LAC_survey

Page 31

user and the poor. It is found that a “pro-poor public official” has no particular personal characteristics that distinguish him from other officials. Reports and contact information: For more information regarding WBI’s diagnostic work in Bolivia contact Maria González de Asís (mgonzalezasis@worldbank.org). Further information on the tool can also be found at http://www.worldbank.org/wbi/governance/bolivia/index.html.

3.2.2 PRMPS Public Officials Survey (1999) Type of tool: Survey of public officials Coverage: Bolivia Source: World Bank Funding: Bank-Netherlands Partnership Program Purpose: Between 1999 and 2001, public officials in 16 countries were surveyed with funds from the Bank Netherlands Partnership Program. These surveys were designed to better understand the ways in which performance of public officials depends on the incentives and constraints provided by their institutional environment. The basic assumption underlying the design of the surveys is that if resources are unpredictable, policies are likely to change, and rules are unlikely to be enforced. As a consequence, “results focus”, accountability and morale will suffer, with adverse effects on agency performance. Survey results help to map the strengths and weaknesses of the public sector and can identify potential pay-offs with reform interventions. Methodology and implementation: The Bolivia survey was carried out from May to August 1999 and administered to 738 public officials in 15 organizations. Of these, 53 were heads of organizations or departments, and 685 were general officials (middle to upper level), working in these organizations. The organizations to be included in the sample were selected by the World Bank manager and the Government of Bolivia counterparts to cover different categories of organizations. The survey firm, Encuestas y Estudios (E&E) stratified each agency by organizational hierarchies (5 levels) and drew a quota sample from each stratum to make the sample similar in appearance to the actual distribution in the population. While drawing the sample, fourteen out of the fifteen organizations identified for the survey did not give E&E a list of employees to construct the sampling frame. As a result, E&E used the 1992 Public Employees Census as an initial database for constructing the sampling frame. As they obtained more accurate (but still partial) lists of agency employees, they corrected the original sampling frame. For this reason, they were not able to draw a probability sample, and therefore resorted to quota sampling (sampling based on certain pre-determined characteristics of the respondents such as gender, ethnicity, etc.). The survey was pre-tested in two rounds, on a total of 31 people. The first round tested the survey on 15 officials. After corrections were made to the questionnaire, a second round tested it with 12 officials. The pre-test was also administered to four CEOs. The pre-test subjects were selected arbitrarily, rather than by random sampling. However, E&E made sure that none were from the agencies to be covered in the main round of the survey, to avoid contamination of the universe, for example by filtration of the questions. The following indicators were used for the survey: • Mission accomplished: Degrees to which agency heads perceive their agency’s mission to have been accomplished. • Vertical Solitude: Degrees to which there is a discrepancy in perceptions about a set of issues between senior managers and middle to lower-ranking staff of the same agencies. This is a proxy

31


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.