Chapter 10 Raising the Wage Floor

From Nominal to Real Wages

The nominal wage is the wage measured in money (dollars in the United States). The real wageThe nominal wage (the wage in dollars) divided by the price level. is the nominal wage in an economy adjusted for changes in purchasing power. It is defined as the nominal wage divided by the general price level:

real wage = nominal wage price level .

Workers care about the real wage, not the nominal wage, because the real wage captures the trade-off between leisure time and goods and services. Firms care about the real wage, not the nominal wage, because it measures the true cost of hiring labor. Figure 10.3 "Real Minimum Wage in the United States" shows the real minimum wage—that is, the minimum wage adjusted for inflation. The real minimum wage is defined as

real minimum wage = nominal minimum wage price level .

The real minimum wage increases in jumps whenever the nominal wage is increased, but it declines over time as it is eroded by inflation. The erosion of the minimum wage by inflation was recognized by President Clinton in the address that we quoted in our introduction. In that request to Congress he also said:

The first increment of the proposal simply restores the minimum wage to its real value following the change enacted in 1989.

If the Congress does not act now, the minimum wage will fall to its lowest real level in 40 years. That would dishonor one of the great promises of American life—that everyone who works hard can earn a living wage. More than 11 million workers would benefit under this proposal, and a full-time, year-round worker at the minimum wage would get a $1,800 raise—the equivalent of 7 months of groceries for the average family.

When President Clinton referred to “7 months of groceries,” he was converting the increase in the minimum wage into real terms, just as our technique for converting to inflation does. Instead of using the bundle of goods that goes into the CPI, however, he was using a bundle of goods representing groceries for the average family.

Nominal and Real Wages in the Labor Market

The challenge when analyzing the minimum wage is that it is set in nominal terms, but workers and firms care about the real minimum wage. To help us understand the difference, we begin with a specific numerical example of the labor market. Suppose we have the following labor supply-and-demand equations, where labor supply and labor demand are measured in hours:

labor supply = 10,000 × real wage

and

labor demand = 72,000 – 8,000 × real wage.

Think of this example as referring to the weekly demand for and supply of unskilled labor in a small city. It is reasonable to think of this as a competitive market, in which market participants will typically agree on a price at or close to the point where supply equals demand. In the supply-and-demand framework, the intersection of the supply and demand curves tells us the equilibrium price in the market and the equilibrium quantity traded. In the labor market, the place where supply and demand meet tells us the equilibrium wage and the equilibrium number of hours worked.

First we solve for the equilibrium in this market. In equilibrium, the quantity of labor supplied equals the quantity of labor demanded, so

10,000 × real wage = 72,000 – 8,000 × real wage.

Add (8,000 × real wage) to each side:

18,000 × real wage = 72,000.

Then divide both sides by 18,000 to obtain

real wage = 4.

If we plug this value of the real wage back into either the supply or the demand equation, we find that the equilibrium quantity of hours worked is 40,000 hours. For example, we might have 1,000 workers, each of whom works a 40-hour week. The equilibrium is illustrated in Figure 10.4 "Labor Market Equilibrium".

Suppose that this example pertains to the base year. From our discussion of correcting for inflation, we know that in the base year we set the price level equal to 1. When the price level is 1, the real wage equals the nominal wage. In the initial year, therefore,

labor supply=10,000×nominal wage 1 =10,000×nominal wage

and

labor demand=72,000−8,000×nominal wage 1 =72,000−8,000×nominal wage.

In the base year, the nominal wage is $4 per hour.

Now imagine we have 10 percent inflation, which means that the price of all goods and services in the economy increases by 10 percent over the course of a year. If a household paid $100 a week for groceries last year, it must pay $110 this year; if a household used to pay $500 a month in rent, it must now pay $550; and so on. Turning this around, a dollar is worth less than it used to be; you need $1.10 to purchase what you could have bought for $1 this year. The price level has increased from 1 to 1.1.

To see what this means in terms of the labor market diagram, think about the situation at a given nominal wage, such as $2.20 per hour. Last year, when the price level was 1, households were willing to supply 22,000 hours (= 10,000 × 2.2). But $2.20 now is worth the equivalent of only $2, so households are willing to supply only 20,000 (10,000 × 2.2 1.1 = 10,000 × 2) hours of labor instead. The same idea applies at every wage; households will supply only the amount of labor that they would previously have supplied when the wage was 10 percent higher.

