Is female labour force participation a good proxy for gendered time use? How do geography and the social institutions of caste and religion interact with the gendered distribution of time within Indian households? In this study, we use gender distance metrics, inspired by distance measures between vectors, to measure the extent to which time allocation within households is gendered. We show that the relationship between gender distance and labour force participation is not monotonic and the linear relationship between the two is not statistically strong. The relationship of caste, religion and region with gendered time use metrics is distinct from their relationship with employment. Interestingly, in contrast to popular hypotheses which suggest North Indian, Muslim, and Upper Caste households are more gender unequal, we only find robust confirmation for the hypothesis related to Islam in our regression framework. To further estimate the direct contribution of caste and religion in explaining the gendered time use gap between groups (as distinct from the contribution of differential distribution of covariates between groups), we supplement our regression results with Oaxaca-Blinder (1973) decomposition and Dinardo-Fortin-Lemeieux(1996) decomposition.
These analyses confirm that caste and religion have complex and unexpected heterogeneous effects on the intensity of gendered time use.
