Class Struggle in America?
In recent years, the New York Times has suffered many blows to its status and reputation. But when it chooses to employ its vast resources toward a specific goal, it can still influence the national discussion on any given issue. In May, the Times undertook to revive the issue of class in America by launching “Class Matters”—a huge, multipart series, more than a year in the making, involving a large team of reporters, specially commissioned polling, and an interactive website.
The gist of the series, conveyed in the very first installment, is that “class is still a powerful force in American life”—so powerful that, in the words of the Times, “it has come to play a greater, not lesser, role,” accounting not only for individuals' success or failure in school but for the increasing isolation of the rich from the rest of American society.
This is hardly a new theme for the Times. To the contrary, one can depend on the paper to report dutifully every new study or statistic purporting to show that the rich have gotten richer in America while the poor have become poorer. Evidence in support of this proposition is, in the Times's view, overwhelming; the only puzzle has been the propensity of voters to endorse and bolster this trend by electing Republicans to national office who are determined to cut the taxes of the rich.
One purported answer to that puzzle can be found in Thomas Frank's What's the Matter With Kansas?, a recent, much-reviewed book that focuses on the many rural and working-class Americans, once a reliable constituency of the Democratic party, who have developed the political tic that so troubles the editors of the Times. According to Frank, what causes such voters to ignore their own economic self-interest, which clearly lies with liberals and Democrats, is their fundamentally conservative position on social issues, from the excesses of Hollywood to abortion-on-demand. This trumps their economic concerns, pushing them into the Republican column.
The Times appears dissatisfied with this interpretation. Instead, it has set out to find a more purely economic explanation for why so many poor and working-class Americans betray their own interests when they enter the voting booth. “Class Matters” is the result.
Before taking up the claims advanced in this series, it is worth glancing back at the rich history of the Times's interest in income inequality. A relevant benchmark is the 1980's, when the paper drew a clear connection between the Reagan tax cuts enacted earlier in the decade and, to the Times, the alarming increase in the share of household income subsequently going to those at the top of the distribution scale.
At first glance, the link drawn by the Times seemed plausible enough. In 1981, according to the Census Bureau, the lowest quintile (20 percent) of households was garnering 4.2 percent of total income and the highest quintile's share stood at 43.8 percent. By the time Reagan left office eight years later, the lowest quintile's take had fallen to 3.8 percent while the highest quintile's had risen to 46.8 percent.
But there were two major problems with these data as the Times presented them. First was the implication of a fixed economic pie, whereby one group's gain necessarily meant another group's loss. In fact, however, the pie itself was growing, so that it was possible for all groups to be better off in absolute terms even if one group's share was falling. And that is exactly what had happened in the 1980's: although those at the top did especially well, all households showed a gain, with the average real income of the lowest quintile rising by almost 10 percent (from $9,348 in 1981 to $10,058 in 1989, in 2003 dollars).
The second problem was that the data measured only before-tax cash income, thus excluding almost everything we have done as a society to equalize incomes—imposing high tax rates on the rich, and providing substantial in-kind welfare benefits for the poor in the form of housing subsidies, food stamps, and much else. To this day, although the Census Bureau produces “experimental” income measures in which these factors are taken into account, it has not yet changed the official measure, which continues to ignore most taxes and benefits. The impact is substantial: for 2001, the latest year available, the inclusion of taxes alone in the Bureau's calculations would raise the lowest quintile's share of income from 3.5 percent to 4.4 percent and lower the top quintile's share from about 50 to 45 percent.
In any event, the Times's heavy focus on income shares tended to leave the impression that the same people were in the same brackets from year to year. Were this the case, there would indeed have been a strong argument for income redistribution. But the reality was, and remains, otherwise, though to prove it one needs data that cannot be derived from the annual Census income-and-poverty reports, which are just “snapshots” of households at a given moment.
To determine mobility, one needs to do longitudinal studies, tracking the same specific individuals and families over time. One of the longest-running projects of this kind is the Panel Study on Income Dynamics (PSID) at the University of Michigan, which has been following several thousand families since 1968. According to a 1984 analysis based on this source, almost half of the families in the bottom income quintile in 1971 had risen to a higher quintile by 1978. Over those seven years, 22 percent had risen one bracket, 9.5 percent two brackets, 7 percent three brackets, and 6 percent all the way to the top quintile of income.
