The $1.2 Trillion Union Wage Projection Is a Hostage of Its Own Assumptions

The Economic Policy Institute has produced a number that is mathematically sound and politically weightless. Its report projects that restoring U.S. union membership to 30% — roughly three times the current 10% rate — would raise the median worker’s pay by 14.5 percent, or $7,700 annually, shifting $1.2 trillion in additional income to workers each year and reversing one-third of the rise in inequality that has occurred since 1979. The arithmetic is not the problem.

The problem is that the report’s own data reveals a self-reinforcing cycle in which the political conditions required to reach 30 percent density are prevented by the low density the projection assumes will be reversed. The report proves the effect of density on wages — union wage premiums historically range from 15 to 20 percent and lift wages for non-union workers through spillover effects — but offers no internal mechanism to produce the density itself. The $1.2 trillion figure is a hostage of its own assumptions.

How the cycle holds

The loop is straightforward and the report names its components. Union density has fallen from more than 30 percent in the 1950s to 10 percent in 2025, passing through 22.2 percent by the 1980s. The decline is attributed to “aggressive corporate opposition and anti-union laws,” including right-to-work legislation in 27 states. Low density weakens labor’s political power, which allows anti-union laws to persist and expand, which keeps density low. Each failed organizing drive reinforces the perception that unions are not viable in a given workplace. Each legislative effort that stalls reinforces the political power of the status quo.

The cycle compounds through a legislative ratchet. Right-to-work laws, once passed, reduce the revenue streams unions use to fund organizing and political activity by prohibiting agency fees. A union organizer in a right-to-work state faces a specific bind: the workers who most need representation are the ones least able to fund it. The states with the lowest density are politically the least likely to repeal the laws keeping density low. This is not a malfunction; it is the system’s self-preservation mechanism.

The same corporate actors whose opposition the report identifies as driving low density also fund the political infrastructure — lobbying expenditures, campaign contributions — that sustains the legislative conditions keeping density low. The material position of the beneficiaries funds the political mechanism that maintains their advantage.

Where the pathway breaks

The report’s policy roadmap is not unserious. It calls for passing the Protecting the Right to Organize Act, revoking right-to-work laws, mandating collective bargaining at companies where the CEO-to-worker pay ratio exceeds 100 to 1, guaranteeing bargaining rights for public-sector workers through the Public Service Freedom to Negotiate Act, and requiring annual raises for newly unionized workers. Each recommendation targets what the report identifies as the structural barrier.

But the report’s own estimate of the most readily achievable reform discloses the loop’s grip. Revoking right-to-work laws and restrictions on public-sector bargaining alone would raise density from 9.9 percent to 14.4 percent — a 4.5-percentage-point gain. The 27 right-to-work states are the states where anti-union political constituencies are strongest and where union revenue to fund organizing has been structurally reduced by decades of those same laws. The remaining 15.6 percentage points require legislative majorities for the PRO Act, sustained organizing capacity in sectors where density has collapsed, and corporate behavioral change — conditions the report does not model as endogenous to the political economy it describes.

The PRO Act has passed the House in multiple Congresses. It has never cleared the 60-vote threshold required to overcome a Senate filibuster. Its co-sponsor count plateaued at 42 — and has not moved. The institutional gap between the report’s recommendations and the political capacity to enact them is not a footnote. It is the central finding the data produces when read against the report’s own framework.

Even if enacted, the PRO Act faces an enforcement vulnerability the report’s framing sidesteps. Enforcement depends on a National Labor Relations Board weakened by funding cuts and hostile judicial appointments. The legal right to organize does not translate into bargaining power if the enforcement apparatus is hollowed out. Legal challenges would delay implementation for years, and the Supreme Court’s current composition makes enforcement uncertain. The risk is acute: without enforcement, the Act functions as a political win that fails to deliver density gains, and a second failed attempt at labor-law reform would further entrench the low-density cycle.

The CEO-to-worker pay ratio mandate is gameable. Firms can restructure compensation through stock buybacks, contractor arrangements, equity deferrals, or stock options to manipulate the measured ratio, or reclassify workers as independent contractors. The gaming mechanisms are known from tax-evasion analogues. The very attempt to avoid the trigger would increase transparency around pay ratios, which may have independent organizing effects, but as a density-raising mechanism the mandate carries high implementation risk.

Who holds the power

The stakeholder map produced by the situation is starkly asymmetric. Large corporations — the parties the report identifies as the primary opponents of organizing — occupy a high-power, high-interest position. They bear the direct cost of higher wages and stronger bargaining, and they have demonstrated organizational capacity to act: corporate spending on union-avoidance consultants and captive-audience meetings is a documented feature of modern organizing drives at firms including Amazon, Starbucks, and Trader Joe’s.

