Woman Labor Force Participation and the Twin Shock: Using a Two-Period Panel CENSUS Data of Low Income Women in a Developing Country

This paper addresses the causal effect of household size on labor participation and other labor outcomes of women receiving a shock of multiple births in their first pregnancy, a situation that isolates the existing endogenous... [ view full abstract ]

This paper addresses the causal effect of household size on labor participation and other labor outcomes of women receiving a shock of multiple births in their first pregnancy, a situation that isolates the existing endogenous component in labor market participation decisions and fertility. We use a two-period panel CENSUS data of low-income households in Colombia that allow us to observe poor women between ages range from 10 to 49 that did not have kids by 2005 but had their first and only child before 2009-10. In addition, unlike previous literature, mothers are observed before their fecundity in their demographic characteristics (baseline) and labor indicators are analyzed in a future period. We compare a large set of baseline characteristics measured in 2005 and we found that observable variables are well balanced between non-twin and twin mothers before the first child was born. The main findings suggest that an increase in family size during the first pregnancy due to giving birth to twins reduces between 4 to 5 ppts female labor supply (14% effect) and formal employment by 2 ppts. Our results are robust to the different specifications and samples considered. We do not find robust effects on formal employment and earnings.