Addressing the Minimum Wage "Debate"

It really bothers me when people talk about the minimum wage. The fact that we even have a minimum wage somewhat indicates the constant pressure internal to the system to depress wages as much as possible. I want to discuss a very annoyingly common talking point propagated by non-economists, to clear up the misinformation: “raise the minimum wage, get inflation and fewer jobs”.

The cliché “raise the minimum wage, get inflation and fewer jobs” is misleading because it treats a conditional result from a narrow textbook model as if it were a universal empirical law. In economics, that is backwards: the prediction depends on assumptions, and the evidence is mixed, context-dependent, and often much less dramatic than the slogan implies.

The slogan mistakes a model assumption for a fact. The clean “higher wage floor → lower employment” result comes from the standard competitive model: firms are wage takers, labor markets clear competitively, firms have free entry/exit, and adjustment happens mainly through reducing labor demand. Dube and Lindner’s 2024 review explicitly frames that result as contingent on those assumptions, and notes that modern evidence increasingly requires models with labor-market frictions, product-market pass-through, and other margins of adjustment.

The labor market often is not perfectly competitive. A large reason the cliché fails is that many low-wage labor markets exhibit monopsony-like features: search frictions, limited mobility, local concentration, and employer wage-setting power. The OECD’s 2022 brief states that minimum wages can be justified in part by “counterbalancing the negative effects of firms’ labour market power,” and says substantial monopsony power exists in many low-wage labor markets across OECD countries. In that setting, a higher wage floor can raise pay with little employment loss, and in some cases even raise employment. (OECD)

The iconic NJ–PA study mattered because it directly contradicted the simplistic story. Card and Krueger’s fast-food comparison found “no indication” that New Jersey’s minimum-wage increase reduced employment; in their conclusion they write that they found “no evidence” the rise reduced employment and that none of their alternative specifications showed a negative employment effect. They also report that employment increased in New Jersey relative to Pennsylvania in their sample.

But the right takeaway is not “minimum wages never reduce employment.” The best reading of the literature is not the opposite slogan. Even David Neumark’s 2021 review opens by noting the literature is summarized in sharply different ways: from “no employment effect,” to “mixed and centered on zero,” to “mostly negative.” Neumark argues the preponderance of U.S. evidence points toward negative effects for less-skilled workers, especially in some study designs. That means the public cliché is wrong as a universal claim, but it is also wrong to replace it with another universal claim.

The central empirical question is magnitude, not sign in the abstract. Recent work has shifted from “is the sign ever negative?” to “how large is the employment response relative to the wage gain?” Dube and Zipperer’s 2024 NBER paper argues the more informative statistic is the own-wage elasticity. Their median estimate for published studies is -0.13, which they interpret as meaning only about 13% of potential earnings gains are offset by job losses, with more recent studies tending to be closer to zero. That is a long way from the popular image of large, mechanical job destruction. (NBER)

Broad low-wage labor markets often show very small net employment effects. Dube and Lindner report that for broader/overall groups, the median own-wage elasticity is about 0.02, and that 90% of those studies are either positive or only slightly negative. Their overall database median is -0.13, while narrower groups (like teens) look somewhat more negative. This is exactly why blanket statements are misleading: results vary by who is studied and how narrow the affected group is.

Some of the apparent “job loss” in narrow samples is composition or substitution, not economy-wide disappearance of work. Studies focusing on teens or tightly defined subgroups can pick up labor-labor substitution—for example, firms favoring slightly older or more experienced workers—without implying large declines in total low-wage employment. Dube and Lindner explicitly note narrower-group estimates can look more negative partly for that reason.

The best modern U.S. evidence often finds reallocation rather than large aggregate job loss. Cengiz, Dube, Lindner, and Zipperer (2019), using 138 prominent state minimum-wage changes, found that the number of low-wage jobs was “essentially unchanged” over five years after an increase: jobs below the new minimum wage disappeared, but were largely offset by jobs bunching at and just above the new floor. That is wage compression and reallocation, not the simple “jobs vanish” narrative. (NBER)

Firms adjust on many margins besides cutting headcount. The cliché pretends employment is the only adjustment margin. The modern literature does not. Dube and Lindner review evidence on prices, profits, productivity, nonwage margins, capital substitution, entry/exit, and worker turnover. Their summary says a “relatively clear picture” has emerged on the importance of price pass-through and productivity-enhancing effects.

Lower turnover matters a lot and softens employment effects. One important offset is reduced quits/separations. Dube and Lindner note that lower turnover can improve productivity and save recruitment/training costs. If a wage hike reduces churn, firms may need fewer new hires, and higher wage costs are partly offset by lower replacement costs and better retention. That is a real economic mechanism the slogan ignores.

