Preston J. Miller - Former Vice President and Monetary Adviser
Published October 1, 1995 | October 1995 issue
Recently, several economists have recommended a hike in the minimum wage. It is tempting to say that these economists must have failed Econ 101, since basic economics textbooks regularly teach the foolishness of a minimum wage policy. Yet, these economists are neither inept nor foolish; they include three Nobel Prize winners and a highly regarded labor economist. These economists must be basing their recommendation on new theory and evidence that textbooks have missed.
Indeed, there is a new economics of the minimum wage. The new economics suggests that a minimum wage policy should not be rejected out-of-hand as foolish. However, it also does not by itself support the economists' recommendation. These economists had to make a number of value judgments in order to reach their conclusion, judgments better made in the political arena. To better understand where their economic analysis ends and personal value judgments begin, one needs to understand the challenge raised by the new arguments for a minimum wage hike.
The old argument about minimum wages which is found in textbooks is quite simple: Government intervention in a competitive labor market is a mistake. If the minimum wage is set at or below the prevailing market wage, it will have no effect. If it is set above the prevailing market wage, the number of workers employed, determined by the firms' demand, will be lower than without a minimum wage. This is a foolish policy, because there conceptually are ways of transferring income to low-paid workers that do not reduce the number of people working.
A study by David Card and Alan Krueger in 1994 challenged the old argument. They found that in New Jersey a higher minimum wage increased employment, rather than decreasing it. Other studies, mainly conducted by at least one of the above researchers in collaboration with others, find similar outcomes.These findings led to development of new theories of minimum wages along two lines. The first, which I refer to as the monopsony theory, is hard to defend. The second, which I refer to as the tax-transfer theory, is quite defensible.
The monopsony theory takes the Card-Krueger findings at face value. It assumes that firms act as monopsonies in hiring low-wage workers; that is, they do not face strong competition for such labor and so have discretion in setting wages. According to this non-competitive theory of the labor market, a higher minimum wage can increase both efficiency and employment.
I find this argument in support of a higher minimum wage weak at every turn. First, the Card-Krueger findings are far too tentative to treat as an observed fact. Their data have been criticized, and researchers working with better data have shown the Card-Krueger findings are reversed. Other studies recently published in the prestigious American Economic Association Papers and Proceedings are also in conflict with the Card-Krueger findings. Moreover, even if the Card-Krueger data were correct, one can argue that their study searched for effects across too few industries and allowed too little time to elapse. Second, evidence about labor markets suggests the low-wage market is better characterized as competitive as opposed to monopsonist. Can Burger King set its wage for hamburger flippers much lower than those for McDonald's flippers, local newspaper carriers or grocery carry-out help without losing them? Finally, even if we accept the monopsony theory, it does not tell us whether or how much to raise the minimum wage. It tells us there is an optimal minimum wage, but the current minimum wage is not zero. Should the optimal wage be higher or lower than the current minimum wage? Is the 90¢ increase proposed by the administration too low or too high? Shouldn't it be set differently in different locations? This theory gives little guidance on these questions.
The tax-transfer theory does not take the Card-Krueger evidence at face value. In fact, as Richard Freeman, the highly regarded labor economist, suggests after reviewing a number of studies on both sides of the aisle, a rise in the minimum wage may lower employment but that effect is modest and much smaller than previously thought.
While a higher minimum wage may entail some modest employment reduction, proponents of a higher minimum wage correctly point out that is not enough to reject it. Benefits also have to be examined. Freeman correctly argues that a useful way of thinking about a minimum wage policy is as a distorting, off-budget tax-transfer. Let's take a closer look at this way of viewing minimum wage policy.To see that a minimum wage policy is a tax-transfer, let us suppose that the government seeks to implement a $5 per hour minimum wage through the budget rather than through mandate. It assigns a government agent, Minnie, to collect taxes each month from Scrooge-like employers and distribute the proceeds to low-wage employees. At each firm Minnie asks for a record of all employee hours worked at less than $5 per hour and the wages actually paid. She then computes the employer's tax bill as the sum of hours worked which were paid at less than $5 per hour times the difference between $5 and the wages actually paid. That is, if the employer had 100 hours of work paid at $3 per hour and 250 hours of work paid at $4 per hour, its tax bill would be computed as: 100*(5 - 3) + 250*(5 - 4) = $450.
