Labor market concentration

From SKAIPedia
Jump to navigation Jump to search


Trends: While the majority of local labor markets are highly concentrated, these markets tend to be disproportionately small and cover a small proportion of jobs or vacancies. In addition, local labor market concentration has been decreasing since the 1970s. As a result, it is not clear that labor market concentration alone could account for a substantial portion of trends in wage stagnation or inequality. However, one notable exception to these trends is the manufacturing sector, where labor market concentration for the average job is high and increasing.

Effect on wages: Estimates of the effect of labor market concentration on wages are consistently negative, and vary from small but statistically significant to substantial. The estimates are sensitive to many data and aggregation choices, especially the choice of measuring markets based on vacancies or employment. Because many local factors could affect both concentration and wages, many papers use national labor market conditions within the same industry or occupation as an instrument for local concentration. The effects on wages appear to be strongest at the very highest levels of concentration and may be heterogeneous across worker demographic groups, but are robust to controlling for productivity and concentration in output markets. There is evidence that concentration negatively affects wages conditional on the job, the composition of jobs, the pass-through of productivity gains to wages, and wage inequality. Finally, worker unions may be able to significantly mitigate the effect on wages.

Other effects: Labor market concentration may have effects on workers other than on wages, especially if there are frictions to adjusting wages downward. There is evidence that labor market concentration also has negative effects on employment-based health insurance. While some papers show evidence that concentration shifts the composition of vacancies towards titles with lower pay and decreases the average human capital level of workers in the market, there is other evidence that concentration increases the number of job postings that require social and cognitive skills. In terms of welfare, labor market power is associated with losses of about 6% of lifetime consumption.

Comparison to national labor market concentration: Though local labor market concentration (the focus of the papers below) has been decreasing over time, national labor market concentration has been increasing, perhaps indicating that a small number of firms are increasingly dominating national industries such that they compete with one another more frequently in local markets.


Concentration is typically measured using the Herfindhal-Hirschman Index (HHI). Local labor markets are usually defined as a location-type pair, where location is a commuting zone, metropolitan area, or county and type is an occupation code (e.g., six-digit SOC), industry code (e.g., 4-digit NAICS) or, for vacancies, job title.

Labor market concentration has been measured using online job boards and administrative data. Online job board data could be from one website (e.g., or aggregated across many websites (e.g., Burning Glass Technologies data). Administrative data such as the Census Bureau's Longitudinal Business Database measures employment. Note that while job boards measure concentration in vacancies, administrative data typically measures concentration in employment. Similarly, posted wages from job boards capture effects on the marginal worker while realized wages from administrative data capture effects on the average worker.

Published papers

Trends & Wage Effects

  • Azar, Marinescu, and Steinbaum (2017)[1] use data from to study the concentration of vacancies. They find that the average labor market is highly concentrated (HHI of 3,157), though larger cities tend to be less concentrated. Going from the 25th to the 75th percentile of concentration is associated with 5% lower posted wages. Using national-level changes in each occupation as an instrument, the authors find that going from the 25th to the 75th percentile of concentration decreases posted wages by 17%. The authors also estimate effects controlling for job title and find smaller effects, indicating that concentration may reduce wages conditional on job title and changes in the composition of jobs towards lower paying ones.
  • Azar, Marinescu, Steinbaum, and Taska (2020)[2] use data from Burning Glass Technologies to study the concentration of vacancies. Compared to Azar, Marinescu, and Steinbaum (2017), this paper studies more job postings and more occupations, as well as the sensitivity of concentration to the definition of a labor market. They find that 60% of labor markets are highly concentrated (HHI > 2,500), and the average labor market has a HHI of 4,378. However, these highly concentrated markets only account for 16% of workers. Labor market concentration is negatively correlated with wages, but not with occupational skill.
  • Rinz (2020)[3] uses employment data from the Longitudinal Business Database and shows that labor market concentration has declined since the 1970s. Using IRS data on earnings linked with demographic data from Census surveys, and instrumenting for local concentration with national changes in concentration for a given industry, the author finds that going from the 50th percentile to the 75th percentile of concentration reduces earnings by about 10%. Furthermore, these effects are heterogeneous across groups of workers; increased concentration leads to greater wage inequality, has no statistically significant effect on wages for Black workers, and increases average wages for women. The author also shows the contrast with national labor market concentration, which has been increasing.

Other Effects

  • Johnson and Lipsitz (2020)[4] provide evidence that looking at wage effects alone may not be sufficient. Standard theory predicts that firms will use noncompete agreements (NCAs) if and only if their benefit exceeds the cost to workers. However, the authors construct a model predicting the use of NCAs when the market-clearing wage is constrained, even when the net benefit is negative. The authors go on to find support for their predictions in a survey of hair salon owners, as well as Census employment data on restaurant workers. Thus, labor market conditions that are favorable to firms can enable them to extract benefit at high cost to workers using non-wage components of the employment contract, particularly when the wage is constrained by policy or other factors.