“It causes inflation” is especially sloppy, because sectoral price pass-through is not the same thing as ongoing macro inflation. Minimum-wage hikes can raise prices in affected sectors, especially labor-intensive services. That is not controversial. Card and Krueger themselves found fast-food prices rose in New Jersey relative to Pennsylvania. But a one-time or localized relative price change is not the same as a broad, persistent inflation process.

The measured price effects are usually modest, not runaway. Renkin, Montialoux, and Siegenthaler (2020) estimate that a 10% minimum-wage hike raises supermarket/drugstore prices by about 0.36%. An earlier survey by Lemos reports estimates of wage/price inflation effects from a 10% increase ranging from 0.2% to 1.8%, with the upper end dropping to 0.37% once the outlier is removed. Those are real pass-through effects, but they are far smaller than the public rhetoric implies.

Even the OECD says the aggregate inflation effect is limited. The OECD’s 2022 policy brief says that because minimum-wage workers are a relatively small share of the workforce, even sizeable minimum-wage increases tend to have a limited impact on inflation. It gives the example that in the UK, where about 5% of workers are at the minimum wage, even a 20% increase would raise inflation by only 0.2%. (OECD)

Mainstream official analysis is nuanced, not sloganistic. CBO’s assessments do not say “minimum wages are harmless,” but they also do not endorse the simplistic public claim. CBO says higher minimum wages would raise earnings and family income for most low-wage workers and lift some families out of poverty, while also reducing employment to some degree. That is a trade-off, not a one-line catastrophe story. (Congressional Budget Office)

The cliché survives because it is rhetorically simple and empirically lazy. It survives because it collapses several distinct questions into one:

• Is the labor market competitive or monopsonistic?
• How binding is the new wage floor?
• Which workers are directly affected?
• Are firms adjusting through prices, profits, productivity, turnover, hours, or staffing mix?
• Are we talking about a small state increase, a large sectoral jump, or a very high national floor? Ignoring those questions makes the claim sound universal when it is not.

The oversimplified claim is wrong because it treats a special-case comparative-statics result as a general law. The actual evidence says: minimum wages usually raise pay at the bottom, sometimes produce small negative employment effects, often show little net employment change in broader low-wage labor markets, and can cause modest sectoral price increases without implying major economy-wide inflation. The serious economic question is not “does Econ 101 say demand slopes down?” It is: under what market conditions, for which workers, and through which margins do firms adjust? On that question, the empirical literature is much richer—and much less slogan-friendly—than public discourse suggests. (NBER)

It’s somewhat interesting when people claim that increasing the minimum wage will destabilize the system, because this is exactly what Leftists/Marxists argue. It admits that capitalist systems rely on low wage earners in order to keep the system afloat. If wages cut into profits enough, the system doesn’t work. This requires the maintenance of a class system, the lower class needing to take cuts in standards of living to maintain higher living standards for capital owners.

From a left / Marxist political-economy perspective, the point isn’t “capitalism needs individual workers to be poor” in a moralized sense; it’s that capitalist reproduction has systemic requirements—profitability, labor discipline, and institutional arrangements—that often make high bargaining power for labor (including very high wage floors) politically and economically destabilizing unless other structures change.

Below is a map of the main literatures that analyze those structural claims and how they connect to wages/profits/employment.

Kalecki: class power, “confidence,” and the political limits of full employment

Kalecki’s “Political Aspects of Full Employment” is the canonical statement that even if the state can technically sustain demand and high employment, capital has political reasons to resist it: full employment undermines the disciplinary function of unemployment, strengthens labor’s bargaining position, and weakens the social authority of business. (5) The system’s “stability” is not only a technical equilibrium; it’s a political equilibrium over who commands production and policy. In Kalecki’s story, persistent full employment endogenously threatens that equilibrium, so capital uses investment strikes, lobbying, and ideology to reassert discipline. (5) This is a foundational piece for later work on the idea that capitalism “needs” slack labor markets—not as a conscious conspiracy, but because tight labor markets shift class power in ways that destabilize capitalist governance and profitability.

 Unemployment as discipline: Marx’s “reserve army,” efficiency wages, and contested exchange

• (a) Reserve army / NAIRU-like logic in Marxian terms: Marx’s “industrial reserve army” (relative surplus population) frames unemployment/underemployment as a structural feature of accumulation that disciplines wage demands and workplace resistance. (6)

• (b) Mainstream microfoundations that rhyme with Marx: The Shapiro–Stiglitz efficiency-wage model formalizes involuntary unemployment as a worker-discipline device: unemployment makes job loss costly, sustaining effort at wages above the market-clearing level. (7) Even though it’s not Marxist, it supplies a bridge: in capitalism, some unemployment can be functionally useful for control and profitability.