The wealthiest 0.1 percent occupy a structurally aligned position through capital allocation — high power, low urgency, because the current equilibrium serves them. Robert Reich, former U.S. secretary of labor, writes in the report’s foreword: “By making it harder and harder for workers to organize and bargain collectively, the rich seized more and more income and wealth, destroying the US middle class.” He adds that the wealthiest 0.1 percent now own more than five times the combined wealth of the entire bottom half of the country — a concentration ratio identifying the funding mechanism for the anti-union campaigns that maintain low density.

Congress holds sole jurisdiction over every legislative lever the report recommends, placing lawmakers in the high-power quadrant. But their interest is mediated by competing pressures: corporate lobbying and campaign contributions sustain the legislative blockade, while constituent demand for organizing access remains diffuse and structurally unconverted into legislative consequence. The Senate filibuster is the institutional gate.

The 50 million workers who report they would join a union if given the opportunity sit in the opposite quadrant — high interest, low structural power. Public approval of unions exceeds 68 percent. But approval and the organizational capacity to convert approval into bargaining power are different things. The gap between 68 percent public support and 10 percent actual density is the signature of a system in which the barrier to organizing is institutional, not attitudinal.

The AFL-CIO occupies a broker position between willing workers and the political system. Liz Shuler, the federation’s president, described the worker sentiment driving the demand signal: “I can’t tell you how many conversations I’ve had with workers, no matter where you go — big city, small town — who basically are saying over and over again: ‘My rent keeps going up, my paycheck does not stretch as far as it used to, I walk into the grocery store and I ask myself, when did shit get so expensive?’ It is just a constant.” But a broker’s leverage derives from the constituency it represents, and at 10 percent density the AFL-CIO’s mobilizational capacity — strike coordination, political endorsements, ground-level organizing infrastructure — is structurally constrained by the same decline it seeks to reverse. The federation needs more workers than workers need the federation.

Why the structural-decline account fails

An alternative reading attributes declining density to deindustrialization, global competition, service-sector and gig-work expansion, and worker choice in a more dynamic labor market. Under this account, legal reform alone cannot restore density because the underlying economic structure has shifted — the jobs most amenable to traditional workplace organizing have shrunk as a share of employment, and the sectors where employment is growing are structurally harder to organize regardless of legal environment.

This framing collapses on its own logic. If deindustrialization were the decisive driver, density would not rise in any state after legal reform — yet the report’s own data shows that revoking right-to-work laws and public-sector bargaining restrictions would raise density from 9.9 percent to 14.4 percent, a measurable gain the structural-decline account cannot explain. The 68 percent approval and 50-million-willing demand signal further undermine the reading: if workers had simply chosen non-union employment, approval would not have risen to 68 percent while density fell to 10 percent. The gap is institutional, not attitudinal.

The structural-decline account also ignores the causal mechanism the report’s own evidence discloses. The self-reinforcing cycle — low density suppresses political power, which sustains anti-union law, which keeps density low — is not an argument about whether workers want unions. It is an argument about whether the political system permits them to have them. The structural-decline reading treats the symptom — declining density in certain sectors — as the cause, while the suppressed-demand reading identifies the mechanism — political suppression — that produces the symptom.

Both framings agree that the 30 percent target is far out of reach under current conditions. The structural-decline account says the target is unachievable because the economy has shifted. The suppressed-demand account says the target is unachievable because the political economy prevents the legal changes required to reach it. The practical conclusion is the same: the $1.2 trillion figure is a measure of lost potential, not a policy forecast. But the causal mechanism matters. If the barrier is structural, no legal reform can solve it. If the barrier is political, legal reform can solve it — if the political conditions can be created. The report’s own data resolves this question: the 9.9-to-14.4-percent gain from legal reform alone is measurable, which means the barrier is at least partially political. The remaining 15.6 percentage points are the question the report cannot answer.

What the frame excludes

Several parties whose outcomes are shaped by density policy sit outside the report’s accounting.

Gig-economy workers, classified as independent contractors and excluded from NLRA coverage, likely make up a portion of the 50 million workers who say they would join a union if able. The report’s policy recommendations do not address the classification barrier that prevents them. Under the suppressed-demand frame, their inclusion inflates the addressable organizing pool: the 50 million figure includes workers who cannot legally join a union under current law, making the demand signal weaker than the headline implies.

Incarcerated workers — roughly 1.2 million people explicitly exempt from the NLRA under the 13th Amendment’s exception — represent a growing share of low-skill manufacturing labor whose exclusion from any wage projection constitutes a structural gap. Their exclusion makes the $1.2 trillion figure an upper bound, not a central estimate.

Downstream consumers, who would bear some portion of $1.2 trillion in additional labor costs through price adjustments, are absent from the report’s accounting. The wage projection does not account for the price pass-through that would offset a portion of the worker income gain.

