đ?&#x2018;˘đ?&#x2018;?đ?&#x2018;Ąđ?&#x2018;&#x2013; =

đ?&#x2018;Ľđ?&#x2018;? 2 ďż˝ â&#x2C6;&#x2019; ďż˝1/đ?&#x2018;&#x203A; đ?&#x2018;&#x2039; 1 â&#x2C6;&#x2019; ďż˝1/đ?&#x2018;&#x203A;

ďż˝â&#x2C6;&#x2018;đ?&#x2018;&#x203A;đ?&#x2018;?=1 ďż˝

Where đ?&#x2018;Ľđ?&#x2018;? represents the value of exports of productđ?&#x2018;?, đ?&#x2018;&#x2039; is the sum of exports of all products and đ?&#x2018;&#x203A; represents the number of products (SITC Revision 3 at 3-digit group level)

for the country đ?&#x2018;&#x2013;.

To measure the impact on growth of the different types of products that a country exports we have disaggregated exports into three categories and construct, as Guaresma and WĂśrz (2005) and Kali et al. (2007), a weighted sector export growth rate:

39

North American Academic Journals, 2014


đ?&#x2018;&#x160;đ?&#x2018;&#x2019;đ?&#x2018;&#x2013;đ?&#x2018;&#x201D;â&#x201E;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2019;đ?&#x2018;&#x2018; đ?&#x2018;&#x2019;đ?&#x2018;Ľđ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;Ąđ?&#x2018; đ?&#x2018;&#x2013;đ?&#x2018;  =

â&#x2C6;&#x2020;đ?&#x2018;&#x2039;đ?&#x2018;  đ?&#x2018;&#x2039;đ?&#x2018;  , đ?&#x2018;&#x2013; = 1, â&#x20AC;Ś ,23, đ?&#x2018;&#x2039;đ?&#x2018;  đ??şđ??ˇđ?&#x2018;&#x192;đ?&#x2018;&#x2013;

đ?&#x2018;  = 1, â&#x20AC;Ś ,3.

The three product-sector categories, s, respect to Food and agricultural exports, Fuel, ores and metals exports and Manufactures exports. Additionally, we have also included a more refined indicator of high value-added exports: â&#x20AC;˘ High technology exports measures the exports of products embedded with high R&D intensity, such as in aerospace, computers, pharmaceuticals, scientific instruments, and electrical machinery, as percentage of manufactured exports.

es

Values of real per capita and level GDP growth rates, population, product concentration

growth

-

were

extracted

from

the

UnctadStat

N

partnersâ&#x20AC;&#x2122;

ot

index, exports by destination - to compute the number of partners, the HHI-destination and

(http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx, accessed in May-June 2012).

an

d

Data regarding gross capital formation, inflation, high-technology exports and product discrimination of exports (Food and agricultural; Fuel, ores and metals; Manufactures) were

rs

extracted from World Development Indicators (WDI), accessed in May-June 2012 at

pe

http://data.worldbank.org/data-catalog/world-development-indicators.

General

descriptive

ic

2.2. Estimation results

Pa

statistics for the sample are presented in Annex B.

on om

Since our cross-section units are not random drawings from a larger sample (our sample covers 23 out of the 27 members of the European Union), the fixed effects model seems more adequate than the random effects model (Gujarati, 2004). In order to estimate the model we use the software Eviews that provides built-in tools for testing fixed effects against random

Ec

effects, and also for testing the joint significance of the fixed effects, cross-section or/and time series. Table 1 and Table 2 below, report the tests made to sustain this choice. Table 1 shows the test for random effects using the â&#x20AC;&#x153;Hausman Testâ&#x20AC;? for the two specifications chosen. The results strongly reject the null hypothesis that individual effects are uncorrelated with the other explanatory variables. Thus, the test points to the option for a fixed-effects model.

Economic Papers and Notes, vol.14, no.1

40


Table 1: Tests on cross-section random effects Specification (I) Hausman Test

Cross-section random

Specification (II)

Chi-Sq. Statistic

Chi-Sq. d.f.

Prob.

Chi-Sq. Statistic

Chi-Sq. d.f.

Prob.

42.790569

11

0.0000

48.491884

10

0.0000

Running the model under fixed-effects, the Eviews provides the test on the nature of the fixed effects (cross-section, period or both). Test results are presented in Table 2, below. Table 2: Tests on cross-section and period fixed effects

ot

Cross-Section/ Period F Cross-Section/ Period Chi-square

22

0.0000

2.887001

(15,320)

0.0003

46.707206

15

0.0000

0.0000

4.360098

(37,320)

0.0000

0.0000

150.224545

37

0.0000

Statistic

5.004684

(22,319)

0.0000

N

Period Chi-square

0.0000

Prob.

109.114206

22

0.0000

2.791687

(15,319)

0.0004

45.389630

15

0.0001

4.207147

(37,319)

146.249470

37

d

5.033668

109.364178

an

Period F

(22,320)

rs

Cross-section Chi-square

Prob.

d.f.

pe

Cross-section F

d.f.

Statistic

Pa

Redundant Fixed Effects Tests

Specification (II)

es

Specification (I)

ic

The first set consists of two tests (“Cross-section F” and “Cross-section Chi-square”) that

on om

evaluate the joint significance of the cross-section effects using sums-of-squares (F-test) and the likelihood function (Chi-square test). The corresponding restricted specification is one in which there are period effects only. The two statistic values (5.00 and 109.11 for specification

Ec

(I) and 5.03 and 109.36 for specification (II)) and the associated p-values strongly reject that cross-section effects are redundant. The next two tests evaluate the significance of the period dummies in the unrestricted model against a restricted specification in which there are cross-section effects only. Both F and Chisquare statistics strongly reject the null hypothesis of no period effects. The remaining results evaluate the joint significance of all of the effects. Both test statistics reject the restricted model with common intercept. Table 3, below, shows the model estimation results for the two specifications chosen (I and II).