• (c) Bowles & Gintis: contested exchange and power in labor contracts: Bowles & Gintis develop “contested exchange” / “discipline” approaches where employment relations are shot through with power and enforcement problems—so labor markets don’t clear like Walrasian markets, and the distribution of rents and the threat of unemployment matter for effort and control. (8) Bowles’s “neo-Hobbesian” and “Marxian” models explicitly treat production as involving coercion/discipline and conflict, not just marginal products. (9)

Class inequality isn’t just “pre-market” or a policy accident; it’s reproduced through the labor contract, monitoring/enforcement, and the macro condition that unemployment (or precariousness) sustains discipline.

Endogenous instability via distributional conflict: the Goodwin “class struggle” cycle

Richard Goodwin’s model is the classic formalization of distributional conflict generating endogenous cycles: high employment → stronger wage growth → falling profit share → lower accumulation → rising unemployment → weaker wage growth → recovering profits, and so on. (10) A 2026 survey paper explicitly interprets the model as aligning with Marx’s view of class struggle driving cyclical dynamics (employment rate vs wage share as state variables). (11)

The wage–profit split isn’t a passive outcome; it feeds back into accumulation and employment, creating endogenous oscillation. That’s a direct analytical route to my point: if wages compress profits beyond a threshold, investment weakens—but the mechanism is dynamic conflict, not a static “minimum wage kills jobs” cliché.

Marxian crisis theory and the “profit squeeze” literature

A major Marxian tradition links crises to distributional struggle: when labor’s bargaining power rises (tight labor markets, strong unions, etc.), wage pressure can squeeze profits, slow accumulation, and trigger downturns that restore profitability and discipline.

• Weisskopf’s “Marxian crisis theory and the rate of profit” situates “profit squeeze” approaches (and related work) within postwar U.S. dynamics. (12)
• Goldstein develops a microfoundation for cyclical profit squeeze theory. (13)
• Kotz (SSA tradition, below) explicitly connects the profit rate and crisis narratives. (14)

Capitalist “stability” often depends on institutions that keep the profit rate high enough for accumulation and keep labor sufficiently disciplined; when those break down, crises are not exogenous shocks but system-restoring episodes.

Neo-Kaleckian distribution-and-growth: wage-led vs profit-led regimes

This is the modern macro-political-economy literature closest to phrasing about “system viability” depending on profits. The Bhaduri–Marglin tradition (and a huge follow-on literature) studies whether higher wage shares are expansionary (wage-led) or contractionary (profit-led) once you include demand and investment responses.

• Stockhammer & Onaran synthesize the framework and argue that neoliberalism raised inequality and lowered the wage share; they also emphasize that many economies are domestically wage-led, while neoliberal growth relied on unstable “debt-led” or “export-led” regimes. (15)
• Oyvat et al. review wage-led vs profit-led growth empirically across countries and explicitly build on Bhaduri–Marglin. (16)
• The PERI overview frames the wage-led/profit-led distinction and the partial-vs-total effects logic that matters for interpreting distributional shifts. (17)

The system’s “stability” depends on the growth regime: some configurations can tolerate (or benefit from) rising wage shares; others are structurally dependent on profitability and investment sensitivity—especially under financialization and global competition. This literature is a good antidote to simplistic “wages up = economy breaks” talk because it replaces it with: which regime, which openness/finance structure, which investment function, which institutions

Institutional “stabilizers” of capitalism: Regulation theory and SSA

These are explicitly about endogenous structural stability: capitalism is not self-stabilizing via markets alone; it is stabilized (for a time) by historically specific institutions that organize production, competition, money, labor relations, and the state.

• Regulation theory (France: Aglietta, Boyer, Lipietz, etc.): Regulation theory distinguishes a regime of accumulation (how production/consumption/distribution fit together) from a mode of regulation (laws, norms, wage relation, state form, money form, etc.) that stabilizes it. Crisis occurs when the mode of regulation no longer supports the accumulation regime. (18)

• Social Structure of Accumulation (SSA) (US: Bowles, Gordon, Weisskopf, Kotz, etc.): SSA theory similarly argues that long upswings require a coherent institutional structure; when contradictions intensify (profitability, conflict, governance breakdown), the SSA decays and crisis follows. (19)

Class inequality is not just a distributional outcome—it is an institutional arrangement (workplace authority, labor law, welfare state, macro policy) that can be part of what makes a given accumulation regime “work” for capital.

