Labor Power
Labor power is the historically dominant counterweight to wealth concentration. The post-1945 Great Compression — the only sustained reduction in wealth concentration in modern economic history — coincided with private-sector unionization rates above 30% in most advanced economies. The post-1980 reversal coincided with a collapse in unionization to under 10% in the U.S. and under 20% in most of continental Europe (though density alone understates the decline; coverage and bargaining centralization fell less in some countries).
The causal evidence connecting unionization to inequality is unusually strong by social-science standards. Farber et al. (2021) use a long historical panel and find that ~1/3 of the post-1980 rise in U.S. wage inequality is attributable to declining union density. Cross-country evidence (Jaumotte & Buitron, IMF 2015) finds similar magnitudes. The mechanism is not only direct wage compression among union members but also “threat effects” on non-union employers and broader political pressure for minimum wages, employment protection, and social insurance.
Modern proposals revive labor power through three channels: traditional union-organizing reforms (PRO Act, EU directive on adequate minimum wages), sectoral bargaining (extending collective agreements to entire industries, as in much of Europe), and codetermination (worker representation on corporate boards, as in Germany and the Nordics).
Mechanisms and evidence
Section titled “Mechanisms and evidence”| Mechanism | Country examples | Evidence on inequality reduction |
|---|---|---|
| Enterprise-level unions | U.S., U.K. (post-Thatcher) | Moderate; declining coverage |
| Sectoral bargaining | Germany, Austria, Nordics, France | Strong; high coverage |
| Codetermination | Germany (1976 Mitbestimmung), Sweden | Moderate; mixed on wages, strong on rent-sharing |
| Statutory minimum wage | All OECD; meaningful U.K., Germany | Moderate at bottom of distribution |
| Living-wage ordinances | U.S. municipalities | Small but positive |
| Just-cause employment | Most of EU | Reduces top-end employer power |
The Farber et al. and IMF cross-country evidence
Section titled “The Farber et al. and IMF cross-country evidence”The empirical case linking unionization to wage compression and reduced inequality has been substantially strengthened by two pieces of work in the past decade. Both are unusual for inequality economics in the strength of their identification strategies.
Farber, Herbst, Kuziemko & Naidu (2021, QJE, “Unions and Inequality over the Twentieth Century: New Evidence from Survey Data”) is the most ambitious long-run U.S. study. The authors constructed a new microdata source on union membership going back to 1936 — primarily ~980,000 Gallup poll responses — to map the joint trajectory of union density and wage inequality across nearly 90 years. Their findings:
- U.S. income inequality has varied inversely with union density across the entire 20th century — including during the 1936–1948 collapse in inequality (the “Great Compression”) and the post-1980 rise.
- The household union wage premium is remarkably stable at 10–20 log points across radically different periods of union strength.
- Instrumental-variable strategies based on the 1935 Wagner Act and WWII War Labor Board find causal effects of unionization on inequality, not just correlation.
- Unions reduce inequality through compositional effects (more low- and middle-skill workers earning the union wage) and through threat effects (non-union employers raising wages to match).
The IMF parallel is Jaumotte & Osorio Buitron (2015, IMF SDN 15/14, “Inequality and Labor Market Institutions”), which extends the analysis cross-country across 20 advanced economies 1980–2010. Their headline finding — particularly notable coming from the IMF, which spent decades pushing labor-market deregulation — is that the decline in unionization is strongly associated with the rise in top income shares, with a magnitude meaningful enough to attribute roughly half of the post-1980 rise in advanced-economy inequality to labor-market institutional change. The IMF paper also documents the mechanism: unionization affects top-end inequality not primarily through direct wage-setting effects but through political pressure on the broader policy regime (taxation, labor protection, social insurance).
The combination of within-country long-run identification (Farber et al.) and cross-country contemporary identification (Jaumotte & Buitron) is unusually strong by social-science standards. The implication is that the post-1980 union collapse is not just a correlate of rising inequality but a substantial cause, with magnitudes that would justify policy interventions even on narrow efficiency grounds.
Why sectoral bargaining coverage matters more than density
Section titled “Why sectoral bargaining coverage matters more than density”A common confusion in U.S. labor-policy discussions is to use union density (share of workers who are union members) as the metric of labor strength. This is misleading because the more important metric in most of Europe is bargaining coverage (share of workers covered by a collective bargaining agreement, regardless of personal union membership), and the two diverge dramatically.
| Country | Union density | Bargaining coverage |
|---|---|---|
| France | ~10% | ~98% |
| Sweden | ~65% | ~90% |
| Belgium | ~50% | ~96% |
| Italy | ~32% | ~80% |
| Germany | ~16% | ~52% |
| Netherlands | ~16% | ~76% |
| United Kingdom | ~23% | ~26% |
| United States | ~10% | ~12% |
Sources: ICTWSS database; OECD Employment Outlook 2024.
