US workers' productivity gains have not translated to wage gains since 1979

The claim

Since 1979, the American economy has grown substantially, and workers have become dramatically more productive — producing far more output per hour than their counterparts four decades ago. Yet median wages have barely moved in real terms. Proponents of the status quo argue this reflects either: (a) workers’ wages accurately reflect their marginal product and the economy has simply shifted toward rewarding capital and high-skill labor, or (b) the divergence is a measurement artifact. The structural claim — which the data strongly supports — is that the decoupling reflects deliberate institutional choices: who has bargaining power, what unions are allowed to do, how corporate governance is structured, and who sets the rules for distributing economic growth.

The mechanism

Standard economic theory predicts that in competitive labor markets, wages should track marginal product, which in turn tracks economy-wide productivity growth. For the first three decades of the postwar period, this is approximately what happened. From 1948 to 1979, net productivity grew 108% and inflation-adjusted median hourly compensation grew 93% — a compression era in which economic growth was broadly shared across the income distribution.

After 1979, the relationship severed. EPI’s analysis using Bureau of Labor Statistics data shows net productivity rose roughly 60% from 1979 to 2019 while median compensation for production and nonsupervisory workers — the bottom 80% of the workforce — rose approximately 15%. This is not a measurement dispute. It is a structural fact about the distribution of economic growth.

The mechanism of the divergence has been formally decomposed by Josh Bivens and Lawrence Mishel at the Economic Policy Institute. They identify two distinct wedges:

The price wedge arises because productivity is deflated by the GDP deflator (the price of everything the economy produces) while compensation is deflated by a consumption deflator (the price of what workers buy). When the prices of investment goods fall relative to consumer goods — as occurred during the technology-driven capital goods deflation of the 1980s and 1990s — measured “real” wages diverge from productivity even if labor’s nominal share of income is unchanged. This wedge accounts for roughly one-third of the total gap.

The inequality wedge accounts for the remaining two-thirds. This wedge measures the difference between median wages and average wages — in other words, how much productivity gains that did flow to workers were captured by workers at the top of the distribution rather than spread broadly. As executive compensation, financial sector pay, and the incomes of the top 1% and top 0.1% grew dramatically faster than median wages, average compensation rose while median compensation stagnated. The inequality wedge has no individual-behavioral explanation: it is the direct consequence of who has bargaining power and how corporate governance distributes rents.

The evidence

The pre-1979 compression era as baseline. The sharpness of the 1979 structural break is essential to understanding the gap as institutional rather than technological. From 1948 to 1979, the US economy sustained near-proportional sharing of productivity gains across the income distribution. Union density peaked above 35% in the private sector. The minimum wage was regularly updated and sat near $12/hour in 2023 dollars at its 1968 peak. Sectoral norms in manufacturing, transportation, and utilities spread union-negotiated wages to non-union workers through pattern bargaining. These institutions created what economists call a “labor-market rent-sharing” mechanism — ensuring that productivity gains translated into broadly distributed wage gains.

The divergence since 1979. Between 1979 and 2019:

• Net productivity (BLS measure, non-farm business sector): +59.7%
• Median hourly compensation, production/nonsupervisory workers (real, CPI-adjusted): +15.0%
• Had median wages tracked productivity growth, the median wage would be roughly $33/hour rather than the approximately $20/hour actually earned in 2019
• The total estimated transfer from labor to capital over this period is on the order of 3–4 percentage points of GDP annually by the late 2010s — roughly $800 billion per year in income that flows to capital rather than to broad-based wages

Labor share of GDP. The most aggregate measure of the productivity-wage relationship is the labor share of net domestic income: what fraction of total economic output goes to workers as wages, salaries, and benefits, versus to capital as profits, interest, and rent. The BLS and BEA data show the US labor share declined from approximately 67% in the late 1970s to around 58% by the late 2010s — a nine-percentage-point shift. Since US GDP is approximately $27 trillion, each percentage point represents roughly $270 billion in annual income distribution. The decline in labor share since 1979 represents one of the largest peacetime transfers of income from labor to capital in recorded American economic history.

Capital share and Piketty’s r>g in practice. Piketty’s central argument in Capital in the Twenty-First Century is that when the return to capital (r) exceeds the growth rate of the economy (g), wealth inequality grows structurally. The US after 1979 provides a near-textbook demonstration. Distributional national accounts constructed by Piketty, Saez, and Zucman show the top 1% income share doubling from 10% to 22% between 1979 and 2019, while the bottom 50% share fell from 20% to 13%. The top 1%’s gains came predominantly from capital income — dividends, capital gains, business profits — not from being exceptionally productive workers. The increase in the capital-to-output ratio in the US since 1980, documented by the Federal Reserve’s Financial Accounts, tracks this shift closely.

The decomposition and what it rules out. The Bivens-Mishel decomposition is important because it forecloses several common alternative explanations. The “skill-biased technological change” (SBTC) hypothesis, associated with Goldin and Katz, argues that technology increased the premium on high-skill workers, explaining wage inequality as a return to education. SBTC is a real phenomenon, but it cannot explain the productivity-pay gap at the aggregate level: it explains within-labor inequality but not labor vs. capital inequality. The gap between average compensation and productivity — the part attributable to capital’s rising share — has no skill-composition explanation. Workers across all education levels saw their wages diverge from economy-wide productivity after 1979.