41

North American Academic Journals, 2014


Table 3: Estimation results Specifications (I) Gross capital formation

(II)

0.256905* (5.085197)

0.267095* (5.181998)

Population growth

-1.450010* (-3.149617)

-1.707803* (-3.733583)

Inflation

-0.057358* (-3.937974)

-0.059466* (-4.166287)

Number of partners

-0.065456* (-3.623746)

-0.061757* (-3.365770)

-17.03450** (-2.175492)

-18.39811** (-2.297597)

1.336077* (6.239717)

1.329256* (6.167562)

es

HHI-destination

N

ot

Partnersâ&#x20AC;&#x2122; growth

0.141395 (0.600801)

d

Food and agricultural exports

an

Fuel, ores and metal exports

rs

Manufactures exports

Pa

High-technology exports

pe

HHI-product

0.093528 (0.388479)

0.181282 (1.167119)

0.158297 (1.018077)

0.119837* (2.812646)

0.136047* (3.099850)

7.964254 (1.612176)

9.280098*** (1.848063)

0.079187*** (1.874670)

on om

ic

No. Observations Adjusted R Squared F-Statistic Prob. (redundant cross-section/period fixed effects)

368 0.776569 27.57434 0

368 0.774125 27.76157 0

Notes: (1) Significant at 1% (*), 5% (**) and 10% (***); t-statistics in parenthesis.

Ec

(2) Estimations made under white-diagonal standard error correction for valid statistic inference.

From the results we conclude that the model delivers a good fit, with the adjusted R-squared around 78% and a high overall significance of the independent variables (F-statistics close to 28). We can also see from the t-statistics (in specification I) that all variables are significant with the exception of Food and agricultural exports and fuel, ores and metals (HHI-product is significant at 10.8% level of significance). Moreover, after controlling for High-technology exports, and with the previous two exceptions, all the variables are significant. Furthermore, with the exception of Number of partners, the signs of the coefficients associated with the independent variables are as expected from the literature. Gross capital

Economic Papers and Notes, vol.14, no.1

42


formation and Population growth have the predicted effect on Real per capita GDP growth, with the first being positive and the second negative (Greenaway et al., 1999; Arora and Vamvakidis, 2004; Arora and Vamvakidis, 2005; Kali et al., 2007; Hesse, 2008). As a measure of macroeconomic stability we use Inflation, which have the predicted negative sign as Arora and Vamvakidis (2004, 2005) find. Higher inflation rates are associated with higher price volatility that causes difficulties to planning and, thus, depresses investment. To analyze the impact of exportsâ&#x20AC;&#x2122; destinations on economic growth we use three indicators: Number of partners, HHI-destination and Partnersâ&#x20AC;&#x2122; growth.

es

The results obtained for the Number of partners suggest a negative impact on growth, a result

ot

strongly robust across all the specifications tested. According to our estimation, an additional trading partner decreases by 6-7 basis points the Real per capita GDP growth rate, keeping

N

other things constant. The empirical literature mostly points to a positive influence to growth

d

from an increasing number of partners but our sample is dominated by developed countries

an

which, during most of this period have increased the number of poor countries as partners.

rs

According to Kali et al. (2007), the number of poor countries trading partners has a

pe

significant negative influence on the growth of rich countries, as technology/knowledge spillovers are rather small (e.g., Coe and Helpman, 1995, Keller, 2004). Moreover, this result

Pa

for Europe apparently supports the view of meaningful transportation costs and or

ic

cultural/social barriers as European countries diversify to new markets.

on om

We now turn the discussion of the effects that trade concentration (HHI-destination) has on economic growth. The estimated coefficients for this variable were negative and statistically significant. Kali et al. (2007) also use this indicator and find differences when they split their

Ec

sample into a sub-sample of poor countries and one of rich countries. In their study this indicator was, for the most of the cases, considered positive and statistically relevant for both the total sample and the poor countries sub-sample. In contrast, for the rich countries subsample, the estimated coefficient was often insignificant and in some cases negative. As our sample is from the European Union, rich countries, the results seems to be consonant with Kali et al. (2007) because since the level of concentration increases as the HII-destination index increases, the results imply that poor countries benefit from more concentrated trade while the evidence for the rich countries is mixed at best. Based on the coefficientsâ&#x20AC;&#x2122; value we can conclude that a variation of 0.1 units in HHI-destination generates a decrease of 1.7 p.p in Real per capita GDP Growth rate, keeping other things constant.

43

North American Academic Journals, 2014


Overall, we argue that the combined results related to the Number of partners and the exportconcentration in partner countries apparently suggest that destination of exports should be diversified enough in order to prevent for asymmetric external shocks on domestic growth, but the enlargement to distant (involving higher costs of transportation, more bureaucratic procedures, adjustment to different economic, social and institutional structures) and less developed trading partners reduces technology and knowledge spillovers. Considering the indicator Partnersâ&#x20AC;&#x2122; growth, the results are in accordance with the literature. We can conclude that a country benefits more from exporting to countries that experience higher real per capita growth rates. This result is expected because the higher the average

es

growth rate of the trading partners, the higher is their demand growth for imports (Arora and

ot

Vamvakidis, 2005). The results show that a percentage point increase in Partnersâ&#x20AC;&#x2122; growth

N

increases by 1.34 p.p. the Real per capita GDP growth rate, keeping other things constant.

d

Besides, establishing trade relations with countries in more advanced stages of development

an

favors the exporting country as it has access to a greater amount of knowledge (Damijan et al., 2004) and may also benefit from the expertise of their buyers (Clerides et al., 1998). We

rs

have controlled for the average level of development of the trading partners (using the

pe

average GDP per capita) but results not reported showed that, across several specifications,

Pa

this variable was highly insignificant and had a substantial negative impact on overall significance. Thus the level of development of the trading partners is not a relevant

on om

ic

determinant of economic growth.

We now pay attention on the product structure of exports. To analyze the impact of different type of products exported, we have disaggregated merchandise exports into three categories: Food and agricultural, Fuel, ores and metals and Manufactures. The results also seem to be

Ec

reasonably in line with the literature. Although Food and agricultural and Fuel, ores and metals exports are not statistically significant, the signs are positive. Since the countries of our sample are not plenty of natural resources and demand for food tends to be income inelastic, itâ&#x20AC;&#x2122;s not surprising that the coefficients on these fail to reach significance (Greenaway et al., 1999). The results are stronger and according to the literature (Greenaway et al., 1999; Herzer et al., 2004) when we consider Manufactures, products with higher value-added. In order to better assess the impact of high value-added exports on economic growth, we add as an explanatory variable the High-technology exports, because many authors defend a

Economic Papers and Notes, vol.14, no.1

44


positive impact of this on economic growth (see for instance McCann, 2007; Guaresma and Wörz, 2005; Spilimbergo, 2000); high value-added exports mainly reflect a more complex product structure which, per se, have stronger effects on growth. Our conclusions reveal to be consistent with the literature indicating that, a 93% confidence interval, a one percent increase in the weight of High-technology exports on Manufactures, increases by 0.08 p.p. the Real per capita GDP Growth rate, keeping other things constant (specification I). According to Guaresma and Wörz (2005) technology-intensive exports have a significant positive effect on economic growth and better performance of high-tech exports is due to the difference in productivity relative to that in the domestic sector.

