Making Sense of the Income Gap Debate
The new economy has brought prosperity to many, but not to all. If the income gap continues to widen, politicians will take action - and business will bear the brunt.
By Frank Levy
Myths of Rich & Poor: Why We're Better Off Than We Think
W. Michael Cox and Richard Alm
Basic Books, 1999
The State of Working America 1998-99
Lawrence Mishel, Jared Bernstein, and John Schmitt
In the midst of the booming U.S. economy, nagging questions remain: How well are average workers doing? Are rich people getting the lion's share of prosperity?
Economists hold fervent opinions about such questions. The intensity of our disagreements isn't surprising since we are talking about money and perceptions of fairness.
Popular beliefs about how the economy distributes its gains will go far in determining public policy. How do Americans feel about raising the minimum wage? Should a CEO's stock options be valued in a company's income statements at the time they are awarded? Should the steel industry be protected from import dumping? Should the capital gains tax be lowered? Many paychecks ride on the answers to those questions. And the answers depend in large part on how we perceive income trends for various groups.
Unfortunately, it's difficult to analyze such trends in an unambiguous way. With so many different numbers available, an analyst has a lot of discretion in deciding how to spin them. Some 270 million people now live in 100 million households, and they make and spend money in many different ways. In assessing economic well-being, should we focus on individuals or households? Consumption or income? We measure income trends with a variety of statistics on wages, dividends, and nonmoney income like fringe benefits and Medicare. Rarely do these statistics fit together well. And since most statistics have no natural yardstick, an analyst can make a particular number seem good or bad through the choice of comparisons.
It follows that someone with a strong point of view, whatever it is, should have no trouble finding supporting data. Two new books, Myths of Rich & Poor and The State of Working America 1998-99, are clear examples. Both describe U.S. income trends over recent decades, but the similarity ends there. Myths of Rich & Poor, written in the style of supply-side optimism, celebrates the economy's abundance. Working America, written in the style of pre-Clinton liberal pessimism, focuses on the uneven results.
Since I too have strong opinions on income trends- I updated my own book on the subject earlier this year- honesty requires me to show my bottom line early. Each book has merit: Myths is the more engaging read, while Working America's data are more reliable. A full understanding of today's economy requires elements from both of these books, supplemented by a few other key ideas- in particular the tightening nexus between education, equal opportunity, and worrisome levels of inequality.
The Bright Side of Free Markets
The same basic history forms the background for both books. From the close of World War II through 1973, the U.S. economy boomed. Spurred by rapid growth in productivity, real median family income increased from $21,000 in 1947 to $42,000 in 1973 (all figures here are in 1998 dollars). But the boom ended badly. Productivity growth began to slow in the late 1960s at the same time that expenditures for the Vietnam War were making labor markets tight. Workers demanded more pay, and the slower productivity growth meant that higher wages were passed along as higher prices. Inflation accelerated after 1972 because of harvest failures in many parts of the world (resulting in the great Russian wheat sale) and the first OPEC oil price shock. By the mid-1970s, inflation was embedded in the economy, and growth in productivity and median family income nearly stopped altogether.
Political leaders responded to stagflation by opening up the economy. >From the Carter to the Clinton administrations, the government aggressively pursued deregulation. It relaxed restraints on corporate mergers. It permitted, and at times advanced, the weakening of unions. It promoted freer trade with other countries. Finally, it abandoned the goal of managing aggregate demand through fiscal policy. Myths and Working America both assess how this reliance on free markets affected incomes.
Myths argues that the deregulation largely succeeded. Written by W. Michael Cox, a vice president at the Dallas Federal Reserve Bank, and Richard Alm, a business reporter for the Dallas Morning News, Myths is a Texas-style supply-side book with a combative, optimistic faith in laissez-faire. Cox and Alm argue that the new regime has created a tide of prosperity that is rising faster than official statistics indicate and that is carrying most people with it.
