Toward an Apartheid Economy?
Fears at the turns of a century
By Richard Freeman
Richard Freeman is the Herbert Ascherman Professor of Economics at Harvard University and program director for labor studies at the National Bureau of Economic Research, both located in Cambridge, Massachusetts. His most recent book is "What Workers Want," Cornell University Press.
(Originally published in Harvard Business Review, Sept/Oct 1996.)
1. Something is wrong
Rising earnings inequality. Stagnant real wages. A declining middle class. High child poverty. Insecure workers. A dying union movement. Homeless people. Bursting jails and prisons. A falling social safety net.
A third world economy heading toward disaster? Or perhaps a bizarre neo-Marxian vision of the future of capitalism? Or is this a description of the United States moving into the 21st century?
Ten to fifteen years ago, you or I would have readily dismissed such talk as soapbox radicalism. Save for the Great Depression, the history of the U.S. has been one in which economic growth benefits all. From 1910 to 1973 the average employee enjoyed substantial gains in real earnings and increases in leisure time. No honest observer could question the success of American capitalism in delivering middle class standards of living to workers. But today, despite the nation's great success in creating jobs and maintaining relatively low unemployment, it is not only radical soap box orators who worry about the ability of our economy to take care of working citizens.
From former Secretary of Labor Robert Reich on the left to Patrick Buchanan on the right, and from the business press to the mass media, comes the message that something is wrong with the U.S. economy. That something is not the country's productivity, technological leadership, or rate of economic growth. That something is an issue normally on the backburner in U.S. public discourse: the distribution of the fruits of economic progress. For many, the rise in AT&T; stock prices after it announced plans to lay off 40,000 employees crystallised the picture of an economy gone haywire, with shareholders gaining and employees losing as a result of innovation and advances in productivity.
Has the distribution of the benefits of economic growth in the United States in fact gone awry? Is the nation heading heading toward an apartheid economy -- one in which the wealthy and powerful prosper while the less well-to-do struggle? What are the facts? What do they mean? Are there real problems -- and can they be solved?
2. The Sinking American WorkerThe United States has always has had a greater degree of inequality than other advanced countries. Inequality, some would say, is the flipside of opportunity and mobility. Some level of inequality is needed to elicit effort, and Americans have never been particularly attracted to egalitarianism. But in the past two decades our normal high level of inequality, except in the category of gender, has jumped:
So what? you may say. Who cares about inequality? As long as living standards are improving for the vast majority, there is no reason to ring alarm bells. Put aside envy or jealousy and enjoy the fruits of progress. But the story of recent decades is not one of economic gains for all. On the contrary, the rising tide of inequality has been accompanied by stagnant real wages.
Until the early 1970s, the history of the U.S. job market in this century was one of rising wages -- about a 2% increase per year. Compounding that figure leads to a doubling of income every 35 years, virtually guaranteeing higher living standards from one generation to the next.
The rate of growth of real wages from 1973 to the mid-1990s is different. The trend is downward for the most widely used statistical measures of real wages -- hourly earnings reported by workers on the U.S. Current Population Survey of households divided by the consumer price index, and earnings reported by employers on the U.S. National Employer Survey, also divided by the CPI. According to the population survey, the median weekly earnings of full-time male workers have dropped by 13% from 1979 through the mid-1990s. According to the employer survey, the average weekly earnings of nonsupervisory private-sector workers (outside of agriculture) have fallen by 12%.
But these measures of real wages are imperfect; for various reaons, they are likely to understate the growth of real earnings. They exclude nonwage compensation -- money that goes into health insurance and pensions, for example -- which has risen as a share of total wages. And the national employer survey obtains earnings solely for nonsupervisory workers. The most serious problem is that the CPI overstate inflation because it does not fully measure improvements in the quality of goods. And if the rise in inflation is overstated, the growth of real earnings is understated. How much the CPI may be in error is a matter of debate.
Regardless of where one comes out in this debate, there is no gainsaying that the growth rate in real earnings has taken a dive. The growth of fringe benefits has slackened over time, so the bias resulting from their omission has become smaller. And the U.S. Bureau of Labor Statistics has continually improved the CPI. Thus the understatement of growth in real earnings was greater in the past than it is now. In fact, better data may demonstrate an even more severe decline in the rate of growth of real wages in recent decades. Whatever the "true" figures are, the problem of stagnant or falling real wages in combination with rising inequality has created a new group of Americans.