After collecting the tax from all employers, Minnie then visits all employees who earned less than $5 per hour. She pays them the difference between $5 and their actual wage times the number of hours they worked in the month.
This hypothetical tax-transfer scheme is identical in effect to a minimum wage policy. The major difference is that the minimum wage mandate removes the government middle-man and moves the policy from on-budget to off-budget for the government.
The tax-transfer scheme which I describe causes distortions. That is, there is an incentive for employers to avoid the tax, and an incentive for individuals to enter the work force to capture the subsidy. These effects are confirmed in recent studies. Employers avoid the tax by substituting machines or higher-skilled workers for their low-wage workers or by changing the composition of the compensation package they offer workers. That is, they raise wages to comply with the mandate but reduce training and keep the total compensation unchanged. Similarly, but on the other side of the labor market, high school students have an incentive to drop out of school to receive the minimum wage subsidy.
While the tax-transfer theory suggests that a minimum wage hike distorts, proponents of a hike would argue that so do all government tax-transfer schemes-contrary to what economic textbooks assume. Based on recent evidence, such as that of Card-Krueger, proponents argue that distortions caused by the minimum wage are no worse than those caused by other tax-transfer programs. They complete their argument that now is a propitious time to raise the minimum wage by making the following observations:
The minimum wage has fallen in inflation-adjusted terms significantly since it was last set, so that distortions from a small increase now should be minimal.
Real wages of low-skilled workers have been declining, so the social benefit of transferring income to this group of workers has increased.
With cuts planned in on-budget tax-transfer programs, such as the Earned Income Tax Credit (EITC), it becomes more desirable to do the tax-transfers off-budget.
Let me comment on these observations in turn. I do not quarrel with the first point, but I would add that while the distortions should be small, so should the benefits. The administration's proposed 90¢ increase over two years would bring the minimum wage up to just $5.15. It's hard to believe many workers would be earning less than this two years from now.
The next two observations, which are crucial to the argument for why an increase is needed now, are based purely on personal value judgments. The judgments may or may not be reasonable. My point is that they do not come from economic analysis.
It is true that real wages of low-skilled workers have fallen relative to those of high-skilled workers. However, one cannot just consider the benefit of giving low-skilled workers more money. One also has to consider who loses in the process. The losers include the owners and customers of the firms that must pay the tax. The losers include the few people who will not be able to get a job at a higher minimum wage, or those who keep their jobs but consequently receive less compensation in other forms, such as training or other benefits. The losers also include the public at large. There is a general consensus that the change noted in relative wages of skilled and unskilled workers is a worldwide phenomenon and is caused primarily by new technologies, such as computers, which simultaneously increase the demand for high-skilled workers and decrease the demand for low-skilled workers. The change in relative wages is the market's way of signaling that more high-skilled workers are needed and fewer lower-skilled workers are required. A higher minimum wage muffles this signal, resulting in the wrong mix of workers and causing a loss of goods and services potentially available to the general public. So yes, the benefits to those that receive higher wages may increase. But a lot of people will lose. It takes a value judgment to say the benefits to the winners outweigh the costs to the losers.
Meanwhile, the suggestion that it is a virtue to have the minimum wage off-budget is a value judgment. These economists are saying they know better what is good for society than elected officials do. It is generally agreed that the EITC dominates the minimum wage as a means of directing income to those who need it. The EITC directs money to low-income families. The minimum wage directs money to low-wage earners. The problem with the minimum wage transfer is that many low-wage earners happen to be teenagers or other secondary workers of middle- and upper-income families. If elected officials determine that there should be reductions in the EITC transfers to low-income families, how can it be consistent with their values to raise the minimum wage?
A decision to raise the minimum wage cannot be made on the basis of economic analysis alone, it also requires the value judgments of elected officials. My recommendation is that the minimum wage be put on budget. This is what Congress decided to do with the mandated employer health care contributions under the proposed Clinton health care policy. If the minimum wage were put on-budget, policymakers could weigh in a consistent manner the desirability of a minimum wage change relative to changes in other tax and transfer policies.