Incoming House Speaker Nancy Pelosi is promising to enact legislation to increase the federal minimum wage from $5.15 to $7.25 per hour within the first 100 hours of the convening of the 110 th Congress on January 4. It would mark the first increase in the minimum wage since 1997 and the twentieth since the Fair Labor Standards Act of 1938 set it at 25 cents per hour. Some argue that this is a long overdue step to help America ‘s poorest workers. Others argue that it will hurt the people it is trying to help. An examination of the existing evidence and who will be affected by the proposed increase suggest that the minimum wage makes for better politics than policy.
“Poppy seed bagel with cream cheese, not toasted.” As the barista prepares your double-espresso, you watch your bagel make a bumpy journey from its wire mesh basket to a counter-long conveyor belt where halfway down it loses its battle with a misfit table-saw blade. Sliced, but not tattered, it skirts the conveyor toaster and comes to rest at the dairy spread wielding hands of your coffee’s caretaker.
Perhaps as much style as substance, the “Bagel Buzzsaw” of a Boston bagel peddler demonstrates that entrepreneurs leave no stone unturned in their efforts to satisfy customers via better products (in this case speedy service) and lower prices. As the incoming Congress will almost certainly enact an increase in the federal minimum wage from $5.15 to $7.25, it also serves as an indelible reminder that businesses face tradeoffs and respond to incentives.
The Economics of the Minimum Wage
The expected increase represents the second largest boost in history (29 percent), and would end the longest period of federal legislative inactivity since its enactment. The minimum wage is periodically increased to offset the deleterious effects of inflation and productivity in order to sustain the purchasing power of low-wage workers. The saw-tooth pattern in Chart 1 shows how the inflation-adjusted value of the minimum wage steadily falls following nominal wage increases – in real terms the minimum wage has not been this low since 1950. Similarly, since average wages in the economy increase with economic growth, the minimum wage as a share of average private sector wages has fallen to an all-time low of 30.5 percent. In other words, the purchasing power of a minimum wage worker today is less than one-third of the typical private sector worker, whereas at the peak in 1968 it was over one-half.
An increase in the minimum wage, like any higher price, leads economic actors to respond if it is above the price which currently prevails in the marketplace (i.e. it is binding). Buyers wish to avoid higher prices and sellers wish to offer more services in light of them. Actors in labor markets respond just as those in goods markets – only that buyers are the employers and sellers are the employees.
There is nothing in the theory that suggests how workers will expand offerings, or how entrepreneurs will respond to higher wages, rather that each will economize in ways most beneficial to themselves. Workers may wish to move from part-time status to full-time status. They might expand the amount of full-time hours they offer. Others may be induced to enter the labor force (perhaps at the expense of additional schooling, or to move off of government assistance). Still others may trade benefits for wages. Employers, at the same time, while commonly thought to reduce the total number of workers employed, could alternatively reduce the amount of on-the-job training offered, reduce non-wage benefit offers, reduce hours of operation, offer fewer services to customers, raise prices, make the workplace less comfortable or safe, substitute now relatively less expensive high-skilled workers and capital equipment for the low-skilled workers, or on other margins that are not readily identifiable. For example, they might induce more effort from existing workers – which raises the question of why the lure of higher wages before the legislation was passed were not sufficient to do so.
The existence of a high enough government mandated price floor on wages, while not necessarily resulting in massive unemployment, will result in an alteration in the direction of resources employed from that which would prevail in an unhampered market – and move us further away from the price mechanism which has been the engine of America’s prosperity for centuries. The degree to which low-wage workers will be net-winners or losers depends on employers’ ability to shift costs in the event of a wage increase. Thus, workers at firms selling products/services for which there are few substitutes; at firms for which capital or high-skilled labor is difficult to substitute with; and/or at firms where labor’s share of total costs is low are more likely to gain from increases in the minimum wage.