Working papers

Trends & Wage Effects

  • Benmelech, Bergman, and Kim (2018)[5] focuses on the manufacturing sector only, using employment data from the Longitudinal Business Database. By focusing on this sector, the authors can control for productivity using the Census of Manufacturers and the Annual Survey of Manufacturers. They find that labor market concentration is high and increasing since the 1970s. When looking at variation across establishments within the same firm, in the same year, controlling for productivity and market size, the authors estimate that a one standard deviation increase in concentration reduces wages by 1-2%. However, they find evidence that the effect is highly nonlinear; most of the decline in wages occurs as markets approach the pure monopsony case. They also find that unionization rates mitigate the wage effects, firms in more concentrated markets are less likely to share gains in productivity with employees through higher wages, and exposure to import competition from China is associated with more concentrated labor markets.
  • Macaluso, Hershbein, and Yeh (2019)[6] look at labor market concentration over time using both vacancies from Burning Glass Technologies and employment from the Longitudinal Business Database. They find that labor market concentration has decreased significantly over time since the 1970s. In the most recent years, 25% of jobs and 5% of postings are in markets with HHI greater than 2,500. A 1% increase in labor market concentration is associated with a 0.14% decrease in realized wages and a 10-13% increase in the number of job postings that require social and cognitive skills ("upskilling"). The authors also show how the decreasing trend in local labor market concentration can be reconciled with increasing national concentration.
  • Qiu and Sojourner (2019)[7] examine the effects of concentration in employment by occupation, but control for output market concentration which they argue is important to avoid omitted variable bias. For example, a small number of nursing homes in a market will have power in the output market (defined by industry), as well as power in the market for registered nurses (defined by occupation). The authors use a database from Dun & Bradstreet with employment data, Census data on occupational distributions, and Census data on worker human capital characteristics. Using national-level changes in each occupation as an instrument, the authors find that going from the 50th to the 75th percentile of concentration implies a 8.7% decrease in wages and a 2 percentage point decrease in the probability of being covered by employment-based health insurance. The wage effect is weaker in industries with higher unionization rates and when output market concentration is higher. Furthermore, this concentration both reduces wages conditional on human capital as well as the average human capital level of workers in the market.

Other Effects

  • Gibbons, Greenman, Norlander, and Sørensen (2019)[8] look at the effect of labor market concentration on workers with guest visas in the U.S. These workers are more concentrated by occupation than the general population, and past work has documented lower wages and frequent labor law violations. The authors decompose the effects of concentration from the frictions that guest workers face, and estimate that an increase from pure competition to pure monopsony lowers wages by 13%. However, simulating a model using estimated parameters predicts a 47% decline in wages. The authors posit that the wage floor that certain guest workers are subject to curtails the wage effect, but that increased labor market power may also lead firms to reduce safety spending or to provide substandard work environments.
  • Berger, Herkenhoff, & Mongey (2019)[9] study the welfare implications of labor market power. The authors develop a general equilibrium model of the labor market with monopsony, and use Census data to estimate key parameters. They find that labor market power is associated with welfare losses of about 6% of lifetime consumption. Consistent with the literature, the authors find that local labor market concentration has decreased over the past few decades, making it an unlikely driver of the declining labor share of income. Furthermore, they find evidence that minimum wages can deliver modest welfare gains, mostly by reallocating workers from smaller to larger, more productive firms.



  1. Azar, José, Ioana Marinescu and Marshall Steinbaum. 2020. "Labor Market Concentration." Journal of Human Resources. Link
  2. Azar, José, Ioana Marinescu, Marshall I. Steinbaum, and Bledi Taska. 2018. "Concentration in US labor markets: Evidence from online vacancy data." Labour Economics. Link
  3. Rinz, Kevin. 2020. "Labor Market Concentration, Earnings, and Inequality." Journal of Human Resources. Link
  4. Johnson, Matthew S., and Michael Lipsitz. 2018. "Why are Low-Wage Workers Signing Noncompete Agreements?" Journal of Human Resources. Link
  5. Benmelech, Efraim, Nittai Bergman, and Hyunseob Kim. 2018. "Strong Employers and Weak Employees: How Does Employer Concentration Affect Wages?" NBER Working Paper. Link
  6. Macaluso, Claudia, Brad Hershbein, and Chen Yeh. 2019. "Concentration in U.S. local labor markets: evidence from vacancy and employment data" 2019 Meeting Papers, Society for Economic Dynamics. Link
  7. Qiu, Yue, and Aaron Sojourner. 2019. "Labor-Market Concentration and Labor Compensation." SSRN Working Paper. Link
  8. Gibbons, Eric M., Allie Greenman, Peter Norlander, and Todd Sørensen. 2019. "Monopsony Power and Guest Worker Programs" IZA Discussion Papers. Link
  9. Berger, David W., Kyle F. Herkenhoff, and Simon Mongey. 2019. "Labor Market Power." NBER Working Paper. Link