HOUSEHOLD WEALTH AND SOCIAL WELFARE MEASURES IN RURAL MALAWI USING PANEL ASSETS DATA AND FIXED EFFECTS ANALYSIS

In order to fill the research gaps in rural wealth data, researchers are increasingly measuring assets as a substitute for monetary data. Most of this research utilizes assets data as an explanatory variable that proxies for socioeconomic status. This paper uses asset data from rural Malawi as an outcome variable. The analysis compares the two most cited methods for calculating the wealth index, 1- the principal component analysis and 2- the unweighted fraction of total assets owned. The paper measures changes to wealth index levels and poverty transitions of 996 households from the 2004, 2006, and 2008 waves of the Malawi Longitudinal Study on Families and Health. A fixed effects model is used to control for unobserved heterogeneity while investigating the impact of household characteristics, including the timing of welfare program participation, on the levels of household wealth. The findings reveal a few substantive differences between wealth index methods. Consistent with the literature, being married, having a larger household and parental schooling are positively associated with the wealth index score. Among welfare programs, participation in an agricultural input subsidy is positively associated with the wealth index score and this association is stronger when using lagged year measures for program participation.