The New York Times

October 20, 2002

For Richer

Paul Krugman

I. The Disappearing Middle

When I was a teenager growing up on Long Island, one of my favorite excursions was a trip to see the great Gilded Age mansions of the North Shore. Those mansions weren't just pieces of architectural history. They were monuments to a bygone social era, one in which the rich could afford the armies of servants needed to maintain a house the size of a European palace. By the time I saw them, of course, that era was long past. Almost none of the Long Island mansions were still private residences. Those that hadn't been turned into museums were occupied by nursing homes or private schools.

For the America I grew up in -- the America of the 1950's and 1960's -- was a middle-class society, both in reality and in feel. The vast income and wealth inequalities of the Gilded Age had disappeared. Yes, of course, there was the poverty of the underclass -- but the conventional wisdom of the time viewed that as a social rather than an economic problem. Yes, of course, some wealthy businessmen and heirs to large fortunes lived far better than the average American. But they weren't rich the way the robber barons who built the mansions had been rich, and there weren't that many of them. The days when plutocrats were a force to be reckoned with in American society, economically or politically, seemed long past.

Daily experience confirmed the sense of a fairly equal society. The economic disparities you were conscious of were quite muted. Highly educated professionals -- middle managers, college teachers, even lawyers -- often claimed that they earned less than unionized blue-collar workers. Those considered very well off lived in split-levels, had a housecleaner come in once a week and took summer vacations in Europe. But they sent their kids to public schools and drove themselves to work, just like everyone else.

But that was long ago. The middle-class America of my youth was another country.

We are now living in a new Gilded Age, as extravagant as the original. Mansions have made a comeback. Back in 1999 this magazine profiled Thierry Despont, the ''eminence of excess,'' an architect who specializes in designing houses for the superrich. His creations typically range from 20,000 to 60,000 square feet; houses at the upper end of his range are not much smaller than the White House. Needless to say, the armies of servants are back, too. So are the yachts. Still, even J.P. Morgan didn't have a Gulfstream.

As the story about Despont suggests, it's not fair to say that the fact of widening inequality in America has gone unreported. Yet glimpses of the lifestyles of the rich and tasteless don't necessarily add up in people's minds to a clear picture of the tectonic shifts that have taken place in the distribution of income and wealth in this country. My sense is that few people are aware of just how much the gap between the very rich and the rest has widened over a relatively short period of time. In fact, even bringing up the subject exposes you to charges of ''class warfare,'' the ''politics of envy'' and so on. And very few people indeed are willing to talk about the profound effects -- economic, social and political -- of that widening gap.

Yet you can't understand what's happening in America today without understanding the extent, causes and consequences of the vast increase in inequality that has taken place over the last three decades, and in particular the astonishing concentration of income and wealth in just a few hands. To make sense of the current wave of corporate scandal, you need to understand how the man in the gray flannel suit has been replaced by the imperial C.E.O. The concentration of income at the top is a key reason that the United States, for all its economic achievements, has more poverty and lower life expectancy than any other major advanced nation. Above all, the growing concentration of wealth has reshaped our political system: it is at the root both of a general shift to the right and of an extreme polarization of our politics.

But before we get to all that, let's take a look at who gets what.

II. The New Gilded Age

The Securities and Exchange Commission hath no fury like a woman scorned. The messy divorce proceedings of Jack Welch, the legendary former C.E.O. of General Electric, have had one unintended benefit: they have given us a peek at the perks of the corporate elite, which are normally hidden from public view. For it turns out that when Welch retired, he was granted for life the use of a Manhattan apartment (including food, wine and laundry), access to corporate jets and a variety of other in-kind benefits, worth at least $2 million a year. The perks were revealing: they illustrated the extent to which corporate leaders now expect to be treated like ancien regime royalty. In monetary terms, however, the perks must have meant little to Welch. In 2000, his last full year running G.E., Welch was paid $123 million, mainly in stock and stock options.

Is it news that C.E.O.'s of large American corporations make a lot of money? Actually, it is. They were always well paid compared with the average worker, but there is simply no comparison between what executives got a generation ago and what they are paid today.

