Unexpected Compression: Competition at Work
David Autor | Episode 106
David Autor joined Markus’ Academy to discuss his recent paper (Autor, Dube and McGrew, 2024). Autor is the Ford Professor in the MIT Department of Economics.
Watch the full talk below and download the slides here.
Highlights
[0:00] Markus’ Introduction. The recent decline in the skill premium could reflect different forms of compensation between blue- and white-collar workers. For blue-collar workers, higher compensation may have taken the form of higher wages, while white-collar workers may have been compensated with greater flexibility in working hours or increased working from home (non-pecuniary compensation). There are also labor supply–side forces such as Long COVID contributing to labor shortages. The evolution of the skill premium might also interact with inflation dynamics. When the economy recovers and workers gain bargaining power, higher interest rates can be used to contain a price–wage spiral. This can be particularly costly for low-income, high–marginal-propensity-to-consume households. On the other hand, higher interest rates can hurt growth/tech stocks and thus higher wage earners, which could also reduce the skill premium. In Germany’s high-inflation period leading into the early 1920s, low-skill workers’ real wages increased relative to higher-skill workers’ wages, but this differentiation disappears under hyperinflation.
[6:04] What does a competitive labor market look like? The post-pandemic labor market has been unexpectedly tight. The question is how this tightness has affected labor market competition. The textbook model of perfect competition is static, with wages adjusting to the marginal product of labor. Empirically, however, employers do not appear to face a perfectly elastic labor supply curve, implying that similar workers can be paid different wages. It is important to distinguish labor market tightness under perfect competition versus imperfect competition. Under perfect competition, increases in tightness reflect labor demand shifting out relative to labor supply (or labor supply shifting in relative to demand), moving the market from one Pareto-efficient equilibrium to another without inherent normative implications. Under imperfect competition, tightness can make labor supply to firms more elastic, reducing employer market power and reallocating workers from less productive to more productive firms, increasing efficiency.
[13:14] Some unexpected facts. Labor force participation and employment rates plummeted during the pandemic, but employment-to-population has almost fully rebounded, and labor force participation has recovered substantially; this is broadly true across education and wage levels. Low-wage, low-education workers experienced the largest drops in labor force participation and employment, but also the steepest recovery. There has been substantial real wage growth over the last 36 months (CPS data through Sep 2022; 36-month window Jul 2019–Sep 2022, annualized), particularly below the median; but in the last 12 months, real wage growth occurs only for the bottom 15% of the distribution. Real wages at the 10th percentile have grown faster than wages at the 50th or 90th percentiles since the onset of the pandemic, compressing the wage distribution. This has coincided with a fall in the Black/Hispanic wage penalty and stronger wage growth among younger workers. The strongest post-pandemic wage gains have come from younger, high-school–educated workers.
[22:40] Distinguishing rising demand from increasing competition: conceptual model. In a perfectly competitive market, labor supply to the firm is perfectly elastic, so when labor supply shifts inward (left/up), the firm raises wages and employment falls. In an imperfectly competitive market where firms face an upward-sloping labor supply curve, the supply curve can rotate and become more elastic, which increases wages. How employment responds depends on the firm’s wage policy: at low-wage firms, a rotation toward greater elasticity can decrease employment, whereas at high-wage firms, employment can increase. The implication is that a more elastic labor supply curve can reallocate workers from low-wage to high-wage employers. Potential reasons labor supply may have become more elastic include: involuntary separations during the pandemic reducing worker–firm attachment; greater household liquidity facilitating job changes by allowing workers to absorb temporary income fluctuations; social learning from coworkers and friends finding better jobs increasing beliefs that better jobs are available; and formal implications from canonical job-ladder models. These models predict that as demand rises or unemployment falls, employment-to-employment (EE) transitions rise, especially at the bottom of the wage distribution, implying that tightness can change the elasticity of labor supply that employers face.
[36:56] Distinguishing rising demand from increasing competition: evidence. Monthly job-to-job transition rates have been higher post-pandemic for young, high-school–educated workers, while remaining relatively stable for older workers and workers with a college degree. Tightness is defined using a function of EE separation rates and the unemployment rate; both components increased sharply post-pandemic. Wage Phillips curves estimated using state-level variation in tightness suggest that wages for workers in the first quartile and for young, high-school–educated workers increased the most in response to increased tightness. Separation-elasticity estimates point in a similar direction, with a disproportionate increase among young, low-skilled workers, though this evidence is suggestive rather than definitive at conventional statistical thresholds. Wage gains are much larger among those who change jobs than among those who stay. The increase in wages for young high-school workers reflects both increased switching rates and increased wage gains conditional on switching. Additional suggestive evidence indicates this wage growth is driven by an increased likelihood that young, high-school–educated workers leave low-wage jobs.
[1:03:05] How much does wage pressure contribute to inflation? Price Phillips curve estimates suggest labor market tightness contributed roughly a 1 percentage point increase in post-pandemic inflation, presented as an implied magnitude rather than a tight causal decomposition. This is about the same magnitude as the effect of tightness on average wage growth, even though tightness is associated with much larger wage growth for young, low-skilled workers.
[1:07:33] Conclusions. For the first time in decades, wage inequality is falling. Real wages are rising among young, low-skilled workers and workers at the bottom of the wage distribution. While it is tempting to attribute the change solely to tighter labor markets, that may be an oversimplification. Evidence suggests competition has intensified, and rising competition implies wages that better reflect productivity and higher aggregate productivity through reallocation to better employers.