Dinâmica da produtividade: uma abordagem por meio de modelo de fatores dinâmicos

Abstract

The purpose of this dissertation is to study the global dynamics of total factor productivity (TFP) through a dynamic factors model and verify the importance of the catch up effect and other determinants of TFP through the System GMM estimator, using data from the Penn World Table 9.0 from 1970 to 2014. Because there’re technological spillovers and international trade, it would be expected to see the existence of common variations in the productivity of different countries. Therefore, we apply a dynamic factor model to these productivities and its growth rates in order to capture these effects. These estimates allow us to extract the common factor and use it as a proxy for the world TFP. It allows us to include the common trend in the dynamic equation in the TFP regression in order to estimate its effects on the TFP. Our estimated common factor captures well the main productivity shocks that occurred in the period, and the sensibility of each country to the common factor seems to be negatively related to its country level of technological development – this correlation is in accordance with the argument that there’s a catch up effect in productivity: countries more distant from the frontier have higher growth rates because there is more room to absorb from the global technology. Moreover, our estimates for the dynamic panel using this estimated factor reinforces this result.