Imperfect Competition in Labor Markets

President Biden’s recent Executive Order on “Promoting Competition in the American Economy” centered employers’ monopsony power and anticompetitive conduct as core components of American competition policy. This is long overdue. At least since Joan Robinson’s revolutionary theorization of imperfect competition in labor markets in 1933, labor economists and social scientists have theorized and empirically documented the extent as well as micro- and macroeconomic effects of employer monopsony over workers. This essay provides an overview of the empirical evidence of imperfect competition in labor markets and documents the scope and substance of early regulatory enforcement efforts — beginning with the Obama Administration on. It then identifies the primary doctrinal and administrative challenges that antitrust enforcers will face in tackling imperfect labor market competition and makes a series of recommendations on how to clarify legal rules, establish enforcement priorities, and ensure an interagency infrastructure and robust agency expertise for successful enforcement.

By Hiba Hafiz1

In her foundational book, The Economics of Imperfect Competition (1933), Joan Robinson made a radical and contrarian claim: the “real world” of labor markets does “not fulfil the assumptions of perfect competition”; labor markets should instead be modeled on imperfect competition. Heterogeneous preferences, information asymmetries, and search and mobility costs naturally generate inelastic demand for workers’ labor, creating employer buyer, or monopsony, power that prevents workers “from reacting to differences in the earnings” they could obtain from alternative sources of employment. The endemic nature of monopsony in labor markets would require an ongoing regulatory response throughout the economy, and she called on economists to “us[e] their tools upon observed facts” to “build up that working model of the actual world which it is their aim to construct” in order to better predict labor market competition outcomes.

But for decades that call went unheeded in traditional Industrial Organization (“IO”) economics that has informed antitrust policy, even if labor, industrial relations, and institutional economists continued to recognize labor market frictions that contribute to employer monopsony. Antitrust agency policy and enforcement has ignored imperfect labor market competition, failing to investigate or challenge employer dominance, anticompetitive agreements and collusion, and they’ve ignored any labor market competition effects in merger reviews. Both the agencies and the courts seemed to presume that labor markets functioned on a model of perfect competition, focusing exclusively on product market competition issues and downstream consumer welfare.

I. EVIDENCE OF IMPERFECT COMPETITION IN LABOR MARKETS

Due in part to an “empirical turn” in antitrust economics and scholarship (as well as changing political winds), the antitrust agencies — and the Executive branch more generally — have recently begun to take seriously the existence and harms of employer monopsony on hiring, wages, and workplace quality, but also on labor market dynamism and growth, workers’ purchasing power, and labor’s share of national income. First, IO economists have shown that, even under traditional antitrust metrics, average labor markets in the United States are highly concentrated at Herfindahl-Hirschman Index (HHI) levels that exceed antitrust thresholds, and workers suffer lower wages as a result. Economists have also shown that labor markets are more inelastic than product markets, with median product elasticity at 4.5 and labor supply elasticities ranging between 0.1 and 1, meaning that the same HHI in a labor market as a product market would result in a larger markdown in the labor market as compared to the corresponding markup in a product market. There is also increasing evidence of employers’ anticompetitive conduct in labor markets, including wage-fixing, reaching no-poach and information-sharing agreements, and pervasively using non-competes that reduce labor market competition.

In addition to traditional antitrust evidence of employer monopsony — and evidence of employer conduct that has emerged from government investigations and private enforcement — labor economists have presented evidence of employer monopsony, even in thick markets and over low-wage workers, due to labor market frictions like search and mobility costs, job differentiation, heterogeneous preferences, and information asymmetries. Research in behavioral economics has also identified sources of employer monopsony in workers’ behavioral assumptions, heuristics, and beliefs about the employment bargain. More broadly, the systemic failure of labor market institutions — whether government institutions and labor law or labor market institutions like unions that support and bolster workers’ countervailing power — contribute to employers’ bargaining leverage relative to workers. Finally, public economists, labor economists, and macroeconomists have identified employer monopsony as contributing to the decline of labor’s share of national income despite increases in worker productivity, exacerbating inequality.