es

Finally, regarding the overall product diversification, the HHI-product captures exports’

ot

concentration in terms of sector or product types. The results underlying the literature on this

N

topic are ambiguous: some authors argue for concentration of exports while others refer that

d

diversification benefits more the growth of the origin country. Our results suggest that, for

an

Europe, exports’ concentration has a positive impact on economic growth: an increase of 0.1 in HHI-product increases by 0.796 p.p. the Real per capita GDP Growth rate, keeping other

rs

things constant. According to Bebczuk and Berrettoni (2006), “the development-export

pe

diversification nexus, though, appears to be governed by a U-shaped pattern, whereby

Pa

diversification increases at low income levels and concentration prevails at high income levels”, which seems consistent with our country sample. Also Hesse (2008) and Imbs and

ic

Wacziarg (2000) conclude that specialization is beneficial for countries in more advanced

on om

stages of development while diversification is a best strategy for developing countries. Since exports are a part of output through the expenditure identity, a positive and significant relationship between exports and output is almost inevitable, even if there are no productivity

Ec

effects from exports. To add robustness to our results, we have run the growth equation under the two specifications but allowing, as in Dreger and Herzer (2012), for non-export per capita output growth as the dependent variable. Results are quite similar, except that weighted exports growth exhibit stronger effects, with Food and agricultural exports becoming statistically significant at 10%. 3. THE STRUCTURE OF EUROPEAN EXPORTS AND POLICY IMPLICATIONS – A NOTE Using the results from the previous section, we intend to assess how the recent evolution of the exports structure has contributed to growth in Europe, aiming to draw a note on export.

45

North American Academic Journals, 2014


Figure 1 shows that in Greece, Italy, France and Spain, for example, the exports represent a small percentage of GDP (between 21% and 26%). The EU23 average is 46.2%, well behind the leading countries - Ireland (86.3%), Estonia (72.3%) and Slovakia (71%).

N

ot

es

Figure 1: EU23 - Exports by country (% GDP), average 1995-2010

d

Source: World Development Indicators, accessed in June 2012 at http://data.worldbank.org/data-catalog/worlddevelopment-indicators.

an

Notes: 1) Exports of goods and services (% GDP)

rs

2) Data refers to simple average across the EU23 countries.

pe

As for the product concentration of exports (Figure 2), the evolution of the HHI for product type in Europe has been irregular and reached a maximum value of 0.13 in 2000, being

Pa

currently around 0.11. Relatively to 1995, there is now a higher value for this index, representing a stronger concentration of exports in certain sectors/products. According to the

ic

results of our model, the trend towards further product concentration, reinforcing comparative

on om

advantages and economies of scale, has contributed positively to European growth.

Ec

Figure 2: EU23 - Product concentration of exports, 1995-2010

Source: UNCTADstat, accessed in June 2012 at http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx. Notes: 1) HII - Product as measured by the Herfindahl-Hirschman index in UNCTADstat. 2) Data refers to simple average across the EU23 countries.

Economic Papers and Notes, vol.14, no.1

46


Disaggregating EU23 exports into Manufactures, Food and agriculture and Fuel, ores and metals (Figure 3), we can see a clear difference between the first sector and the other two. Manufactures exports represent about 80% of the merchandise exports, but have been exhibiting a declining trend in recent years due to the increase in the exports of Fuel, ores and metals.

an

d

N

ot

es

Figure 3: EU23 - Exports structure by sector, 1995-2010

Source: Own calculations based on data from World Development Indicators, accessed in June 2012 at http://data.worldbank.org/data-catalog/world-development-indicators.

pe

rs

Note: Data refers to simple average across the EU23 countries.

In particular, within exports of Manufactures, the high-technology exports have reached the

Pa

highest shares in 2000 and 2006 accounting for about 14% of Manufactures exports in 2010 – see Figure 4. According to the results of our model, the larger the growth/weight of

ic

manufactured exports and the larger the high-technology component of manufactured exports,

on om

the larger the economic growth. Thus, the EU23 countries should consider increasing exports in these sectors by focusing on products for which they have increased competence.

Ec

Disaggregating the European exports by product (Figure 5), we can confirm that, on average, Machinery and transport equipment has the greatest weight (representing around 45% of Manufactures exports), followed by Medicinal and pharmaceutical products (8.50%). Given stronger comparative advantages in Machinery and transport equipment, products within these groups should be the main engine for exports growth, namely through technological reinforcement (recall the “product proximity” idea in Bonaglia and Fukasaku, 2003).

47

North American Academic Journals, 2014


es

Figure 4: EU23 - High-technology exports (% Manufactures exports), 1995-2010

ot

Source: World Development Indicators, accessed in June 2012 at http://data.worldbank.org/data-catalog/worlddevelopment-indicators.

N

Note: Data refers to simple average across the EU23 countries.

on om

ic

Pa

pe

rs

an

d

Figure 5: EU23 - Structure of manufactured exports, 2010

Ec

Source: UNCTADstat, accessed in November 2012 at http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx. Note: Product classification based on SITC, Rev.3.

In particular, Machinery and transport equipment includes Road vehicles (27.41%), Electrical machinery, apparatus and appliances (26.78%), Other industrial machinery and parts (14.29%) and Power generating machinery and equipment (9.04 %) â&#x20AC;&#x201C; see Figure 6.

Economic Papers and Notes, vol.14, no.1

48


d

Source: UNCTADstat, accessed in November 2012 at http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx.

N

ot

es

Figure 6: EU23 - Structure of machinery and transport equipment, 2010

an

Note: Product classification based on SITC, Rev.3.

rs

Besides the diversification of destinations (on average, countries have more 20 trading

pe

partners in 2010 relative to 1995) 5, Europe has also diversified the volume of exports within trading partner as Figure 7 shows a decreasing path of the average HHI-destination. Thus

Pa

Europe has been heading towards a larger diversification of export destination â&#x20AC;&#x201C; either

ic

through new exporting markets or through reorganizing their export volumes across existing

on om

partners. In particular, the latter trend is consistent with greater economic growth, reducing the dependency relative to idiosyncratic shocks. Figure 8 confirms the recent tendency to diversification within partners, by reducing the

Ec

weight of some of the major markets such as Germany, France, United States and the United Kingdom, while increasing the exports to countries like China, Poland, Russian Federation and Czech Republic. Moreover, positive contribution to growth was reinforced as this shift was towards markets with greater potential demand growth (see Figure 9).

