Over the past half century, the US economy has expanded substantially, but the typical worker has captured little of the gains. Adjusting for changes in workforce composition, real hourly earnings roughly doubled between 1940 and 1970, but have barely risen since 1980 (Figure 1). Explaining this slowdown has become a central question in economics. Much of the literature emphasises technological change (Acemoglu 2002), globalisation (Autor et al. 2013), declining unionisation (Farber et al. 2018), and the real erosion of the federal minimum wage (Card and DiNardo 2002). Our recent paper (Engbom et al. 2026) highlights a complementary and less studied mechanism: a weakening job ladder has reduced workers’ opportunities to move to better-paying firms.

Figure 1 Real hourly wages in the US, 1939–2024

Measuring upward job mobility

Direct evidence on job-to-job transitions in the US back to the 1980s is limited. We therefore develop a new measure of upward job mobility using Current Population Survey data and a canonical job-ladder model. In these models, the wage distribution for all workers lies to the right of the wage distribution for workers hired from nonemployment, because workers gradually move to better-paying jobs. Figure 2 shows that this gap shrank markedly between the 1980s and 2010s. Based on this evidence, we estimate that upward job mobility fell by about half over the past 40 years. The decline is broad-based across gender, race, and education groups, and it has not been offset by greater mobility along non-wage margins.

Figure 2 Wage and offer distributions

Could this shrinking gap reflect something other than weaker upward mobility? We test three alternatives: faster wage growth within jobs, unobserved differences between recent hires from nonemployment and other workers, and measurement error in reported employment status. None changes the main result. Returns to tenure are modest and stable over time, and allowing for worker heterogeneity or classification error mainly affects the level of mobility, not its decline. Evidence from the National Longitudinal Survey of Youth points in the same direction: conditional on observables, most post-unemployment wage growth reflects movement to better jobs, and the 1997 cohort experienced both weaker upward mobility and smaller gains from job switching than the 1979 cohort.

Why has the job ladder weakened?

We next ask why the job ladder has weakened. To guide the discussion, we embed the model in general equilibrium. In this setting, a decline in upward mobility could reflect less efficient matching between searching workers and firms, lower vacancy creation, or changes in workers’ job-acceptance behaviour. We show that such factors affect the job finding rates from nonemployment and employment proportionally, whereas the data display a sharp decline in job finding from employment but only a modest fall in job finding from nonemployment. This pattern is hard to reconcile with a broad deterioration in labour demand or matching efficiency, and instead points to forces that have specifically impaired the efficiency of on-the-job search.

We first examine the role of two such forces highlighted by the literature: housing lock-in associated with changes in mortgage rates (Ferreira et al. 2010) and dual-career constraints (Costa and Kahn 2000). Estimating the model separately for homeowners and renters, and for single- and dual-career households, we find larger declines in upward mobility among renters and single households. These patterns suggest that housing lock-in or dual-career constraints are not the main cause of the declining efficiency of on-the-job search.

Motivated by recent work, we next turn to the role of rising employer concentration (Berger et al. 2023) and the increasing use of noncompete agreements (Gottfries and Jarosch 2023). Using cross-state variation, we show that states with larger increases in concentration and higher noncompete prevalence experienced larger declines in the efficiency of on-the-job search. The evidence is consistent with both forces limiting workers’ ability to shop around for better jobs. Quantitatively, our estimates suggest that rising concentration and noncompetes together may account for about 60% of the national decline in the efficiency of on-the-job search between the 1980s and the 2010s.

How much has the weaker job ladder mattered for wages?

To quantify the wage effects, we use an extension of the model with heterogeneous firms and posted wages. When on-the-job search becomes less effective, firms face less competition for workers because employees are less likely to leave for better-paying jobs. Firms therefore post lower wages, with especially large effects at large, high-paying employers. Cross-state evidence supports this mechanism: states with greater worker mobility have higher offered wages, a higher labour share, and especially high wages at large, high-paying firms.

Our preferred estimate implies that declining upward job mobility driven by observed changes in labour market structure has lowered annual real wage growth by about 0.7 percentage points between the 1980s and 2010s, accounting for roughly one-third of the post-1980 slowdown. About one-third of that effect reflects the mechanical consequence of reduced movement up the job ladder, while the remaining two-thirds arises from general-equilibrium responses as firms post lower wages when worker mobility falls (Figure 3). The largest wage effects are concentrated among large, high-productivity, high-wage firms, consistent with the fall in the firm size-wage premium over this period.

Figure 3 The impact of the decline of the job ladder on aggregate wage growth

Conclusion

Mobility to better-paying jobs is a central source of wage growth. Our results suggest that roughly one-third of the US wage-growth slowdown since 1980 reflects a decline in the rate at which employed workers receive better-paying outside offers. Rising employer concentration and the spread of noncompete agreements appear to be important contributors. Any policy agenda aimed at reviving wage growth therefore needs to take seriously the long-run weakening of competition for labour.

References

Acemoglu, D (2002), “Technical Change, Inequality, and the Labor Market”, Journal of Economic Literature 40(1): 7–72.

Autor, D H, D Dorn, and G H Hanson (2013), “The China Syndrome: Local Labor Market Effects of Import Competition in the United States”, American Economic Review 103(6): 2121–2168.

Berger, D W, K F Herkenhoff, A R Kostøl, and S Mongey (2023), “An Anatomy of Monopsony: Search Frictions, Amenities and Bargaining in Concentrated Markets”, NBER Working Paper 31149.

Card, D and J E DiNardo (2002), “Skill-Biased Technological Change and Rising Wage Inequality: Some Problems and Puzzles”, Journal of Labor Economics 20(4): 733–783.

Costa, D L and M E Kahn (2000), “Power Couples: Changes in the Locational Choice of the College Educated, 1940–1990”, The Quarterly Journal of Economics 115(4): 1287–1315.

Engbom, N, A Baksy, and D Caratelli (2026), “The Long-Term Decline of the U.S. Job Ladder”, NBER Working Paper 34981.

Farber, H S, D Herbst, I Kuziemko, and S Naidu (2018), “Unions and Inequality Over the Twentieth Century: New Evidence from Survey Data”, Working Paper.

Ferreira, F, J Gyourko, and J Tracy (2010), “Housing Busts and Household Mobility”, Journal of Urban Economics 68(1): 34–45.

Gottfries, A and G Jarosch (2023), “Dynamic Monopsony with Large Firms and Noncompetes”, Working Paper.

Gregory, V, G Menzio, and D G Wiczer (2025), “The Alpha Beta Gamma of the labor market”, Journal of Monetary Economics 150: 103695