Indirect inference estimation of dynamic panel data models

Abstract

This paper proposes an estimator for higher-order dynamic panel models based on the idea of indirect inference by matching the simple within-group estimator with its analytical approximate expectation. The resulting estimator is shown to be consistent and asymptotically normal. For the special case of first-order dynamic panel, the estimator yields numerically the same result from an existing procedure in the literature, but the inference to follow differs and this paper examines the differences and implications for hypothesis testing. Monte Carlo simulations show that the proposed estimator is virtually unbiased, achieves usually lower root mean squared error than competing estimators, and delivers very reliable empirical size across various parameter configurations and error distributions. This new estimator is used to estimate the convergence parameter in an inequality measure among 63 countries during 1985–2015. It shows strong evidence of convergence over long test horizons but much weaker evidence over a 5-year horizon for developing countries.