Does Family Planning Policy Matter? Dynamic Evidence from China

This paper, among the first, dynamically estimates the effects of China’s family planning policy on fertility with a micro-level panel data, transformed from the cross-sectional birth history data of the China Health and Nutrition Survey. This paper applies a mixed proportional hazard duration model where the unobserved individual heterogeneity is non-parametrically specified, recommended by Heckman and Singer (1984), and extends the improved policy measures from Wang (2012)’s static analysis to dynamic analysis. It’s found that, the one-child policy, the most recent and ongoing period of China’s policies, has remarkably negative effects on the probabilities of having the second and third child, but shows little impact for other birth orders. The effects of earlier periods of policies are simply trivial through all birth orders. Therefore, family planning policies didn’t play a major role in China’s fertility transition which mainly happened before the one-child policy period. Further, a more-educated woman tends to respond to policy shocks at lower birth orders in contrast to a less-educated woman, because the former desires a smaller family. Moreover, the model without heterogeneity would strikingly overestimate the policy effects and a parametric heterogeneity might also bias the estimation.