Analysing Heterogeneity in Global Production Technology and TFP: The Case of Manufacturing

In this paper we ask how technological differences in manufacturing production across countries can best be modeled when using a standard production function approach. We emphasise the importance of allowing for differences in the impact of observables and unobservables across countries, as well as of data time-series properties. Our novel twostage estimator, similar to the Pesaran (2006) Common Correlated Effects approach, can account for these matters. Empirical results and diagnostic testing provide strong support for technology heterogeneity. Results are supported when we adopt an econometric approach allowing for reverse causality. In the light of these findings the interpretation of regression intercepts as TFP levels collapses and we introduce an alternative measure.