5

Source: http://data.worldbank.org/data-catalog/world-development-indicators.

49

North American Academic Journals, 2014


es

Figure 7: EU23 - Destination concentration of exports, 1995-2010

N

ot

Source: Own calculations based on data from UNCTADstat, accessed in June 2012 at http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx. Notes: 1) HII - Destination as measured by the Herfindahl-Hirschman index.

an

d

2) Data refers to simple average across the EU23 countries.

Ec

on om

ic

Pa

pe

rs

Figure 8: EU23 - Relative importance of main trading partners, 1995-2010

Source: UNCTADstat, accessed in November 2012 at http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx.

Economic Papers and Notes, vol.14, no.1

50


an

d

N

ot

es

Figure 9: EU23 - GDP growth forecasts for main trading partners, 2011-2017

CONCLUSIONS

pe

4.

rs

Source: World Economic Outlook, WEO, accessed in November 2012 at http://www.econstats.com/weo/V002.htm

Pa

After reviewing, both empirically and theoretically, the channels through which exports affect economic growth, especially through product structure and destination, we have assessed how

ic

these dimensions impinge on the economic growth of the EU. We have estimated a Solow-

on om

decomposition growth model augmented with several dimensions capturing the literatureenlightened aspects of product structure and destination of exports. The model is estimated using annual data for a panel of 23 EU countries across 1995 to 2010. Relative to existing

Ec

literature, our model improves on including, simultaneously, several dimensions of both product and destination structure of exports and also in focusing in the EU set of developed countries. Our results report a rather well-specified and robust model which delivers a strong relationship between real exportsâ&#x20AC;&#x2122; growth and real output growth. The results suggest that where to and what to export do matter for the EU growth dynamics. In particular, our results lend support to that rich/developed countries should export more value-added products, with special focus on high technology exports. Better economic growth performance is also enhanced if countries specialize rather than export a large set of products, a result in line with

51

North American Academic Journals, 2014


the comparative advantage hypothesis. Moreover, we find evidence that higher growth is fostered by export diversification across partners while enlarging the portfolio of partners, mainly to less developed and more distant countries, has negative impacts on European growth. Unambiguously, and as expected, relative concentration of exports should be directed towards the trade partners that exhibit higher potential growth rates: the larger the weighted average growth rate of trading partners, the stronger the leverage effects to economic growth. Given these conclusions, the European countries should support high technology exports and should reinforce the exports of Machinery and Transport Equipment, Medicinal and pharmaceutical products and Iron and Steel. Moreover, a move towards more diversification

es

among trade partners is desirable, namely from the most representative in the export portfolio

d

REFERENCES

N

potential such as China, Poland and Russian Federation.

ot

– Germany, France, the UK and the US - to the less representative and with higher growth

an

[1] Acemoglu, D. and Zilibotti, F. (1997), “Was Prometheus Unbound by Chance? Risk, Diversification and Growth”, Journal of Political Economy 105 (4): 709-751.

pe

rs

[2] Alexander, C. and Warwick, K. (2007), “Governments, Exports and Growth: Responding to the Challenges and Opportunities of Globalization”, The World Economy, 30: 177-194.

Pa

[3] Al-Marhubi, F. (2000), “Export Diversification and Growth: An Empirical Investigation”, Applied Economics Letters, 7: 559-562.

ic

[4] Amable, B. (2000), “International Specialisation and Growth”, Structural Change and Economic Dynamics, 11 (4): 413-432.

on om

[5] Amador, J. and Opromolla, L. D. (2008), "Product and Destination Mix in Export Markets", Working Paper nº17, Banco de Portugal. [6] Andraz, J. and Rodrigues P. (2010), “What Causes Economic Growth in Portugal: Exports or Inward FDI?”, Journal of Economic Studies, 37 (3-4): 267-87.

Ec

[7] Arora, V. and Vamvakidis, A. (2004), “The Impact of U.S. Economic Growth on Other Countries: How Much Does It Matter?", Journal of Economic Integration, 19 (1) [8] Arora, V. and Vamvakidis, A. (2005), "How Much Do Trading Partners Matter for Economic Growth?", IMF Staff Papers, 52 (1): 24-40. [9] Awokuse, T. (2008), “Trade Openness and Economic Growth: Is Growth Export-Led or Import-Led?”, Applied Economics, 40: 161-163. [10] Balassa, B. (1978), “Exports and Economic Growth: Further Evidence” Journal of Development Economics, 5: 181–189. [11] Baliamoune-Lutz, M. (2011), “Growth by Destination (Where You Export Matters): Trade with China and Growth in African Countries”, African Development Review/Revue Africaine de Developpement, 23 (2): 202-218.

Economic Papers and Notes, vol.14, no.1

52


[12] Bebczuk, R.N. and N.D. Berrettoni (2006), “Explaining Export Diversification: An Empirical Analysis”, CAF Research Program on Development Issues. [13] Bonaglia F. and Fukasaku, K. (2003), “Export Diversification in Low-Income Countries: An International Challenge after Doha”, OECD Development Centre Working Papers No 209, OECD. [14] Clerides. S. K., Lach, S. and Tybout, J. R. (1998), “Is Learning by Exporting Important? Micro-Dynamic Evidence from Colombia, Mexico, and Marocco”, The Quarterly Journal of Economics, 113 (3): 903-947. [15] Coe, D. and Helpman, H. (1995), "International R&D Spillovers," European Economic Review, 39: 859–887.

es

[16] Damijan, J., Polanec, S. and Prasnikar, J. (2004), “Self-Selection, Export Market Heterogeneity and Productivity Improvements: Firm-Level Evidence from Slovenia”, LICOS Discussion Paper 148

N

ot

[17] De Loecker, J. (2007), “Do Exports Generate Higher Productivity? Evidence from Slovenia”, Journal of International Economics, 73: 69-98.

an

d

[18] Dornbusch, R., Fischer, S. and Samuelson, P. (1977), “Comparative Advantage, Trade and Payments in a Ricardian Model with a Continuum of Goods”, American Economic Review 67: 823-839.

rs

[19] Dreger, C. and D. Herzer (2012), “A further examination of the export-led growth hypothesis”, FIW Working Paper series, No. 84, FIW.

pe

[20] Feder G. (1983), “On Exports and Economic Growth”, Journal of Development Economics, 12: 59-73.

ic

Pa

[21] Frankel, J., Stein E. and Wei S. (1995), "Trading Blocs And The Americas: The Natural, The Unnatural, And The Super-Natural," Journal of Development Economics, 47 (1): 6195.

on om

[22] Ghosh, A. and Ostry, J. (1994), “Export Instability and the External Balance in Developing Countries”, IMF Staff Papers 41: 214-235. [23] Giles, J. and Williams, C. (2001a), “Export-Led Growth: A Survey of the Empirical Literature and Some Non-causality Results. Part I”, Journal of International Trade & Economic Development, Taylor and Francis Journals, 9(3): 261-337.