II. REGULATING IMPERFECT COMPETITION IN LABOR MARKETS

The Obama Administration was the first to respond to these labor market realities, taking a more aggressive approach to increasing awareness and developing policy to respond to imperfect labor market competition. President Obama appointed as Chairs to his Council of Economic Advisors (“CEA”) Alan Krueger and Jason Furman, two leading economists that highlighted employer monopsony and anticompetitive conduct—and their adverse micro- and macroeconomic effects — as subjects of national attention and as policy priorities within the Executive Branch and its agencies. Between 2015 and 2016, President Obama issued Executive Order 13,725 on “Steps to Increase Competition and Better Inform Consumers and Workers to Support Continued Growth of the American Economy” and his Administration issued White House Reports, a State Call to Action, CEA Issue Briefs, and Treasury Department Office of Economic Policy Reports highlighting, analyzing, and announcing policy to challenge employer monopsony and labor market competition issues, concentrating on employer collusion, pervasive use of non-competes, and the abuses of occupational licensing. During this period, the Department of Justice (“DOJ”) and Federal Trade Commission (“FTC”) also issued their Antitrust Guidance for Human Resource Professionals (“HR Guidance”) to alert regulated parties of their position that wage-fixing, no-poaching agreements, and information-sharing about wages and other terms and conditions of work can violate the antitrust laws. The DOJ only began launching investigations and filing criminal cases challenging “hard-core” wage-fixing and market allocation agreements between employers during the Trump Administration, but the Trump DOJ also intervened in private actions challenging the use of no-poaching agreements in franchising to argue that such vertical agreements should be evaluated under the rule of reason.

The Biden Administration dramatically accelerated the pace of challenging employer monopsony, starting with its July 2021 issuance of Executive Order 14,036 on “Promoting Competition in the American Economy” (“Biden EO”). The Biden EO establishes broad policy goals and a “whole-of-government” approach to combat “excessive market concentration” that “threatens” worker welfare. It focuses on the three labor market abuses that the Obama Administration had targeted: occupational licensing restrictions, no-poaching agreements, and non-compete provisions in employment contracts, “encourag[ing]” the DOJ and FTC to strengthen its HR Guidance, presumably with regard to non-competes which the HR Guidance had not addressed. The EO also encourages the FTC to engage in rulemakings on non-competes and occupational licensing restrictions and asks the DOJ and FTC to review and consider revisions to their Merger Guidelines. Finally, the EO establishes a White House Competition Council to review and take on a wider set of labor market competition issues, including implementing the EO, developing “procedures and best practices for agency cooperation and coordination,” and identifying further agency and legislative actions that might be needed to further the EO’s policies. Biden’s DOJ has since filed the first criminal no-poaching case and has taken on an active amicus strategy to clarify its legal positions in existing private enforcement actions on vertical franchising agreements and franchisor-franchisee collusion.

In all, since 2016, the antitrust agencies have shown a commitment to enforcing and engaging in policymaking against employers’ anticompetitive agreements, but they have focused on “hard-core” wage-fixing, market allocation, and no-poaching agreements, not agreements that may be subject to “quick look” or rule of reason analysis. While they have publicly stated that they intend to review the labor market effects of mergers (and did so in cursory form in the Sprint/T-Mobile merger), they have issued no clear guidance and taken no clear positions on exactly how they’ll assess the labor market effects of mergers, how they’ll balance labor and product market effects, if at all, or what theories of harm they may consider beyond hiring and wage reductions, like reduced bargaining leverage. Nor have the agencies waded into enforcing against or intervening in unlawful monopsony cases or commented on state-level antitrust reforms targeting employer dominance. Finally, the agencies have not been clear on the nature and extent of expertise they intend to bring to their policymaking and enforcement, including the methods they will apply when analyzing labor market competition issues and whether they intend to look beyond IO economic modeling and evidence to incorporate labor economics, behavioral economics, industrial relations expertise, or labor agency expertise from the National Labor Relations Board, Department of Labor, and Equal Employment Opportunity Commission.

Still, the Executive Branch has by far been the leading government regulator of imperfect labor market competition. Congress has remained on the sidelines with regard to tackling labor antitrust reforms, even if the House and Senate have put forward a number of non-antitrust reform bills to increase worker power, ease union organizing, and set higher wage floors and social benefits to strengthen workers’ outside options to monopsonistic employers. While the courts have begun to come around to labor antitrust claims against unlawful employer agreements and even employer dominance, their reasoning and analysis has been confused and incoherent at best. For example, in the context of unlawful agreements as well as merger and monopsony challenges, courts have applied a consumer welfare standard, allowing employers to successfully defend labor market restraints or product market restraints with labor market effects based on downstream consumer efficiencies. In the most recent high-profile consideration of how to review labor market restraints, the Supreme Court in NCAA v. Alston punted on deciding the question of whether and how to weigh labor market and consumer effects of labor market restraints.