Also in 1984, the Census Bureau initiated its own Survey of Income and Program Participation (SIPP), allowing it, too, to study the same people over time.1 The results confirmed the PSID findings. According to a 1989 Census report, almost 20 percent of people in the lowest quintile in 1984 had risen to a higher quintile by the following year, while an equivalent number of those in the top quintile had fallen to a lower one.
As one would expect, extending the time period yields an even higher degree of mobility. Using the PSID data, a 1995 study by the Federal Reserve Bank of Dallas found that in the decade-and-a-half between 1975 and 1991, fully 95 percent of those in the lowest income quintile had risen to a higher quintile, while about 40 percent of those in the top quintile had fallen to a lower one. Similarly, the most recent SIPP report, looking at families in the three-year period between 1996 and 1999, found almost 40 percent of those in the lowest quintile rising to a higher one, with almost the same number falling from the highest quintile to a lower one. As for poverty in the same period, only 2 percent remained officially poor for very long.
The issue of income inequality hit a peak during the 1992 presidential campaign. Bill Clinton made a point of talking about it at almost every campaign stop, promising that, if elected, he would force the rich to pay their fair share. In formulating his case, he relied heavily on research by the economist Paul Krugman, as channeled through Sylvia Nasar, a reporter for the New York Times. One particularly inflammatory piece appeared on the front page of the Times on March 5, 1992. Citing Krugman as her source, Nasar wrote that during the 1980's, the top 1 percent of American families had received an astonishing 60 percent—three-fifths—of the aggregate national gain in after-tax income.
At least, that is what appeared in the edition of the Times distributed in the Washington, D.C. area—the one seen by Clinton and the one that caused him, according to his campaign press secretary Dee Dee Myers, to go “crazy.” Duly cranking up the rhetoric, he hammered George H.W. Bush mercilessly on this theme.
The only problem was that Nasar or Krugman had thoroughly misread the data. The top 1 percent had not come anywhere near three-fifths of the aggregate income gain in the 1980's. Had they done so, that would have left the lower 99 percent of the population dividing the other two-fifths, with highly disconcerting implications. Rather, the increase in their average income was 60-percent greater than the increase in the income of the rest of the population. But as I have already noted, the real income of every income class increased in the same period.
Republicans were furious with the Times report. But the paper refused to admit error or even to issue a clarification. Only now, if one goes to the Times online archives, will one find a (silently) corrected version of the Nasar article. This is what is known at the paper as a “rowback”: fixing an error without ever acknowledging it.
A few months after the Nasar story ran, the Bush administration sought to retaliate by issuing a Treasury Department study that utilized tax records unavailable to private researchers.2 Matching tax returns by the same filers ten years apart, Treasury economists found that 86 percent of those in the bottom quintile in 1979 had moved to a higher quintile by 1989, with 15 percent rising all the way to the top. Similar findings were reported at about the same time by the Urban Institute—not known as a haven for conservative Republicans. Comparing the same families in 1977 and 1986, it too found a high degree of income mobility, and it also confirmed that the income gains of the 1980's were broadly shared. Indeed, the real average family income of those in the bottom quintile had risen by fully 77 percent between 1977 and 1986, while, in stark contrast to the New York Times portrait, those in the top quintile saw an increase of only 5 percent—the lowest of any income group.
But then came the Clinton years, a period in which the Times developed a quite different perspective on the whole question of income distribution.
Between 1992 and 2000, according to the Census Bureau, the share of aggregate income going to the top quintile rose from almost 47 to almost 50 percent—an increase larger than the one during the Reagan “decade of greed.” Similarly, the top 5 percent of households saw their share rise by 3.3 points, while the comparable increase during the Reagan years had been just 1.8 points. Yet the Times found nothing particularly objectionable about this.
To the contrary: by 1995, in the Times Magazine, the paper's editorial writer Michael Weinstein, in an odd flip, was celebrating the gains of the ultra-rich. Defending Michael Milken's $500-million paycheck from Drexel Burnham Lambert in 1987, and the $200 million earned by Michael Eisner from Disney in 1993, Weinstein reminded his readers that these men had taken risks, made key corporate decisions, and were paid out of an increase in profits that they themselves had been largely responsible for. So why attack them?