Ec

[24] Giles, J. and Williams, C. (2001b), “Export-Led Growth: A Survey of the Empirical Literature and Some Non-causality Results. Part I”, Journal of International Trade & Economic Development, Taylor and Francis Journals, 9(4): 445-470. [25] Greenaway, D., Morgan, W. and Wright, P. (1999), “Exports, export composition and growth”, Journal of International Trade and Economic Development, 8 (1): 41-51. [26] Grossman, G.M. and Helpman, E. (1991), “Trade, Knowledge Spillovers, and Growth”, European Economic Review, 35 (3): 517-526. [27] Guaresma, C. J. and Wörz, J. (2005), “On Export Composition and Growth”, Review of World Economics/Weltwirtschaftliches Archiv, 141 (1): 33-49 [28] Gujarati, D. (1995), Basic Econometrics, 3rd Edition, New York: McGraw-Hill

53

North American Academic Journals, 2014


[29] Hausmann, R. and Rodrik, D. (2003), “Economic Development as Self-Discovery”, Journal of Development Economics, 72 (2): 603–633. [30] Hausmann, R., Hwang, J. and Rodrik, D. (2007), “What You Export Matters”, Journal of Economic Growth, 12: 1-25 [31] Herzer, D. and Nowak-Lehmann D., (2006), “What Does Export Diversification Do for Growth? An Econometric Analysis”, Applied Economics 38: 1825-1838. [32] Herzer, D. and Nowak-Lehmann, F.; Siliverstovs, B. (2004): “Export-Led Growth in Chile: Assessing the Role of Export Composition in Productivity Growth”, Ibero America Institute for Economic Research (IAI) Discussion Papers No 103 [33] Hesse, H. (2008), “Export Diversification and Economic Growth”, Working Paper Nº 21, Commission on Growth and Development.

es

[34] Hummels, D. and Klenow, P. (2005). “The Variety and Quality of a Nation’s Exports”, American Economic Review, 95(3): 704–723.

N

ot

[35] Imbs, J. and Wacziarg, R. (2000), “Stages of Diversification”, Research Papers Series No 1653, Graduate School of Business, Stanford University.

d

[36] Kali, R., Mendez, F. and Reyes, J. (2007); “Trade Structure and Economic Growth”, Journal of International Trade and Economic Development, 16 (2): 245-269.

an

[37] Keller, W. (2004), “International Technology Diffusion”, Journal of Economic Literature, 42 (3): 752-782.

pe

rs

[38] Kónya, L. (2006), “Exports and growth: Granger causality analysis on OECD countries with a panel data approach”, Economic Modelling, 23 (6): 978-992.

Pa

[39] Laursen, K. (2000), “Do Export and Technological Specialisation Patterns Co-evolve in Terms of Convergence or Divergence?: Evidence from 19 OECD Countries, 1971-1991”, Journal of Evolutionary Economics, 10(4): 415-436.

on om

ic

[40] Lederman, D. and Maloney, W. F. (2003), “Trade Structure and Growth”, Working Paper Nº 3025, World Bank Policy Research [41] Lorde, T. (2011), “Export-led Growth: A Case Study of Mexico”, International Journal of Business, Humanities and Technology, 1 (1)

Ec

[42] McCann, F. (2007) “Export Composition and Growth”, School of Economics University College Dublin [43] Michaely, M. (1977), “Exports and economic growth: an empirical investigation”, Journal of Development Economics, 4, 49-53. [44] Peneder, M. (2002), “Industrial Structure and Aggregate Growth”, WIFO Working Paper 182. [45] Ramos, F. F. R. (2001), “Exports, imports and economic growth in Portugal: evidence from causality and cointegration analysis”, Economic Modelling, 18 (4): 613-623. [46] Soukiasis, E. and Antunes, M. (2011), “Is foreign trade important for regional growth? Empirical evidence from Portugal”, Economic Modelling, 28: 1363-1373 [47] Spilimbergo, A. (2000), “Growth and Trade: The North Can Lose,” Journal of Economic Growth, 5: 131-146.

Economic Papers and Notes, vol.14, no.1

54


[48] Stokey, N. (1991), “The Volume and Composition of Trade Between Rich and Poor Countries,” Review of Economic Studies, 58: 63-84. [49] Vacek, P. (2010), “Productivity Gains From Exporting: Do Export Destinations Matter?”, IES Working Paper 18, IES FSV, Charles University. [50] Vernon, R. (1966), “International investment and international trade in the product cycle”, Quarterly Journal of Economics, 80 (2): 190-207 [51] Ziramba, E. (2011), “Export-Led Growth in South Africa: Evidence from the Components of Exports”, Journal for Studies in Economics and Econometrics, 35 (1): 113 WEBSITES:

es

http://data.worldbank.org/data-catalog/world-development-indicators

Ec

on om

ic

Pa

pe

rs

an

d

N

http://unctadstat.unctad.org/ReportFolders/reportFolders.aspx

ot

http://www.econstats.com/weo/V002.htm

55

North American Academic Journals, 2014


Annex A - Partnerâ&#x20AC;&#x2122;s growth index (%)

Economic Papers and Notes, vol.14, no.1

2005 1.95 1.64 1.84 1.95 3.65 2.76 1.71 2.37 2.48 1.91 1.48 2.13 4.22 4.41 1.21 2.11 1.95 2.10 2.47 2.10 1.19 1.96 1.81

2006 3.68 2.90 3.71 3.18 4.96 4.09 2.94 3.27 3.67 4.00 2.20 3.38 5.83 6.06 2.68 3.84 3.00 3.33 4.06 3.65 2.40 2.87 2.67

ot es

2004 2.38 3.12 2.15 2.63 4.29 3.46 2.08 2.83 3.22 2.43 2.19 2.70 4.25 4.64 1.92 2.74 1.92 2.60 2.67 2.58 1.93 2.57 2.35

