In his State of the Union address, President Obama proposed to increase the federal minimum wage, from its current $7.25 per hour to $9.00 per hour. So, in the words of the old union organizing song, “Which Side Are You On?”
It is an easy call if you are either (a) a strict libertarian or (b) an enthusiastic advocate of the less fortunate with limited concern about the scarcity of resources. (If you belong to both of those groups, there is little advice that I can offer.) However, in between those poles of opinion, things become rather murky, rather quickly.
In fact, the opening question emphasizes that policy analysis is hard to convey using spirit-rousing songs. The reflex thought of what music would characterize analysis of this issue led me to Béla Bartók’s “Music for Strings, Percussion and Celesta,” which is typically played with two small string orchestras – divided by the percussion instruments – sitting on the opposite sides of the stage, spending about half of the piece shouting back and forth at one another. (This is not to put you off if you have not heard the piece. I enjoy it – which may be a predisposition, in that Bartók attended high school in the town where all of my grandparents were born and grew up, and at about the same time.) As in the minimum wage debate, there are arguments on both sides, and the argument often becomes quite intense.
To start with the philosophical debate, but postponing the empirical question of results: Libertarians tend to oppose restraints on voluntary agreements among responsible persons. If a prospective worker finds it advantageous to offer labor at a low wage, and a prospective employer is willing to accept that offer, there is no reason for government to interfere. Those who take the other side argue that such agreements are the result of unequal power, and facilitate injustice and abuse.
The best attempt at resolution of this difference that I have ever seen comes from the late economist Arthur Okun, who wrote in his classic, Equality and Efficiency: The Big Tradeoff, under the heading “Bans on Exchange:”
Once political and civil rights are seen as integral to human dignity, it becomes clear why they shouldn’t be bought or sold for money. If someone can buy your vote, or your favorable draft number, or a contract for your indentured service, he can buy part of your dignity; he can buy power over you. By prohibiting your sale of rights, society is encroaching on your freedom, but it is also protecting you from others who might want to take your rights away. Your creditors cannot make you part with your dignity. They cannot force you into trades that are made as a last resort, which could not be fair trades and which would be distorted by vast differences in the bargaining power of the participants and by the desperation that spawns them. Any rational person who would sign a contract for indentured service must be in desperate straits…
Whenever the law bans trades of last resort, it shuts some potential escape valve for the person in desperate straits. In shutting the valve, society implies that there must be better ways of preventing or alleviating that desperation. When, for example, child labor was restricted, widowed mothers and disabled fathers were deprived of the opportunity to make ends meet out of the earnings of their young children. When the battle over child labor was fought in Great Britain, the proponents of the ban viewed it as part of a larger program in which society would provide the disadvantaged with aid in another and better form.
Minimum-wage laws and work-safety legislation can be viewed most fruitfully as further examples of prohibitions on exchanges born of desperation, extending the logic of the ban on indentured service. Some economists strain to understand the sources of minimum-wage laws: Are they justified as an offset to monopoly power in hiring labor? Do they emerge out of a conspiracy by skilled workers to reduce the job opportunities of the unskilled? Or are they urged by the skilled on the premise that wages will be raised all along the line as customary differentials are preserved? Are they well-intentioned but misguided efforts to help the poor? Similarly, some economists wonder whether work-safety legislation is warranted by lack of information about on-the-job dangers.
As I read the laws, they declare that anyone who takes an absurdly underpaid or extremely risky job must be acting out of desperation. That desperation may result from ignorance, immobility, or genuine lack of alternatives, but it should be kept out of the marketplace. Recognizing that objective still leaves plenty of room for debate about the proper scope of these laws. With these bans, society assumes a commitment to provide jobs that are not excessively risky or woefully underpaid. That commitment is often regrettably unfulfilled, and perhaps, if it were fulfilled, the bans would be unnecessary. Still, closing a bad escape valve may be an efficient way of promoting the development of better ones through the political process.
Of course, if you find persuasive the libertarian absolute prohibition of restraints of voluntary exchange, Okun’s niceties are beside the point. But even if you believe that Okun’s reasoning opens the door to a minimum wage, it does admittedly leave “plenty of room for debate about the proper scope” of an increase to it, as well as society’s obligation to make available jobs above that minimum. The public debate leaves a strong sense that some are not concerned about that obligation; they believe that the minimum wage has little ill effect on the availability of jobs. Let’s survey the arguments, on both the minimum wage itself, and its effects.
