Program impact evaluation using a matching method with panel data

Difference-in-differences with matching is a popular method to measure the impact of an intervention in health and social sciences. This method requires baseline data, that is, data before interventions, which are not always available in reality. Instead, panel data with two time periods are often collected after interventions begin. In this paper, a simple matching method is proposed to measure the impact of an intervention using two-period panel data after the intervention. The method is illustrated by the measurement of the effect of health insurance in Vietnam using household panel data.