Executive Summary

In fact, Secretary Reich grossly overstates both the number of poor people earning the minimum wage and the number of households dependent on a minimum wage worker's income.

• Only 3.7 percent of hourly wage earners are paid the minimum wage and most of those are not poor.
• A majority of minimum wage workers are either young persons living in nonpoor families or a second or third earner in a household - not the primary breadwinner.
• In 1992, only 198,500 of the nation's 4.7 million minimum wage earners were adult householders.
• Only 1.2 percent of all minimum wage workers (about 58,600) were adult heads of households with less than $10,000 of income.

Supporters of a higher minimum wage also frequently imply that a large portion of minimum wage workers are single mothers for whom welfare is an alternative to work. However, this belief is also disproven by the facts.

• Single parents, male and female, make up only 6.5 percent of the minimum wage workforce.
• Only about half of them (155,900) of a total 311,600) work full time. The number of poor people earning the minimum wage is small in part because most poor people of working age are not working.
• Only 9.2 percent of poor people of working age have full-time jobs.
• About 60 percent do not work at all.

Thus the minimum wage increase proposed by President Clinton would do little to reduce poverty. Instead, it would cause real hardship for some low-income Americans, the very people it is designed to help. A large majority of scholarly studies demonstrate that increasing the federal minimum wage causes higher unemployment. Those who suffer are most likely to be teenagers, racial minorities and low-skilled workers.

• Teenage unemployment rose sharply when the minimum wage was increased from $3.35 to $4.25 in 1990 and 1991.
• Furthermore, the extent to which teenage unemployment exceeds that of the whole population tends to increase in direct proportion to increases in the minimum wage.

All the evidence shows that the job-killing impact of the minimum wage is worse for blacks than for whites.

• Prior to the imposition of the minimum wage, the unemployment rates for blacks and whites were very similar.
• Today, however, the unemployment rate for nonwhites is about twice the rate for whites, and changes in the unemployment rate for nonwhites closely parallels changes in the real minimum wage.
• For example, the 1990 and 1991 increases in the minimum wage were accompanied by rising nonwhite unemployment, which reached a peak of 13.1 percent in June 1992.
• The minimum wage reduces on-the-job training opportunities that allow low-skilled, low-income workers to rise up the job ladder out of poverty.

An increase in the minimum wage would also shock the labor market and might trigger a recession, especially since this is a time of economic uncertainty. In the past, increases in the minimum wage have triggered recessions or prolonged depressions. For example:

• Evidence demonstrates a link between minimum wage increases and the recessions of 1990-91 and 1974-75.
• A good case also could be made that both the 1979-80 economic downturn and the recession that began in late 1981 were exacerbated by minimum wage increases, but this link is less obvious because inflation and other factors were involved
• A minimum wage requirement mandated by the National Industrial Recovery Act in 1933 halted an increase in employment, lengthening and deepening the Great Depression.

According to the president, a higher minimum wage is needed in order to make work more attractive than welfare. But if that is the goal, why stop with a $5.15 wage? To support a family of four at an income above the poverty level, a worker needs to earn more than $7.00 an hour.1 And if a $7 wage would lift everyone out of poverty, why stop there? Why not a $10 or $20 minimum wage? The fact that Clinton's proposal stops at $5.15 implies that even he is aware there are negative consequences from raising the minimum wage.

Actually, the harmful effects are quite severe - so much so that the minimum wage causes real hardship for some low-income Americans, the very persons it is designed to help. This is true for two reasons.

First, the minimum wage causes unemployment and reduces job opportunities for people for whom entry-level jobs are essential. The denial of an entry-level job keeps people from getting the on-the-job training that leads to higher labor productivity and higher income. Thus a higher minimum wage will keep many young Americans from taking the first step up the job ladder.

Second, minimum wage increases cause shocks to the labor market that can trigger recessions or prolong recessions or depressions already under way. By pushing labor costs up, an increased minimum wage prevents labor markets from creating jobs that routinely result from the dynamic changes in our market economy. A higher minimum wage in 1995 increases the risk of a recession in 1996.

Putting the two laws of economics together, higher wages increase the number of workers willing to work but decrease the number of workers employers will hire. Artificially raising wages by governmental edict creates a surplus of labor, better known as unemployment. Despite the appearance of some recent studies to the contrary, a multitude of evidence has led most economists to accept this conclusion. The magnitude of the unemployment effects of minimum wage increases can be questioned, but the existence of those effects cannot.