Small and mid-size employers, whose thin margins create different competitive pressures than large corporations face, are folded into a unitary “corporate opposition” category that obscures the variation within it. The suppressed-demand frame assumes higher mandated labor costs can be absorbed by firms with pricing power, but small and mid-size employers may close rather than absorb cost increases, reducing the employment base the density projection depends on.

The judicial system operates as a background condition that can shift the pathway for every other party through NLRB-precedent narrowing and constitutional challenges to union-security clauses, without being directly mobilized by any of them. Political parties, as coalition-management bodies distinct from Congress as an institution, mediate between electoral incentives and legislative action in ways that dilute pressure to act.

These absences do not invalidate the report’s central finding that union wage premiums are real and that density correlates with inequality reduction. They constrain the projection’s completeness as a guide to what tripling density would mean across the economy, and they narrow the $1.2 trillion headline figure toward a more realistic ceiling.

The probable future

Scenario projections weighted for probability suggest the most likely outcome is not the report’s headline figure but something considerably smaller.

A Sticky Floor scenario — density drifting in the 10-to-14-percent range through 2040 as the self-reinforcing cycle holds — carries a 50-to-58 percent probability. Median wage gains would reach $2,000 to $4,000 annually, far short of the $7,700 headline. Downside risk is real: density could fall to 8.5 percent by 2028 if employer anti-union spending intensifies and a Supreme Court ruling narrows NLRB jurisdiction over worker-classification disputes; the PRO Act could stall in committee; right-to-work expansion could spread to additional states. Recoverability from that position is low — rebuilding density from a lower base requires more political capital than maintaining it, and the wage gains that would anchor broader public support are absent.

A Legislative Inflection scenario — the PRO Act passes and right-to-work laws are revoked in key states, with the 15-to-20 percent union wage premium lifting non-union wages through spillover effects — carries a 27-to-32 percent probability and would produce density of 18-to-25 percent by 2035, with median wage gains of $5,000 to $10,000. A Platform-Union Model — digital-first organizing outside traditional NLRB election structures — carries a 20-to-30 percent probability but has no historical precedent for scale. A Cascading Organizing Wave triggered by a recession or sustained inflation shock that activates latent demand from 50 million willing workers carries 12-to-15 percent probability. A Federal Labor-Law Overhaul from a political realignment sits at 5-to-15 percent.

The probability-weighted central tendency across scenarios yields an illustrative annual wage transfer of roughly $400 to $600 billion — approximately one-third of the headline projection. That figure applies the report’s stated density-wage methodology to the probability-weighted median density outcome; it is not a sourced EPI figure, and it depends on the report’s elasticity assumptions, which the available source summary does not fully specify.

The pre-mortem failure pathways are sobering. In one, corporate anti-union spending intensifies, the PRO Act stalls in a divided Senate, and density drifts below 10 percent by 2032 — the $1.2 trillion projection becomes a textbook counterfactual, economically rigorous and politically inert. In another, the PRO Act passes but implementation is delayed by corporate litigation for years; density barely moves from 10 percent to 13 percent; the wage projection stalls at roughly $400 billion, meaningful but a third of the headline figure. In a third, a recession causes mass layoffs, organizing drives stall, and density drops back below 10 percent by 2031.

Each pathway demonstrates that the partial-density scenario carries a specific risk: a small density increase may be absorbed by the existing structural loop without disrupting it, producing just enough wage effect to relieve political pressure for deeper reform while leaving the underlying power asymmetry intact.

The thesis

The $1.2 trillion union wage projection is analytically sound as a counterfactual and politically inert as a forecast. The report’s own data reveals a self-reinforcing cycle in which the political conditions required to reach 30 percent density are prevented by the low density the projection assumes will be reversed. The report proves the effect of density on wages but offers no internal mechanism to produce the density. The probability-weighted outcome — roughly one-third of the headline figure — renders the full projection a measure of lost potential, not attainable gain.

The 50 million workers who would join a union if able and the 68 percent public approval represent genuine demand. The institutional barriers to converting that demand into organizing capacity are the same barriers that suppressed density over the past half-century. The report’s policy roadmap — the PRO Act, right-to-work revocation, CEO-to-worker pay-ratio mandates — is the correct diagnosis, but the political economy it describes is the reason the diagnosis cannot be enacted.

The gap between what workers want and what the labor market delivers is not a market failure. It is a political failure — one that the report documents with precision and the political system reproduces with equal precision. Until the self-reinforcing cycle is broken, the $1.2 trillion figure remains a measure of what was lost, not what can be recovered.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Stakeholder Mapping

Charts the parties to a situation — their interests, power, and alignments.

Wicked Futures

Explores a long-horizon, deeply entangled future with no clean resolution.

Wicked Problems

Treats a problem as wicked — no stopping rule, no clean test of success, every attempt consequential.