Ghana's poverty monitoring system

Africa Region. Number 64. May 2001 Findings Infobriefs reports on Good Practice in ongoing operational, economic and sector work carried out by the World Bank and its member governments in the Africa Region. It is published monthly by the Knowledge and Learning Center on behalf of the Region. The views expressed in Findings are those of the author/s and should not be attributed to the World Bank Group. Ghana's Poverty Monitoring System 22105 Poverty monitoring in most parts of the developing world is substantially inadequate. The importance of building a sound and sustainable poverty and welfare monitoring system has acquired more significance in recent years in the context of the Poverty Reduction Strategy Paper (PRSP) and Heavily Indebted Poor Countries (HIPC) initiatives, amongst others. The problem is particularly serious in Sub-Saharan Africa where the systematic collection of data on socio-economic issues is constrained by the lack of capacity and resources. In such an environment, the case of Ghana where the first Living Standards Measurement Survey in 1987 is particularly noteworthy. The Government of Ghana, with the assistance of the IMF/World Bank, embarked on a Economic Recovery Program in 1983. To monitor progress on the welfare of households, the Ghana Statistical Service (GSS), in collaboration with the World Bank, implemented the first round of the Ghana Living Standards Survey (GLSS1) in 1987/88. Since then, the GSS has conducted four rounds of GLSS - GLSS1 in 1987-88, GLSS2 in 1988- 89, GLSS3 in 199 -92 and GLSS4 in 1998-99. According to the current plan, a GLSS5 will be carried out in 2003- 04. The GLSS series provides data on a wide range of areas including demographic characteristics of the population; education: health; employment and time use; migration; housing conditions; household agricul- ture; non-farm businesses; the state and use of facilities and infrastructure in rural communities; and the prices of selected basic commodities. The GSS has systematically analyzed these data and published reports on each round of the GLSS as well as two poverty reports covering the periods 1987-92 and 1992-98. Key changes in GLSS4 implementation Over the years, the collection and analysis of the GLSS data have substantially improved. The gestation period from data collection to final report publication was reduced from three years in 1992 to one year in 1998. Similarly, while the poverty report for 1992 was published in 1995, the poverty report for 1999 was published just six months after the survey was completed. Thus, the GLSS4 poverty report became the centerpiece of Consultative Group (CG) meetings in November 1999. The rapid completion of data collection and analysis is undoubtedly a "good practice" that needs to be documented and disseminated as other countries in SSA develop their own poverty monitoring systems. Comparing the GLSS3 experience with that of GLSS4, clearly GLSS4 was much more efficient than GLSS3 in all respects. Some of the improvements are described below. The "Good Pra,i,, Inthiicf t erie iis cJred b% P C. %t,ohan. Rm. i-Ii. Ki. .,iede a.-d LarningL Ceniler. World Bank. 18S18 H Street NW. Washington DC 20433. Tel. (202147--4114, e-mail: pm,oltan@u orldbank or Data entry: Data entry was mostly performed in the capital, Accra, for GLSS3, while for GLSS4 it was done entirely in the regional offices. This not only eliminated transport time but also reduced errors - any anomalies in the data were immediately resolved by enumerators. Data cleaning and editing: Data cleaning for GLSS3 was done at the end of the survey by a consultant based abroad. The cleaning thus took much longer as errors could not be corrected immediately. With GLSS4; not only was there an error correction system built into the data entry process, but the data was also cleaned first after six months of data collection and then again at the end of the survey. This reduced the time needed to produce the report and improved the quality of the data. Data analysis: During GLSS3, since only very few GSS staff were proficient in the Statistical Analysis System (SAS), data analysis was slowed down substantially and this contributed to the hiring of an external consultant to do most of the analysis. However, in the case of GLSS4, careful pre-planning led to over twenty people being trained by the World Bank to undertake this duty. As a result, they were able to work with the external consultants in the analysis. Management The involvement of GSS management in past GLSS surveys provided them with extensive experience. Consequently, coordinating people, skills, and equipment throughout the country over the twelve- month period was well managed. In addition, management ensured that the timing and inputs of external consultants did not hinder progress. Computers/vehicles: The fast computers that were put in each of the regional statistical offices expedited the data entry and data cleaning process. In addition, a vehicle was assigned to each of the GSS regional offices, helping ensure that work did not stop due to transportation problems. Data dissemination: The GLSS3 dissemination process was delayed by the slow data set production (three years) and the unavailability of complete documentation. These problems were addressed upstream in the GLSS4 by improving the data set construction process and producing final documentation as soon as the survey was over. The information was assembled within a year. The data are available to users on CD-ROM and through the Living Standards Measurement Survey (LSMS) site on the web. These two channels greatly reduced delays and data transfer problems that afflicted the GLSS3 survey. Lessons for the future Trends that facilitated the process: * upstream planning in data entry, data cleaning, and data analysis; * building capacity and training of individuals prior to survey launch; * procuring goods and services early to reduce delays during survey implementation; * holding regular management meetings to identify and resolve problems; and * all work being carried out in-country. Areas for further improvement: * Procurement/disbursement: The procurement of goods and services was hampered due to various institutional bottlenecks and delays in the release of counterpart funds. * Overcrowding of surveys: Census work immediately after the GLSS4 fieldwork meant that very few staff were available for data analysis and report writing. * Continuing lack of local capacity: While the situation has improved between GLSS3 and GLSS4, over-reliance on external consultants who are not always available at short notice, continues to be an issue. This Infobrief was written by Sudharshan Canagarajah, Task Team Leaderfor Poverty Monitoring work in Ghana. For more information, please e-mail:scanagarajah@worldbank.org; phone:202-473-3515; fax: 202-522-3252. Persons accessing the Bank's External Website can access information on Poverty by clicking on Topics and Sectors, then Poverty. World Bank staff can access the necessary information by clicking on Sectors and then on Poverty.