Taboos, Agriculture and Poverty

We study the impact of work taboos (fady days) on agriculture and poverty. Using cross-sectional data from a national household survey for Madagascar, we find that 18 per cent of agricultural households have two or more fady days per week and that an extra fady day is associated with 6 per cent lower per capita consumption and 5 per cent lower rice productivity. To address the possible endogeneity of fady days, we present instrumental variable estimates and heterogeneous effect regressions using village fixed effects. We find that smaller households and those with less education employ less labour in villages with more fady days.