The comparative role of determinants of household-level consumption expenditure inequalities (henceforth, inequalities) in rural India between two sub-periods, 1994–2005 and 2005–12 are examined, using three rounds of the National Sample Survey Consumer Expenditure Survey. The changes in the components of consumption expenditure and population characteristics are explored that explain inequalities during the two sub-periods, which represent distinct policy environments. We use both a priori and regression-based decomposition methods for the analysis. We find that there is a complete reversal of the role of education in explaining inequalities. It shifted from being an inequality-increasing factor during 1994–2005 to an inequality-equalising factor during 2005–12. This reversal is induced by decreasing consumption returns to education due to the depressed job market. The role of locational factors has increased in explaining the increase in inequalities over time. The non-food components induce an increase in the overall inequalities via an increased expenditure on durables. The within-group component contributes the most to the level of and change in inequalities.