A careful Marxist/left reading of the best political-economy literature is:

1. Capitalism requires profits for accumulation, so distribution matters for investment and growth (neo-Kaleckian wage-led/profit-led frameworks). (15)
2. Capitalism also requires labor discipline and workplace control, which is supported by unemployment/precarity and by institutional power relations (Kalecki; reserve army; efficiency wages; contested exchange). (5)
3. Stability is historically contingent: institutions can temporarily reconcile mass consumption with profitability (Fordism), or shift the burden onto debt/export reliance (neoliberal regimes)—but those fixes are often unstable (regulation theory; SSA; Stockhammer/Onaran on debt-led/export-led regimes). (18)
4. Distributional conflict can generate endogenous cycles: tight labor markets raise wages, squeeze profits, slow accumulation, raise unemployment, and restore profits/discipline (Goodwin; profit-squeeze crisis theory). (10)

So it’s not simply “the system needs poverty,” but rather that many capitalist configurations are structurally reliant on unequal bargaining power and on institutions that prevent labor’s share from rising “too far” without changing ownership/control, investment behavior, and state structures.

Not all capitalist configurations have the same dynamics. “Capitalism” isn’t one invariant machine with one invariant wage–profit–employment response. Different institutional configurations (labor law, bargaining structures, welfare state, finance, trade openness, corporate governance, monetary regime) produce different dynamics—including different limits on how far wages can rise before profitability/accumulation/political coalitions react.

1. Regimes of accumulation are historically specific: Regulation theory’s states that a regime of accumulation (how production, demand, distribution fit together) can be stable for a time because a matching mode of regulation (institutions, norms, state forms) supports it—but that match can break, producing crisis and transition. So Fordism, postwar social democracy, neoliberal financialization, export-led models, etc., are not just “variations”; they change the underlying feedback loops. (18)

2. Growth can be wage-led or profit-led depending on structure: Neo-Kaleckian work makes the point very explicitly: whether higher wages support growth (via demand) or choke it (via investment/profit sensitivity, external constraint, markups) depends on openness, financialization, sector mix, investment behavior, and policy. That’s why you can’t treat “wages rise → employment falls” as a structural law. (15, 16, 17)

3. Class conflict dynamics differ with bargaining institutions and discipline mechanisms: Goodwin/Kalecki-style conflict stories depend on how tight labor markets translate into wage pressure (union density, bargaining coverage, strike capacity), and how firms/state respond (policy tightening, investment strikes, offshoring threats, austerity). The “reserve army” mechanism exists across capitalism, but its strength and the channels through which it operates vary. (5, 10)

4. Even the “profit squeeze” story is configuration-dependent: Profit-squeeze crises are more plausible when labor has strong bargaining power and when investment is very profit-sensitive; less so when labor is fragmented, when finance substitutes for demand, or when firms can restore profit share through global supply chains, monopoly power, or policy regimes. (12, 14)

Some capitalist configurations are stabilized by high wages + mass demand + coordinated bargaining (classic Fordist/social-democratic stories). Others are stabilized by profit restoration + labor discipline + financial/external demand channels (many neoliberal “debt-led” or “export-led” patterns). And some are simply unstable, lurching between distributions via crises and political realignments. Capitalism has family resemblances (private ownership, wage labor, profit-driven accumulation), but it doesn’t have one universal wage/employment/inflation dynamic. The dynamics are endogenous to the institutional and political structure of the particular capitalist regime. (18, 15)

Broad-brush, the U.S. is usually classified as a liberal market economy (LME): coordination happens largely through markets (rather than industry-wide bargaining, patient capital, or corporatist institutions), and firms face strong pressures to maintain profitability because finance/corporate governance are geared toward shareholder-value disciplines. (20)

In Marxist terms, the U.S. configuration is a relatively “pure” form of capitalist social relations: production is privately organized, coordination happens through markets and corporate hierarchy, and the state’s core macro role is to secure the conditions for accumulation (property rights, contract enforcement, a credible monetary regime) rather than to institutionalize class compromise. What VoC calls an LME (20) reads, in Marxist terms, as a low-institutionalization class settlement: capital does not bind itself to broad wage norms or corporatist bargaining; it prefers flexibility and the ability to reorganize production rapidly. What that means in practice—as a configuration that shapes wage/profit/employment dynamics:

Decentralized wage-setting and weak collective bargaining institutions

Compared with coordinated market economies, the U.S. has low union density and low collective bargaining coverage, and coverage is close to membership because there’s little sectoral/extension bargaining. OECD’s ICTWSS country sheet reports U.S. union density around 9.9% (2024) and very low adjusted bargaining coverage. (21) This implies that wage growth for low/middle workers relies more on tight labor markets, statutory floors, and firm-level conditions than on coordinated bargaining that can “internalize” macro tradeoffs.

