Child Discipline, Domestic Violence and Social Programs: Evidence from Colombia ∗ Preliminary Version-Please do not cite

This paper examines how interventions during early childhood affect disciplining methods in Colombia, where poor households are eligible for a number of social pro-grams targeted to young children based on a proxy means index. These programs include child care options, nutritional programs and health checks. I analyze whether these programs affect parents ’ disciplining methods through two different identification strategies. I implement a regression discontinuity design exploiting the discontinuity on the probability of benefiting from these programs, as a function of the proxy means index used for targeting. I also implement a propensity score matching using differ-ences in length of exposure to one of these programs, a subsidized child care option. Under both approaches, I find evidence of a change in parents ’ disciplining practices due to social programs targeted to young children. Parents of children, who attend the health checks regularly, are 20 percentage points less likely to use physical punishment, as compared to parents of children who attend less frequently. Mothers of children, who attend at least 16 months to the subsidized child care option, use less physical punishment and more frequently pedagogic punishment, as compared to mothers of children who only attend 2-15 months. Impact is focused mainly on young children, children 0-5 years of age. This evidence suggests that early childhood interventions can be important vehicles for changing parents ’ disciplining behavior