Like much supply-side writing, Myths is strongest when it emphasizes the economy's dynamics. Dazzling new products are emerging and finding places in households. More and more goods are being traded globally, leading to improvements in both quality and availability. And individuals are increasing their incomes as they gain experience and move up the promotional ladder- or move to different industries altogether. This emphasis is all to the good. It is hard to look at today's economy without sensing a transformation in how we produce and consume. The transformation may not be as large as those caused by railroads or electricity, but it is generating enormous opportunities.
Does all this innovation show up in workers' purchasing power? Most measures indicate that median wages were stagnant or declining from 1973 into the 1990s. Median family income, now at $45,300, is only $3,700 higher than it was in 1973 despite the growth in two-income families. But Cox and Alm believe that workers have been benefiting from the dynamic economy for a long time. They point to surveys of consumption among other indications that life has improved a great deal. The average household now owns many items, from air conditioners to microwave ovens, that would have been considered luxuries in the 1970s, if they were available at all.
There is some truth to this argument. The consumer price index and other inflation measures do have difficulty incorporating the value of new products like VCRs and Viagra and the subtle benefits of mass customization. As a result, official statistics probably overstate inflation and understate the real value of wages.
But because Cox and Alm are so intent on proving their point, they overrun their data. For example, they say that statistics showing stagnant wages understate progress by omitting the value of fringe benefits and other nonwage sources. They suggest instead that we focus on the rapid growth of a different statistic: per capita income. In their words, "A simple division of total output by the number of people, per capita income isn't skewed by changes in the way we work, the way we live, how we're paid, and what we produce."
This statement is just wrong. Per capita income is skewed by how we work and live- in particular, how the population divides between workers and dependents. When more of the population is receiving paychecks, income per capita (per man, woman, and child) rises even when wages are stagnant. This has happened in recent decades. Young baby-boom teenagers became adult workers. Baby-bust birth rates meant the number of children declined. Large numbers of married women moved into the labor force. In 1970, there were 1.5 dependents for every worker; by the late 1980s, there was only one dependent per worker. We could afford all the VCRs and balsamic vinegar not because average wages were rising briskly- they weren't- but because each paycheck was divided among fewer people.
The authors' discussion of inequality shows insight, this time combined with a more troubling error. They begin with census statistics on income inequality among families. (These numbers aren't comprehensive- they report pretax money income only, with no reductions for taxes paid and no additions for fringe benefits or nonmoney income such as Medicare coverage; still, they tell a clear story.) In 1947, the richest fifth of families received $8.60 for every dollar received by the poorest fifth. Inequality narrowed over the next quarter century; by 1969, the ratio stood at $7.25 to $1. Then inequality began to grow again, and it now stands at $11.24 to $1.
Myths correctly argues that these data are limited in an important way- they are a series of point-in-time snapshots that say nothing about the potential for income mobility, the way in which a family might move up as its breadwinners gain experience and receive promotions. If this mobility were substantial- if most families occupied the upper reaches of the distribution sometime during their careers- then higher point-in-time inequality would be less of a concern.
Cox and Alm estimate the extent of this income mobility with data that track the earnings of a sample of people from 1975 through 1991. They use the data set, from the University of Michigan's Panel Study of Income Dynamics, to track the experiences of individuals rather than families. Most of these people saw their incomes rise a great deal. Among the poorest fifth of earners in 1975, only 5% remained in the lowest fifth in 1991. The rest had moved up, including 29% who went all the way to the highest fifth of earners. The authors imply that similarly impressive mobility rates apply to the income distribution for entire families. Not very likely.
Cox and Alm first published these mobility results several years ago. Because their estimates of mobility were so much higher than other estimates, they drew both attention and reanalysis. As Peter Gottschalk of Boston College has pointed out, the poorest fifth of Cox and Alm's sample started with an average 1975 income of only $3,000 (again in 1998 dollars). Although 1975 was a recession year, it is hard to imagine that so many people in a national sample of workers had incomes that low. The explanation, Gottschalk notes, is that the sample included part-time workers and persons as young as 16. Under these parameters, the "poorest" fifth of workers would be dominated not by heads of families but by teenagers selling fast food after school. These people could become upwardly mobile simply by moving into the full-time labor force as they aged.