A minimum wage has the greatest impact on those with low skills whose normal wage would be less than or near the legally established minimum. However, as minimum wages are raised secondary effects tend to increase the wages of workers who earn more. Most products can be produced in different ways. For example, crops can be harvested with stoop labor or by skilled workers using expensive machinery. As a result, unskilled labor often competes directly or indirectly with semiskilled and skilled labor. This explains why the AFL-CIO regards an increase in the minimum wage as so important, even though almost all AFL-CIO workers earn far more than the minimum. Increasing the minimum wage is organized labor's attempt to price its competition out of the market.

Early History of the Minimum Wage Law.

• Needlework exports from Puerto Rico, where the law also applied, declined about 70 percent after the enactment of the Fair Labor Standards Act.2
• Nationwide, a government estimate concluded that between 30,000 and 50,000 jobs were lost as a direct result of enforcing the minimum wage.3
• Since the labor force has more than doubled over the years, that would be the equivalent of perhaps 100,000 jobs lost today.

Interestingly, the provisions of the Fair Labor Standards Act recognized the possibility that minimum wages could have employment effects. The basic legislation invoked the phrase -without curtailing employment- several times in describing options available to those administering the law. For example, a 25-cent-an-hour minimum was to begin October 24, 1938. This was to be in effect for one year, followed by a 30-cent minimum for the next six years and a 40-cent rate in 1945.4 The minimum could rise more rapidly towards the 40-cent level if administrators felt that this was possible without increasing unemployment.5

After 1945, the minimum wage was not increased further until January 25, 1950. At that point, it rose to 75 cents an hour. Prior to this increase, the 40-cent minimum represented about 30 percent of the average hourly wage. After the increase, it was over half the average hourly wage.

Economists Reach a Consensus. In a classic article published in 1946, Nobel Laureate George Stigler argued that minimum wage regulations reduced employment. But empirical evidence was still sparse.6 In subsequent years, though, as the federal government repeatedly increased the minimum wage, the opportunities to explore the empirical relationship between minimum wage changes and employment levels multiplied. In the third of a century following Stigler's article, a professional consensus developed around his basic thesis.7 This consensus was firmly established before 1980.8 In 1982, a survey of the economics literature concluded that minimum wage increases have significant negative effects on employment.9

At about the same time, both a Minimum Wage Study Commission established by Congress and the conservative American Enterprise Institute conducted additional investigations of the issue. Despite these groups' distinctly different policy perspectives, their findings were so similar that ...if one did not know which study had been funded by which group, one could not guess from the results. ...
The vast bulk of the research studies funded by the two groups show modest/moderate impacts consistent with the professional consensus.10

Recent Heresies.

By then, the orthodox view of the employment impact of minimum wages was being challenged. In 1992, Princeton economist David Card disputed this view.11 Lawrence Katz and Alan Krueger, the former and present chief economists of the U.S. Labor Department, sided with Card.12 Since then, others have joined and extended the debate.13 The popular press has cited these challenges to orthodoxy as a justification for increased minimum wages.14

Card, Krueger and Katz rely largely on evidence from changing state minimum wage laws to reach their conclusions that the unemployment effects of minimum wages are very small. However, Ronald Ehrenberg of Cornell has pointed out that these studies ignore the fact that raising the minimum wage forces some firms out of business.15 Finis Welch of Texas A&M University believes the authors have ignored the impact of other factors that might have increased employment had the minimum wage not been increased.16 Those who challenged the conventional wisdom have fundamentally violated the basic ceteris paribus (hold other factors constant) assumption critical to economic analysis.17

Further, when Card and Krueger studied low-wage jobs in the New Jersey and Pennsylvania fast-food industry, they maintained that the number of such jobs actually went up in New Jersey when that state raised its minimum wage, while the number of jobs decreased in neighboring Pennsylvania, where the minimum wage was not increased. However, Linda Chavez, a former staff director of the U.S. Commission on Civil Rights, says that teenage employment overall fell 28 percent in New Jersey while it was falling only 9 percent in Pennsylvania.18

The arguments advanced by Card et al. seem to have been answered adequately by David Neumark and William Wachser.19 Also, in a recent study of the restaurant industry, using a different methodology than that of most minimum wage studies, we obtained results that are strongly consistent with the earlier findings.20 Moreover, even the authors of other new studies do not appear to advocate minimum wage hikes. In fact, Krueger said, "I want to emphasize that my comments should not be interpreted as support for the position that increasing the minimum wage is sound public policy."21

• Teenage unemployment rates fell from the 22-23 per cent range in early 1983 to less than 15 percent by the beginning of 1990, at the same time that the real minimum wage was falling from about $5.15 to under $4.00 (in late-1994 dollars).
• The minimum wage rose in two steps in 1990-91 and teenage unemployment rose with it, peaking a few months after the new minimum peak in the second quarter of 1991.
• Subsequently, teenage unemployment rates fell, although not to the low levels of 1989 or early 1990, coinciding with the decline in the real minimum wage because of inflation.