First, a couple of broad points. Some apparently see an increase in the minimum wage as an economic stimulus: The spending of the higher pay of those who benefit will generate greater economic activity. That isn’t quite right. That higher pay has to come from somewhere – and heaven is not a potential source. It might come from lower earnings of the small businesses, or lower dividends to the shareholders of larger businesses, that hire the low-wage workers. At least part of those earnings or dividends would have been spent, too. Yes, it is likely that the minimum-wage workers would spend more of their wages, faster, than would their employers. But the margin between the two is a sliver of the gross, not the entire gross.
Another possibility – see below – is that the higher wages come out of higher prices for the product or service. If so, then the economic stimulus comes from the difference between the consumption of low-wage workers, and the consumption of people who buy the product of low-wage work. That margin is probably even smaller.
So in short, an increase in the minimum wage is a poor, poor substitute for real macroeconomic stimulus. Those who yearn for the latter should not seek comfort in the former (though they may still favor it for other reasons).
An argument that you hear occasionally on the other side is that the minimum wage provides support for the unworthy. But we can probably discard that one, too. To earn the minimum wage, you have to work. And outside of Lake Woebegone, where everyone is above average, many who work at the minimum wage probably are making the most of the opportunities they have under what for many others would be demoralizing circumstances. Are there some with significant talent and training who choose to accept the minimum of responsibility at the minimum wage? Probably. But the nation cannot make policy for the exception or the extreme, instead of the rule or the mean.
So if we should choose to try to improve the minimum circumstances of those who find themselves working at the minimum wage, is an increase in the minimum wage the best tool – and even if the best, is it also an acceptably good one? Cue Bartók’s two string orchestras.
Background: Research on the effects of increases in the minimum wage is highly controversial these days. Straightforward economic theory would say unambiguously that increasing the price of low-skilled labor would price some of it out of the market, and reduce the availability of such jobs. However, attempts at empirical measurement of job changes in past minimum-wage increases seem to be fighting to a close margin around a draw – very small decreases, but even sometimes very small increases. How can that be, when the theory seems so clear?
This question would get much less attention if low-wage labor were used in the same proportions across all industries. If it were, its impact on the cost structure of each firm would be thin. An increase in the minimum wage would hardly be felt. Instead, however, low-wage labor is used disproportionately in service industries, of which fast food is the poster child. Much of the recent economic research has focused on fast food.
In a 1992 “natural experiment,” when New Jersey raised its state minimum wage while Pennsylvania did not, minimum-wage employment in fast food in New Jersey relative to Pennsylvania was apparently close to unaffected. Two sets of researchers are still battling over a narrow margin of statistical results, on one side or the other of net zero.
There are some possible reasons why the impact of this minimum wage increase was apparently small. One is that there is a significant difference between one employer raising its entry-level wage, and all employers complying with an increase in the statutory minimum wage. In the former case, the employer might suffer an increase in costs relative to competitors, and be forced to cut back. In the latter case, all employers would stay in roughly the same competitive position, and might find ways to accommodate the higher wage, including efficiencies elsewhere in the operation, or even efficiencies because workers at the higher wage turn over less frequently and therefore reduce hiring costs, or are more loyal and efficient employees.
Another potential offset for the impact of a minimum wage increase is an increase of prices. Even though the fast food industry makes intensive use of low-wage labor, some researchers of the New Jersey minimum-wage increase calculated that the increase in the wage bill even if all employees were retained at the higher wage would have been offset by a three-percent price increase. That seems small, and on a fast-food restaurant tab, it might not be noticed by some. (And again, all such employers would be affected about equally.) But at a time of rapid inflation, it could be critical. And it might not be an option for a small independent operator competing with large chains that have deep pockets and access to capital for labor-saving investment.
But this instance raises another issue, which is that a minimum-wage increase that would reduce employment, all else equal, might have no measurable impact because all else is not equal. The economy grew strongly but with low inflation after 1992. Growing demand might have helped employers to maintain hiring at the minimum wage, while small price increases in fast food and similar industries simply were lost in a much larger sea of overall price stability. Clearly, from the standpoint of inflation, the best time to increase the minimum wage is when prices are generally stable. But if prices are stable because demand is weak, it might be the worst time to increase the minimum wage from the perspective of employment, because businesses that are barely surviving could be pushed over the edge by an increase in the cost of labor. Today’s situation is clearly closer to the latter scenario.