Besides, Weinstein went on to explain, the gains of the super-rich did not themselves cause an increase in income inequality. “There are simply too few super-rich to matter,” he wrote, and “even if large companies were to eliminate compensation for CEO's, the return to shareholders would rise by a measly 6 cents for every $100 they've invested.”
By 1999, even Sylvia Nasar had formulated a more nuanced view (partially, perhaps, because she herself had by then earned a handsome return from her best-selling book, A Beautiful Mind). In an April 14 article, she correctly pointed to the need to take account of mobility when thinking about income distribution:
Growing inequality could have devastating effects if it convinced those at the bottom that efforts to move up are doomed to failure. But largely the opposite has happened. . . . Most ordinary Americans . . . seem to feel that, whatever has happened to the income distribution, opportunities abound—and not just for the rich.
Another factor that may have had some influence on the Times's treatment of these issues was the growing presence of women in the labor force, and especially at the highest levels of wage earners. As Diana Furchtgott-Roth of the Hudson Institute has pointed out, because income data are typically compiled for families or households rather than for individuals, and because so many women now work outside the home, the simple act of marriage tends to “worsen” the distribution of income in an upward direction. A man and woman who separately would be in the middle quintile are likely to jump into the top quintile once married. Similarly, the economists Katherine Bradbury and Jane Katz have shown that families with an employed woman move up the distribution scale faster and more frequently than those in which the wife does not work outside the home, and are less likely to move down.
For some years, then, little was to be seen in the Times about the maldistribution of income. Even its in-house scold, Paul Krugman, who was appointed a columnist in 1999, had little to offer on the subject. For their part, the news pages were relatively free of the sorts of de-facto editorializing that had been so common in the 1980's.
But by 2001 the younger Bush was in the White House, and the times, and the Times, were changing again. Krugman, for instance, returned to the fray with a Times Magazine piece in 2002 charging that America had entered a “new Gilded Age,” and in August 2004, smack in the middle of last year's presidential campaign, a report by Timothy Egan charged that the American middle class was disappearing. Accompanying Egan's article was a chart that showed a fall in the proportion of households with incomes (in 2003 dollars) in the $25,000-$75,000 range: from almost 52 percent in 1980 to just under 45 percent in 2003.
Note the start date—1980, the year Ronald Reagan was elected—and the finish date—2003, on George W. Bush's watch. If, as seems clear, the point was to put the blame for the sufferings of the middle class on Republican rather than Democratic policies, this was an absolutely necessary move. For had the calculation been done with a start date of 1992, the year Bill Clinton was elected, and a finish date of 2000, the year he left office, the results would have shown the exact same trend: among those earning $25,000-$75,000, a drop of almost three percentage points.
But the most serious defect of this particular article was its implied message that a significant percentage of those defined as “middle class” had become poor. This was totally untrue. In fact, the ranks of the poor (those earning under $25,000, in the Times's definition) fell during the same two-plus decades from 33.1 percent of the population in 1980 to 29 percent in 2003. Looking at the data from the other end, we see that the percentage of those making more than $75,000 rose in the same period from just under 15 percent to just over 26 percent.
In other words, the ranks of both the poor and the middle class shrank for one reason only—more of them were well off. And more of them, one might add, were black. To be precise, the proportion of black households making less than $25,000 (in 2003 dollars) fell from 53.8 to 43.4 percent in the period 1980-2003, while the ranks of the black middle class increased from 40.5 to 42.9 percent and black families earning over $75,000 rose from 5.8 to 13.7 percent.
To turn all this good news into bad, the Times had to go to extraordinary lengths indeed.
In the event, of course, neither John Kerry nor the Times gained any traction with the class-warfare issue in 2004. Many of those who were supposedly suffering under George W. Bush's economic policies seemed to vote for him anyway, a possibility the Times no doubt anticipated in undertaking its major research effort into class in America, to which we may now return.
The series that opened on May 15 would deal with the influence of class across a host of related topics: health, marriage, religion, education, immigration, and more. But the essence was captured in the first article, an overview entitled “Shadowy Lines that Still Divide.” Its central thrust was to downplay the role of the one enduring economic factor that has time and again undercut the Left's efforts to incite class warfare—namely, mobility. To do this, the Times writers had to get around all the data I have already discussed, and to redefine mobility itself.
