Journal of Statistical and Econometric Methods

A Monte Carlo Analysis of Robustness of the Synthetic Control Method and Dynamic Panel Estimation: A Comparative Case Study of a Policy Intervention.

• [ Download ]
• Times downloaded: 9148

• Abstract

   

  In comparative case studies, by solving an optimization problem, the synthetic control method provides a point estimate for an intervention effect and it suffers from lack of considering an asymptotic distribution of the estimator. On the other hand, we can benefit from such considerations while working with a regression framework; and many studies have been done and many methods have been offered in order to overcome the potential shortages of a traditional regression framework in such case studies. In this paper, we use Monte Carlo simulation to compare the robustness and sensitivity between the synthetic control method and a dynamic panel data regression framework. Empirical work in based on a suitable case of a policy intervention and a comparative case study: sanctions on Iran. We conclude that the dynamic panel data model seems to be performing well with the macro level aggregate data and a comparative case study scenario, and the assumptions are appropriate. However, for the synthetic control method we observe large standard errors in the estimated values which result in insignificance of the point estimates. We also take advantage of the replicated trials, and we analyze and compare the sensitivity of the synthetic control method and the dynamic panel data model to the choice of the donor pool and the treatment assignment.

  JEL classification numbers: C15, C33, C5

  Keywords: Synthetic Control Method, Panel Data Model, Monte Carlo Simulation, Comparison