Female Employment and Violence in the Household: Evidence from the Latin American Demographic and Health Surveys

Abstract

The relationship between female earnings and violence in the household is negative in the US but not in many other contexts. Recent studies from developing countries suggest that, despite greater financial autonomy, rising female employment is associated with more violence. A generally positive conditional correlation between female employment and violence is here documented using available Demographic and Health Surveys from Latin American countries. These surveys do not suggest that violence is associated with women having less say in decisions regarding resource allocation decisions in households. As well, although levels of violence are higher among less educated women, violence is still very prevalent among women with completed secondary school. These stylized facts suggest that bargaining power and earnings potential are only two of several factors important to the explaining the incidence of violence. Historical data from the 1971 Nicaragua census is here combined with survey data to find that a rise of 0.1 in women's employment propensities is associated with an increase in violence propensities of about 0.01-0.04 in both 1998 and in 2011-12. In Nicaragua, this positive association appears to be concentrated among less-educated women. The results are consistent with violence rising with women's employment because of associated changes in the nature of cooperation between spouses in the provision of household public goods.