All of these developments make clear that while we are in a revolutionary moment in taking on imperfect labor market competition, we are still in the early days of building an infrastructure for policymaking and enforcement and constructing a labor antitrust canon — there remains considerable uncertainty. So, what should we expect from antitrust enforcement and developing doctrine moving forward in order to most effectively ensure labor market competition?

Given our current empirical knowledge about imperfect labor market competition, even in unconcentrated markets, as well as the level of legal uncertainty surrounding labor antitrust claims, there is both a substantial risk of false negatives — or erroneous exonerations — and high micro- and macroeconomic costs of mistakenly concluding that employers’ anticompetitive exercise of their wage-setting power is not in fact anticompetitive. This heightened risk should instruct our analysis of the appropriate legal rules to apply to labor competition issues to avoid unnecessary, resource-intensive economic inquiries and reduce error costs in enforcement decisions and judicial outcomes.

III. ANTITRUST ENFORCEMENT: EMPLOYER AGREEMENTS

The antitrust agencies have been most proactive in challenging per se unlawful employment agreements to wage-fix, allocate markets, and not poach each others’ workers. Where the agencies have not taken clear or consistent positions, and where the courts have also been equivocal and even ill-reasoned, is in evaluating employer agreements that do not amount to “hard-core” unlawful conduct, like vertical agreements in the franchising context, information-sharing agreements, and non-compete agreements. With regard to these types of agreements, there remain a number of issues that could be clarified, like:

• Whether the standards for inferring agreements should be the same in labor market as in product market cases;
• Whether employers’ use of non-compete agreements can contravene the antitrust laws;
• Whether plaintiffs must define a relevant market or prove employers’ market power, especially where they present direct evidence of employers’ wage-setting power;
• Whether such agreements should be subject to “quick look” or rule of reason analysis;
• What kinds of procompetitive benefits are cognizable for labor market restraints; and
• When such agreements will be found unlawful based on balancing their anticompetitive effects and procompetitive benefits.

The choice of default rules and presumptions regarding employer agreements subject to “quick look” or rule of reason analysis should be sensitive to the empirical realities of imperfect competition in labor markets and the substantial risk of false negatives. So, for example, when inferring agreements in concentrated labor markets, we may reconsider the assumptions that have informed our current resolution of the “oligopoly problem” — in other words, erring on the side of rejecting conscious parallelism as tacit agreements unless the evidence shows more likely than not that purported colluders reached an agreement. Labor markets are unique in facing two-sided matching costs, so heterogeneous preferences by both employers and employees affect the terms of the employment bargain. Compounding this fact, search costs and imperfect information are endemic to labor markets’ functioning. So when oligopsonistic employers share information or do not compete on hiring each others’ workers through mechanisms not shared or made available to workers as well — like, say, through incorporating no-poach provisions in agreements with franchisees that employees are not parties to — they reduce their uncertainties in matching while increasing information asymmetries with workers and workers’ search costs in a way that exclusively advantages their cartel and disadvantage workers. For example, if employers aggregate even anonymized data on labor costs in a way not made available to counterparties — their workers — they can stabilize a cartel in a way that disadvantages workers and increases their monopsony power. Thus, it may be that, in those settings, tacit agreements should be presumed. Further, due to employers’ pervasive use of non-compete agreements — even in low-wage employment — and empirical evidence of their adverse effects on job mobility and wages, they should be banned or at least presumed to be unlawful.

Additionally, there are strong reasons to allow plaintiffs that present direct evidence of employers’ wage-setting power to obviate the need to define a relevant market or rely on indirect or circumstantial evidence of market power. The Supreme Court’s decision in AmEx suggested that enforcers must define the market before establishing defendants’ market power in vertical agreements cases. But its reasoning was based on the assumption that “[v]ertical restraints often pose no risk to competition unless the entity imposing them has market power.” That assumption can be rejected where enforcers provide direct evidence of employers’ wage-setting power in the form of artificially suppressed wages or reduced hiring. So a rule requiring market definition or indirect evidence of market power just adds unnecessary costs to enforcement without any offsetting benefits.

