The Productivity Mirage
American workers are more productive than at any point in history. American workers have not seen meaningful wage gains in decades. These two facts are not a paradox. They are a policy choice.
There is a question that American economics education consistently fails to prepare people to ask: when productivity increases, who gets the gains? The standard introductory treatment presents productivity growth as the engine of general prosperity — workers become more productive, the economy grows, wages rise, living standards improve. This is described as a mechanism, practically a law of nature, the way economists sometimes talk about comparative advantage or diminishing returns. The history of the last fifty years is a sustained empirical refutation of the mechanism, and the refutation has barely penetrated the conversation.
Between 1948 and 1973, productivity and compensation in the United States grew at nearly identical rates. The Bureau of Labor Statistics data, analyzed by Lawrence Mishel and his colleagues at the Economic Policy Institute, shows a tight coupling: productivity up 96.7 percent, hourly compensation up 91.3 percent. This was not inevitable. It was the result of specific institutional arrangements — strong unions that could bargain for wage increases, a corporate governance culture that treated workers as stakeholders, a progressive tax structure that limited the accumulation of after-tax gains at the top, and a regulatory environment that prevented the most aggressive forms of labor market manipulation. The gains from productivity growth were distributed across the economy, not because markets required this but because political institutions structured markets to produce it.
After 1973, the coupling broke. From 1979 to 2022, productivity grew 64.7 percent. Compensation for the typical worker grew 14.8 percent. The gains from a generation of productivity growth went somewhere. The question of where is not an economic mystery. It is a political history.
The Decoupling and Its Causes
The mechanisms of the productivity-wage decoupling are, at this point, well-documented in the economic literature, even if they remain underrepresented in public economic commentary. The primary drivers are three, and they are not independent of each other.
The first is the destruction of organized labor. Union membership in the private sector, which stood at approximately 35 percent in the mid-1950s, fell to below 6 percent by 2023. The decline was not organic. It was the product of specific policy choices, legal interpretations, and corporate strategies. The Reagan administration's decision to fire striking air traffic controllers in 1981 — and the NLRB's subsequent acquiescence in a decade of increasingly aggressive union-busting — signaled to private employers that coordinated opposition to organizing was acceptable, even encouraged. The Taft-Hartley Act's prohibition on secondary boycotts had already weakened labor's ability to apply systemic pressure; the 1980s further hollowed out the NLRB's capacity and willingness to enforce labor law.
The consequence was the removal of the primary mechanism by which workers had captured productivity gains in the postwar period. Unions do not just negotiate wages for their own members; they create wage pressure that spills over into non-union labor markets, both directly (through the threat of unionization that motivates employers to pay competitive wages) and indirectly (through political power that translates into labor-protective legislation). When union density falls to six percent, both channels close. Employers can capture productivity gains because workers have no credible collective mechanism to claim them.
The second mechanism is the shift in corporate governance ideology. The shareholder primacy doctrine, articulated academically by Milton Friedman in 1970 and codified in managerial practice by the 1980s leveraged buyout wave and the proliferation of stock-based executive compensation, reorganized the objectives of the large corporation. The postwar corporation — described by John Kenneth Galbraith, critiqued by the Chicago school — treated itself as an institution with obligations to multiple stakeholders: shareholders, employees, communities, customers. The shareholder primacy doctrine held that only one obligation was legitimate: maximizing returns to shareholders. Everything else was either a means to that end or an illegitimate diversion.
The practical consequence was a dramatic shift in how productivity gains were allocated within firms. Where the postwar corporate governance culture had reinvested productivity gains in workers, in capital investment, in research, and in reserve, the shareholder primacy era redirected them toward dividends, stock buybacks, and executive compensation indexed to share price. The gains still existed; they were just flowing through a different channel, one that led to the top of the income distribution rather than to the middle.
The third mechanism is tax policy. The post-1980 compression of marginal income tax rates, the reduction in corporate tax rates, the preferential treatment of capital gains income relative to labor income — these were structural changes that, in combination, dramatically increased the after-tax returns to capital relative to the after-tax returns to labor. When a productivity gain can be captured as capital income taxed at 20 percent, or as wage income taxed at 37 percent, the incentive structure for how a firm structures the distribution of its gains is clear. Tax policy did not cause the decoupling by itself, but it amplified and reinforced the other mechanisms.
