Link Search Menu Expand Document
Bibliography

Why is productivity correlated with competition?

NBER link

Wiley Link

BibTeX 

@article{backus2020productivity,
  title={Why is productivity correlated with competition?},
  author={Backus, Matthew},
  journal={Econometrica},
  volume={88},
  number={6},
  pages={2415--2444},
  year={2020},
  publisher={Wiley Online Library}
}

Abstract

The correlation between productivity and competition is an oft observed but incompletely understood result. Some suggest that there is a treatment effect of competition on measured productivity, for example, through a reduction of managerial slack. Others argue that greater competition makes unproductive establishments exit by reallocating demand to their productive rivals, raising observed average productivity via selection. I study the ready-mix concrete industry and offer three perspectives on this ambivalence. First, using a standard decomposition approach, I look for evidence of greater reallocation of demand to productive plants in more competitive markets. Second, I model the establishment exit decision and construct a semiparametric selection correction to quantify the empirical significance of treatment and selection. Finally, I use a grouped instrumental variable quantile regression to test the distributional predictions of the selection hypothesis. I find no evidence for greater selection or reallocation in more competitive markets; instead, all three results suggest that measured productivity responds directly to competition. Potential channels include specialization and managerial inputs.

Notes and Excerpts

Closest to this work is a subset of the literature that has focused on industries where we can measure TFPQ rather than TFPR. Notable contributions include Foster, Haltiwanger, and Syverson (2008), using data on several industries where measures of physical output are available, Collard-Wexler and De Loecker (2015) on steel, and Asker, CollardWexler, and De Loecker (2019) on oil production. The leading example of such an industry, however, is ready-mix concrete. This paper builds on the pioneering work of Syverson (2004), which studied the negative correlation of productivity dispersion with competition.

I use U.S. Census of Manufacturers (CMF) data for the ready-mix concrete industry (SIC (standard industrial code) 3273) from the years 1982, 1987, and 1992

I follow Syverson (2004) in using the 1995 Component Economic Areas (CEAs), which are constructed by the Bureau of Economic Analysis.

Reduced form:

  • Regresses plant productivity on density of plants.
  • But productive firms could discourage competitors from entering, so instruments on density of building permits and government road expenditures.

borrows a static decomposition of productivity from Olley and Pakes (1996), where output-weighted productivity in a market is the sum of unweighted average productivity and of productivity-output covariance.