3. Prime victims: less-educated menIn the United States, the big losers in the earnings distribution are men, particularly young or less-educated men. If the fall in earnings greatly increased their employment, the decline in real earnings would have produced at least one benefit: more jobs. But instead, their prospects for work have deteriorated. As inequality in earnings has risen, so has inequality in the hours people worked over the year. The yearly earnings of the low paid and less skilled have fallen by a larger percentage than their hourly pay. In addition, the likelihood that employers will provide low-skilled workers with fringe benefits such as health insurance or pensions has also dropped. The fall in the real earnings of the low paid means that a substantial number of working men earn less than their fathers did two or three decades ago.
Furthermore, low-paid American men are losing ground in comparison with their counterparts in other advanced countries -- even as the United States remains the most productive economy in the world. In Western Europe, a male worker in the bottom 10% of the earnings distribution earns 68% of the median worker's income. In Japan, that male worker earns 61% of the median. In the United States, he earns 37% of the median. And a comparison of purchasing power reveals that U.S. workers in the bottom 10% earn less real pay than their counterparts in other advanced countries. Low-paid German workers earn 2.2 times more than low-paid Americans; low-paid Norwegian workers, 1.85 times more. Comparisons of wages based on exchange rates, which are appropriate for business costs, show even greater differences.
It is not just the bottom 10% of U.S. workers that trail the workers in other advanced countries, however. Approximately one-third are paid less, in purchasing-power units, than comparable workers overseas. Not until the third or fourth decile of earnings do Americans do better than Europeans. That so many American workers fare poorly compared with their peers in other countries shows that the problem of low pay is not simply a matter of low-skilled immigrants or poorly educated minority youths. It is a problem of the overall distribution of income.
Because taxes are higher in many countries, comparisons of pay before taxes may exaggerate the disadvantage of low-paid Americans. That would surely be the case if taxes gave nothing back in value, but in most countries they pay for health insurance, pensions, housing subsidies, and other social benefits -- the bulk of which Americans purchase from their earnings. Comparisons of wages after taxes are thus likely to understate the U.S. disadvantage in pay among low-paid workers.
Nor can U.S. workers depend heavily on the government when they fare poorly in the marketplace. In much of Western Europe, economic assistance programs have created a huge welfare state, which many rely on for years. In the United States, such programs provide an effective safety net provided that workers rebound quickly from their economic problems. But the safety net is weak and has been shrinking in recent years. Declining pay and hours worked translate more commonly into poverty than is the case in other advanced economies. Low-paid Americans who lose their jobs may end up in big trouble: they often lose health insurance, have their cash flow dry up, and have difficulties paying their rent and consumer debt.
The principal way in which poverty is reduced in the United States -- except for the elderly who are supported by Social Security -- is through higher wages in the job market. As real earnings have stagnated, the downward trend in the poverty rate has slackened, and the rate has increased for families of young workers and for children. Many children are brought up in single-parent families whose welfare benefits have been reduced and whose heads have not had much success obtaining work, though various workfare-style welfare reform proposals may alleviate that problem. But single-parent families on welfare are just one part of the poverty probem. Compared with most other advanced countries, the United States has a higher rate of fully employed working poor. In the 1980s, poverty rates in Canada dropped below those in the United States, not because the Canadian economy boomed relative to the U.S. economy (it didn't) but because Canada has stronger safety net programs.
The absence of a significant safety net has some advantages. Taxes and fiscal deficits are lower than in most other advanced countries, and citizens have greater incentives to work. But it also means that Americans must rely more heavily on the labor market for their well-being. When the job market turns bad, as it did for the less skilled and the low paid in the 1980s and 1990s, people suffer.
4. Who's Responsible?Nearly everyone has a list of culprits. Pat Buchanan's list focuses on trade and immigration. The AFL-CIO's list includes declining unionization. The Clinton Administration stresses technological change. The Republican Congress blames the administration, high taxes, and excessive government spending. If we could identify a single culprit, we'd have a good television drama. But the story of rising income inequality is more complex. Several factors have been at work, and honest analysts weigh their importance differently. Let's examine each factor in turn.
Trade has contributed to the problems of low-wage workers. Losses in pay or employment by those who make products in competition with foreign imports, particularly imported goods from less-developed countries, are the flip side of the much larger benefits that trade brings the nation. Ross Perot and Sir James Goldsmith claim that trade is the main story, but most studies disagree. The estimated effects of trade on the demand for low skill labor as opposed to high skill labor simply are not big enough for it to dominate the distribution of income in the U.S.: less-skilled women are most affected by imports from third-world countries, and they have not suffered the losses of less-skilled men. Most Americans work in service sector jobs that are not directly affected by trade, and the estimated effects from manufacturing and other traded-goods sectors on the rest of the economy are not large enough to account for the huge rise in inequality.
