Research has shown that labor markets with higher levels of labor market concentration have lower wages. It does not necessarily follow that regulators should limit merger activity because of labor market concentration.
Labor market concentration and wages are negatively related. This relationship has led some to call for changing the way the FTC reviews and regulates merger and acquisition activity (Abdela and Steinbaum, 2018; Azar et al., 2017; Naidu et al., 2018). In fact, this changing regulatory standard may already be happening. In response to questioning from Senator Richard Blumenthal (D-CT) at the Senate Antitrust Committee’s oversight hearing of October 3, 2018, FTC Chairman Joseph Simons testified that “[FTC management has] told staff that they’re supposed to look at potential effects on the labor market with every merger they review.” But should they block mergers because of these labor market concerns?
Public policy and regulation are meant to ameliorate problems and the suggestion to regulate mergers based on labor market concentration is no different. My own working paper (Lipsius, 2018) shows that local labor market concentration, however, is not responsible for the falling labor share, for stagnating wages, or for increasing income inequality. I show that in order to drive any of these phenomena, the employment-weighted average labor market concentration1)Employment-weighted average labor market concentration is the average labor market concentration weighted by how many employees work in that labor market. Mathematically, it is where L is total employment, lm is the employment in market m and hm is the concentration in market m. would have to have increased since the 1980s. Instead, it has decreased. Moreover, falling local concentrations are remarkably robust to various labor market definitions (Qiu and Sojourner, 2019; Rinz, 2018).
These findings run counter to several recent, high-profile studies. These other studies, however, are limited in their ability to shed light on these important issues either because they do not cover a long enough time frame (Azar et al., 2017, 2018) or because they only examine limited, non-representative sectors of the US economy (Benmelech et al., 2018). An additional branch of literature (Barkai, 2016; David et al., 2017) connects the falling labor share to increasing concentration measured at a national level. These nationally-calculated concentrations should not be confused for local labor market concentrations. Finally, other forms of employer labor-market power may be increasing even if the labor market concentration, on average, is not (Bivens et al., 2018).
While labor market concentration has not driven these societal ills, one might argue that regulations based on labor market concentration could possibly help alleviate them. In order for this sort of regulation to effectively counteract these trends, however, it would have to impact the employment-weighted average labor market concentration because the employment weighting links local markets to national aggregates.2)The employment-weighted average labor market concentration is the level of labor market concentration faced by the average US employee as opposed to the level of concentration in the average labor market, a subtle but important difference. Azar et al. (2018) presents a heat-map of the most heavily concentrated US commuting zones that shows that highly concentrated labor markets tend to be rural–that is, the labor markets most likely to fail a concentration-level test of competitiveness are exactly the labor markets least likely to impact aggregates such as the labor share or aggregate wages (because they have so few employees and thus a very low employment weight), making it highly unlikely that introducing this new review can reverse the trends in the aggregates.
While this new regulation likely cannot affect aggregate trends, it could still potentially affect local wages in specific labor markets. It is unclear, however, if this sort of regulation will end up helping workers. A first issue with blocking takeovers is that larger firms tend to pay their workers more. Oi and Idson (1999) provide a review of the literature around this so-called firm-size effect up to 1999. More recently, Bloom et al. (2018) show that, while the firm-size effect has been shrinking, employees of 10,000-employee firms still make 20 percent more than similar employees of 100-employee firms. More research is needed to understand how labor market concentration and the firm-size effect interact with each other and which effect is larger. Blocking a merger because of local concentration considerations without taking into account that this new, larger firm could end up paying more than the previous smaller firms might hurt the very workers this regulation intends to help.
Similarly, current research relating to labor market concentration and wages has shown that, all else equal, labor markets with higher levels of labor market concentration have lower wages. A merger or an acquisition, however, does not leave all else equal. A productivity-enhancing merger could raise wages even as it raises levels of concentration. How to weigh the wage gains of productivity increases against the wage declines from increased concentration isn’t clear in the literature. In fact, the relationship between labor market concentration and productivity is still unclear. It is plausible that wages are negatively correlated with labor market concentration because of the relationship of each with the market’s productivity i.e. unproductive areas have low wages and high concentrations because they are unproductive. Allowing productivity-enhancing mergers could help increase an area’s productivity and decrease long-run concentration while increasing short-run concentration.These proposed regulations eliminate this possibility.
A third problem with introducing this new regulation is the likely response of large companies. If profitable mergers are blocked because of small operations in rural areas, the most obvious reaction for able companies is to close their rural operations or move operations to larger cities where labor market concentration is not a concern. Large companies abandoning rural areas will not only deprive these areas of better paying jobs (see the firm-size effect above) but will also deprive these areas of important services provided by less efficient vendors—possibly raising prices. Lowering wages and raising prices is exactly the opposite of the regulation’s intended effect and more research is needed to understand the long-run, general equilibrium effects of a new policy such as this.