This they accomplished in two ways. First, “Shadowy Lines that Still Divide” suggested that, although individuals may appear to demonstrate income mobility over the course of their lives, eventually they end up in the same place they started—which is to say, in the same income class as their parents. In other words, the principal standard for measuring mobility should be intergenerational mobility. Second, whatever the degree of mobility in American society, it is no greater now than it has been in the past, nor is our society any longer to be distinguished in this respect from Europe.
Recent academic research, much of it financed by the Russell Sage Foundation in New York, supports the Times's twin theses.3 But there is an equally large body of research showing that mobility has in fact risen here, and that it is still higher than in Europe. A highly technical debate on these issues has been conducted for some time in academic journals like the Journal of Income Distribution, the Review of Income and Wealth, and the Journal of Human Resources. The Times, however, mostly cites only one side of the debate—and specifically those who claim to have found a sharp decline in inter-generational mobility in recent years. So let me fill in the other side.
The economist Dean Lillard of Cornell is among those who think the data suggesting a decline in mobility are flawed. In his estimation, they ignore a substantial population of men who are intermittently unemployed for a year or more. Including this population, he shows in an article co-authored with Kenneth Couch of the University of Connecticut, causes the correlation between a father's and son's earnings largely to disappear. The same finding is confirmed by Angela Fertig of Indiana University, who cites data demonstrating a rise over recent decades in intergenerational mobility and opportunity. One's family, Fertig concludes, does not predetermine one's position in the income distribution.
When it comes to comparative mobility between and among countries, the data are especially difficult to assess. One complicating factor is that in some European countries, heavier taxes and more generous welfare benefits than ours have the effect of compressing the distribution of income. In Sweden, for instance, the spread between the highest and lowest quintiles is much smaller than it is here. This means that it is easier to move from one quintile to another, but also that such movement is less meaningful for one's living standard.
In the end, what really matters is not whether there has been some small change in the degree of mobility in America but whether the level of mobility remains high. After all, it cannot rise infinitely—if it did, at some point income distribution would become purely random, with no linkage whatsoever to family, education, or anything else. Complaining that mobility is not rising is akin to lamenting that baseball batting averages are no higher today than they were a generation ago. There is no reason to think they should be.
What is truly important for a society is how people in general perceive their circumstances. In this respect, it is clear that Americans continue to believe in the Horatio Alger model, according to which everyone, no matter how humble his beginnings, has a chance to make it big. Interestingly, the Times itself provides the strongest evidence for the perduring strength of this belief in the poll data it compiled for its series and has made available on its website.4
Perhaps the most valuable aspect of this research is that some figures are given over time. Thus, in 1983, people were asked: “Do you think it's still possible to start out poor in this country, work hard, and become rich?” At the time, only 57 percent of respondents agreed with this proposition, while 38 percent disagreed. By the 1990's, the percentage of positive responses had risen to 70. As of March 2005, 80 percent said it was possible for anyone to become rich in America, and only 19 percent said it was not.
The Times poll also provides a useful perspective on intergenerational mobility. In one question, people were asked: “Compared to your parents when they were the age you are now, do you think your own standard of living now is much better, somewhat better, about the same, somewhat worse, or much worse than theirs was?” In 1994, the first time the question was asked, 31 percent answered “much better,” 32 percent “somewhat better,” 21 percent “about the same,” 11 percent “somewhat worse,” and 3 percent “much worse.” In 2005, the corresponding figures were 39 percent, 27 percent, 20 percent, 9 percent, and 4 percent.
In other questions, people were asked to compare their social class today with the one in which they grew up. Across the board they responded the same way, reporting that they lived in a higher social class today.
On balance, then, it appears that most people believe themselves to be living better than their parents. What is more, 40 percent of them believe that, as compared with the 1970's, the likelihood of one's moving up in social class is greater today, while 35 percent think it is the same; only 23 percent think otherwise.