Families, by contrast, usually have at least one full-time worker at any time, so they can't make such big jumps. Yet a reader of Myths won't know about the limitations of the sample because the book (unlike the authors' earlier article on this research) omits a full description of it. When medical researchers exercise this much discretion with data, they find themselves in court.
In fact, Cox and Alm's exaggeration is unnecessary. Moderate mobility does exist in the family-income distribution, and it does make the income picture look better than census snapshots of inequality suggest. The authors' apparent reason for exaggerating is that, as they say in their preface, they are writing not just to present a true picture of the economy but also to defend free markets against recent books that paint the economy in pessimistic tones. But when an economy produces both good and bad outcomes, as the U.S. economy has over the last quarter century, no book can achieve both goals. An accurate picture must include pessimistic developments and so give comfort to the enemy.
A Glass Half Empty
The preface to Myths includes a list of the sort of books the authors want to counter. The books were selected for their pessimistic titles- Jeffrey Madrick's The End of Affluence is one example- so the mildly titled State of Working America 1998-99 is not among them. But if content rather than title had determined the list, Working America certainly would have had a place. This biennial volume is written by economists at the Economic Policy Institute in Washington, D.C. The institute was founded in the 1980s in part to provide a liberal counterweight to supply-siders' claims that President Reagan's policies were increasing incomes without adding to inequality.
The new edition of Working America fully maintains the tradition. Lawrence Mishel, Jared Bernstein, and John Schmitt argue that the greater reliance on free markets has failed. Targeting those like Cox and Alm who see a bountiful economy for nearly everyone, they focus on official statistics showing that while unemployment is down and GDP is up, productivity and average income are barely growing. According to these data, workers are seeing only greater inequality and less job security.
Working America has a second target as well. Statistics show that wages of less-educated workers have declined sharply, particularly since the late 1970s. Many economists contend that wage declines among these workers are an unfortunate but inevitable by-product of technological change: unable to keep up with the demands of new produc- tion systems, undereducated workers have been shunted into lower-paying jobs. But Working America says there is little evidence that technological change accelerated after the 1970s. Instead, the authors argue that the eroding minimum wage, weaker unions, and shifts in production toward lower-wage countries are the main drivers for declining pay. The dispute is hardly academic- technological change is unavoidable, but government can do a lot to influence the minimum wage, the strength of unions, and foreign trade.
While Myths tells an engaging story, Working America is mainly a narrative guide to a great variety of data, most of it carefully presented with clear source notes. But like Myths, it suffers from the authors' desire to counter an opposing view. Accordingly, the authors discuss the strong economy of the last two years only in grudging tones- for instance, they acknowledge that wages of the lowest-paid workers are now rising but point out that those wages didn't rise for most of the 1990s.
Similarly, Working America gives only passing notice to the age-related mobility that Cox and Alm emphasize. The book often mentions "stagnant wages" but doesn't explain that this condition still allows for workers to receive higher pay as they age: If in 1988 and 1998, adjusting for inflation, 30-year-old men earned an average of $30,000 and 40-year-old men earned an average of $40,000, wages were stagnant under the definition used by the authors (and by many other people). But men still could see rising wages as they aged from 30 to 40. Presumably, Working America downplays age-related mobility for the same reason it avoids discussing the convenience of ATM machines and the possibilities of the Web: to avoid softening the book's message that freer markets are benefiting only a few.
On the connection between technological change and inequality, Working America overstates a valid point. It's true that economists have too readily accepted the idea that change in technology is the only reason for the decline in the wages of less-educated workers. But technological change surely plays a role. About half of all 17-year-olds still read below a ninth-grade level, yet most well-paying jobs in industry now demand a much higher level of literacy. In automobile repair, a field I've been examining, the introduction of electronics into engines and transmissions has sharply increased the frequency with which technicians must read diagnostic instructions rather than fix problems by executing routines known by heart.
Read Myths of Rich & Poor to appreciate how new technology and expanded trade are now important engines of economic growth. Read State of Working America 1998-99 to appreciate how growth is generating benefits very unequally. If one book portrays the U.S. economy as Silicon Valley writ large, the other sees it as Oakland.