Immigration has also contributed to the problems of low wage native-born Americans. Fully one third of employed high-school dropouts are foreign-born. If they had not immigrated, the wages and employment of their native-born counterparts surely would be higher. But our best estimates suggest that the increased flow of low-skilled immigrants is not the major cause of falling earnings for less-skilled Americans, although it is probably a more important factor than trade. Much of the rise in inequality in the United States has occurred among workers with vocational skills that are unaffected by immigration, such as those of mechanics or lawyers.
The decline of trade unions has contributed to inequality. Unions reduce inequality by: standardizing pay among workers, bringing the pay and benefits of their members closer to that of higher-paid executives and professionals, and inducing nonunion companies to raise pay or benefits to avoid unionization. Imagine if 30% of the private sector were unionized, as it was in the 1960s, instead of 11%, as it is today. Would there be less inequality in earnings? Yes. Studies suggest that about one-fifth of the rise in inequality is due to the decline of unions.
The fall in the real value of the minimum wage has also contributed to rising inequality. However, the impact of the minimum wage is felt mostly in the bottom decile, particularly by women and teenagers. In the U.S., changes in the minimum wage do not spill over to as many workers as such changes do in France and other countries..
The Clinton Administration has favored technology as the culprit, perhaps because it seems to let policymakers off the hook. No one knows how to predict or control the direction of technological change. There is evidence that comcomputerization has adversely affected low-skilled workers, but the case for the technology as the man culprit is underwhelming. Why hasn't the same technology produced massive inequality in other advanced countries? Why, if we have entered a brave new world of accelerating technology, have increases in productivity been moderate? Every generation deals with new technology and often blames it for whatever ills occur: in the 1960s, we had the great automation scare, for example, which blamed rising unemployment on machines and robots.
The influx of women into the job market may also have contributed to the problems of low wage workers: an increase in the supply of labor drives wages down. But because women's pay has risen relative to that of men, and because women and men generally are employed in different industries and occupations, it is difficult to see how women's increased presence in the workforce adversely affects men. Not even the U.S. Senate's Whitewater Committee can blame Hillary Clinton for this one.
Another factor has been the slowdown in the rate of growth of the supply of college graduates. In the 1960s, Jan Tinbergen, the Nobel prize-winning Dutch economist, wrote that inequality was a race between the increasing demand for highly educated workers due to technology and the increasing supply of these workers due to expanded public education. For most of Western economic history, the increase in the supply has offset the increase in the demand for them. In the 1970s, the increase in supply in advanced countries dwarfed the increase in demand, reducing the economic value of education. In the United States, higher education lost some of its attractiveness, and the rate of growth of the college work force decelerated in the 1980s, contributing to the rise in the earnings advantage of the more educated. In countries where the educated work force grew more rapidly in the 1980s, pay differentials did not widen much, if at all. In some cases, they even fell. But again, the 1970s slowdown in the flow of young men into college is not sufficient to explain the massive rise in inequality.
The debate over how much weight to give to these various factors has enlivened economic conferences and journals. It is an interesting debate, but it is largely irrelevant to the policy question of how, if at all, the nation should respond to the rise in inequality. That is because there is no necessary link between the causes of the problem and potential cures. When someone has myopia, a largely genetic disease, we cure it with glasses or contact lenses. We do not mess with the genes, although they may be the root cause. When you have a headache, you take aspirin to cure it rather than go into a long exegesis about where the headache came from.
The same reasoning applies to the problem of rising inequality and stagnant real wages. If a cause of the rise of inequality is trade, should we build a tariff wall around the country? No. Protective tariffs are one of the most inefficient ways to redistribute income that the mind of man has ever conceived. If the United States wanted to help citizens suffering from foreign trade, it could do so with other tools. And if the cause of rising inequality is technological change, should we become Luddites, smash our PCs, and march on MIT with pitchforks? The heart of the matter is not how we got into this fix, but whether we should care about it and what we can do about it.
5. Should We Care?Rising inequality and falling real wages reduce the living standards of individuals and families on the lower rungs of the income distribution. Many of those affected will feel alienated from society and behave accordingly.
For people with high incomes, however, inequality has some benefits. If I am rich and you are poor, I can hire you cheaply as my gardener, maid, or nanny. Not surprisingly, the personal services sector has grown with the rise of inequality. Moreover, if the middle class shrinks and buys fewer tickets to basketball games and concerts, there will also be more places for the well-off at these events.