Of course, the general standard of living has risen over time: the typical middle-class family today occupies better housing, drives nicer cars, and owns things (like computers) that would have been luxuries inconceivable even to the very rich 30 years ago. Acknowledging this point, the Times spins it so as to demean, if not to deny, its significance:
Factories in China and elsewhere churn out picture-taking cellphones and other luxuries that are now affordable to almost everyone. Federal deregulation has done the same for plane tickets and long-distance phone calls. Banks, more confident about measuring risk, now extend credit to low-income families, so owning a home or driving a new car is no longer evidence that someone is middle class.
About this, the paper is flat wrong: there is no more definitive symbol of middle-class status than owning one's home. But one can say much more: in recent years, home ownership in America has helped measurably in reducing inequalities in the distribution of wealth. Rich people have most of their wealth in financial assets, which have done poorly of late; by contrast, most of the wealth of the middle class is in housing, which has witnessed a vast run-up in value. The net effect is that aggregate wealth has become much more evenly distributed.
The role of housing is also important in interpreting income data. The Times poll found that 25 percent of the entire American population own their homes outright, with no mortgage. Among retirees, according to the Bureau of Labor Statistics (BLS), three-fourths are homeowners, and an amazing 86 percent of these own their homes free and clear. In effect, all such people have substantial income—what economists would call imputed rent, or rent they do not have to pay—that never shows up in the Census data, making them appear less well off than they really are.
The BLS data offer another insight into the question of living standards. In 2002, the bottom quintile of consumers, with a before-tax income of $8,316, were spending well over twice that amount ($19,061); the top quintile, with an income of $121,367, were spending only $79,199. Thus, if one bases one's calculations on expenditures rather than on income, living standards are much more equal among rich and poor.
How can this be? The main reason is that there are many retirees in the bottom quintile, people with low incomes but substantial savings. Those savings are being drawn down to supplement their incomes and maintain their standard of living. Similarly, the bottom quintile includes many people temporarily unemployed because of a job loss or some other setback. With every reason to expect that their income will bounce back, they may be borrowing to pay for consumption at, to use another term of economists, their permanent income level.
As for why the wealthy are consuming much less than their income, the answer is that they are saving and investing—adding to the national seed corn, so to speak, and providing the capital that will eventually raise productivity and wages across the economy. If the rich did not save so much voluntarily, the nation as a whole would be that much poorer.
In the end, luckily, one doubts that the Times's effort to revive the idea of class consciousness in America will bear fruit. Regardless of what some academic research may show, people simply do not perceive reality as the Times construes it.
And this brings us back to the subject of tax cuts for the rich. A whopping 76 percent of those polled by the Times favor abolition of the estate tax—which is paid only by the richest 2 percent of decedents. Going a step beyond Thomas Frank in What's the Matter with Kansas?, Michael Graetz and Ian Shapiro, both of Yale University, argue that these Americans are victims of false consciousness, having been tricked into going against their own self-interest by the self-interest of a small but determined group of wealthy families.5
More likely, however, is that the hoi polloi are indeed motivated by their perceived self-interest. Rightly or wrongly, many of them believe that they or their children will one day find themselves wealthy enough to incur the estate tax. Once again the Times poll inadvertently provides confirming evidence: 11 percent reported it was very likely that they would some day be wealthy, and another 34 percent thought it was somewhat likely. Only 22 percent said it was not at all likely.
In short, it is going to take a lot more than a series of New York Times articles to convince most people that they have no chance of success and will die in the same social class into which they were born. This is hardly to suggest that the Times series is therefore a harmless exercise; one can be sure that its findings will be cited as indisputable and solidly grounded fact in many a political and ideological campaign to come. Still, much more newsworthy would be an inquiry into why so many wealthy Americans betray their supposed class interest by voting Democratic. So far, that does not appear to figure on the Times's agenda.
1 I was randomly chosen to participate in the SIPP program for several years in the mid-1990's.
2 The study was initiated by Congressman Dick Armey (R-TX), who then had to pull teeth to get it released.
3 See, for example, Unequal Chances: Family Background and Economic Success, ed. Samuel Bowles, Herbert Gintis, and Melissa Osborne Groves (Princeton University Press, 2005) and Generational Income Mobility in North America and Europe, ed. Miles Corak (Cambridge University Press, 2004).
5 Death by a Thousand Cuts: The Fight over Taxing Inherited Wealth (Princeton University Press, 2005).