On the whole, the teenage unemployment rate moves in tandem with changes in the real minimum wage.
A criticism of the reasoning behind Figure II is that the changes in the teenage unemployment rate reflect in large part changes in overall unemployment. The fall in teenage unemployment in the middle and late 1980s, for example, might be expected since unemployment in general was falling, probably for reasons largely unrelated to the minimum wage laws.

To deal with this, we calculated what might be termed the "excess unemployment" of teenagers, determined by subtracting the overall unemployment rate from the teenage rate. This calculation measures the extent to which teenage unemployment exceeds the norm for the whole population. Figure III shows the strong correlation between the real minimum wage and excess teenage unemployment. When the real minimum wage fell, the disparity between teenage and "normal" unemployment also fell; when the real minimum wage rose in 1990-91, the disparity grew, only to decline again in the past two years in response to the fall in the real minimum wage.

A picture is worth a thousand words. Figure III shows that teenage unemployment has been raised by increases in the real minimum wage over the past dozen years. The use of alternative statistical techniques (e.g., regression analysis) confirms the finding, and the relationship is highly significant in a statistical sense. While a more elaborate analysis incorporating a number of control variables would be needed to measure the unemployment effects with any precision, it is clear from this analysis that the loss of jobs from a 90-cent minimum wage hike would be measured conservatively in the hundreds of thousands.

• Beginning with the immediate post-Civil War era, nonwhite family income has remained quite stable as a percentage of white income.
• Representative estimates are 59 percent in 1880, 60 percent for 1907-09, 51 percent in 1947, 54 percent in 1964 and 58 percent for 1992.

Various explanations for these differences have been advanced: lack of education, limited skill levels and/or labor market discrimination. Whatever their cause, their existence indicates that employment levels of nonwhites are more likely to be affected by changes in the minimum wage. This is confirmed by Figure IV, which shows that the unemployment rate for nonwhites closely parallels the real minimum wage.

It is also worth noting that the unemployment rate for nonwhites typically is twice or more that for whites.24 This was not true prior to the widespread adoption of a minimum wage in the United States.

• In the three decennial censuses prior to the enactment of the Fair Labor Standards Act in which unemployment data are available (1890, 1900 and 1930), the white unemployment rate averaged 5.82 percent and the nonwhite 5.90 percent.

• Only after the enactment of a national minimum wage law did nonwhite and white unemployment begin to diverge.

Second, in the 1930s many felt that underspending had caused the Great Depression. Some wage-increasing legislation, notably the National Labor Relations Act of 1935, explicitly incorporated this view. The argument was that increases in wages would lead workers to spend more, reducing the unemployment rate that remained in double digits for a decade.

Most modern economists reject the notion that demand stimulus is likely to have any long-term impact on output or employment. A significant number of economists doubt that demand stimulus even has a positive short-run effect. Even if raising wages did stimulate demand, however, the economy is currently operating at about the lowest levels of unemployment since the 1970s.26

What about the effects of the minimum wage on poverty? Poverty today is a much smaller problem and a different kind of problem than in the Depression era. While a significant amount of poverty persists in America today, most poor persons of working age are not working, and most workers who earn the minimum wage are not poor.

Economic Finding: Most Poor People Are Not Working. According to the latest data from the Current Population Survey:27

• Only 9.2 percent of poor people of working age have full-time jobs. [See Figure V.]
• Even if part-time work is included, very few heads of poverty households are in the labor market.

  Thus the problem is more a lack of labor force effort than of low compensation. Even if a higher minimum wage lifted every full-time worker above the poverty level, this would affect less than 10 percent of all poor families. Today, a majority of women work, so even where income is constrained by low wages, the presence of second or even third earners increases household income, often to well beyond the poverty level.

