Water, walls, and bicycles: Wealth index composition using census microdata

Abstract

In this study, we produce a valid and consistent variable for socioeconomic status (SES) at the household level with census microdata from ten developing countries available from the Integrated Public Use Microdata Series - International (IPUMS-I), the world's largest census database. We use principal components analysis to compute a wealth index based on asset ownership, utilities, and dwelling characteristics. We validate the index by verifying socioeconomic gradients on school enrollment and educational attainment. Given that the availability of socioeconomic indicators varies considerably across samples of census microdata, we implement a stepwise elimination procedure on the wealth index to identify the conditions that produce an internally consistent index. Using the results of the stepwise methodology, we propose which indicators are most important in measuring household SES. The development of the asset index for such a large archive of international census microdata is a very useful public resource for researchers.