This keeps labor’s collective capacity to claim surplus relatively weak. That’s not just “fewer unions”; it’s a different class relation: the working class is less able to act as a class, which helps stabilize profit claims and managerial authority. The OECD bargaining/union facts are the surface expression (21); the Marxist reading is “a regime of labor discipline with limited organized counterpower.”

A “fissured” workplace and fragmented employment relations

A widely used lens is David Weil’s “fissured workplace”: lead firms outsource/subcontract, shifting risks and labor standards to smaller suppliers/contractors, often weakening worker bargaining power and enforcement. (22) This implies that wage standards (including minimum wages) interact with subcontracting, compliance, and market power in supply chains—so the adjustment margins aren’t just “hire fewer workers,” but also contracting structure, hours, pricing, and enforcement.

It’s a reorganization of the relations of production that: weakens workers’ ability to target the real centers of surplus appropriation (lead firms), disperses responsibility for wages/conditions, turns labor standards enforcement into a chronic collective-action problem, and increases precarity—strengthening the “reserve army” effect even when headline unemployment is low.

Finance-heavy capitalism and demand management through credit cycles

A big chunk of macro/finance literature emphasizes how household credit expansions can support short-run demand but raise medium-run fragility (growth reversals, financial instability). IMF and BIS work summarize this “short-run boost, medium-run drag” pattern for rising household debt. (23) This implies that distributional outcomes (wage share vs profit share) can be “patched over” for a while via credit-supported consumption, but that can also make the regime more crisis-prone.

In the U.S., profitability isn’t merely what firms hope to earn; it’s enforced through capital markets and corporate governance norms (buybacks, performance pay, activist pressure). Marxist translation: finance acts as a command-and-control layer over productive capital, pushing management to protect surplus extraction and maintain “credibility” with owners and creditors.

What this configuration tends to produce

• More inequality and wage dispersion than coordinated systems (because bargaining is decentralized and labor protections are weaker). (20)
• A stronger role for statutory policy (minimum wage, EITC, labor standards enforcement) to move the lower tail of the distribution, because bargaining coverage doesn’t do it. (21)
• Greater sensitivity of firm behavior to profitability and capital-market pressures than in “patient capital” settings—one reason the same wage shock can play out differently in, say, Germany vs the U.S. (20)

In this regime, wage gains are more likely to trigger “restoration” mechanisms rather than negotiated adjustment

In a corporatist model, higher wages can be absorbed via coordinated price-setting, productivity deals, or macro policy aligned with a wage norm. In the U.S., the typical adjustment channels are more conflictual and decentralized: price pass-through where possible, speedup/intensification, scheduling cuts, subcontracting, automation, relocation/offshoring threats, and political pushback. Marxist shorthand: capital restores the conditions of exploitation through multiple margins, not just layoffs.

“Profit-led” tendencies are more plausible under U.S.-style financialization and openness

Neo-Kaleckian work emphasizes that whether higher wages are expansionary depends on the growth regime (15). Marxist translation for the U.S.: with strong finance discipline and global competitive pressures, capital is quicker to treat rising labor costs as a threat to profitability and investment—so policy and firm behavior often bias toward preserving profit claims. That doesn’t mean wages can’t rise; it means the political economy of investment is structured to resist sustained redistribution without institutional change.

The U.S. “mode of regulation” stabilizes accumulation by weakening labor’s claim on surplus Regulation theory and SSA say stability is institutional and contingent (18). Marxist translation: the U.S. mode of regulation—weak bargaining coverage, fissuring, finance dominance, credit-based consumption smoothing—has (for decades) been a way to stabilize accumulation without strong egalitarian class compromise, at the cost of higher inequality and periodic crisis tendencies (debt and asset cycles).

The U.S. configuration is an LME in comparative terms (20) and, in Marxist terms, a regime of accumulation stabilized by weak institutionalized class compromise and strong labor discipline. Low bargaining coverage (21) and workplace fissuring (22) fragment worker power and diffuse responsibility for wages and conditions, while financialized governance and household-credit dynamics (23) allow demand to be propped up without a durable rise in the wage share—“private Keynesianism” substituting for a robust social wage. In this configuration, distributional conflict is managed less through coordinated negotiation and more through decentralized restoration mechanisms (prices, intensification, fissuring, relocation, political reaction), making Kalecki/Goodwin-style class-power feedbacks especially salient (5, 10), and making the stability of the regime more dependent on institutions that preserve profitability and discipline than on institutions that embed egalitarian wage growth (15, 18).