Misallocation and Productivity Growth: a Meta-analysis
We use a meta-analysis to quantify the impact of misallocation of production factors on aggregate productivity. A key estimate in empirical studies on misallocation is the implied aggregate total factor productivity (TFP) loss due to the sub-optimal allocation of resources across firms. In our meta-analysis, we correlate this effect size with various study characteristics. First, we find that the TFP growth effect size is smaller than the level effect size. Second, we make a distinction between studies following a direct or indirect approach, where the former relates misallocation to one or more specific factors while the latter quantifies the overall effect of all possible sources. We find that studies following a direct approach generally report a smaller TFP loss than those using an indirect approach. Third, we find that the extent of misallocation and the corresponding productivity loss depends on the country of analysis. In particular, there is a negative orrelation between TFP loss and the level of income.
Keywords: meta-analysis; misallocation; productivity
JEL codes C40; D24; O47
Working paper no. 774
Research highlights :
- We use a meta-analysis with 72 primary studies to quantify the impact of misallocation on aggregate total factor productivity (TFP).
- A key estimate in the primary studies is the aggregate TFP loss due to the sub-optimal allocation of resources across firms.
- In the meta-analysis, we correlate this productivity loss with various study characteristics.
- We find that misallocation is persistent over time and depends on the country of analysis. In particular, we find a negative correlation between the TFP loss and the level of income.
- The productivity loss from misallocation due to financial frictions is relatively large.