A similar logic applies to the demand for labor. The increase in the price level means that firms get 10 percent more dollars for the goods that they sell. As a consequence, the labor performed by workers generates more dollars than it used to. If it was worth paying $7 for an hour of work before, it is now worth paying $7.70 for that same hour of work.

In terms of real wages, however, nothing has changed. The equilibrium real wage is still $4, as it was before. But because

real wage = nominal wage price level ,

the nominal wage must increase by 10 percent to match the increase in the price level. The equilibrium in the labor market is shown in Figure 10.5 "Labor Market Equilibrium after 10 Percent Inflation". It is no coincidence that this diagram looks exactly the same as Figure 10.4 "Labor Market Equilibrium"; that is the point. An increase in the price level is matched by an increase in the nominal wage, and nothing changes in terms of the real wage or the real equilibrium quantity of labor.

The Imposition of a Minimum Wage

When the government imposes a minimum wage, firms are not permitted to pay less than the amount that the government mandates. Suppose we are again in the base year, so the price level is 1. Imagine that the market equilibrium wage is $4 per hour, but the government now passes legislation stating that all firms must pay at least $5 per hour. At this wage, supply does not equal demand. Figure 10.6 "Labor Market with a Minimum Wage" illustrates what happens.

Markets are based on voluntary trades. In Figure 10.6 "Labor Market with a Minimum Wage", we see that sellers (the workers who supply labor) would like to sell 50,000 hours of labor to the market at the set minimum wage—that is, 250 more people would like to have a 40-hour-a-week job when the wage increases from $4 to $5. But firms wish to purchase only 32,000 hours of labor—firms want to hire 200 fewer workers (8,000 fewer hours). In a market with voluntary trade, no one can force firms to hire workers. As a result, the equilibrium quantity of labor traded in the market will be determined by how much the firms wish to buy, not how much workers want to sell.

We can now answer our first motivating question of the chapter: what is the consequence of imposing a minimum wage? Two things happen when the government imposes a minimum wage:

1. The amount of labor hired in the market decreases. In our example, the number of unskilled workers employed decreases from 1,000 to 800. Thus while those who have jobs earn a higher wage, there are now some individuals who no longer have jobs. Employment has decreased.
2. At the government-imposed wage, there are more people who want to work than are able to find jobs. Thus the minimum wage has created unemployment. Because 1,250 people would like jobs at a wage of $5 but only 800 jobs are available, 450 people are unemployed; they would like a job at the prevailing wage, but they are unable to find one.

The number of unemployed workers is 450, even though employment decreased by only 200 workers. The difference comes from the fact that the higher wage also means that more people want to work than before. In this case, the higher wage means 250 more people would like a job.

We have assumed in this discussion that everyone works for 40 hours, in which case the number of people employed must decrease by 200. Another possibility is that everyone who wants a job is able to get one, but the number of hours worked by each individual decreases. Because there are 32,000 hours of work demanded and 1,250 people who want jobs, each worker would work 25.6 hours a week. In this situation, we say that there is underemployment rather than unemployment. Yet another possibility is that, after the introduction of the minimum wage, the number of people employed stays the same as before (1,000), but those individuals are allowed to work only 32 hours per week. In this case, we have both underemployment (of the previously employed) and unemployment (of the extra workers who want a job at the higher wage). In real-life situations, there may be both unemployment and underemployment.

Inflation and the Minimum Wage

Although inflation made no difference to our basic analysis of the labor market, it does change our analysis of the minimum wage. Minimum wages are fixed in nominal terms and do not automatically change when there is inflation. So if the minimum wage is set at $5 and the price level increases from 1 to 1.1, the real minimum wage declines. Looking back at the definition of the real minimum wage, we find that

$$\text{real minimum wage = }\frac{\text{nominal minimum wage}}{\text{price level}}\text{ = }\frac{\text{5}}{\text{1}\text{.1}}\text{ = 4}\text{.55}\text{.}$$

The effect of a reduction in the real minimum wage is shown in Figure 10.7 "A Reduction in the Real Minimum Wage". At the lower real wage, firms are willing to hire more workers. Employment increases from 32,000 hours to 35,600 hours: 90 more people can find jobs.

In Figure 10.7 "A Reduction in the Real Minimum Wage", the real minimum wage of $4.55 is still higher than the equilibrium wage of $4.00. To put the same point another way, the equilibrium nominal wage has increased to $4.40, but this is still below the nominal minimum wage of $5.00. However, if the price level were to increase by 25 percent or more from the base year, the minimum wage would become completely irrelevant. The minimum wage would be below the market wage. Economists say that the minimum wage would no longer be “binding” in this case.