And this raises yet another question: Assuming that the nation wants to act to help low-wage workers, is the minimum wage the best way to do so? We have another tool: the earned income tax credit (EITC), which was enacted in this country in 1975. The EITC is a federal-government-financed wage supplement, delivered as an income tax credit to each qualifying worker. It is refundable, meaning that if the credit exceeds a taxpayer’s federal income tax liability, the taxpayer receives the excess in cash. Because the EITC is funded by all taxpayers, not just the employer of a particular low-wage worker, its cost is spread across all industries – not concentrated on those that employ disproportionate amounts of low-wage labor. And because it does not add specifically to labor costs, it does not tend to accelerate inflation. It has been targeted most heavily to workers who support children; some criticize the minimum wage because it raises the pay of, for example, teenagers from relatively affluent families. (In fact, about 20 percent of all minimum-wage workers are teenagers.)
Those are the EITC’s pluses. Its minuses include that it is “clawed back” (to use a graphic British term) as the worker’s income rises, to limit the budgetary costs. The credit is recaptured by in effect subjecting the worker’s wages to a higher marginal income tax rate – which is feared by some to be a significant work disincentive. Some also fear that the EITC puts a substantial administrative burden on the tax collectors, whose operation is designed to take in money, not to give it out. From the EITC’s earliest days, there were fears that low-wage workers would in effect “do each other’s laundry” to increase their wage income just to the point where the EITC would reach its largest possible value. And in fact, there have been prosecutions of willful EITC fraud. But at the same time, the complexity of the EITC – including factors such as determining the eligibility of divorced parents, and offsetting for income from capital rather than from labor – has led to frequent errors that have resulted in underpayments, as well as overpayments, of the credit. And the delivery of the tax credit upon filing a return after the end of the income year means that the EITC is not available for the day-to-day needs of the neediest workers.
Some experts would argue that, because of these administrative complexities, it would be difficult to expand the EITC very much beyond its current levels, and therefore that additional support for low-wage workers today would best come by raising the minimum wage. But by no means would all experts agree. Some would say that greater use of electronic technologies such as “smart cards” might help to get around these administrative constraints and make an expansion of the EITC feasible – and indeed preferable to an increase in the minimum wage that would threaten the viability of some marginal firms and possibly increase inflation.
But the EITC and the minimum wage have a problem in common – namely, geographic variations in living costs and incomes. The federal government never has stepped down the slippery slope of adjusting income tax liabilities for local income levels and costs of living. But as most Americans well know, a $200,000 income in Manhattan yields a very different living standard from a $200,000 income in Mississippi. Likewise, the federal minimum wage applies uniformly across the country, and a $9 per hour wage in Manhattan would again be quite different in effect from a $9 per hour wage in Mississippi.
That is why 19 states and the District of Columbia now have enacted their own minimum wages greater than the federal $7.25 per hour level. (The federal minimum wage law has exceptions for some industries and other work conditions; four states have set minimum wages below the federal level that might apply in the instances of the federal exceptions, but those state laws have exceptions as well; five states set no minimum wage.) Thus, raising the federal minimum wage pulls up from the bottom, but not necessarily all the way to the top. However, a $9 federal minimum wage would be higher than the current levels of all but one state (Washington state).
So action by the federal government applies in a sense selectively, to those states that choose not to set their own higher minima. To the extent that states with higher-than-average wages and costs of living tend to set their own higher minimum wages (which appears generally true; but New York, for example, currently follows the federal minimum), then raising the federal minimum wage will tend to raise labor costs in the states with comparatively lower wages and incomes, and therefore might be more likely to price low-wage labor out of those markets.
This raises a big, broad question: Given the diversity across states (and even within states; there are substantial differences in housing costs and prevailing wages within, to be obvious, New York and California), could it be advantageous to back off on the minimum wage and allow the market and the states to choose the right wage level? Some would argue that this could open up large numbers of jobs in service firms that would choose to hire much more low-wage labor and change fundamentally their interactions with their customers. These additional workers would get an introduction into the habits and requirements of successful work. It raises subsidiary questions: Would substantial additional labor supply be forthcoming at lower wage rates? How much of that supply would come from immigration – or could that flow be controlled? Could such a process become a “race to the bottom” on wages, in which experienced older workers with family responsibilities would be underbid by teenagers – perhaps reminiscent of the concerns that Okun raised?
This is a complex area, especially if one considers potential quantum changes in the norms under which the labor market has operated for many years. The arguments and counterarguments do recall the disputes between Bartók’s two string orchestras. On first hearing, few would say that by the end of that piece those disputes were resolved. The differences over the minimum wage are unlikely to be resolved by the end of this debate, either.