A Mixed Bag
A full picture of today's economy would show that it has been very good indeed. Inflation and unemployment are low, and while average incomes have probably grown very slowly in recent decades, we're now beginning to see faster gains. But more important for the long run, we're now seeing signs of a return to higher productivity growth. Since the early 1970s, productivity has grown at an anemic 1% per year. A return to the historical average of 2% per year would mean a big improvement in the growth rate of total family income.
I emphasize total income because faster productivity growth does not guarantee that all family incomes will rise. The question is whether this should concern us. Some observers argue that we have lived with much worse: that family-income inequality is lower today than it was in the 1920s and not much higher than in 1947. But point-in-time comparisons obscure the issues that Cox and Alm raise- the prospect of mobility within the income distribution and how that prospect may have changed. In 1947, most of the poorest families were farmers, with their very low cash income, and the elderly. Where are these two groups today? Over time, farm families have become a tiny share of the population, while social security and private pensions have boosted elderly incomes. Based on these two movements alone, income inequality should be lower today than before, not higher.
One reason for higher inequality is the growing diversity of family structures. With more families headed by single women at one end of the distribution and more two-income families at the other, some divergence was inevitable. But a larger factor is the greater educational requirements for high-paying jobs. When President Kennedy spoke of a rising tide lifting all the boats, the economy was evolving in ways that generated demand for both better-educated and less-educated workers. Since the 1980-1982 blue-collar recession, the relative demand for less-educated workers has declined significantly. In 1979, a 30-year-old man with only a high school diploma earned an average of $32,000 if he worked full time. Today, his 30-year-old counterpart (adjusting for the greater fraction of young people going to college) averages about $5,000 less, while the income of college graduates has stayed constant at Reserve Board reports that in 1995, the latest year for which these data are available, the richest 1% of households (averaging about $7.75 million) owned about a third of all net worth; the next richest 10% of households ($823,000) owned approximately another third; and the remaining households ($77,000) owned the rest. More households participate in the stock market today than in 1995, but the change is too small to move these figures significantly.
More important, these factors have restricted the chances for mobility. In 1947, much of the inequality reflected geographical differences between urban and farm incomes. A 20-year-old man in a depressed area could usually get on a bus and go to a city where he had the basic skills to get a factory job at higher pay. Today, poor education plays the role that geography did in the past. A 20-year-old without a job can still take important steps, such as going to an inexpensive junior college. But as the economist James Heckman noted in a recent speech, many aspects of what we call the ability to learn may be largely set by the time a young person is 15 or 16. Whatever one's definition of equal opportunity, it does not apply.
The last two decades of economic change have inflicted a heavy blow on less-educated men and women. Their paychecks have suffered the greatest impact, and now they will have to struggle for the educational resources to make sure that their children don't repeat the cycle. While many people in the U.S. are now registering impressive gains, many others are not. In the future, those on the wrong side of the educational divide will find it harder and harder to climb from low income to high income.
The Political Challenge
Economic theory generally addresses efficiency, not equity. From a narrowly economic perspective, growing inequality is not much to worry about. As long as markets freely reward greater effort and ability, income disparities can even serve as spurs to greater effort, as people on the low end of the scale see the riches to be made. In their concluding chapter, not surprisingly, Cox and Alm argue that the key to future prosperity is further acceptance of free markets.
But free markets require that the government do more than simply get out of the way. Markets work only when powerful institutions provide clear and equitable rules for commerce. The legal infrastructure needs to be strong enough for people to have the confidence to invest. These institutions, in turn, depend on public support. If the majority of people feel that markets are operating to their benefit, they'll support this infrastructure while leaving the markets alone.
In this context, prolonged and high inequality is a great danger. If the results of free markets do not accord with widespread conceptions of fairness- if a growing number of citizens feel they are left out of the American dream of advancement- they'll make government do more than take the wageworker's side in occasional policy decisions. Left unchecked, inequality could lead to a renewal of intense industry regulation and trade restrictions. In that case, we'll all be poorer.