N

2003 1.03 2.40 0.83 1.20 3.01 2.23 1.03 1.53 1.80 0.95 1.09 1.37 3.30 3.15 0.55 1.43 0.75 0.99 1.19 1.43 0.47 1.21 1.03

d

2002 1.07 1.88 0.85 1.10 2.44 1.64 0.89 1.24 1.65 0.88 1.00 1.13 2.48 2.92 0.69 1.20 0.94 1.02 1.15 1.39 0.57 1.00 0.91

an

2001 1.82 2.10 1.57 1.52 2.69 1.92 1.49 1.41 1.70 1.58 1.39 1.40 2.93 3.45 1.32 2.12 1.51 1.41 2.01 1.72 1.38 1.45 1.16

rs

2000 3.49 3.99 3.18 3.79 4.61 4.36 3.55 3.73 4.16 3.44 3.43 3.61 4.55 5.08 3.40 3.66 3.45 3.80 3.72 3.60 3.38 3.57 3.75

pe

1999 2.37 2.38 2.33 3.00 3.55 3.19 2.73 3.03 2.73 2.57 2.95 2.79 2.89 3.14 2.73 2.44 2.97 1.94 2.43 2.35 2.61 2.87 3.35

Pa

1998 2.28 1.72 2.49 2.54 2.99 2.55 2.79 2.75 2.43 2.16 2.70 2.46 2.59 2.03 2.49 1.75 2.76 2.06 1.77 2.63 2.87 2.69 2.52

ic

1997 2.31 1.35 2.81 2.87 3.80 3.26 2.73 2.89 1.57 2.18 2.71 2.63 3.54 3.93 2.49 1.93 2.57 2.43 1.43 3.88 2.54 3.26 3.22

om

1996 1.31 0.19 2.08 1.81 1.52 1.78 1.79 2.01 1.25 1.25 1.82 1.63 0.33 0.12 1.25 0.57 1.76 1.12 2.09 3.52 1.63 2.31 2.11

on

1995 1.94 1.83 2.32 2.21 1.18 2.22 2.12 2.34 2.73 1.59 2.05 1.87 -0.35 -1.60 1.90 1.13 2.11 1.75 3.09 2.59 1.92 2.35 2.43

Ec

Country / Year Austria Cyprus Czech Republic Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania The Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden United Kingdom

2007 3.43 3.26 3.82 3.11 5.41 3.78 2.86 3.19 3.49 4.03 2.21 3.23 5.96 5.80 2.54 3.63 3.24 2.76 3.45 3.41 2.22 2.95 2.44

2008 1.22 0.29 1.38 0.03 0.48 0.77 0.18 0.65 1.60 1.71 -0.60 0.75 1.06 1.15 0.05 0.90 0.57 0.70 1.42 1.24 -0.19 0.23 -0.43

2009 -4.51 -3.78 -4.25 -4.40 -8.02 -4.41 -4.04 -3.50 -4.24 -4.90 -4.32 -3.74 -8.44 -6.81 -4.35 -5.40 -4.14 -4.93 -4.60 -4.66 -3.84 -3.98 -4.14

2010 2.75 0.48 2.78 2.79 2.83 3.10 2.46 2.48 1.93 2.31 1.81 2.35 2.82 2.95 2.26 2.50 0.87 2.46 2.31 1.96 1.81 2.05 2.09

56


Average across countries

Std. Dev. across countries

Max

Min

Obs.

2.69

3.72

13.06

-17.37

368

22.74

4.63

40.39

10.61

368

Population growth

0.28

0.62

2.18

-1.79

368

Inflation

5.12

10.5

154.76

-4.48

368

190.12

24.42

218

91

368

Partnersâ&#x20AC;&#x2122; growth

1.87

2.05

es

Annex B - General descriptive statistics

6.06

368

HHI - product

0.12

0.05

High technology exports

10.07

7.7

Manufactures exports

75.24

11.91

ot 0.29

0.04

368

41.84

0.36

368

90.27

44.56

368

0.95

10.75

-2.43

368

0.31

Fuel, ores and metals exports

0.31

1.04

7.67

-7.01

368

0.09

0.03

0.2

0.04

368

Ec

on om

ic

Pa

pe

rs

Food and agricultural exports

HHI - destination

57

-8.44

N

Number of partners

d

Gross capital formation

an

GDP growth

North American Academic Journals, 2014


The Observational Equivalence of Commons and Anti-Commons *

**

N

ot

es

Bingyuan Hsiung

an

d

Abstract

The idea of anti-commons (re)-introduced by Heller (1998) has initiated numerous subsequent

rs

studies and the concept has been used to cover scenarios of a different nature, including those

pe

of commons. Specifically, the empty stores in Moscow are better interpreted as the result of rent seeking, with each and every approval-granting bureaucrat seeking favors and thus

Pa

depleting the pool of potential resources (potential bribes). This is actually a case of commons, as opposed to the case of Indian reservations where a large number of owners makes it

ic

difficult to utilize the resources collectively and efficiently. Implications of this arguably more

Ec

on om

realistic interpretation are discussed.

Keywords: commons, anti-commons, rent seeking, property rights, observational equivalence. JEL classification: K10, K11, K19, D62, D70.

*Support by a research grant from the College of Social Sciences, National Taiwan University is gratefully acknowledged. Kao Tze Han provided valuable research assistance. ** Chair Professor and Director of Law and Economics Research Center, Zhejiang University, Hangzhou, People’s Republic China. Email: hsiung@zju.edu.cn

Economic Papers and Notes, vol.14, no.1

58


1. Introduction The tragedy of the commons is a well known phenomenon in social sciences in general and in economics in particular (Hardin, 1968). The idea of anti-commons (re-)introduced by Heller (1998) has generated numerous subsequent studies.