The dramatic French case — 10% density, 98% coverage — reflects a specific institutional feature: collective bargaining agreements negotiated between sectoral employers’ associations and union confederations are then extended by law to all employers and workers in the sector, whether or not they are union members. The extension mechanism turns union negotiations into industry-wide minimum standards.
This matters because inequality outcomes track coverage, not density. France has lower wage inequality than the U.K. despite having a third of the U.K.’s union density, because France’s sectoral coverage produces wage compression across the entire workforce while the U.K.’s enterprise-level bargaining only affects unionized workers. The implication is that the most cost-effective labor-power intervention may be expanding bargaining coverage, not direct organizing.
The U.S. legal framework is hostile to extension — the National Labor Relations Act presumes enterprise-level bargaining and individual representation elections — which is one reason that even highly successful U.S. organizing campaigns (Amazon Labor Union, Starbucks Workers United) have produced limited inequality effects relative to their visibility. The PRO Act and analogous reforms would change this on multiple margins (see below).
Codetermination evidence
Section titled “Codetermination evidence”Codetermination — formal worker representation on corporate supervisory boards and at the workplace level — is the third major labor-power institution alongside unions and sectoral bargaining. The German Mitbestimmung system (Codetermination Act 1976; Works Constitution Act 1972) is the canonical example, with parity board representation in firms above 2,000 employees and works councils with information, consultation, and codetermination rights at the workplace.
The empirical evidence is more mixed than for unions and bargaining. The most rigorous recent study, Jäger, Schoefer & Heining (2021, NBER WP 28921, “What Does Codetermination Do?”), uses German variation in board-level representation requirements as a quasi-experiment and finds:
- Wage effects: small. Board-level codetermination produces zero to ~1.5% wage increases. No significant rent-sharing effect.
- Productivity effects: slightly positive. Small increases in productivity, slightly larger capital stocks, no reduction in revenue.
- Employment effects: stabilizing. Reduced involuntary separations (layoffs), with possible offsetting reductions in hiring.
- Profitability effects: neutral to positive. Codetermination does not measurably hurt firm performance.
The Jäger et al. result is important because it counters two opposing narratives. Conservative critics have argued that codetermination would damage profitability and competitiveness; the data say it does not. Progressive advocates have argued that codetermination would substantially compress wages; the data say the direct effect is small. The honest reading is that codetermination produces modest improvements in worker voice and job stability without measurable downsides — valuable but not transformative.
The larger effects of the German system come from the combination of codetermination with strong sectoral bargaining: codetermination alone is modest; sectoral bargaining alone produces wage compression but not workplace voice; together they form a coherent industrial-relations regime that is meaningfully more productive of equality than either component alone. This complementarity is why the U.S. PRO Act incorporates not just NLRA reform but also research mandates on codetermination feasibility.
Current legislative vehicles
Section titled “Current legislative vehicles”The most consequential current labor-reform vehicles are the U.S. PRO Act and the EU Adequate Minimum Wages Directive, with several smaller country-level efforts also active.
| Vehicle | Scope | Status (2026) |
|---|---|---|
| PRO Act (Protecting the Right to Organize) | U.S. NLRA reform | Passed House 2021; Senate filibuster blocked; reintroduced |
| EU Adequate Minimum Wages Directive | EU 27 member states | Adopted Oct 2022; transposition deadline Nov 2024 (8 states missed) |
| U.S. FTC non-compete ban | All U.S. workers | Final rule April 2024; enjoined by Texas court Aug 2024; on appeal |
| Sectoral wage councils (state-level) | California, New York, others | California Fast Food Council operating since 2023 |
| U.K. fire-and-rehire reform | U.K. employment law | Code of Practice in force July 2024 |
| German “Tariftreuegesetz” expansions | Germany federal/state | Public-procurement collective-agreement requirements |
The PRO Act would substantially expand U.S. labor rights on three margins: (1) outlawing right-to-work laws, (2) banning permanent striker replacements, (3) creating ABC-test independent contractor classification, and (4) authorizing card-check union recognition. The Senate filibuster is the binding constraint; the substantive support is roughly bipartisan but the procedural path is unclear.