Still, many people in higher income brackets are upset when fellow-Americans and their families struggle in a land of plenty. Those distressed by the poverty of others range from bleeding-heart liberals and religious people to the philanthropists and volunteers whose contributions to charities give the United States the largest charitable sector in the world.
Personal feelings aside, inequality can also impose costs on the entire society, including the rich who don't give a hoot about how their poorer neighbors fare. One such cost is the high rate of crime, which has generated a growing prison population and huge expense for the criminal justice system. By 1995, more than 2% of the U.S. male workforce was incarcerated, and nearly 7% was under the supervision of the criminal justice system. Mass incarceration may be the reason the crime rate is finally falling, but it is an incredibly expensive way to deal with the problem. Incarcerating a criminal costs about as much per year as sending someone to Harvard. In 1995, the state of California budgeted more for prisons than for higher education. While there is no smoking gun linking the high crime rate to falling real wages or employment, the economic logic is clear: whatever makes legitimate work less attractive to young men makes crime more attractive. Most studies support that logic.
Another cost is found in the sense of insecurity at work. When inequality is high, people worry that they make have to work for less pay if they lose their current job. That may underlie the great concern over downsizing among middle aged men, who are typically unemployed for for many months if they are laid off and take a 20% or so wage cut when they finally land a new job. Insecure workers may be less productive, less willing to share their knowledge, and less likely to talk about problems with supervisors than more secure workers. The recent repudiation of the virtues of downsizing byStephen Roach, the chief economist at Morgan Stanley, shows that even proponents of lean companies recognize the cost to productivity of a workforce frightened about its future.
Falling real eanrings of the low paid may also breed antagonisms between groups as well as social instability. Animus against CEOs and Wall Street is a logical consequence of rising inequality. When my earnings are falling despite my hardest efforts, whom shall I blame? If year after year, you are getting wealthier and I am getting poorer, who knows what demogoguery might appeal to me?
Inequality has consequences for family structure, as well. There is nothing more anti-family than the falling level of earnings and job prospects of less-educated young men. If a young father cannot earn enough to support a family, he is unlikely to try to form one. If he engages in crime and ends up incarcerated, one result is more fatherless children. Hundreds of thousands of children in single-parent homes in the United States have fathers in jail or prison.
Inequality also affects business in various ways. Some businesses -- those providing services to the rich -- may gain; others -- those producing for the great middle class -- may lose. Consumer credit defaults may rise. Savings by ordinary citizens, already low, may fall. As the divide between the bosses and subordinates grows, many may come to view business unfavorably and support policy interventions harmful to the nation's economy. How much are you making, Mr. CEO?
The list of costs -- crime, insecurity at work, stress on families, skewed business decisions, and political turmoil -- is neither exhaustive nor definitive. The magnitude of the costs is not well known, because there is little in our history to provide guidance. But the ultimate cost of increasing inequality lies in the potential for an apartheid economy, one in which the rich live aloof in their exclusive suburbs and expensive apartments with little connection to the working poor in their slums. Just as many South African whites were blind to the plight of nonwhites, so too in an apartheid economy will many of us be blind to citizens of lower economic status -- if we are not blind to them already. When was the last time you were shocked by the homeless?
6. Preventing an Apartheid EconomyIf you are happy about this picture of the emerging U.S. economy -- if you see the benefits of inequality as greater than the costs -- stop reading. The remainder of this article considers ways to reduce inequality and rebuild the American middle class. It is a menu of possibilities, not a sales pitch for any specific policy. The economy is a highly complex system that continually surprises us; no one can be sure what will work best to alleviate the trend toward inequality. Still, it is important to put on the table at least some policies that might prevent the development of an apartheid economy.
For starters, many policies on the front burner today won't do much to raise up those at the bottom of the earnings distribution. For instance, some supply-siders believe that faster economic growth by itself will cure the woes of the low paid, and that tax cuts for the wealthy (embodied in a flat tax) or reduced capital-gains taxes are the way to spur that growth. Such views fly in the face of recent experience. The sad lesson of the expansion of the 1980s and 1990s is that even even very healthy growth may not improve the pay or income of those in the lower parts of the income distribution. The Reagan-Bush recovery from the 1981-1982 recession was one of the longest and most impressive in U.S. history: employment and income grew significantly. If divided evenly, the $2 trillion increase in the GDP from 1980 to 1994 would have been enough to give roughly $2,000 more to each American family. That's not what happened. Median family income was stagnant. Only the upper 20% had healthy gains in real income, with most of those gains concentrated in the top 5%. The proportion of aggregate income going to the top 5% rose from 15.3% in 1980 to 19.1% in 1993, while the proportion going to 80% fell. Men and especially women had to work more to maintain their standard of living. Nearly a decade of gains in the GDP and in employment were accompanied by falling incomes for the bottom half of American families! Growth alone is no cure for the problems of rising inequality and stagnant real wages.
