Reallocation, Firm Turnover, and Efficiency: Selection on Productivity or Profitability?

Abstract

A pervasive finding in the burgeoning literature using business microdata is that firm turnover is high and that this churning process contributes substantially to aggregate (industry) productivity growth, as more productive entrants appear to displace less productive exiting businesses. A limitation of this research is that establishment-level prices are typically unobserved, resulting in within-industry price differences being embodied in productivity measures. If prices reflect idiosyncratic demand shifts or market power variation, high "productivity" businesses may not be particularly efficient. In this case, the literature's findings might be better interpreted as evidence of entering businesses displacing less profitable, but not necessarily less productive, exiting businesses. This distinction is important not only for the sake of understanding the positive features of selection, but the normative ones as well; whether selection is driven by efficiency or market power differences has important welfare implications. In this paper, we investigate the nature of selection using data from industries where we observe both establishment-level quantities and prices. We find that, as has been found in the preceding literature for revenue-based TFP measures, physical productivity and prices also exhibit considerable within-industry variation. We also show that while physical productivity shares common traits with revenue-based measures, there are important differences. These involve the productivity levels of entrants relative to incumbents and the size of the impact of net entry on productivity aggregates. Furthermore, we characterize the dimension(s) of selection and show that both idiosyncratic productivity and demand (price) conditions affect businesses' survival probabilities.

Agglomeration economies with consistent productivity estimates

“…On the one hand, it confirms the theoretical (Melitz 2003) and empirical (Foster et al 2008) finding that highly productive establishments set lower prices. This, on the other hand, stresses the importance of separating price effects from true productivity.…”

“…Foster et al (2008) make use of a dataset with information on output in physical quantities and revenues, confirming that revenue based productivity estimates embody price variation. Consequently, inference from revenue based and physical productivity estimates is different.…”

Recall and Unemployment

“…The employer-level (log) TFP estimated by Foster, Haltiwanger, and Syverson (2008) has exactly the same persistence, 0.7 annual, and higher volatility, 21 percent over five years. Given possible sources of measurement error, we find this to be in the ballpark; our model does not require implausibly volatile or transient measures of productivity.…”

Quantifying Confidence

“…The figure shows that under a plausible value forσ a , the range of values for σ ξ that would be consistent with the restrictions imposed by a common-prior specification is very large. 33 Empirical estimates of the volatility of firm-level productivity suggest settingσa between 0.2 to 0.43 (Abraham andWhite, 2006, Foster, Haltiwanger, andSyverson, 2008). In a similar setting as ours, Huo and Takayama (2015b) use a value of 0.14.…”