Created 5/23/1997
Go to
Brad DeLong's Home Page

The Minimum Wage: The U.S. Experience


J. Bradford DeLong

Professor of Economics, U.C. Berkeley


May 1997

forthcoming in Parliamentary Brief

Sophisters, Calculators, and Economists

On January 25, 1995 Secretary of Labor Robert Reich debated the conservative U.S. Senator Robert Bennett--from Utah, the state that contains the worldwide headquarters of the Mormon Church. The topic? Whether the U.S. government should raise the minimum wage, then set at $4.25 an hour.

Bennett's position was that "the minimum wage should be abolished. If someone isn't worth $4.25 an hour, he should be paid less." Reich's response was: "I completely disagree. Every hard-working American is worth at least a wage that lifts a family out of dire poverty."

In the end the left won this debate, at least on the level of political symbolism: on July 15, 1996 the final block to President Clinton's minimum wage proposals lifted as the U.S. Senate voted overwhelmingly--76 to 22--to raise the minimum wage from $4.25 and $5.15 an hour.

As Reich looking back on his tenure at the Labor Department comments (in his recent memoirs, Locked in the Cabinet) on the minimum wage debate:

"Note the key word: worth. [Bennett] used it first. It's a moral concept as well as an economic one. Can someone's labor really by worth less than $4.25 an hour? In purely economic terms, surely it can. But in moral terms, the answer is far from clear.

"And herein lies the importance of having this debate: It crystallizes a much larger debate about whether Americans are mere participants in an impersonal market, or are members of a common culture and society. Raising the minimum wage is a good thing to do. But quite apart from the wisdom of raising it, having a sharp public discussion about it is itself worthwhile. It helps Americans clarify their beliefs about what we owe one another as members of the same society."


For Robert Reich--I dare say for Tony Blair--the minimum wage is a way of pledging allegiance to social democratic values. It is a symbolic statement that the U.S. Democratic Party and the British Labour Party think that you are worth something, even if your market wage is low. It is, in a sense, a gesture of chivalry.

And they are not wrong. The minimum wage is a symbolic declaration of social-democratic fraternity: our mutual recognition of one another's social worth. The politics of symbols is an important part of reminding us of who we as societies and nations are. There is a place in politics for gestures of chivalry.

Yet the Age of Chivalry has been over for more than two centuries. The Age of Sophisters, Calculators, and Economists has succeeded it. I do not think that as a result the glory of Europe is extinguished forever. But I am an economist. And I am tired of symbolic political victories won without counting the substantive cost. Sometimes such symbolic political victories are truly Pyrrhic.

So--leaving aside the political symbolism--what were the substantive economic consequences of the Clinton Administration's successful campaign to raise the minimum wage in the U.S.? What would the substantive economic consequences be of a Blair Administration's introduction of the minimum wage in Britain?

The U.S. Minimum Wage

In the U.S. there is a rough consensus about the costs of recent increases in the minimum wage: the consensus is that the substantive costs have been very small, in large part because legislated increases in the minimum wage since 1970 have failed to keep pace with inflation. The level of the $5.15 U.S. minimum wage today is low relative to the distribution of wages in the U.S. economy: not many workers are directly affected by the minimum wage.

Economists' standard analysis of changes in the minimum wage is straightforward: Workers whose wages are less than the new minimum are fired, and cannot be reemployed because their marginal economic product is less than the government-mandated minimum cost. Workers whose wages are greater than the new minimum see their wages go up, as employers substitute higher-priced more-skilled labor for the low-wage labor they no longer find it cost-effective to employ.

The peak economic impact of the minimum wage in the U.S. came under President Richard Nixon, when the minimum wage was raised to a level that corresponds to some $7.50 an hour in dollars of today's purchasing power. Since 1970 the inflation-adjusted purchasing power of the real wage has steadily eroded. By 1980 the minimum wage was down to some $6.10 an hour dollars of today's purchasing power. By 1986 the minimum wage was down to some $5.10 an hour in dollars of today's purchasing power. Since 1986 the inflation-adjusted value of the real wage has remained constant: the increases in 1991-2 and 1996-7 keeping pace with but not increasing the minimum wage faster than inflation.

While the real value of the hourly minimum wage has been falling, the hourly productivity of the average American worker has been rising. The average American worker today is some 46% more productive than the average American worker of 1970. The minimum wage today is some 31% less in real terms than in 1970. In 1970 the minimum wage was set at a level of perhaps half the economic productivity of the average American worker. Today the minimum wage is set at less than a quarter of the economic productivity of the average American worker.

Substantive Consequences of the U.S. Minimum Wage

Thus it should be no surprise that with the minimum wage so low it is hard to find evidence of destructive effects. Economists David Card and Alan Krueger have recently carried out seven studies of various changes in the minimum wage, focusing on teenagers, restaurant employees in general, and fast-food employees in particular. In all seven cases increases in the minimum wage have led to positive and statistically significant increases in the earnings of workers. In none of the seven cases have increases in the minimum wage led to estimated decreases in employment.

Economists used to reliably find, looking at how teenage unemployment varied over time, that an increase in the minimum wage led to an increase in teenage unemployment. But the long decline in the real value of the minimum wage in the 1980s was not associated with a similar fall in teenage unemployment. And it now seems more likely than not that the association between higher minimum wages and higher teenage unemployment that used to be strongly visible over time in the U.S. economy was the result of other factors than the minimum wage increase.

