The significance of structural transformation to productivity growth.

This paper critically discusses the most common methodology for decomposing productivity change into inter- and intra-firm effects. It is argued that the methodology can be improved to explicitly take the role of structural transformation into account, and by so doing, a potential source of bias in the results is corrected. This requires the use of a tool from the field of theoretical evolutionary biology: Price's equation. A review of a sample of studies that apply decomposition analyses shows that the methodology is best suited for studies of the evolution of labour productivity and the reallocation of labour. Based on Danish data for 1992-2010, it is then demonstrated how the results of decomposition analyses can be considerably improved by the explicit inclusion of levels in the selection process. In the specific analysis conducted by the current paper, economic selection among industries is included. It is found that the structural transformation of the economy has a large impact on the results of decomposition studies, not least on the magnitude of the inter-firm selection effect. Structural transformation from capital-intensive and thus high-labour-productivity manufacturing towards labour-intensive and thus low-labour-productivity services entails that the traditional methodology is biased downwards in its measure of economic selection. Finally, it is demonstrated that the length of the interval studied, while often determined by data limitations, has a significant but predictable effect on the results, and it is tentatively demonstrated that economic selection tends to be stronger in the trough of the business cycle. [ABSTRACT FROM AUTHOR]