It is exactly this process of increasing prices that lies behind Figure 10.3 "Real Minimum Wage in the United States". As the price level increases, the minimum wage becomes worth less in real terms (and has less of an effect on employment). Eventually, Congress acts to increase the minimum wage to bring it back in line with inflation—although, as Figure 10.3 "Real Minimum Wage in the United States" shows, Congress has allowed the real minimum wage to decline substantially from its high point in the late 1950s. The reduction in the real minimum wage also leads to a reduction in unemployment, as shown in Figure 10.8 "Effects on Unemployment of a Reduction in the Real Minimum Wage".

Efficiency Implications of a Minimum Wage

Markets are a mechanism that allow individuals to take advantage of gains from trade. Whenever a buyer has a higher valuation than a seller for a good or service, they can both benefit from carrying out a trade. This is how economies create value—by finding opportunities for mutually beneficial trades.

The minimum wage interferes with this process in the unskilled labor market. It reduces employment, which is the same as saying that fewer transactions take place. Because each voluntary transaction by definition generates a surplus, anything that reduces the number of transactions causes a loss of surplus. In economists’ terminology, the minimum wage leads to a departure from efficiencyThe basis that economists use for judging the allocation of resources in an economy.. We can represent that inefficiency graphically. Figure 10.9 "Deadweight Loss from Minimum Wage" shows the effect of the minimum wage, using the ideas of buyer surplus and seller surplus.

In part (a) of Figure 10.9 "Deadweight Loss from Minimum Wage", we see the market without the minimum wage. In the labor market, it is the firm who is the buyer. The total buyer surplus is the profitRevenues minus costs. that firms obtain by hiring these workers; it is the difference between the cost of hiring these workers and the revenues that they generate. Graphically, it is the area below the labor demand curve and above the market wage. The total sellersellers’ surplus#8217;s surplus is the benefit that accrues to workers from selling labor time. Sellers of labor (workers) receive surplus equal to the area below the market wage and above the supply curve.

In part (b) of Figure 10.9 "Deadweight Loss from Minimum Wage", we show the effect of the minimum wage. As we already know, the higher wage leads to a reduction in employment. Fewer transactions occur, so the total surplus in the market is reduced. Economists call the lost surplus the deadweight lossThe economic surplus that disappears as a result of any policy that distorts a competitive market and thus causes the total volume of trade to differ from the equilibrium level. from the minimum wage policy.

The most obvious cost of the minimum wage is this loss of surplus. But there may be other hidden costs as well. Whenever people are prevented from carrying out mutually beneficial trades, they have an incentive to try to get around these restrictions. Firms and workers may try to “cheat,” conducting hidden transactions below the minimum wage. For example, a firm might pay a worker for fewer hours than he or she actually worked. Alternatively, a firm might reduce other benefits of the job. Such cheating not only subverts the minimum wage law but also uses up resources because the firm and the worker must devote effort to devising ways around the law and ensuring that they do not get caught.

The loss in surplus could also be greater than is shown in Figure 10.9 "Deadweight Loss from Minimum Wage". The figure is drawn under the presumption that the trades taking place in the labor market are the ones that generate the most surplus. But suppose that the minimum wage is $5.00. It is possible that someone who would be willing to work for, say, $2.00 an hour loses her job, whereas someone willing to work for, say, $4.50 is employed.

Remember that if there were no restrictions in the labor market, the wage would adjust so that anyone wanting to work could find a job. This means that both the person willing to work for $2.00 and the person willing to work for $4.50 could both find a job as long as the wage is above $4.50. If the equilibrium wage were $4.00, then the person willing to work for only $4.50 would not be employed. In either case, this is the efficient outcome, consistent with obtaining all the gains from trade.

The Effect of a Minimum Wage Increase on Employment and Unemployment

Figure 10.10 "Effects of Increasing the Real Minimum Wage" amends our view of the labor market to show an increase in the minimum wage from $5 to $6. (We suppose that the price level is constant, so an increase in the nominal minimum wage implies an increase in the real minimum wage.) The increase in the minimum wage leads to a reduction in the level of employment: employment decreases from 32,000 to 24,000. Labor is now more expensive to firms, so they will want to use fewer hours. At the same time, the higher minimum wage means that more people would like jobs. The increase in the amount of labor that people would like to supply, and the decrease in the amount of labor that firms demand, both serve to increase unemployment.