To the extent that the new federal minimum wage is binding, the cost increases must be paid for somehow – a point lost on many of those who confuse the mixed evidence of existing economic research with low-wage workers eating a free lunch. Who is the Tooth Fairy that hands entrepreneurs the funds to pay for increased wages? There’s a reason why we have not seen serious proposals to increase the minimum wage to say, $15.00 per hour – particularly since the employer portion of the Social Security and Medicare tax, and payments for unemployment insurance and workman’s compensation have increased markedly since 1968. These costs are tied to wages, and are estimated to add 13.3 cents for every additional dollar wages are increased.
Don’t expect your local coffee shop to replace its baristas with “buzz saws” anytime soon, however. The proposed minimum wage increase will lack teeth. Though the federal minimum increase of $2.10 appears large, it is not likely to be binding in many places. First, twenty-eight states plus the District of Columbia already have legislated state minimum wages in excess of the federal level – many of these well above $7.25. Second, the structure of the labor market is different today than in the past – with hourly workers making up a smaller share of the workforce, and a substantial decline in the share of hourly workers earning the minimum wage. Third, the equilibrium wage for low-wage workers in many areas is far in excess of $5.15. This does not mean that there will not be negative impacts within some populations – why else would the Illinois legislature, for example, insert a 50 cent allowance for teenagers (i.e. a lower minimum) into their recently passed legislation?
Reframing the Question
Ostensibly, the minimum wage is popular among the general public because of a feeling that responsible people who work hard and play by the rules should be in jobs that pay a “living wage” or that pay “enough to support a family.” What would constitute such an amount is highly subjective and, in particular, the notion that a family should be able to live with a single breadwinner is of relatively recent provenance – for most of human history everyone worked. Moreover, the complexity of determining what is “enough” or “fair” is mind-boggling. Incomes cannot be raised by dictate to whatever is necessary to live in some locale. For example, should workers have enough to live in Ashtabula , OH ? What about Monterey , CA or Newport , RI ? And, if merely identifying a problem with a price is sufficient to justify legislative action to change that price, why should this action be restricted to labor hours?
Even if one were able to demonstrate that minimum wage increases result in no deleterious employment or other effects for low-wage workers, the foregoing discussion is of little consequence because it overlooks a fundamental issue. If a higher minimum wage is supposed to alleviate poverty, particularly among working families, then it is essential to understand: 1. Who makes the minimum wage? 2. Who is poor (and why)? 3. How much overlap is there between the first two questions?
Nobel Laureate George Stigler in 1946 was among the first to recognize that the connection between raising the minimum wage and alleviating poverty was tenuous – and this at a time when the link was much stronger than it is today. He pointed out that, “The connection between hourly wages and the standard of living of a family is remote and fuzzy. Unless the minimum wage varies with the amount of employment, number of earners, non-wage income, family size, and many other factors, it will be an inept device for combating poverty even for those who succeed in retaining employment.”
How accurate is the portrait of the low wage worker that is painted in the press and by politicians? Are they hard working single mothers trying to support several children by working multiple substandard jobs? In chart two we provide some insight by looking at the characteristics of workers that are likely to be directly affected by a new minimum wage law – non-supervisory production workers earning hourly wages of $7.15 or less (the Bureau of Labor Statistics data does not track wage categories to match up with proposed minimum wage increases). Right-leaning think tanks tend to report data only for people making the minimum wage or less. Left-leaning think tanks tend to report data for people earning much more than even the $7.25 per hour increase (claiming “positive spillover” effects).
What the chart shows is that less than nine percent (approximately 11.3 million workers) of the overall working population earned low-wages in 2005. Of these low-wage workers, 60 percent were women (despite women being substantially less likely to be in low-wage jobs in 2005 as compared to 1979), 79 percent were white (compared to 80% of overall population), 15 percent were black (compared to 13% of overall population) and 21 percent were Hispanic. Thus while women are more highly represented in the low-wage population than in the population at large, the ethnic composition of the low-wage workforce is nearly identical to that which prevails at large. Over half of the low-wage earning population is under the age of 25 – and strikingly, within this young population, nearly 70 percent earn more than $7.15 per hour. Furthermore, 56 percent of all low-wage workers work part-time, with an average work-week below 30 hours – yet even only a quarter of all part-time workers do so for low-wages. Most part-time work is chosen voluntarily – in 2005, 86 percent of part-time work was selected by employees for non-economic reasons. Conversely, 99 percent of all full-time workers earn more than the minimum wage (not depicted).