The "Skills Gap" Evasion
For thirty years, the dominant response to evidence of wage stagnation from the center and center-right of the economics profession has been the skills argument: wages are stagnant at the median because technological change has made low- and medium-skill labor less valuable, and the appropriate response is education and training that upgrades the workforce's capabilities. This is a theory with some empirical content. Technology has genuinely changed the labor market; there is real evidence for skill-biased technological change in some sectors.
What the skills argument cannot explain is why the decoupling between productivity and compensation affected the broad middle of the wage distribution — including workers with college degrees — while returns accrued primarily to the top one percent. If the premium for skill is the explanation, you would expect to see rising wages for college graduates broadly, not just for the executives and financial professionals who sit at the very top. The actual data shows that college wage premiums, after rising in the 1980s and 1990s, stagnated for the 2000s and 2010s. The gains at the top are not primarily a return to education or skill. They are a return to power — the power to negotiate the distribution of gains from a position that workers, absent collective organization or labor market tightness, do not have.
The skills argument also has a moral dimension that deserves to be named directly. It locates the responsibility for wage stagnation in the inadequacy of workers — their failure to acquire the right skills, their choices about education, their immobility across regions or sectors. This framing is both empirically dubious and politically convenient for those whose policy preferences produced the decoupling. It transforms a question about the distribution of power — who has the leverage to claim gains from productivity — into a question about individual human capital investment. The solution it implies (more education, more retraining, more workforce development) has the virtue, for its proponents, of not requiring any redistribution of economic power.
Where the Money Went
It is worth being specific about the destination of the productivity gains that did not flow to workers, because the abstraction "capital" can obscure what is actually a very concrete transfer to very specific people.
The Federal Reserve's distributional financial accounts show that the share of financial assets held by the top one percent of American households increased from approximately 30 percent in 1990 to 38 percent in 2023. The share held by the bottom 50 percent remained below 3 percent throughout this period. The mechanism was straightforward: the productivity gains that had previously been distributed through wages were instead distributed through corporate profits, which flowed to shareholders, who were overwhelmingly concentrated at the top of the wealth distribution. Stock buybacks — illegal under SEC rules until 1982, when the agency adopted a safe harbor provision under SEC Rule 10b-18 — became the primary mechanism by which corporations returned capital to shareholders rather than investing it in workers or productive capacity. In 2023, S&P 500 companies spent approximately $800 billion on buybacks, a figure that dwarfs total private-sector union wage gains in the same period.
The financialization of the corporate sector — the shift in corporate strategy from production-oriented to finance-oriented — is the single most important structural change in the American economy since 1980, and it is the change that most directly explains the productivity-wage gap. When a corporation's primary objective is the maximization of returns to financial stakeholders, and when financial stakeholders are concentrated at the top of the income distribution, the distribution of productivity gains follows directly. This is not a mystery or a paradox. It is arithmetic.
The Political Economy of the Correction
The productivity-wage decoupling is not a law of nature, and the period during which it held was not an inevitability. It was a regime — a set of political, legal, and institutional arrangements that structured the labor market and the corporation in specific ways, with specific distributional consequences. Regimes change.
There is evidence that the labor market tightness of 2021-2023, produced by the combination of pandemic-era fiscal stimulus and supply disruptions, generated real wage gains at the bottom of the wage distribution for the first time in decades. The lowest-paid workers saw their wages rise faster than the median during this period, suggesting that when labor markets are tight enough, the power balance shifts. This is an empirical confirmation of the union argument applied to labor market conditions: what matters for wage distribution is the relative bargaining power of workers, and tight labor markets produce bargaining power in the absence of unions the same way that unions produce it in the absence of tight markets.
The policy toolkit for correcting the decoupling is not secret or experimental. It consists of labor law reform that makes organizing easier and union-busting harder; corporate governance reform that reintroduces obligations to non-shareholder stakeholders; tax policy reform that treats capital income and labor income equivalently; and antitrust enforcement that limits the monopsony power of large employers in concentrated labor markets. None of these are radical proposals in international context — most wealthy democracies have some version of most of them. All of them have been vigorously opposed by the interests that have benefited from the existing arrangement, which is why they have not been implemented.
The vocabulary of economic inevitability — the skills gap, the technology premium, the global labor supply shock — serves a specific political function: it presents distributional outcomes that result from political choices as the results of impersonal forces, not amenable to policy intervention. This presentation is wrong, and the evidence that it is wrong has been available for decades. The productivity-wage gap is a policy choice that was made, through specific legislative and regulatory decisions, by specific political actors responding to specific organized interests. It is a policy choice that can be unmade through the same process, if the political will to do so develops. The first step is being honest about what the choice was.