  Economic Finding: Most People Who Earn the Minimum Wage Are Not Poor. Only 3.7 percent of workers paid hourly earn the minimum wage or less and most of them are not poor. A majority of minimum wage workers are either young persons living in nonpoor families or they are the second or third earner in a household - not the primary breadwinner. Many come from prosperous middle-class homes. For example, as Figure VI shows:28
  

• More than 44 percent of part-time workers in early 1992 receiving a wage of $5 an hour or less were under age 21.
• Less than 46 percent of those making $5 an hour or less were working full time.
• Even among adults, in 1992 more were members of families with incomes over $20,000 than with incomes less than $10,000.29
• In short, a majority of workers making $5 an hour or less do not live in poverty. Secretary of Labor Robert Reich claims that nearly two-thirds of minimum wage workers are adults and that more than one-third are their families' sole breadwinners.30 The first assertion is correct; the second is not. In 1992, there were 4,735,000 minimum wage workers and 3,047,000 were over age 20. However, there were only 198,500 minimum wage-earning non-teenage householders. Further, about three-fourths of teenage minimum wage workers lived with their families. The numbers simply do not add up.31 As Figure VII shows:32
• Only 1.2 percent of all minimum wage workers (about 58,600) are adult heads of household with less than $10,000 of income.
• Of the 198,500 adult heads of household, 70,700 are in households with incomes of $10,00-$20,000, and 69,200 are in households with incomes exceeding $20,000. Single parents make up 6.5 percent of the minimum wage workforce. Only about half of them (155,900 of a total 311,600) work full time.33 [See Figure VIII.]

Does the Minimum Wage Increase Poverty? To get some insight into this question, we looked at the poverty rate in the last full calendar year before each minimum wage increase since 1974 and compared it to the first full calendar year after each increase. From 1974 through 1981, the minimum wage was increased every year, so the comparison is between 1973 and 1981 (the increase in the latter year was on January 1).

After an increase in the minimum wage to $3.80 in 1990 and $4.25 in 1992, the poverty rate climbed from 12.8 percent to 14.5 percent. [See Figure IX.]

In both cases, the poverty rate rose after the minimum wage increase was enacted. This is consistent with the evidence that minimum wages only marginally touch poor Americans and that the unemployment they create worsens the economic status of many low-income workers. While this simple comparison ignores other determinants of poverty, it strongly suggests that the minimum wage is more likely to create poverty than to eliminate it.

Do Minimum Wage Increases Trigger Recessions or Prolong Depressions?

Minimum wage increases introduce wage shocks into the overall economy. The impact of such disturbances is illustrated by recent business cycles.

The 1990-91 Recession. In 1989 the American economy was prosperous and was enjoying the longest peacetime expansion in American history. The expansion continued into 1990, with unemployment rates staying remarkably stable at about 5.3 percent. Yet labor markets began to weaken significantly in the second quarter of 1990. It is not a coincidence that the minimum wage was raised by 45 cents (more than 13 percent) on the first day of that quarter. Figure X shows that while more than 800,000 jobs were created in the civilian economy in the six months prior to the minimum wage increase, over 350,000 jobs were lost in the first six months afterward.

Within three months of the wage increase, the unemployment rate began rising from the 5.3 percent level, reaching an average of nearly 6.5 percent in the first quarter of 1991. Natural market forces were ready to start lowering unemployment, but they were thwarted by a second 45-cent increase in the wage, to $4.25, in April of 1991. Rather than falling, as often happens a year after a recession begins, the unemployment rate continued to rise, reaching 7.0 percent by the end of the year.

The 27 percent rise in the minimum wage had a general wage-increasing impact, since an increase in the minimum wage increases the demand for skilled and semiskilled workers who indirectly compete against unskilled labor. Compensation per hour in the business sector, which had risen at a 3.5 percent annual rate in 1989, rose at an extraordinary 8.4 percent annual rate in the second quarter of 1990, the biggest increase in many years.34 This wage shock priced many workers out of the market and led to the end of the great job explosion of the Reagan era. By early 1991, the annual rate of compensation increases had fallen to below 3 percent on an annual basis, and the market was showing signs of adjusting to the wage shock of April 1990. However, in the second quarter of 1991 compensation increases rose to 4.6 percent on an annual basis, almost certainly a consequence of the April 1991 increase in the minimum wage from $3.80 to $4.25. This led to a further delay in job recovery.

The 1974-75 Recession. The 1990-91 recession is not unique in being at least partially triggered by a minimum wage shock. A good case can be made that both the short 1979-80 downturn and the recession beginning in late 1981 were exacerbated by minimum wage increases in 1979, 1980 and 1981.35 However, other factors were at work and inflation made this period somewhat atypical. Perhaps a better example is the 1974-75 downturn, which clearly was affected by minimum wage increases.

