Estimating the Effect of Fertility on Poverty in Vietnam: An Example of Causal Inference with Multilevel Data in Demographic Research

Bruno Arpino, Universitat Pompeu Fabra

Multilevel structured data are very common in demographic research: e.g., individual clustered in households; households clustered in communities. In this paper we focus on the role of the community context in the estimation of the causal effect of fertility on poverty in Vietnam. The contextual characteristics can strongly influence both poverty and fertility and therefore it is crucially important to account for this in order to draw valid causal inference. The multilevel dimension introduces statistical complications and stimulates interesting research questions. From the methodological point of view, we use multilevel techniques in the propensity score matching implementation and a weaker version of the traditional SUTVA assumption. We find a negative and substantial effect of fertility on wellbeing; this effect is stronger in high-level fertility communities. On the contrary, fertility measured at the community level does not have a significant effect per se on households’ economic wellbeing.

  See paper

Presented in Session 137: Causal Inference and Experimental Designs