Who benefits

Shareholders captured the bulk of the productivity-wage gap. The shift from labor to capital income since 1979 directly enriched equity holders. Corporate stock buybacks, which were legally constrained before the SEC’s Rule 10b-18 guidance in 1982 and which exploded thereafter, transferred corporate cash to shareholders rather than workers or investment. S&P 500 companies spent more on buybacks than on capital investment in most years after 2010. The Business Roundtable, the Chamber of Commerce, and the National Association of Manufacturers lobbied successfully for the Reagan-era NLRB appointments that tilted labor law enforcement, and have consistently opposed minimum wage increases, union certification rules, and card-check legislation.

Private equity, which expanded dramatically in the 1980s under the leveraged buyout model, explicitly arbitraged the gap between what workers produced and what they were paid — buying companies, cutting labor costs, and distributing the gains to limited partners. Academic research by Acharya et al. (2013) and Appelbaum and Batt (2014) documents systematic wage suppression following private equity acquisitions.

The think-tank infrastructure that provides theoretical support for the divergence — the American Enterprise Institute, Cato Institute, and Heritage Foundation — has received substantial funding from Koch Industries-affiliated foundations, financial sector donors, and large-cap corporate interests with a direct stake in maintaining capital’s share of income. The Peterson Institute for International Economics, while more centrist, has historically framed the productivity-wage gap as primarily a measurement or technology story, downplaying the institutional-bargaining-power explanation.

The counter

The most serious challenge to the structural account is the measurement critique. Robert Lawrence and others have argued that the price wedge alone — the divergence between GDP and consumption deflators — explains much of the apparent gap, and that when productivity and wages are deflated consistently, the gap narrows substantially. This is partially correct: the price wedge is real and the Bivens-Mishel decomposition acknowledges it. But it accounts for only about a third of the total gap, and it does not explain why the labor share of income fell, which is a nominal-share question not subject to deflator disputes.

A second serious argument is that benefits — employer-provided healthcare and pension contributions — grew faster than wages, and that total compensation rather than wages is the correct comparison. This is also acknowledged in the EPI analysis. When the comparison uses total compensation (including benefits) deflated by the GDP deflator, productivity still outpaced median compensation by roughly 40 percentage points between 1979 and 2019. The benefits argument is also somewhat circular: rising employer healthcare costs reflect a dysfunctional healthcare market, not a benefit to workers — workers would prefer $5,000 in wages to $5,000 in healthcare premiums that cover a less-than-comprehensive plan.

The skill-biased technological change literature (Goldin & Katz, 2008; Acemoglu & Autor, 2011) provides a genuine partial explanation for within-labor inequality growth. The returns to college education did rise after 1979, and automation has hollowed out middle-skill routine occupations. But SBTC is a complement to the structural-bargaining-power account, not a substitute: it explains the shape of wage inequality within the labor share, not the decline in the labor share itself.

References

Acemoglu, D., & Autor, D. (2011). Skills, tasks and technologies: Implications for employment and earnings. In O. Ashenfelter & D. Card (Eds.), Handbook of Labor Economics (Vol. 4B, pp. 1043–1171). Elsevier. https://doi.org/10.1016/S0169-7218(11)02410-5

Appelbaum, E., & Batt, R. (2014). Private equity at work: When Wall Street manages Main Street. Russell Sage Foundation.

Bivens, J., & Mishel, L. (2015). Understanding the historic divergence between productivity and a typical worker’s pay: Why it matters and why it’s real (EPI Briefing Paper No. 406). Economic Policy Institute. https://www.epi.org/publication/understanding-the-historic-divergence-between-productivity-and-a-typical-workers-pay/

Goldin, C., & Katz, L. F. (2008). The race between education and technology. Harvard University Press.

Lawrence, R. Z. (2015). Recent declines in labor’s share in US income: A preliminary neoclassical account (NBER Working Paper No. 21296). National Bureau of Economic Research. https://doi.org/10.3386/w21296

Mishel, L., & Kandra, J. (2020). CEO compensation surged 14% in 2019 to $21.3 million: CEOs now earn 320 times as much as a typical worker (EPI Report). Economic Policy Institute. https://www.epi.org/publication/ceo-compensation-surged-14-in-2019/

Mishel, L., & Wolfe, J. (2019). Top 1.0% of earners see wages up 157.8% since 1979 (EPI Report). Economic Policy Institute. https://www.epi.org/publication/top-1-0-of-earners-see-wages-up-157-8-since-1979/

Piketty, T. (2014). Capital in the twenty-first century (A. Goldhammer, Trans.). Harvard University Press.

Piketty, T., Saez, E., & Zucman, G. (2018). Distributional national accounts: Methods and estimates for the United States. Quarterly Journal of Economics, 133(2), 553–609. https://doi.org/10.1093/qje/qjx043

Stansbury, A., & Summers, L. H. (2020). The declining worker power hypothesis: An explanation for the recent evolution of the American economy (NBER Working Paper No. 27193). National Bureau of Economic Research. https://doi.org/10.3386/w27193