Our model generates a qualitative prediction: an increase in the minimum wage will decrease employment and increase unemployment. At the same time, the wage increase will ensure that those with jobs will earn a higher wage. So we can see that there may be both advantages and disadvantages of increasing the minimum wage. To go further, we have to know how big an effect such a change would have on employment and unemployment—that is, we need the quantitative effects of a higher minimum wage.

To understand the quantitative effects, we want to know when to expect big or small changes in employment or unemployment—which depends on the wage elasticity of labor demandThe percentage change in the quantity of labor demanded divided by the percentage change in the wage. and the labor supply. Remembering that the wage is simply the price in the labor market, the wage elasticity of demand is an example of the price elasticity of demandThe percentage change in the quantity demanded in the market divided by the percentage change in price. in a market:

$$\text{wage elasticity of labor demand = }\frac{\text{percentage change in labor demand}}{\text{percentage change in real wage}}\text{.}$$

From Figure 10.10 "Effects of Increasing the Real Minimum Wage", we can see that the wage elasticity of labor demand tells us everything we need to know about the effects of a change in the wage on employment. If the demand curve is relatively elastic, then a change in the minimum wage will lead to a relatively large change in employment. If the demand curve is relatively inelastic, then a change in the minimum wage will lead to a relatively small change in employment. This is intuitive because the elasticity of labor demand tells us how sensitive firms’ hiring decisions are to changes in the wage. An elastic demand for labor means that firms will respond to a small change in the wage by laying off a large number of workers, so the employment effect will be large. The elasticity of labor supply is not relevant if we are concerned only with employment effects. This is illustrated in Figure 10.11 "The Employment Effect of a Change in the Minimum Wage" and summarized in Table 10.1 "Employment Effects of a Change in the Real Minimum Wage".

If we are interested in the effect on unemployment, however, we must look at both demand and supply. A worker is counted as unemployed if he or she is looking for a job but does not currently have a job. The labor supply curve tells us how many workers are willing to work at a given wage; those who are not employed are looking for a job. To understand the effects of the minimum wage on unemployment, we need to look at the mismatch between supply and demand at the minimum wage, so we must look at the supply of labor as well as the demand for labor. The price elasticity of supplyThe percentage change in the quantity supplied to the market divided by the percentage change in price. measures the responsiveness of the quantity supplied to a change in the price: in the case of the labor market, we obtain the wage elasticity of labor supplyThe percentage change in the quantity of labor supplied divided by the percentage change in the wage.:

$$\text{wage elasticity of labor supply = }\frac{\text{percentage change in labor supply}}{\text{percentage change in real wage}}\text{.}$$

The more elastic the labor supply curve, the bigger the change in labor supply for a given change in the real wage (Figure 10.12 "The Unemployment Effect of a Change in the Minimum Wage"). A bigger change in labor supply means a bigger change in unemployment. Combining this with Table 10.1 "Employment Effects of a Change in the Real Minimum Wage", we get the results summarized in Table 10.2 "Unemployment Effects of a Change in the Real Minimum Wage". If demand and supply are both inelastic, the change in the minimum wage has little effect on unemployment. The higher wage does not make much difference to firms’ hiring decisions (inelastic demand), and it does not induce many additional workers to look for a job (inelastic supply). The overall effect on unemployment is small. By contrast, if both curves are elastic, then an increase in the wage will lead to a big decrease in the number of jobs available and a big increase in the number of job seekers. If we can find good estimates of the elasticities of labor demand and supply, we will be able to make good predictions about the likely effect of an increase in the minimum wage.

Winners and Losers from the Minimum Wage

We know that the working poor are made better off by an increase in the minimum wage—after all, the whole point of the increase is to ensure that these individuals earn more. We can see this gain in Figure 10.9 "Deadweight Loss from Minimum Wage": it is the extra surplus that sellers obtain at the expense of buyers. Remember, though, that some of the working poor will lose their jobs as a result of the higher minimum wage. So our first conclusion is that those members of the working poor who keep their jobs are unambiguously made better off.

The increase in the minimum wage means that there are more people who are unemployed. Those who are already unemployed are not directly affected by the increase in the minimum wage. These unemployed individuals may be indirectly affected, however, because it becomes harder for them to find jobs.

