Essays in Productivity

Abstract

This dissertation is a collection of three essays linked through the common theme of the aggregation of productivity in motion. The first essay examines productivity trends in European Manufacturing through the movement of activity between firms and the growth within and between firms. The second essay examines firm-level productivity growth through the movement of workers between firms. The third essay examines the underlying measurement of productivity at the firm level. The common strand in the first two chapters is the role of distortions at the micro level in explaining productivity outcomes at the macro level.

In the first essay, we examine the changes in productivity, the labour share, and concentration in Europe manufacturing over the past decade. We provide evidence that lobbying decreases aggregate productivity growth by limiting potential positive reallocations. We further show that lobbying is related to declining business dynamism and increasing movements in activities to high markup and low labour share firms in the long run.

In the second essay, we examine worker-embodied technological spillovers in South African manufacturing and find evidence of both positive and negative productivity spillovers. While positive worker-embodied technological spillovers have been found for other countries, the negative spillover is unique. While we fail to find evidence for positive productivity spillovers for exports, imports, or R&D intensity we do find positive productivity spillovers for training intensity. These results provide evidence of substantial mismatches in skill demand and supply in South African Manufacturing.

The third essay improves the standard estimation implementation used to estimate the factor elasticities required to impute productivity. We show that by directly estimating the productivity shock process while varying step size we can obtain estimates inside the unit interval more consistently than standard implementations. We show that our results are generally more robust to the measurement of labour and capital than standard approaches.