1

While discussions on

anti-commons have been numerous, it will be argued below that the concept remains a thorny one, as it seems to cover a wide range of phenomena with different natures. 2 One important reason is that in the seminal paper Heller himself applied the concept in such a manner. Specifically, the case that many street front stores remain empty in Moscow and the case that lands with numerous owners in many Indian reservations are managed inefficiently both

es

imply resources being underutilized. However, the main reason that the stores are closed is

ot

due to the bureaucratic system, as the right to run a store has not been granted, i.e., the right

N

has not been recognized by the legal system. Alternatively, the major reason for lands being mismanaged in some Indian reservations is that a property has too many owners and it is

d

difficult to reach a collective decision concerning its usage. The property right itself is clearly

an

and specifically defined. While on the surface the two scenarios may be similar

rs

observationally, the underlying problems are obviously different. Moreover, it will be suggested below that the closed stores in Moscow, the motivation behind Heller’s important

pe

paper, is better interpreted as a case of not anti-commons but commons.

ic

Pa

2. Commons and Anti-commons

on om

The concepts of commons and anti-commons both come from observations about the real world. Heller mentioned in his article first the closed street front stores, and then suggested that certain Indian reservations have lands owned by dozens or even hundreds of people, after generations of inheritance; he also discussed the case that in Moscow it is often

Ec

difficult to tear down old apartment buildings occupied by aging occupants who are unwilling to evacuate. Heller and Eisenberg (1998) argue that bio-chemistry patents for inter-related inventions have often been granted to different individuals and the patent holders may face great difficulties in combining the patents so as to produce commercially viable products. In these scenarios the allocation of resources has to be approved of by many parties concurrently and the disapproval by a single party will bring a halt to the plan. As such, the common thread 1

Heller indicated that the term anticommons was first introduced by Michelman (1982); it is generally agreed that Heller’s paper touched off the subsequent studies. In addition to commons and anti-commons, the case of semi-commons has been discussed in the literature (Smith, 2000; Fennell, 2011). It refers to the situation between private property and commons in which there is some range of interacting uses. This is an interesting topic but is not directly related to the present inquiry. 2 For instance, Schulz et al. (2002) suggest that “Despite the growing significance of the concept and the issue of anticommons in both economic theory and law and economic shcolarship, the notion still lacks a generalized formation in the literature.”

59

North American Academic Journals, 2014


of all these cases seems to be that each and every approval granting unit exerts veto power and enjoys exclusive rights with respect to the use of valuable resources.

2.1 Anti-commons and Rent Seeking Concerning the closed street front stores, a deeper and arguably more convincing explanation can be obtained by analyzing the scenario from the rent seeking perspective. In particular, to get all the approvals needed to open a store in Moscow, one is subject to the under-the-table demands of the bureaucrats who will take the opportunity to seek personal favors. Therefore, the true reason, or at least a major reason, that there are street front stores remaining closed is that there may be too many agencies and too many bureaucrats seeking

es

favors. The “tolls” needed may be such that the total costs in terms of time and money are too

ot

much to bear. An alternative route is available, however; Heller mentioned that in nearby vacant lots and with make-shift tents and other temporary structures, business is thriving!

N

That is, as each and every bureaucrat along the normal application process is likely to ask for

d

bribes or other benefits, the total burden may become too great, to the point that the potential

an

applicant’s willingness to provide favors has been depleted. That is, the source of bribes has dried up and no bribes are forthcoming. In an abstract but real sense, this is exactly the

rs

tragedy of the commons—each and every bureaucrat in the process cares only for him/herself

pe

and takes his/her cut, but as a whole the resources are eventually depleted. It is essentially a direct application of the rent seeking insight, 3 and in this interpretation the term

Pa

anti-commons Heller (1998) used is arguably a misnomer.

ic

Moreover, the inappropriate use of terminology is evident from a different perspective.

on om

Specifically, the closed street front stores—the stores and/or the businesses---are the property of the owner or the tenant of the store and not the bureaucrats or the regulatory agencies. The bureaucrats have the powers to grant approvals to the applicant only but do not have powers to open or run the business themselves, barring the more complicated case in which they ask

Ec

to share the revenues with the business owners. Consequently, they are not owners of the property in any sense of the word. As such, Heller’s use of the term “anti-commons,” indirectly referring to the case of commons where sheep owners collectively own and use the commons, is not consistent with a normal interpretation of the idea of property. By contrast, if the pool of potential bribes are seen as a commons and the bureaucrats’ taking their cut individually finally depletes the pool, the use of terms would be more consistent with ordinary usage, as the bribes are indeed the property of the bureaucrats. Consequently, it is clear that

3

With each and every bureaucrat seeking favors the stock of favors, or bribes, are finally used up. This is exactly the case of the tragedy of the commons and not anti-commons per se. However, the present study follows the literature but modifies the term to “rent seeking type anti-commons.” Alternatively, Fennell (2011) states that “The anti-commons is an assembly problem, nothing more and nothing less.” This remark might be appropriate for CPR type anti-commons but it is evidently problematic concerning RS type anti-commons.

Economic Papers and Notes, vol.14, no.1

60


Heller used the term anti-commons to cover both commons (the closed stores) and anti-commons (some Indian reservations), and the case of commons is actually rent seeking. 4

2.2 The Observational Equivalence It is interesting to note that the Indian reservations commons and the closed street front stores in Moscow share observational equivalence, for they have been studied under the same rubric of anti-commons and resources are underutilized in both cases. More precisely, the equivalence is on two different levels. First, consider the case of the closed stores in Moscow. It is evident that there are two possible explanations for why the street front stores remain closed--one is that the administrative agencies are rent seeking so the potential entrepreneurs

es

look for alternatives, and the other is that each and every agency exerts veto power and it is

ot

difficult to get approved by all the agencies, thus driving away potential entrepreneurs. This implies that different causes would result in observationally equivalent phenomenon. Which

N

of these two explanations is more persuasive seems to be an empirical issue.

d

Secondly, for the Indian reservations the key of the problem is that a certain property is

an

simultaneously owned by numerous persons, a case termed fragmentation of property by Schulz et al. (2002). As a result, even though there are potential gains to be captured, e.g.,

rs

selling or developing the lands, the potential gains are not realized because the transaction

pe

costs involved (mainly the negotiation and enforcement costs) are often prohibitive. The

Pa

nature of the problem is the logic of collective action, and it is common when dealing with public goods in general (Demsetz, 2011). By contrast, for the closed street front stores in

ic

Moscow the problem is that the right to open the store has not been granted by all of the

on om

relevant administrative agencies, and resources stay in an unutilized and inefficient state. The nature of the problem is that the administrative system is not functioning satisfactorily and the legal system is not well developed. As such, while both cases suggest underutilized resources, the underlying reasons are very different. But they are observationally equivalent

Ec

nevertheless.

