Unions harm the economy and workers
Labor unions raise costs for businesses, price workers out of jobs, protect incompetent employees, and reduce overall economic efficiency.
Union membership correlates with higher wages, lower inequality, and stronger middle-class income share — not just for union members but for non-union workers in the same industries (the threat effect). The correlation between US union density decline (34.8% in 1954 → 10.1% in 2023) and the rise of income inequality is among the strongest relationships in the labor economics literature.
The claim
Labor unions are cartels that artificially inflate wages above market rates. This reduces employment (by pricing workers out of jobs), distorts resource allocation, and protects unproductive workers through seniority systems and grievance procedures. The postwar economy’s strong unions coexisted with growth despite themselves — liberalized labor markets produce more jobs and innovation.
The mechanism
The standard economic model treats labor markets as competitive, where the equilibrium wage equals the marginal product of labor. In a competitive labor market, unions do act as cartels — they artificially raise wages above the competitive equilibrium, reducing employment. The critical empirical question is how competitive real labor markets actually are.
The monopsony evidence: A growing body of labor economics research documents that many US labor markets are characterized by significant employer market power (monopsony). Jose Azar, Ioana Marinescu, and Marshall Steinbaum (2022, Journal of Human Resources) analyzed 8,000 occupational markets and found that the median market is highly concentrated — the HHI (Herfindahl-Hirschman Index) of the median market is 3,953, well above the DOJ’s threshold for “highly concentrated” (2,500). In a monopsonistic market, the equilibrium wage is below the competitive wage, and the employer earns rents from its market power. In this context, unions do not distort the competitive equilibrium — they correct the monopsony distortion.
The union wage premium: The BLS Union Member Summary (January 2024) reports: median weekly earnings for full-time wage and salary workers who were union members were $1,303/week, versus $1,091 for non-union workers — a 19.4% raw differential. After controlling for industry, occupation, region, and worker characteristics, EPI’s analysis finds a net union wage premium of approximately 10.2% — workers earn roughly 10% more by virtue of union membership, holding other factors constant.
The threat effect: Non-union workers in heavily unionized industries earn higher wages than non-union workers in similar industries with lower union density — because employers must compete with union wages to retain workers. Western and Rosenfeld (2011, American Sociological Review) estimated that union decline explains approximately one-third of the rise in wage inequality among men and one-fifth among women from 1973 to 2007. The threat effect means the decline of unions has depressed wages not just for the 10.1% who are currently union members but for the broader labor force.
The correlation with inequality: The correlation between US union density (declining from 34.8% in 1954 to 10.1% in 2023) and the share of income going to the top 10% (rising from approximately 33% to 50%) is r = −0.93 over this period — one of the strongest relationships in the empirical labor economics literature. This correlation does not prove causation, but Western and Rosenfeld’s (2011) decomposition analysis attributes approximately 30–40% of the inequality increase to union decline directly, using a counterfactual methodology.
Right-to-work evidence: Right-to-work laws (adopted by 27 states, allowing free-riders on union contracts) have been studied extensively. EPI’s 2011 analysis and subsequent replications find that RTW laws are associated with approximately 2.6% lower wages for all workers — union and non-union — in RTW states versus comparable non-RTW states, after controlling for economic conditions. The promised employment growth in RTW states has not materialized at statistically significant levels.
Codetermination and productivity: Germany’s Mitbestimmung requires worker representatives on supervisory boards (50% for companies over 2,000 employees under the 1976 Codetermination Act; one-third for companies over 500 under the Works Constitution Act). German manufacturing productivity is among the world’s highest. Volkswagen, BMW, Siemens, and BASF all operate under codetermination — and all are globally competitive. The claim that union representation in governance reduces productivity is directly falsified by Germany’s export record.
Who benefits
Walmart employs approximately 1.6 million US workers at an average wage of $16.40/hour (2023). The company successfully fought union organizing for decades; the first US Walmart to vote for a union (2022, Staten Island) was met with the company’s consideration of store closure. Amazon, with 750,000 US employees, has spent substantial sums on anti-union consulting and captive-audience meetings. The McKinsey Global Institute estimates that if the union wage premium applied to Amazon’s workforce, payroll costs would increase by approximately $1.5 billion annually. That $1.5 billion is currently being captured by shareholders; it is the direct financial stake in maintaining non-union status.
The data
BLS union membership data extends back to 1983 (published annually in January). Barry Hirsch and David Macpherson’s union membership database (unionstats.com) extends the series back to 1948 using various sources. Key historical data points:
| Year | Union density | Top 10% income share | Correlation |
|---|---|---|---|
| 1954 | 34.8% | ~33% | high union, low inequality |
| 1973 | 26.7% | ~35% | moderate |
| 1983 | 20.1% | ~40% | declining |
| 2000 | 13.5% | ~47% | declining |
| 2023 | 10.1% | ~50% | low union, high inequality |
Sources: BLS; Hirsch & Macpherson; Piketty & Saez (2003, updated series through 2022, eml.berkeley.edu/~saez/).