Do U.S. economists believe that an increase in the minimum wage has no effect reducing employment? No, they do not. Ask economists to predict what the impact of doubling the minimum wage would be, or of even larger increases that would take the minimum wage up to the average wage, and they will reply that at such high levels the loss of jobs for currently low-wage workers from the government's mandating higher wages would be substantial. But ask U.S. economists today about the effects of a small marginal increase in the minimum wage, and the answer you will get is that a 10% boost to the minimum wage will reduce employment among affected workers by 0.5% to 1.0%--a loss of employment opportunities that is small relative to the boost in earnings received by those low-wage workers who remain employed.

The low relative level of the minimum wage in the U.S. today has also led to a rough consensus that that the substantive benefits of changes in the U.S. minimum wage from a distributional point of view are relatively small. Card and Krueger have calculated that a $0.90 an hour increase in the minimum wage transfers some $5.5 billion a year in earnings to low-wage workers. Compared to the roughly $40 billion a year spent on Temporary Assistance to Needy Families, the $35 billion a year tax expenditure of the Earned Income Tax Credit, the $30 billion a year in Food Stamps, the $70 billion a year in Workers' Compensation Insurance, or the $170 billion a year spent on Medicaid (medical treatment for the poor and the disabled), a $0.90 an hour increase in the minimum wage does not loom large as a component of the American social insurance state. The fact that many minimum wage workers are adolescents living in middle-class households means that President Clinton's welfare reform did more to reduce the total income of poor Americans than the minimum wage increase did to raise it.

The European Unemployment Dilemma

At the end of the 1970s Milton Friedman and his monetarists offered the governments of the industrial west, then beset by rising inflation of ten percent per year or so, a bargain: Governments would instruct their central banks to restrain the rate of growth of the money stock. Unions and businesses would see the central bank's commitment to slow growth of the money stock, and recognize that, if they did not restrain wage and price increases, their organizations would die. Inflation would then be brought under control.

There would, monetarists said, be a bad year or two or three of high unemployment and deep recession after the policy of tight money had begun and before unions and businesses had recognized reality. But that would pass. In the long run the unemployment rate would return to what Milton Friedman called the "natural rate of unemployment" as unions and firms recognized reality. A permanent reduction in inflation would be obtained at the price of a temporary, relatively short (if sharp) rise in unemployment.

The governments of the industrial west took the bargain.

In the United States the monetarists delivered, not in the sense that there turned out to be one particular measure of the money stock that was the magic guide to economic policy, but in the sense that tight money ended stagflation and gave the U.S. an economy with low inflation and not significantly higher unemployment by the mid 1980s.

In Europe the reduction in inflation took place as promised, but the return of unemployment to its pre-inflation crisis levels did not. According to the OECD, unemployment in its European members averaged 4% in 1974, 6% in 1978, 10% in 1985, 8% in 1990, and 11% in 1996. The monetarist response was that the natural rate of unemployment in Europe must have changed. But no one had forecast that such a big change in the natural rate would follow a campaign against inflation.

European governments faced an insoluble problem. The example of Mitterand's failed attempt at Keynesian expansion in the early 1980s convinced them that such policies would generate falling exchange rates, capital flight, and high inflation--but not rising employment. And no one else had any suggestions for how to reduce unemployment, because no one understood why fighting inflation had led the natural rate of unemployment to increase by so much.

Economists came up with a multi-syllable word--"hysteresis," a borrowing from electrical engineering--to name the problem, but a name is not an explanation. Economists built theoretical models in which high unemployment led to an increase in the proportion of long-term unemployed, and in which the long-term unemployed lose the skills that had made them employable and lose their ability to look for jobs; sociologists built theoretical models in which high unemployment changes society's attitudes and makes it socially acceptable to become unemployed; political scientists built theoretical models in which high unemployment calls forth higher unemployment benefits and higher wage taxes to fund them. But no one could produce evidence to support their particular explanation. And in the absence of knowledge of what, exactly, had caused higher unemployment, it was hard for governments to determine what to do to try to reduce it--other than to occasionally redefine the unemployment statistics to make the problem look smaller.

On the other side of the Atlantic, there were calls for a "two-handed approach": governments would commit to sharp cuts in employment taxes, mandates, and social insurance benefits while central bankers would commit to expansionary monetary policies. The reforms of the social insurance state would boost the potential labor force by enough to allow the expansionary monetary policy to produce higher employment and production rather than higher inflation. Central bankers can be reasoned with. After all, President Clinton had made (or Federal Reserve Chair Alan Greenspan had imposed) a bargain in 1993: deficit reduction for monetary expansion.

But European governments feared that social insurance reform without prior demand expansion would simply transform structural unemployment into Keynesian unemployment and rip holes in their safety nets, causing human and electoral disaster. Central bankers feared that demand expansion without prior radical social insurance reform would destroy their credibility as inflation fighters and lead to immediate price rises.

A Minimum Wage for Britain?

This is the context of proposals for a minimum wage for Britain. And to me at least the four main implications are pretty clear:


Created 5/23/1997
Go to
Brad DeLong's Home Page

Professor of Economics Brad DeLong, 601 Evans
University of California at Berkeley; Berkeley, CA 94720-3880
(510) 643-4027 phone (510) 642-6615 fax