Are there any effects on the relatively rich? The answer is yes. The increase in the minimum wage means that firms will earn lower profits. We can see this because buyer surplus is reduced in Figure 10.9 "Deadweight Loss from Minimum Wage". Although firms are just legal entities, they are owned by individuals. When a firm earns lower profits, the shareholders of that firm receive lower income. The working poor and the unemployed are not, for the most part, individuals with portfolios of stocks; the shareholders of firms are the relatively rich. Thus the relatively rich are made worse off by the increase in the minimum wage. Our broad conclusion is therefore that the working poor benefit from an increase in the minimum wage, but everybody else in society is made worse off.

There is another concern when we think about the distribution of income. One consequence of the minimum wage is that jobs become a scarce resource: more people want jobs than there are jobs available. We must consider how this scarce resource is allocated. Do workers line up outside factory gates? In this case, the time that they spend waiting in line is an additional cost of the minimum wage. Does some individual control who gets hired? Then there is the potential for corruption, whereby jobs are sold, meaning that the gains from the minimum wage flow not to workers but to this individual instead.

Does the Minimum Wage Benefit Unskilled Labor as a Whole?

We have concluded so far that the minimum wage benefits the working poor but at the cost of creating unemployment: some people who used to have jobs will lose them as a result of the minimum wage. Because of the flows between unemployment and employment in a dynamic labor market, it does not really make sense to think of the unemployed and the employed as different people. If we instead think about unskilled labor as a whole—a group that includes both those with jobs and those unable to find jobs—what can we conclude about the effects of an increase in the minimum wage? There are a few different ways of looking at this question.

Estimates of Labor Demand Elasticity

Labor economist Daniel Hamermesh summarizes the findings from numerous studies in his book, Labor Demand.Daniel S. Hamermesh, Labor Demand (Princeton, NJ: Princeton University Press, 1996). Based on his review of all these studies, Hamermesh argues that a good estimate of −(elasticity of labor demand) is about 0.3. This means that if we increase the minimum wage by 10 percent, employment will decrease by about 3 percent. With these, we would conclude that labor demand is relatively inelastic, and the employment and welfare implications of the minimum wage are not that large. Labor supply also tends to be relatively inelastic, partly because the income effectsWhen a good decreases in price, the buyer can afford more of everything, including that good. and substitution effectsIf one good becomes cheaper relative to other goods, this leads the buyer to purchase less of other goods and more of that particular good. of changes in the wage tend to be offsetting. It certainly seems possible, then, that an increase in the minimum wage will raise expected wages.

We must be careful, though. Hamermesh notes that labor demand becomes more elastic as the skill level of workers decreases. It is difficult to find substitutes for workers with specialized skills. If you need an electrician, then you must hire an electrician, even if plumbers are much cheaper. If you need an airline pilot, you can’t hire a computer programmer. If you are running a store, however, and you find that labor is becoming more expensive, you might be able to upgrade your cash registers and inventory control systems and get by with fewer employees. We might expect the elasticity of demand at the minimum wage to be significantly higher.

It is therefore useful to look at studies that focus directly on minimum wage changes. One approach is to look at the relationship over time between the minimum wage and the employment experience of groups that are most directly affected by the minimum wage. Alternatively, one can look across groups that are differentially affected by the minimum wage to gauge its effect.

A reminder before we proceed. The point of going through this material is not only to help you understand the effects of a change in the minimum wage but also to provide you with a glimpse of how economic research proceeds so that you can do a better job of evaluating evidence that economists compile in all sorts of areas.

Economists have focused particular attention on teenage workers because that group is typically unskilled and is likely to be subject to the minimum wage. (This is not to say that the effects of the minimum wage are primarily on teenagers. Indeed, about two-thirds of minimum wage earners are adults.) Figure 10.18 "Teenage Unemployment in the United States" graphs the teenage unemployment rate over the period from 1956 to 2007 along with the real minimum wage for this period. We would like to know if the minimum wage has an impact on the teenage unemployment rate. This figure is suggestive of a relationship, particularly in the last couple of decades. For example, the real minimum wage fell during the 1980s, and the teenage unemployment rate also fell during that time. The teenage unemployment rate also often seems to increase around the time that the minimum wage increases. On the other hand, teenage unemployment also fell substantially during the 1960s at a time when there were several increases in the minimum wage.

