Does Raising the Minimum Wage Kill Jobs?

Three economic frameworks predict different outcomes. In the competitive model (Stigler 1946), a binding wage floor above the market-clearing rate reduces labor demand, producing unemployment. In monopsony/search-friction models (Manning 2003; Card & Krueger reinterpretation), employers have wage-setting power, so a higher floor can raise pay with little employment loss or even gains. In hybrid/dynamic adjustment, firms respond on multiple margins: prices, hours, automation, worker composition, and productivity. Which framework applies depends on the local labor market.

The federal minimum wage has been $7.25 since July 24, 2009, the longest period without an increase in its history. Using CPI-U, the 1968 federal minimum of $1.60 was worth about $15.02 in February 2026 dollars. Today’s $7.25 is roughly 48% of the 1968 peak. Its real value is lower than at any point since 1949, when the floor was $0.40; the $0.75 set in January 1950 was worth about $10 in today’s money, and the floor has not been worth less than $7.25 since. In 2024, BLS counted about 82,000 hourly workers paid exactly $7.25 and another 760,000 paid less, together 842,000 or 1.0% of hourly workers, partly because 30+ states now set higher floors.

Sources: DOL FLSA wage history; BLS CPI-U via FRED; BLS Characteristics of Minimum Wage Workers 2025.

Source: DOL nominal series; CPI-U deflator via FRED. Real values in February 2026 dollars (CPI-U 326.785), deflating each year’s wage by that year’s annual-average CPI-U.

Card & Krueger (1994) surveyed ~410 fast-food restaurants in New Jersey and Pennsylvania around NJ’s April 1992 increase from $4.25 to $5.05. Using difference-in-differences with PA as a control, they found no employment decline and possibly a small increase. This paper changed the field because it replaced older national time-series with a credible natural experiment. Neumark & Wascher (2000) reanalyzed with payroll records and found small negative effects; Card & Krueger (2000) replied using BLS administrative data, again finding no systematic loss.

Dube, Lester & Reich (2010) compared all contiguous U.S. county pairs straddling state borders from 1990–2006, especially in restaurants. This “border-county” design controls for local economic conditions that national panels miss. Result: strong earnings gains with no detectable employment loss. Later critique by Jha, Neumark & Rodriguez-Lopez (2024) using commuting zones instead of county pairs flipped the sign negative, illustrating how sensitive results are to control-group choice.

Cengiz, Dube, Lindner & Zipperer (2019) used a bunching estimator across 138 state-level increases (1979–2016) and found that jobs below the new minimum disappeared but were largely offset by jobs bunching at or just above the new floor, leaving overall low-wage employment essentially unchanged over five years.

Jardim et al. (University of Washington; NBER working paper 2017, revised October 2017; AEJ: Economic Policy 14(2):263–314, 2022) studied Seattle’s phased increases using Washington State administrative payroll records. For the $13 phase the revised working paper found wages up about 3% and hours down 6–7% in low-wage jobs, lowering pay by about $74 a month per job; the first June 2017 draft had said 9% and $125, and those figures still circulate. The published version reports aggregate employment elasticities from −0.2 to −2.0, concentrated in hours rather than headcount and largest among inexperienced workers, and says in its own abstract that the aggregate analysis likely overstates the employment effect because some of the decline is jobs moving above the wage threshold rather than disappearing. Restaurant headcount was unaffected, but total hours in low-wage work fell. The Berkeley/CWED team looked at food services specifically and found no effect: illustrating that the answer depends on sector choice, control group, and whether you measure headcount vs. hours.

Sosinskiy & Reich (2025) studied California’s $20 fast-food minimum wage, which took effect in April 2024 for chains with 60 or more locations. In the May 2025 version of their IRLE paper, using administrative wage and employment data, Glassdoor pay reports and scraped menu prices from over 2,000 restaurants, covered workers’ average weekly wages rose 12% with no reduction in employment, and prices rose 2.1% two quarters after the policy, 8 cents on a $4 item, with around 60% of the higher labour cost passed to consumers. Earlier drafts had put the wage gain at 18% and the price rise at 3.7%, so the figures have moved as data accumulated. Dube (2026, NBER working paper 35171) revisited the same policy with Quarterly Workforce Indicators and found a sharp fall in separations alongside the wage gains, which he uses to reconcile somewhat more negative employment estimates in the QCEW data.

This is the core methodological issue: national/state panels (Neumark-style) capture correlated shocks and often find negative effects. Border-county/local designs (Dube-style) better control local conditions and usually find near-zero effects, but may miss spillovers. Magnitude matters enormously: modest, gradual hikes show near-zero effects; large, rapid hikes (Seattle $13+) show more negative results, especially on hours. Outcome choice matters: headcount employment can look stable while hours fall. Who you study matters: teens and entry-level workers show more consistent negative effects than the overall workforce.

Sources: Card & Krueger (1994), Dube/Lester/Reich (2010), Neumark & Wascher (2007), Cengiz et al. (2019), Jardim et al. (2022), Dube OWE (2024). “Elasticity” = % change in employment per % change in wages induced by policy.

Minimum wage increases reliably raise wages for affected workers. Real wage growth at the 10th percentile has outpaced higher wage groups since 2019, especially in states with hikes. But the poverty effect is more complicated: many minimum-wage workers are not in poor households (teens, secondary earners), and many poor households have no workers at all.

Dube (2019, AEJ:Applied Policy) found long-run minimum wage elasticities of the non-elderly poverty rate ranging from −0.22 to −0.46, real but modest reductions. CBO (2023) projected that a $17-by-2029 federal minimum would lift ~0.4 million from poverty (range: near-zero to 0.9 million) while boosting cumulative pay for affected workers by $238 billion (2024–2033), offset by $86 billion in lost pay from job losses, a net increase of $151 billion.

Firms do pass higher labor costs to consumers in labor-intensive sectors. Aaronson, French & MacDonald (Fed Chicago) found a 10% minimum wage increase raises restaurant prices by roughly 0.6–0.7%. Ashenfelter & Jurajda (2021) found near-full pass-through at McDonald’s: a price elasticity of ~0.14 with respect to minimum wage increases. Sosinskiy & Reich (2025) found California’s $20 fast-food minimum raised prices by about 2.1% (8 cents on a $4 item), with 60–70% pass-through.

These are real effects, but they are sector-specific (restaurants, fast food, retail) and represent one-time adjustments, not ongoing inflationary pressure. No credible study finds that minimum wage hikes cause broad, sustained inflation.

Rao & Risch (QJE 2026), using matched universe tax-return data on independent businesses, found that higher minimum wages raise wage bills but the average exposed firm raises about 3.3% more revenue four years later, with profits left intact on average. They also find slower entry in highly exposed industries rather than a broad sectoral collapse.

Wursten & Reich (2023) used 30 years of Census data with a stacked event-study estimator and found no detectable disemployment effects in small businesses across low-wage industries. Luca & Luca (2017) found that a $1 minimum wage increase raises the exit rate of median (3.5-star) restaurants by about 10%, while higher-rated restaurants are unaffected. Marginal firms are more vulnerable, but the effect is not universal.

Higher wages increase the incentive to substitute capital for labor. Evidence exists but is more limited than headlines suggest. Ashenfelter & Jurajda (2021) did not find a clear link between minimum wage hikes and touchscreen ordering adoption at McDonald’s. Lordan & Neumark (2017) found minimum wages reduce employment in “automatable” occupations. The measured effect appears mainly as slower hiring and occupational reshuffling, not mass immediate layoffs.