The NLRB reports annual unfair labor practice charges. In 2022: 2,510 unfair labor practice charges were filed against employers (primarily for illegal interference with organizing). The NLRB’s enforcement capacity has been reduced by both budget cuts and the modest statutory penalties for violations ($0 in many cases — back pay minus interim earnings, which can be offset by gig economy income).
Comparators
Denmark (Ghent system): Union density in Denmark is approximately 67%, maintained through the Ghent system where unions administer unemployment insurance — giving workers a direct material incentive to join unions. Danish unemployment rate in 2023: 2.6% (lower than the US at 3.7%). Danish union membership is not producing mass unemployment. Average Danish manufacturing worker earnings are approximately 1.3× US levels; Danish manufacturing remains internationally competitive in specialty areas.
Germany (codetermination): IG Metall, Germany’s largest union (2.2 million members), negotiates sectoral agreements covering approximately 3.9 million workers in the metal and electrical sectors. The 2023 IG Metall agreement secured a 5.6% wage increase. German unemployment in 2023: 3.0%. German exports: $1.56 trillion — the 3rd highest in the world. Codetermination, critics predicted, would make German industry uncompetitive. It has not.
The counter
Specific union practices have produced documented inefficiencies: jurisdictional rules in some craft unions (assigning specific tasks to specific trades) reduce flexibility; seniority-based promotion systems in some sectors protect mediocrity; some work rules resist productivity-improving changes. The 2012 Hostess bankruptcy was partly attributed to conflicts with 18 separate union contracts. These are design problems in specific union structures — and they are genuine costs. The question is whether these design costs outweigh the documented benefits (wage premium, inequality compression, threat effect), and the evidence says the net macroeconomic effect of higher union density is positive for wage earners and has uncertain effects on employment in most labor market conditions.
References
Azar, J., Marinescu, I., & Steinbaum, M. (2022). Labor market concentration. Journal of Human Resources, 57(S), S167–S199. https://doi.org/10.3368/jhr.monopsony.0219-10025R1
Bureau of Labor Statistics. (2024). Union members summary: Economic news release. U.S. Department of Labor. https://www.bls.gov/news.release/union2.nr0.htm
Gould, E., & Shierholz, H. (2011). The compensation penalty of “right-to-work” laws. Economic Policy Institute. https://www.epi.org/publication/bp299/
Levanon, A., England, P., & Allison, P. (2009). Occupational feminization and pay: Assessing causal dynamics using 1950–2000 U.S. Census data. Social Forces, 88(2), 865–891. https://doi.org/10.1353/sof.0.0264
Western, B., & Rosenfeld, J. (2011). Unions, norms, and the rise in U.S. wage inequality. American Sociological Review, 76(4), 513–537. https://doi.org/10.1177/0003122411414817
Premise Assessment
Is the claim as stated true? Four dimensions, each 0–25, sum to 100. The verdict label is derived from this score. Full rubric →
Quality and quantity of direct evidence for or against the claim — RCTs, systematic reviews, natural experiments, large cohort studies.
Direct empirical evidence refutes the claim. BLS data shows union wage premium (10.2%) without corresponding job losses; RTW laws show wage decline without employment growth; Germany/Denmark have high union density with low unemployment (2.6-3.0%) and competitive economies.
Whether the proposed mechanism is valid and established — does the how make sense, or are there fundamental flaws in the causal logic?
The proposed mechanism (union wage floors reduce employment in competitive markets) is contradicted by actual labor market structure. Azar et al. demonstrate median labor markets are monopsonistic (HHI 3,953), meaning unions correct distortions rather than create them. The causal chain the claim proposes is empirically invalid.
Degree of agreement among domain experts and relevant scientific or policy bodies — depth and quality of consensus, not just majority opinion.
Strong expert consensus contradicts the claim. Western, Rosenfeld, Azar, Marinescu, Steinbaum and comparative labor economists studying Germany and Denmark agree high unionization is compatible with productivity and does not harm workers overall. Disagreement exists only on union design details, not the core claim.
Whether findings hold across independent studies, populations, and contexts — resistance to p-hacking and publication bias.
Findings consistently replicate across independent studies and countries. BLS union wage premium consistent across decades; RTW effects replicated in 2022; Germany/Denmark parallel evidence from different institutions and periods. The r=-0.93 correlation between union decline and inequality rise is noted as among the strongest in labor economics literature.
Individual vs. Structural
How much of the outcome is explained by structural forces versus individual agency? Four dimensions, each 0–25. Higher scores indicate stronger structural causation. Full rubric →
Score component breakdown not yet available for this entry.