Drinking water and off†farm labour supply: between†gender and within†gender bias

In this paper, we use the China Health and Nutrition Survey (CHNS) panel data to analyse the impact of drinking water on off†farm labour supply. A two†stage least squares (2SLS) multivariate Tobit regression model with random effects was applied. We find that impacts of drinking water conditions on off†farm labour supply may be greater for women than men but depends on the specific family role or family structure. A strong within†gender effect exists in households. For example, daughters are not sensitive to water access nor water quality, but householder's spouses are sensitive to water access, and daughters†in†law are sensitive to water quality. Our findings suggest that infrastructure development in improved access to safe water has contributed positively to reductions in traditional gender biases, evening the playing field between daughters, daughters†in†law, mothers and mothers†in†law. We also find that water the infrastructure program may actually encourage off†farm labour mobility, reducing the supply of agricultural labour and the share of household labour on the farm. Thus, a broader approach to water policy should also include public investment in achieving greater labour efficiency and productivity.