Measuring Income and The Potential For Poverty Reduction In Rural Kenya

Abstract

Agricultural policy in Kenya often is made with little reference to statistics and hard data. Tegemeo Institute of Egerton University has been filling part of the information gap in the agricultural sector for the last decade, largely through sector specific pieces of work. Beginning in 1997 Tegemeo, in collaboration with a new partner, Michigan State University, began undertaking large household surveys. Michigan State has a long tradition of undertaking large household surveys to inform policy makers in different parts of Africa. Household survey work by Tegemeo in Kenya began with two surveys, in 1993 and 1995 focusing on urban household consumption. These surveys focused on measuring the impact of maize market liberalization on urban consumption patterns. These surveys were followed with a large rural household survey in 1997. The present paper presents results from a 2000 update of the same survey. The household surveys allow the calculation of household incomes in 9 representative zones of rural Kenya. The data also allow the decomposition of that household income into its key constituent parts such as income from crops, livestock, off-farm salary and informal business income. The spatial distribution of the study sites, and the fairly large sample within the zones also allow for the sampled households to be analyzed and differentiated based on zone and income, but also according to education and gender, for example, of the household head. It is a rich data set whose surface is only scratched in this single paper. Large household surveys are expensive to mount, and very involving. But it is possible to generate the majority of key indicators using a Proxy Methodology that is introduced in this paper. Using that method, good econometrically derived estimates of indicators of interest can be generated without the full cost and logistical problems of collecting full information from thousands of households. Proxy methods are particularly useful as they can be updated annually. This is the kind of information that Kenya will need to generate regularly in order to fully monitor the outcomes of its poverty reduction strategy. Monitoring poverty cannot only be done through household surveys. In this paper Tegemeo will also present in summary form, a methodology called the PAPPA - Policy Analysis for Participatory Poverty Alleviation - that combines the numerical and statistical techniques of the large or proxy survey, with participatory techniques. The combination of the two methods provides more insight into poverty, and how to deal with it, than either the either statistical or participatory approaches on their own. This paper begins by presenting the findings of the 2000 survey in terms of household income and income sources. The proxy methodology is then presented followed by an introduction to PAPPA. The paper closes by linking the different methods used in Tegemeo to the monitoring and evaluation work that government along with the private sector and civil society will be undertaking as part of the Poverty Reduction Strategy - PRSP- process.