On May 1, 1974, the minimum wage was increased for the first time in over six years. The increase was a very steep 25 percent, as the wage went from $1.60 to $2.00 per hour. Special minimum wages then in effect for some employees who had begun to be covered after 1966 (e.g., farm workers) were increased roughly proportionately. The unemployment impact was almost immediate, as Figure XI shows. In the four months prior to the enactment of the minimum wage, the unemployment rate was virtually constant at around 5 percent. In the month the minimum wage increase was enacted, the unemployment rate began to rise. By the end of the year, the unemployment rate was at 7.2 percent.36

Compounding the problem, on January 1, 1975, the minimum wage rose another 5 percent, and proportionately more for workers temporarily covered by lower federal minimum wages. The unemployment rate soared in January 1975, and by spring of that year was nearing 9 percent. In the year between the first quarter of 1973 and the first quarter of 1974, compensation of all workers in the private economy had risen 7.3 percent. In the next six months, the annualized rate of increase in compensation rose to about 12 percent. A wage explosion occurred. A sharp slowdown in the rate of compensation increase that normally would have occurred was delayed by the further 1975 wage increase. Again, low-skilled labor was priced out of the market.

The Great Depression: Wage Increases Under the NIRA

. A powerful example of the impact of the minimum wage comes from the pre-Fair Labor Standards Act portion of the 1930s. In fact, instead of describing the early 1930s as the Great Depression, we could describe them as a period of high wages and low employment.37 Much of the public policy analysis of the time focused on the importance of maintaining purchasing power by keeping wage rates high.38 Herbert Hoover argued vigorously against any reduction in money wages at the outset of the Great Depression. At the conclusion of its renowned "first hundred days," Franklin Roosevelt's New Deal made a strong commitment to the same principle. The National Industrial Recovery Act, passed in June 1933, required that a minimum wage be included in any industrial code. As it evolved, the actual minimum wage was generally about 40 cents an hour.39 This was remarkable, since a 40-cent-an-hour minimum wage represented more than 90 percent of the average hourly wage.40 The impact on wage rates was dramatic, driving them upward by almost 20 percent in the last half of 1933.41 The timing of this surge was unfortunate. From March to July, the unemployment rate had fallen by a full 5 percentage points, indicating that an economic recovery had begun.42

These data understate the impact of the codes on employment. In addition to the minimum wage provisions, the codes contained maximum hours requirements. They generally were set at 40 hours per week, below the average workweek of the time.43 Consequently, as unemployment rates stopped declining, the average workweek fell by 13 percent between June and December 1933.44 Over the next 15 months, unemployment rose slightly, standing at 23.5 percent in October 1934. The following year saw a slight improvement, but unemployment still measured nearly 22 percent in October 1935. By this time, the National Industrial Recovery Act had passed from the scene, having been declared unconstitutional earlier in the year by the Supreme Court. In the absence of the act's minimum wage provisions, employment conditions improved dramatically. By May of 1937, the unemployment rate had fallen to almost 12 percent. Again, all the evidence points to the same conclusion: If we introduce a wage shock in the form of a hike in the minimum wage, unemployment rises; if we allow the minimum wage to fall from the effects of inflation or court rulings, unemployment falls.45

Conclusion

President Clinton has called for an increase in the minimum wage in order to alleviate poverty. Since a very large proportion of the working-age poor do not work, and many minimum wage workers are young persons who are second or third family earners, it is questionable whether the minimum wage is an effective way to combat poverty - even in the absence of any employment effect. However, the evidence shows a very clear correlation between increases in the federal minimum wage and the level of unemployment. In particular, such increases cause greater joblessness among some of the most helpless and disadvantaged members of our labor force, including teenagers and members of racial minorities.

But that is not all. By trying to evade the labor market imperatives imposed by supply and demand, minimum wage laws cause labor market disruptions that can and do trigger economic downturns. The 1974 and 1990 recessions coincided with large minimum wage increases. Recovery from the Great Depression was slowed by the imposition of an early version of the federal minimum wage.

A higher minimum wage sounds like the right thing to do. It seems a reasonable attempt to redistribute income and reduce poverty. But a higher minimum wage means fewer jobs for low-income Americans and may trigger a recession. On balance, the evidence suggests that a higher minimum wage tends to increase rather than reduce poverty.

Moreover, the national income data show that Americans are extremely mobile with respect to income status, with a large percent of the working poor of one year being in the middle class a few years later.46 This happens in part because holders of low-paying jobs work their way up, becoming more productive through on-the-job training. For example, based on 1992 data we estimate that relatively uneducated (less than ninth-grade education) male workers aged 50-54 are more than 60 percent more productive than their counterparts aged 18 to 24.47 By denying workers initial employment opportunities, the minimum wage thwarts the potential for economic improvement that comes with hard work and learning. In short, it thwarts the American Dream.

---