Detailed analysis of the government micro-data would be required to address this issue in finer detail. While we remain skeptical of the reports from many that have done so, some useful information can nonetheless be gleaned from them. For instance, a prominent right-wing think tank claims that low-wage workers account for just over one-third of the income in their households and that one-fifth of affected families rely on low-wage earners as the sole breadwinner; a prominent left-wing think tank reports, using the same data, that families with affected workers rely on those workers for over half their earnings and that nearly half of affected families rely on these workers as the sole breadwinners. Therefore, we are confident that the likely numbers are somewhere in between – families in which a low-wage worker is present probably rely on them for something near 40 percent of their total annual income, and that about one-third of families with low-wage workers rely on them as the sole breadwinners (in fact, the Congressional Budget Office just released a study putting this figure at 26 percent). Estimates of family incomes for those families with minimum wage workers cluster around $40,000 – a figure confirmed in the new Congressional Budget Office study.
In short, the image of the struggling minimum wage worker is hyperbole. Few minimum wage workers are sole earners working full-time to support a poor family. We are not saying that trying to raise a family on $5.15 an hour is easy, or even possible in many places, what we are saying is that very few Americans find themselves in such a position. Moreover, to the extent that the depictions are not exaggerations, the notion that those workers currently in low-wage positions are condemned to them for life is grossly inaccurate. With just a little bit of training and experience, the vast majority of low-wage workers move out of these positions in a short period of time and onto the escalator of income. Data from two different government agencies demonstrate this. The Bureau of Labor Statistics reports that 63 percent of all minimum wage workers receive raises within their first year of working – with a median increase of 14 percent. The Congressional Budget Office reports that workers at the bottom tenth in the hourly wage distribution in 1996 experienced a 32 percent wage increase by 1999 and an additional 19 percent by 2003 – compared to 0.6 percent and minus 1.9 percent for workers in the top tenth and 5.2 percent and 3.8 percent for workers in the middle.
For those who work and are still in poverty, we already help them with programs such as food stamps and Medicaid – and as their wages rise this help is withdrawn. For example, a family of four in Massachusetts with one parent working full-time at the current minimum wage would see wages increase by about $4,000 under the new law. They would, however, see a decline in food stamp assistance of $1,000. Economist and New York Times columnist Paul Krugman has pointed out that only about twenty percent of the mandated increase in wages and benefits would get manifested in disposable income – the rest is taken away as benefits decline (according to Krugman, such families receive $9,700 in total government benefits). Thus, targeting the low-wage population with this social program will only reach a small portion of those that are materially deprived. The correct question then is, can we do better?
The answer begins with understanding who is poor and why they are poor. Poverty in America is generally not the result of people earning too small a wage because most people in poverty are not fully connected to the labor market. The majority of poverty cases stem from workers unable to work enough hours, unable to find a job, or simply that they are unable to produce something that others are willing to pay for (often through no fault of their own). It is also important to understand that workers are not independent economic entities, they are part of households, and that is the relevant unit. Thus, it is the total income and benefits in households, not just the hourly wage rates of one worker, that affect worker well-being. Hence, focusing on the low-wages of my one-year old daughter will paint a starkly misleading picture of how well-off she actually is.
To illustrate, the 37 million citizens in poor households (as defined by the Census Bureau) receive 57 percent of their total income from wages, while non-poor households rely on wages for over 90 percent of income. Seventy-three percent of adults in poverty did not work full-time while 71 percent have a high school degree or less. Only 2.9 million adults worked full-time for twelve months. Looking at the poorest 20 percent of households shows that 64 percent of adults did not work at all and that 59 percent of households had no wage earners at all (compared to 15 percent of middle class households).