If the only cause of changes in teenage employment were minimum wage changes, Figure 10.18 "Teenage Unemployment in the United States" might give us lots of answers. But this is a very big “if.” In terms of our analysis of supply and demand, it would amount to saying that the supply and demand for labor didn’t change over the entire period. In fact, you can make an enormously long list of things that might have shifted the supply curve, the demand curve, or both. Examples include whether the economy was in a boom or a recession, changes in tax rates, technological advances, and population growth. As we saw in Figure 10.17 "Difficulties Inferring Labor Demand from Data", it is difficult to disentangle the effects of a changing minimum wage from the effects of other changes. Thus we cannot isolate the effects of the minimum wage just by looking at diagrams like Figure 10.18 "Teenage Unemployment in the United States".

We need some way to take into account these other factors so that we can focus on the effects of the minimum wage. This is a complicated statistical problem that arises time and again in economics, and the ways of dealing with it go far beyond this textbook. Indeed, this problem is one of the hardest things about studying economic data. But even though the statistical details are complex, there are three simple ideas that are important to understand:

1. The basic idea of these statistical techniques is that economists attempt to include in their analysis—as best they can—all the factors that shift the demand and supply curves. Then they can predict what would have happened if there had been no change in the minimum wage. Once they have done this, they can then determine the effect of minimum wage changes.
2. When different economists study the same problem, they do not always reach the same conclusion. The reason that they disagree is usually not because they have different ideas about economic theory. Almost all economists have the same general framework for understanding labor markets, for example. Economists disagree because they use different approaches to take into account all the other factors.
3. If economists could conduct direct experiments, such as those performed in the physical sciences, they would not have this problem. They would be able to completely control the stimulus (minimum wage changes), and thus get an accurate estimate of the response (changes in employment). Instead, the best we can do is to look at past minimum wage changes and try to determine what happened.

Studies from the 1970s and early 1980s found some evidence that increases in the minimum wage reduced the employment of 16- to 19-year-olds. According to economists David Card and Alan Krueger, the average estimate is that a 10 percent increase in the minimum wage would reduce employment by about 1.5 percent.David Card and Alan Krueger, Myth and Measurement: The New Economics of the Minimum Wage (Princeton, NJ: Princeton University Press, 1995). So the implied value of −(elasticity of labor demand) is about 0.15. With an average employment rate of about 50 percent, this means that a 10 percent increase in the minimum wage would reduce the rate of teenage employment by about 0.75 percentage points.

It is striking that this estimate is lower than the estimate from looking at labor demand as a whole. What is more, Card and Krueger note that more recent studies produce even smaller estimates of these effects.David Card and Alan Krueger, Myth and Measurement: The New Economics of the Minimum Wage (Princeton, NJ: Princeton University Press, 1995), Table 6.2 "Calculating Revenues". In their own work, they find that the apparent negative effects of the minimum wage on employment are statistically insignificant. This result holds not only for all 16- to 19-year-olds but also when the sample is split by race and sex.

The evidence from teenage employment studies suggests that, as an empirical matter, the effects of the minimum wage on employment may be very small. This result has surprised many economists. Although economic theory did not suggest actual magnitudes for the labor demand elasticity, the existence of substitutes for unskilled labor did suggest that labor demand would be at least somewhat elastic.

Not surprisingly, there have been many other studies of the minimum wage. Some researchers have found larger employment elasticities for teenage employment than those reported by Card and Krueger. Sometimes we are simply unable to give a definitive answer to empirical questions in economics. This can be frustrating for both students of economics and practitioners—but we are not going to pretend that the world is simpler than it actually is.

Cross-Section Studies

The studies we have just discussed analyzed the minimum wage by looking at minimum wage changes over time. Another approach to analyzing the effects of the minimum wage is to take advantage of differences over individuals rather than variations over time. Economists call these cross-section studies because they look at a cross section of different individuals or firms at a point in time. Many of these studies look at the effects of minimum wage changes at the level of an individual worker. Others exploit differences in minimum wage laws across states. Such differences across states give rise to a natural experimentAn exogenous change that can be associated with a shift in either the demand curve only or the supply curve only. because they can substitute, at least in part, for economists’ inability to conduct experiments in which only one thing changes at a time.

