Consumption, Income and Wealth of the Very Poor: An Empirical Investigation of Cross-sectional and Lifetime Inequality in Sub-Saharan Africa

While household-level inequality is well-documented for a large set of industrialized economies, little is known about the same distributional facts of consumption, income and wealth in developing countries, in particular, for Sub-Saharan Africa (SSA). This paper works toward closing this gap. First, using new and unique panels of cross-sectional nationally representative data from the Integrated Surveys on Agriculture (LSMS-ISA), we analyze in-country cross-sectional inequality in five Sub-Saharan countries: Malawi, Tanzania, Uganda, Nigeria and Niger. We also study lifetime inequality and the ability to smooth consumption. We find that, after controlling for household composition and life expectancy, consumption inequality in urban areas tracks income inequality fairly well over the life cycle. Further, consumption and income display patterns of inequality that, while increasing with age, show a steeper profile for income than for consumption suggesting some degree of partial insurance similar to that in the U.S.. However, in rural areas---where the vast majority of households live---consumption inequality tends grow more than income inequality over the life cycle which suggests rural households have a very poor ability to insure. Our results are robust to alternative measures of disposable income that include private and public transfers. We also find that only 88% borrowed in the last year. These households also face considerably higher interest rates. There is also a large number of loans with interest rates equal to zero even for households at the top of the income distribution. The shocks that affect rural households are mostly due to weather and prices, but death and illnesses are also important for the poorest households. We show that 70% of rural households report having no response to their most severe shock. The most common response is to use their own savings - this is true even among the poorest households - suggesting savings for precautionary reasons as an important smoothing mechanism. The second source of response to shocks is the help or relatives, which stresses the role of transfers as a way to try and smooth consumption. These findings directly contrast with the relatively large ability to smooth consumption reported in other studies for developing countries of other world regions. Further, variance decompositions suggest that while between-group inequality determined by education, sex, societal systems, marital status and health behavior explain xx% of lifetime inequality, it is the within-group (residual) inequality what determines the rise in inequality over the life cycle, even after controlling for demographics. This suggest the role of idiosyncratic shocks is relevant to understand inequality over the life cycle in these low-income environments. We discuss how simple modifications of the standard PILCH theory can, potentially, reconcile some of these facts.