On applying “quick look” or rule of reason analysis to employer agreements, case law from Philadelphia National Bank on has been clear that courts should not engage in cross-market balancing — or weighing anticompetitive effects against procompetitive benefits in different antitrust markets. But courts have nevertheless done so in labor antitrust cases. NCAA v. Alston failed to squarely resolve the question of whether such balancing is appropriate in such cases. But for the reasons stated in both case law and the antitrust agencies’ Merger Guidelines — the complexity of comparing effects across markets, administrability challenges, and the difficulties and political judgments involved in trading off welfare between workers and consumers or between classes of consumers — courts should not balance harms to labor markets against benefits to consumers nor should they require a showing of harm to downstream consumers where harm to workers in labor markets is shown.

IV. ANTITRUST ENFORCEMENT: MERGERS

While the agencies have publicly stated their intent to review the labor market effects of mergers, they have never blocked a merger under Section 7 on grounds that it may substantially lessen labor market competition. Moving forward, key unresolved questions on the agencies’ merger enforcement concern whether and how they intend to review the labor market effects of mergers and whether they intend to amend their Merger Guidelines to formally alert regulated parties of their policy positions. There are a number of areas for the agencies to clarify, including:

• Whether they intend to maintain the same HHI thresholds for triggering merger enforcement in labor markets as they do in product markets, given the level of imperfect competition in even unconcentrated labor markets
• Whether they will engage in cross-market balancing when analyzing the anticompetitive effects and procompetitive benefits of mergers in labor and product markets; and
• If so, whether and when product market effects can trump labor market effects for the purposes of their enforcement.

There are compelling reasons for the agencies to revise their Horizontal Merger Guidelines to lower thresholds that trigger labor market enforcement because buyer power can be more harmful than seller power and wage suppression has multiplier effects impacting the entire economy. Further, the agencies should incorporate into their modeling of HHI thresholds IO and labor economics innovations in determining labor market concentration levels, including incorporating labor market frictions into traditional IO HHI analyses, looking at job-to-job flows across industries pre- and post-merger, combining flows across firms with shares of payroll, and measuring concentration based on wage bargaining models in relation to workers’ outside options.

With respect to cross-market balancing and allowing product market effects to trump labor market effects, for statutory, administrative, and public policy reasons, the agencies should block mergers that lessen labor market competition or create local monopsony power and use conduct remedies to ensure that, even where product market effects favor approving a merger, labor market competition is protected. Section 7 of the Clayton Act prohibits merger and acquisitions “in any line of commerce” where their effect “may be substantially to lessen competition, or tend to create a monopoly,” so adverse labor market effects are alone sufficient to block a merger under the statute. The agencies have made clear in their Merger Guidelines that they will “assess competition in each relevant market affected by a merger independently and normally will challenge the merger if it is likely to be anticompetitive in any relevant market,” but they reserve prosecutorial discretion to “consider efficiencies not strictly in the relevant market” if they are “inextricably linked with it that a partial divestiture or other remedy could not feasibly eliminate the anticompetitive effect in the relevant market without sacrificing the efficiencies in the other market(s).” Again, in light of the complexity and administrability concerns of engaging in cross-market balancing, current labor market realities, the high risk of false negatives, and broader federal labor and economic policy goals of ensuring equal bargaining power between employers and employees, the agencies should exercise their prosecutorial discretion to ensure that any merger or acquisition they approve does not lessen labor market competition.

V. ANTITRUST ENFORCEMENT: EMPLOYER MONOPSONY

The antitrust agencies have yet to bring an enforcement action for unlawful monopsony despite empirical evidence of its prevalence. In addition to bringing enforcement actions, the agencies could take significant steps to clear up doctrinal questions that remain thorny with regard to establishing unlawful employer dominance, including:

• The thresholds for proving employer monopsony; and
• The range of anticompetitive conduct in labor markets that can subject employers to liability.

Similar to the merger context, there are strong arguments favoring lower thresholds for establishing monopsony power in labor markets than current thresholds understood to be sufficient for establishing monopoly power in product markets. As discussed, empirical evidence of labor market inelasticity relative to product markets, institutional and social constraints on workers’ ability to switch jobs, and labor market frictions allow significant monopsony power even without barriers to entry or employer collusion. Combined, these labor market realities favor finding market share thresholds lower than 50 percent sufficient to establish monopsony power. Additionally, as in the agreement and merger contexts, the agencies and courts should review direct evidence of wage-setting power as well as innovative IO and labor economics metrics for discerning market power in their analysis, including evidence of labor market frictions that contribute to monopsony power.