Here is an example. Recall that in 1938 a minimum wage of $0.25 per hour was put into effect under the Fair Labor Standards Act. In the United States, this minimum wage was about 40 percent of the average manufacturing wage. However, the law also applied to Puerto Rico, which was much poorer, where this minimum wage was about twice the average factory rates. In a book published more than 25 years ago, researchers John Petersen and Charles Stewart noted that the increase in the minimum wage led to numerous factory closings, and there was a dramatic decrease in output and employment.John M. Peterson, and Charles T. Stewart Jr., Employment Effects of Minimum Wage Rates (Washington: American Enterprise Institute, 1969). In the case of Puerto Rico, the introduction of the minimum wage apparently had a large adverse effect on employment.

Peterson and Stewart also provide an extensive account of early studies of minimum wages that looked at employment and wages at individual production sites. These studies compared employment before and after a change in the minimum wage in an attempt to infer the effects of the policy. A study of the seamless hosiery industry from 1938 to 1941 is of particular interest. This period saw the introduction of the $0.25 per hour minimum wage, followed by an increase to $0.325 per hour in September 1939. A researcher named A. F. Hinrichs looked at 76 different plants and divided them into two groups: those that paid high wages and those that paid low wages. We would expect the minimum wage to have a much bigger effect at the low-wage plants. Between September 1938 and September 1939, the low-wage plants had employment losses of 12 percent. Employment at high-wage plants actually expanded by 23 percent, perhaps in part because workers who lost jobs in the low-wage plants became part of the labor supply for the previously higher-wage plants. A similar pattern was noted for the period from 1938 to 1940.

A much more recent study by Card and Krueger is another example of this approach. They studied employment patterns in fast-food restaurants in New Jersey and Pennsylvania. The key to their research was that, during the period of study, the minimum wage was increased in New Jersey but not in Pennsylvania. (Remember that individual states sometimes set minimum wages above the federal minimum.) From this natural experiment, Card and Krueger found that the increased minimum wage in New Jersey actually seemed to have increased employment.

The evidence from other countries (both cross-section studies and studies over time) is likewise mixed.A recent Organisation for Economic Co-operation and Development report has a summary of minimum wage studies from different countries: “Making the Most of the Minimum: Statutory Minimum Wages, Employment and Poverty,” accessed March 14, 2011, http://www.oecd.org/dataoecd/8/57/2080222.pdf. One study of Greek labor markets, for example, found a negative effect for men but a positive effect for women. Another study found negative effects for Mexico but not for Colombia. Different researchers in France have come to different conclusions about the effects of the minimum wage there; researchers in New Zealand likewise disagree; and so on.

Beyond Employment Effects

We started with what seemed to be some simple questions about the minimum wage. The answers turned out to be quite complex. Empirical research does not deliver a definitive answer about whether minimum wages have a big effect on employment. This leads to some disagreement among economists, particularly because the minimum wage is a politically charged issue. From the perspective of policymaking, the lack of a consensus creates difficulty in formulating good policy. On the other hand, the lack of a consensus provides a stimulus for continued work on these important issues.

Though we have emphasized the employment effects of minimum wage changes, there are other effects of minimum wages as well. First, remember that the main argument in favor of minimum wages is that they are a vehicle for redistributing income toward the working poor. Card and Krueger present a detailed analysis of the types of individuals most likely to be directly affected by minimum wage changes. Although empirical work often focuses on the employment of teenage workers, young workers are not the only group in the labor market that is paid close to the minimum wage. About 50 percent of the workers affected by the April 1990 increase in the minimum wage were older than 24 years old, for example.

How much income then flows to these workers as a consequence of an increase in the minimum wage? Card and Krueger conclude that the increase in the minimum wage during 1990 and 1991 had only a tiny effect on the distribution of income. They calculate that the minimum wage increase from $3.35 to $4.25 transferred about $5.5 billion of income to low-wage earners. This amounts to about 0.2 percent of family earnings. The host of transfer programs in place in the United States swamps the effects of the minimum wage on the redistribution of income. The evidence from other studies and countries is broadly in line with this conclusion: several studies find some effects of minimum wages on income distribution, but these effects are typically small.

Second, we can think about the effect of the minimum wage on firms. An increase in the minimum wage increases firms’ marginal costThe extra cost of producing an additional unit of output, which is equal to the change in cost divided by the change in quantity. of production. As a consequence, firms will increase their prices and sell less output. Because of this, increases in the minimum wage reduce profits, so we might expect to see this reflected in the share prices of firms that employ minimum wage workers. Relative to the large empirical literature on employment effects, the implications for employers have been largely neglected. Card and Krueger survey the evidence and find relatively small effects on the stock market value of firms.