Some believe that further deregulation of the economy will create more and better jobs. This is the standard solution of the Paris-based Organization for Economic Cooperation and Development for nearly any economic problems. But the United States has the least regulated economy among advanced countries. Its problem is not creating jobs but making work pay. There are reasons to cut regulations, but preventing an apartheid economy is not one of them.
Others believe that cutting the federal deficit will improve the economy so much that real earnings will begin rising rapidly once the markets are assured that deficit reduction is for real. I know of no historical support for that claim, even if cutting the deficit in fact increases investment and growth. There are reasons to cut the deficit, but boosting the earnings of the low-paid is not one of them.
Education and job-training programs for the less skilled are popularly suggested solutions to the problem. In the long run, education and training will help reduce inequality by increasing the number of people in higher-paying jobs and by diminishing the number competing for low-wage jobs. But, on their own, such supply-side changes have not yet shown themselves capable of offsetting the market forces that favor the highly paid and skilled.
Many economists expected that the wages of young workers would rise once the baby boomers aged. The logic for this expectation was impeccable: the decline in the number of young people after the baby boom would cause a shortage of young workers in the market -- and shortages raise prices. But wages did not rise; instead, the economic position of young job seekers worsened in the 1980s and 1990s. In response, more young people flocked to colleges and universities than in the past. Eventually, the growing number of college graduates should reduce the difference in pay between graduates of high school and college, but for now the gap remains high. As for government-funded training programs designed to improve the skills of those who do not go to college, even the best such programs increase wages or employment only modestly and cannot conceivably restore losses of earnings of 20% to 30%.
Will higher tariffs, repeal of NAFTA, or other protectionist measures cure the inequality problem? The economic costs of protectionism so dwarf the potential benefits that they make the cure worse than the disease. What about policies to reduce immigration? They would cut the supply of low-paid workers, but if the massive decrease in young Americans entering the job market after the baby boom did not improve their economic position, why should modest changes in immigration do the trick?
The problem with many of the policies described above is that they approach low wages and rising inequality indirectly rather than head-on. The topic of income distribution has never been popular in the United States. In a society where economic growth benefits everyone, it is sensible to worry about efficiency and growth first and distribution last. In other words, let the Swedes worry about whose slice of the pie is bigger -- we'll worry about baking a bigger pie.
But new problems require new solutions, and it behooves us to examine more carefully the redistributive tools at our disposal if we wish to prevent an apartheid economy. The earned income tax credit, which gives cash payments to low-wage workers, tackles the problem of low incomes head-on. At one time, both conservatives and liberals favored it, and the Clinton Administration increased the credit in its first year in office. Another redistributive tool is the minimum wage. Like the earned income tax credit, is is not perfect -- it is not well targeted at the poor and cannot be raised too high for fear of increasing unemployment -- but it does buttress the bottom parts of the earnings distribution.
And we should look at the payroll tax that funds social security: the tax is a high, regressive levy that does nothing to increase national capital investment. Perhaps we should combine privatization of social security with cuts in the payroll tax for low-paid workers in forming a new national retirement plan. Finally, a more lively and successful trade union movement -- led by a truly new AFL-CIO that succeeded in winning a raise for workers -- would help reduce inequality. Perhaps business should rethink what the "union free" private sector toward which we are headed really means for society.
Having warned you that no one can be sure of the best way to reduce inequality and that some policies won't work much, if at all, I would be insulting your intelligence and mine if I did more than list possibilities. A judicious mixture of policies might prove effective. Perhaps another mixture might backfire. You may have better policies to suggest. Perhaps, magically, the invisible hand will resolve the problem without our help -- although if I thought that likely, I would have written a very different article. At this stage, we need to recognise that the country has an inequality problem based on falling real wages for low-paid workers that is unparalleled at least since the Great Depression. We need to begin debating ways to prevent the United States from developing an apartheid economy. We have a great record solving problems once we recognize and put them on the national agenda. Rising inequality and falling wages for those at the bottom of the income distribution are real problems. Let's start reasoning together about how to solve them.
Copyright © 1996 Richard Freeman. Reproduced by permission of the author.