The EU Adequate Minimum Wages Directive (Directive 2022/2041) is the most ambitious cross-national labor reform in a generation. It requires member states to maintain “adequate” statutory minimum wages (with the indicative benchmarks of 60% of gross median wage and 50% of gross average wage) and to promote collective bargaining coverage to 80% of workers in member states currently below that threshold. The November 2024 transposition deadline saw 8 member states miss the deadline, but the longer-term institutional pressure is meaningful — the directive establishes a one-way ratchet toward higher wage floors and broader coverage.
Interaction with antitrust via monopsony
Section titled “Interaction with antitrust via monopsony”The labor-power and antitrust agendas converge through the labor-market monopsony evidence developed in Antitrust and Competition. Azar, Marinescu & Steinbaum (2022) document that ~80% of U.S. workers face meaningful employer monopsony power (HHI > 2,500 in their local labor market), with wages depressed ~17% from the 25th to 75th percentile of concentration. Yeh, Macaluso & Hershbein (2022, AER) extend this with Census micro-data and estimate average wages at ~35% below marginal product in U.S. manufacturing.
These findings change the labor-power intervention calculus. In monopsonistic labor markets, raising the wage floor (via minimum wages or sectoral bargaining) does not produce the disemployment effects that competitive labor-market models predict — and may actually increase employment by reducing employer wage-setting power. The empirical record on minimum-wage employment effects (Cengiz, Dube, Lindner, Zipperer 2019; Card & Krueger 1995) is consistent with this prediction.
The complementarity is direct: antitrust enforcement against monopsony reduces employer wage-setting power; collective bargaining coverage raises the wage floor; codetermination provides workplace voice that constrains the conduct margin. The four-pillar strategy in Synthesis: Core Strategy treats these as coordinated rather than as substitutes.
The non-compete ban (FTC final rule April 2024) is the cleanest example of an antitrust-labor-power convergent intervention. Non-competes are a contractual mechanism for employer monopsony — they bind workers to a specific employer by prohibiting outside offers — and the FTC’s authority to ban them rests on antitrust law. The rule was enjoined in 2024 and is on appeal; whatever the final legal outcome, the analytical framework that labor mobility is a competition question has reshaped both the antitrust and the labor-policy agendas.
Objections it must clear
Section titled “Objections it must clear”Sectoral bargaining and stronger unions meet the price-of-growth objection — that labor-market “rigidity” sacrifices efficiency — and the mechanical-not-extractive objection that market wages already reflect productivity. Both assume a competitive labor market in which pay equals marginal product. Where employers have wage-setting power (monopsony), that assumption fails: wages sit below marginal product, employment below the competitive level, and collective bargaining moves both toward the efficient point rather than away from it. On this reading higher bargained wages are not a deadweight cost imposed on a well-functioning market but a correction to an existing market failure — the same monopsony the antitrust lever attacks from the other side.
Distributional effects: gender and race/ethnicity
Section titled “Distributional effects: gender and race/ethnicity”Unions are among the most racially and gender-equalizing institutions in the wage distribution, and their decline is a documented driver of widening gaps. Farber, Herbst, Kuziemko and Naidu (2021), reconstructing a century of survey data, show that unionization compressed the wage distribution and that the union premium was largest for less-educated and non-white workers; Rosenfeld’s What Unions No Longer Do (2014) traces how de-unionization since the 1970s reopened those gaps. The union wage premium is proportionally larger for Black and Hispanic workers than for white workers, so union coverage narrows the Black-white pay gap even as it lifts the floor (EPI, Gould & McNicholas 2017).
Sectoral bargaining would bite hardest where it is most needed distributionally: the low-wage care, health-support, and service occupations that sectoral mechanisms are designed to cover are disproportionately staffed by women and workers of color (BLS union and occupational data). Because these sectors are both low-paid and highly segregated by gender and race, extending bargaining coverage to them is simultaneously a wage-floor policy and an occupational-equity policy, reaching the compounded low-wage position — women, and disproportionately women of color — that a race-blind or gender-blind wage policy passes over. The point generalizes beyond the United States. The same care and domestic work sits at the bottom of labour markets worldwide and is done overwhelmingly by women, including the roughly three-quarters of the world’s migrant domestic workers who are women and who bargain from the intersection of gender and migration status (ILO 2024). The intersectional version of this design test is developed in the comprehensive strategy.