Empirical evidence on the impact of the minimum wage on employment is mixed. Pessimistic studies have found that a minimum wage hike of the size proposed would result in a 12 percent increase in unemployment within the youth population (those most at risk for job loss). Optimistic studies, namely the work of economists David Card and Alan Krueger in 1995, have found no disemployment effects for youths in the short-run. Studies looking at the impact on other outcomes (such as school dropout rates, long-run skill acquisition) are less sanguine, but still not definitive. Despite these mixed findings, we have yet to find studies that show that the minimum wage has had any positive impact on poverty – presumably the most important metric upon which it should be evaluated.
Ardent supporters of the minimum wage understand its feebleness as a poverty reduction tool. Estimates of the benefits of a minimum wage increase peg the gains to wealthy families equal to the gains of poor families. A now famous letter sent to Congress on behalf of 650 prominent economists (including five Nobel Prize winners) who support a small increase in the minimum wage reveals that their support for an increase is for symbolic and not economic reasons. While it might make sense to support the minimum wage on grounds that it is a second-best policy in the face of politically impossible expanded cash payments to the poor or a well designed negative income tax, it is resoundingly not the case that it is second best in comparison with the Earned Income Tax Credit (EITC). Already expanded substantially since its enactment in 1975, a considerably larger share of the poor population would qualify for EITC programs than for minimum wage increases. Furthermore, if “we as a society” want to raise hard-working families out of poverty, then the costs of doing so should be widely shared by society at large – precisely what the EITC does (it works through the progressive tax system) and the minimum wage does not. The EITC is also better for long term economic growth because it is less distortionary to promote low-tax rates on a wide population than it is a high tax rate on a small subsection.
Estimates of the cost of the minimum wage increase, assuming no deleterious labor market effects, are that firms paying the minimum wage will expend an additional $30 billion per year. If not by the workers, these costs will be borne either by the owners of such firms or its customers. Contrary to popular opinion, the majority of firms paying low-wages are not the Fortune 500 behemoths, nor do the wealthiest consumers purchase the low-cost amenities that are likely to be affected. Alas, the benefits are not well targeted, and the costs fall disproportionately on smaller businesses and customers of bowling alleys, campgrounds, driving ranges, etc.
Wages are a function of a worker’s ability to produce goods and services valued by consumers. But, most poor families are not poor due to low wages; and most minimum wage earners are not poor. Legislating wages by fiat does not change these conditions. Nonetheless, if the effects of the minimum wage are likely to be small, what is the harm in making a symbolic gesture?
Symbolism is not benign. Making a statement about priorities is inferior to finding ways to actually help the poor, as the case of the EITC clearly demonstrates. But support for the EITC and other better programs is complicated by political grandstanding. Both Democrats and Republicans agree that the EITC is a superior program – but advocating the EITC would force politicians to abandon the positions on which their reputations and careers are staked on. The political game also prevents many poor from getting a foothold in the labor market on their own. Occupational licensing restrictions (in the name of safety and fairness) disproportionately hurt the poor workers who wish to enter these professions, and the poor consumers who must suffer the higher prices caused by these restrictions. The case of African hair braiders is the starkest example. Recently, California had required that African hairstylists spend nine months (1,600 hours) and at least $5,000 at a government-approved cosmetology school before sitting for the state licensing examination, which allowed braiders to legally practice their craft. But none of these government-mandated classes actually taught students how to braid hair. In fact, they taught techniques that were especially damaging to African hair.
Symbolism obscures the damage done by excessive government intervention in the workings of the marketplace. Symbolism leads voters to believe that the wave of a magic wand will solve any of society’s deepest problems and it obscures from view the real causes of poverty. Symbolism makes it easy to show compassion for the poor without being held accountable for doing anything about it – it is easy to be compassionate when spending other people’s money. Symbolism allows politicians to effectively play politics without having to act in ways that best serve their poorest constituents. It is far from clear that the goal of minimum wage legislation is to help poor families.