In 1979, China imposed a one-child policy. Evaluating the impact of this policy has presented challenges, particularly in identifying a clear control group. This made it particularly challenging to establish robust methods to identify the effect of the one-child policy on various family outcomes.
In “Estimating Treatment Effects of the One-Child Policy: A Self-Report Approach,” authors Fangzhu Yang and Yingyao Hu employ a self-reported survey measure, which captures couples’ preferences regarding the ideal number of children. This approach is broadly applicable to heterogeneous treatment effect models, as individuals will possess the only information about their preferences, which makes self-reported preferences a valuable tool for uncovering treatment effect heterogeneity across subpopulations.
In a 2014 version of the survey, respondents indicated how many children they would want to have regardless of the policy restriction; another survey was conducted in 2018, two years after the policy was ended. The authors use this unique structure to estimate the number of children couples would have had in 2014 without the policy, thereby identifying the causal impact of the policy. By leveraging the unique feature of the survey that asks about couples’ preference, they estimate the counterfactual number of children couples would have had. This method provides a broadly applicable framework for evaluating policy effects where such data are available.
Access the article here (subscription required).