Beyond questions surrounding the efficacy of the proposed new regulations, there are measurement issues, left largely unaddressed by the literature, that call into question the practicability of such a regulation. Most importantly among these questions are: What is a labor market and how do you measure concentration within an appropriately defined labor market?
Different studies have used different definitions of the labor market, but the basic approach is to create some definition of “local” and some definition of economic activity. The interaction of these two components is the market. While the negative relationship between labor market concentration and wages is robust to these definitional differences, the exact measure, and therefore the review of any particular merger, depends heavily on which definition is chosen.
Within the literature, the definition of economic activity is even bigger than the controversy surrounding the definition of “local”. Some studies use classifications based on what industries people work in (industry codes) while others use classifications based on what people do (occupation codes). Which is right, is unclear, but important for proposed regulation. To help address this question, it is useful to think about what is a labor market and what is being measured. Within a labor market, firms are looking for employees and people are looking for jobs. The relative importance of industrial experience versus occupational experience is an important open question. Sullivan (2010) suggests that the answer depends on the particular job and industry. Because some jobs require more occupational experience and some require more industrial experience and people can look to change jobs between industries, occupations or both, neither using occupational codes nor using industrial codes covers all situations.
The implication of all of this is not that this sort of measure can never be useful for regulation because of measurement issues, but rather that more research and a better understanding of the particulars of the markets in question and the deals under review are needed to ensure that policy is not counterproductive.
Labor market concentration and wages are negatively related. “Regulators should therefore limit merger activity because of labor market concentration” does not follow. There are too many unknowns and too many negative potential unintended consequences to changing the regulatory environment for it to be a good idea. A better understanding of rural (and purportedly concentrated) labor markets, of firms’ and employees’ dynamic responses to changes in environment and in concentration level, of concentration measurement, and of the underlying data are all necessary to make good policy. More research is needed before new regulations can be safely implemented.
Ben Lipsius is a PhD candidate in Economics at the University of Michigan.
José Azar, Ioana Marinescu, and Marshall I. Steinbaum. “Labor Market Concentration.” Technical report, National Bureau of Economic Research, 2017.
José Azar, Ioana Marinescu, Marshall I. Steinbaum, and Bledi Taska. “Concentration in US Labor Markets: Evidence From Online Vacancy Data.” Technical report, National Bureau of Economic Research, 2018.
Efraim Benmelech, Nittai Bergman, and Hyunseob Kim. “Strong Employers and Weak Employees: How Does Employer Concentration Affect Wages?” Technical report, National Bureau of Economic Research, 2018.
Nicholas Bloom, Fatih Guvenen, Benjamin S. Smith, Jae Song, and Till von Wachter. “The Disappearing Large-Firm Wage Premium.” In AEA Papers and Proceedings, volume 108, pages 317–22, 2018.
Ben Lipsius. “Labor Market Concentration Does Not Explain the Falling Labor Share.” Available at SSRN 3279007, 2018.
Walter Y. Oi and Todd L. Idson. “Firm Size and Wages.” Handbook of labor economics, 3:2165–2214,1999.
Kim Parker, Juliana Horowitz, Anna Brown, Richard Fry, and D’Vera Cohn. “What Unites and Divides Urban, Suburban and Rural Communities: Amid Widening Gaps in Politics and Demographics, Americans in Urban, Suburban and Rural Areas Share Many Aspects of Community Life.” Pew Research Center, 2018.
Yue Qiu and Aaron Sojourner. “Labor-Market Concentration and Labor Compensation.” Available at SSRN 3312197, 2019.
Kevin Rinz et al. “Labor Market Concentration, Earnings Inequality, and Earnings Mobility. Technical report, Center for Economic Studies,” US Census Bureau, 2018.
Paul Sullivan. “Empirical Evidence on Occupation and Industry Specific Human Capital.” Labour economics, 17(3):567–580, 2010.
For more on this, check out the following episode of the Capitalisn’t podcast:
Disclaimer: The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.
References [ + ]
|1.||↑||Employment-weighted average labor market concentration is the average labor market concentration weighted by how many employees work in that labor market. Mathematically, it is where L is total employment, lm is the employment in market m and hm is the concentration in market m.|
|2.||↑||The employment-weighted average labor market concentration is the level of labor market concentration faced by the average US employee as opposed to the level of concentration in the average labor market, a subtle but important difference.|