Additionally, the range of conduct that may subject dominant employers to antitrust liability is highly underdetermined, and both agency and court guidance is needed. At the very least, agencies and courts should make clear that where employers unlawfully acquire or maintain their monopsony power through mergers and acquisitions, collusion with other employers through information-sharing or other schemes, using mobility and litigation restrictions in employment contracts (including employment contracts within their supply chains), restricting worker information-sharing on wages and benefits, committing work law violations, shutting down or displacing work from unionized plants and divisions in favor of non-union production, permanently replacing striking workers, or committing other actions that reduce worker bargaining leverage but that are currently lawful under the National Labor Relations Act, they can violate anti-monopsony law.

VI. GOVERNMENT ADMINISTRATION OF LABOR MARKET REGULATION

Finally, government enforcement against unlawful employer monopsony and anticompetitive conduct could substantially benefit from a stronger administrative infrastructure and expertise at the agency and interagency levels. As an initial matter, as have agencies like the Consumer Financial Protection Bureau, the DOJ and FTC should expand their expertise and stakeholder engagement — in this case, with non-IO economists, workers, and labor organizations — to inform their enforcement. They could first expand the ranks of their Economic Analysis Group and Bureau of Economics, respectively, to include a broader range of social scientific experts such as experts in labor economics, behavioral science, industrial relations, and even economic sociology and organization studies. Introducing more expansive methodologies and metrics of detecting and assessing employer buyer power will be critical for the agencies’ work in the investigative, liability, and remedial phases of antitrust enforcement. The agencies could also rely on and solicit labor market expertise and worker participation in Tunney Act proceedings as part of courts’ “public interest” review of their settlements with employer defendants. Such expertise would better inform the agencies’ use of structural and conduct remedies to ensure workers’ countervailing power. The agencies could even establish Remedial Task Forces with worker representation to monitor the administration of consent decrees. The dramatic decline of union density in the telecommunications industry following the breakup of AT&T and the Bell System — from 56 percent then to 14.3 percent today — ought to be an abiding lesson: firm breakups, or structural relief, must be fortified by aggressive government involvement maintaining workers’ countervailing power, including through creative use of conduct remedies to, for example, establish card-check neutrality agreements with labor organizations, timelines for reaching first collective bargaining agreements that bind subsidiaries and divisions, reporting requirements for work law violations like worker misclassification, posting notice of employee rights, and establishing rules for labor organization access to employer property and employee lists.

Second, while the Biden EO’s “whole-of-government” approach to combatting employer monopsony proposed specific priorities for the antitrust agencies and a role for the Labor and Treasury Departments in setting economic policy and encouraging labor market competition, it could better institutionalize and integrate labor agency involvement into anti-monopsony enforcement. The EO envisioned no role for the National Labor Relations Board (NLRB) beyond “encourag[ing]” its compliance with the EO and invited no role by its lead officials in competition policymaking, even though the NLRB is the agency tasked with ensuring workers’ equal bargaining power with employers. There is currently no memorandum of understanding establishing interagency coordination between the antitrust and labor agencies to share information, gather labor market data and analysis, or engage in joint investigations, joint enforcement, and joint policymaking regarding monopsonistic employers and their conduct or merger reviews. As I’ve argued elsewhere, the antitrust agencies should collaborate with the NLRB in their analysis of the labor market effects of mergers and in designing and monitoring remedies to employers’ anticompetitive conduct that reduces workers’ bargaining power.

To conclude, we are in a transformational moment of labor market regulation where, for the first time, antitrust enforcement has taken on the challenge of imperfect labor market competition as a core component of competition policy. While there are many doctrinal and administrative challenges and uncertainties moving forward with labor antitrust enforcement, clarifying substantive legal rules and standards, legal decision rules, and ensuring a robust and expert agency and interagency infrastructure will be critical next steps for antitrust enforcers and the courts.

1 Hiba Hafiz joined BC Law as an Assistant Professor of Law in 2018. She teaches and writes in the areas of labor and employment law, antitrust law, and administrative law.