TRADE LIBERALIZATION AND FIRM PRODUCTIVITY: ESTIMATION METHODS MATTER.

In this paper, we show that the relationship between trade liberalization and firm productivity is sensitive to the method used to estimate the production function. We estimate the productivity of Colombian manufacturing plants using the methods of Levinsohn and Petrin, Ackerberg et al., and Gandhi et al. and at times come to surprisingly different conclusions about firm productivity growth after the liberalization. Results from a growth decomposition exercise and from a quantile regression model reinforce the dissimilarity of results across methods. [ABSTRACT FROM AUTHOR]