3. Discussion and Conclusion By comparing the two opposite cases of the closed street front stores in Moscow and the Indian reservations it is easy to see that as economic development progresses, administrative efficiency will improve and the legal system will similarly assume a normal trajectory. Therefore, the scenario as represented by the closed street front stores will gradually decrease in importance and may even disappear eventually. This means that these phenomena would 4

The earliest literature on rent seeking includes Krueger (1974), Tullock (1989, 1993) and Bhagwati (1982). All these studies were related to observations in developing societies in which regulations and bureaucrats are omnipresent. This implies that the bureaucrats are in a good position to seek personal favors by way of carrying out his/her official duties.

61

North American Academic Journals, 2014


probably exist only in a certain stage of the economic development process, and that the phenomena will rarely be found in a highly developed modern society. By contrast, the case of the Indian reservations has its problem rooted in the nature of public goods and the logic of collective action. The degree of economic development will not change the nature of the problem and the phenomenon would exist in even the most advanced society. It may assume different configurations, e.g., developing the Indian reservations and renovating apartment buildings, but the essence of the problem remains the same, regardless of the passage of time or of economic development. Alternatively, the rent seeking explanation of the closed stores in Moscow suggested above is evidently different from the current explanation in the anti-commons literature. The

es

potential theoretical conflict is best resolved by empirical evidence. One possible way to

ot

proceed is to find direct or indirect evidence to illustrate the degree of rent seeking involved.

N

Moreover, concerning the decision rules used, the commons correspond to the any-person rule as anyone can decide to make use of the resources, and the anti-commons correspond to the

d

unanimity rule where any negative vote would veto the use of resources. Examining the

an

commons and anti-commons from the perspective of voting rules is somewhat different from the existing theoretical models, where a Nash-Cournot cooperative game has often been

pe

rs

employed. 5 A comparative study along this line seems to promise fruitful results. To conclude, it was suggested above that the closed stores in Moscow are better

Pa

interpreted as the result of rent seeking, with each and every approval granting bureaucrat seeking favors and thus depleting the pool of potential resources (potential bribes). This is

ic

actually a case of commons, as opposed to the case of Indian reservations where a large

on om

number of owners makes it difficult to utilize the resources collectively and efficiently. As such, anti-commons and commons could be observationally equivalent, even though the

Ec

underlying causes are drastically different.

5

See for instance, the first formal model introduced by Buchanan and Yoon (2000) as well as subsequent expansions and elaborations by Schulz et al. (2002) and Parisi et al. (2004, 2005).

Economic Papers and Notes, vol.14, no.1

62


References [1] Bhagwati, Jagdish N., “Directly Unproductive, Profit-Seeking (DUP) Activities,” Journal of Political Economy, 90(5): 988-1002, 1982 [2] _________., Brecher, Richard. A. and Srinivasan, T.N. “DUP Activities and Economic Theory,” in D.C. Colander, ed., Neoclassical Political Economy, pp. 17-32, Cambridge: Ballinger, 1984. [3] Buchanan, James M., and Yoon, Yong J., “ Symmetric Tragedies: Commons and

es

Anticommons, ” Journal of Law and Economics, 43(1): 1-13, 2000. [4] Demsetz, Harold, “The Problem of Social Cost: What Problem? A Critique of the

ot

Reasoning of A.C. Pigou and R.H. Coase,” Review of Law & Economics, 7(1): 1-13,

N

2011.

[5] Hardin, Garrett, “The Tragedy of the Commons,” Science, 162 (3859): 1243–1248, 1968.

an

d

[6] Heller, Michael A., “The Tragedy of the Anticommons: Property in the Transition from Marx to Markets,” Harvard Law Review, 111(3): 621-688, 1998.

rs

[7] Heller, Michael A., and Eisenberg, Rebecca S., “Can Patents Deter Innovation? The

pe

Anticommons in Biomedical Research,” Science, 280 (5364): 698-701, 1998. [8] Krueger, Anne O., “The Political Economy of the Rent-Seeking Society,” American

Pa

Economic Review, 64: 291-303, 1974.

[9] Michelman, Frank I., “Ethics, Economics and the Law of Property,” in J. Roland Pennock

on om

NYU Press, 1982.

ic

and John W. Chapman, eds., Nomos XXIV: Ethics, Economics and the Law, New York: [10] Parisi, Francesco, Schulz, Norbert, and Depoorter, Ben, “Simultaneous and Sequential Anticommons,” European Journal of Law and Economics, 17(2): 175-190, 2004. [11] __________, “Duality in Property: Commons and Anticommons,” International Review

Ec

of Law and Economics, 25(4): 578-61, 2005.

[12] Schulz, Norbert, Parisi, Francesco, and Depoorter, Ben, “Fragmentation in Property: Towards a General Model,” Journal of Institutional and Theoretical Economics, 158(4): 594-613, 2002. [13] Tullock, Gordon, The Economics of Special Privilege and Rent Seeking, Boston, MA: Kluwer Academic Publishers, 1989. [14] __________, Rent Seeking, Brookfield, VT: Edward Elgar, 1993.

63

North American Academic Journals, 2014


Call for Papers North American Academic Journals The Editorial Office at North American Academic Journals is accepting manuscripts for review and consideration for publication. North American Academic Journals publishes the North American Review of Finance and Economic Papers and Notes. Both journals offer publishing opportunities to academic researchers from around the world in an open access format and are indexed in the American Economic Associationâ&#x20AC;&#x2122;s electronic bibliography (EconLit). The North American Review of Finance is a peer-reviewed journal that publishes original research articles, as well as review articles in all areas of financial economics. The Review publishes concise, high-quality papers, in both established and newly developing fields including: Asset Pricing, Corporate Finance, Derivatives, Financial Management, Risk Management, Investments, Alternative Investments, Banking, Behavioral Finance, Case Studies, and Finance Education. Economic Papers and Notes is a peer-reviewed journal that publishes short accounts of new original research and encourages discussion of previously published papers in other journals. Papers submitted to the Economic Papers and Notes are generally applied in nature, but may also include discussions of theory, principles, education, econometrics, and or other economic practices. North American Academic Journals maintains an open call for papers policy and publishes on an accelerated timescale which offers rapid publication opportunities for authors. Paper submissions should not have been previously published or be currently under consideration for publication with any other publishers. For full consideration, all papers must be submitted electronically, are subject to a peer review process, and must be written in English. PAPER SUBMISSION PROCEDURE: Articles for consideration should be submitted via email to the Editor at the following address: editor@najournals.com. There is no submission fee. There is a publication fee for accepted manuscripts. If you have any questions regarding the submission process please send an email to the Editor at editor@najournals.com. North American Academic Journals P.O. Box 891 Urbana, IL 61803 http://www.najournals.com


EPN 2014 1