Bangladesh Reality Check 2008: listening to poor people's realities about primary healthcare and primary education
Health researchers often wish to study the impact of wealth on health outcomes. To do this they must collect data on social and economic factors. However, the collection of detailed data on income and expenditure is rare in health studies in developing countries. Instead, researchers generally adopt more rapid procedures based on survey methodology. Increasingly this has included the use of Principal Components Analysis (PCA) to generate a number of separate indicators of welfare. An alternative approach is to utilise participatory wealth ranking to generate a measure of household wealth. The aim of this paper was to compare the results of Participatory Wealth Ranking (PWR) and an indicator-based survey methodology conducted within a health research programme (the IMAGE study) in rural South Africa. The data point to widespread and severe poverty among the study population, including indications of high unemployment, reliance on wage remittance and state grants, poor access to clothing, and fragile food and educational security. Household wealth indices were created from both techniques, using PCA to combine the survey data. Data from both techniques was available from 1467 households. There was a high level of internal of consistency in the participatory wealth ranking data However there was only moderate agreement between the ranking of householdsÆ relative wealth from the two techniques, although both techniques identified similar factors as of importance in determining wealth. The reasons for the discrepancy are unclear, but are likely to include methodological and conceptual factors inherent in both techniques. PWR may be a useful tool for the assessment of relative wealth in health studies in developing countries.