Over the past 30 years most people have seen only modest salary increases: the average annual salary in America, expressed in 1998 dollars (that is, adjusted for inflation), rose from $32,522 in 1970 to $35,864 in 1999. That's about a 10 percent increase over 29 years -- progress, but not much. Over the same period, however, according to Fortune magazine, the average real annual compensation of the top 100 C.E.O.'s went from $1.3 million -- 39 times the pay of an average worker -- to $37.5 million, more than 1,000 times the pay of ordinary workers.

The explosion in C.E.O. pay over the past 30 years is an amazing story in its own right, and an important one. But it is only the most spectacular indicator of a broader story, the reconcentration of income and wealth in the U.S. The rich have always been different from you and me, but they are far more different now than they were not long ago -- indeed, they are as different now as they were when F. Scott Fitzgerald made his famous remark.

That's a controversial statement, though it shouldn't be. For at least the past 15 years it has been hard to deny the evidence for growing inequality in the United States. Census data clearly show a rising share of income going to the top 20 percent of families, and within that top 20 percent to the top 5 percent, with a declining share going to families in the middle. Nonetheless, denial of that evidence is a sizable, well-financed industry. Conservative think tanks have produced scores of studies that try to discredit the data, the methodology and, not least, the motives of those who report the obvious. Studies that appear to refute claims of increasing inequality receive prominent endorsements on editorial pages and are eagerly cited by right-leaning government officials. Four years ago Alan Greenspan (why did anyone ever think that he was nonpartisan?) gave a keynote speech at the Federal Reserve's annual Jackson Hole conference that amounted to an attempt to deny that there has been any real increase in inequality in America.

The concerted effort to deny that inequality is increasing is itself a symptom of the growing influence of our emerging plutocracy (more on this later). So is the fierce defense of the backup position, that inequality doesn't matter -- or maybe even that, to use Martha Stewart's signature phrase, it's a good thing. Meanwhile, politically motivated smoke screens aside, the reality of increasing inequality is not in doubt. In fact, the census data understate the case, because for technical reasons those data tend to undercount very high incomes -- for example, it's unlikely that they reflect the explosion in C.E.O. compensation. And other evidence makes it clear not only that inequality is increasing but that the action gets bigger the closer you get to the top. That is, it's not simply that the top 20 percent of families have had bigger percentage gains than families near the middle: the top 5 percent have done better than the next 15, the top 1 percent better than the next 4, and so on up to Bill Gates.

Studies that try to do a better job of tracking high incomes have found startling results. For example, a recent study by the nonpartisan Congressional Budget Office used income tax data and other sources to improve on the census estimates. The C.B.O. study found that between 1979 and 1997, the after-tax incomes of the top 1 percent of families rose 157 percent, compared with only a 10 percent gain for families near the middle of the income distribution. Even more startling results come from a new study by Thomas Piketty, at the French research institute Cepremap, and Emmanuel Saez, who is now at the University of California at Berkeley. Using income tax data, Piketty and Saez have produced estimates of the incomes of the well-to-do, the rich and the very rich back to 1913.

The first point you learn from these new estimates is that the middle-class America of my youth is best thought of not as the normal state of our society, but as an interregnum between Gilded Ages. America before 1930 was a society in which a small number of very rich people controlled a large share of the nation's wealth. We became a middle-class society only after the concentration of income at the top dropped sharply during the New Deal, and especially during World War II. The economic historians Claudia Goldin and Robert Margo have dubbed the narrowing of income gaps during those years the Great Compression. Incomes then stayed fairly equally distributed until the 1970's: the rapid rise in incomes during the first postwar generation was very evenly spread across the population.

Since the 1970's, however, income gaps have been rapidly widening. Piketty and Saez confirm what I suspected: by most measures we are, in fact, back to the days of ''The Great Gatsby.'' After 30 years in which the income shares of the top 10 percent of taxpayers, the top 1 percent and so on were far below their levels in the 1920's, all are very nearly back where they were.

And the big winners are the very, very rich. One ploy often used to play down growing inequality is to rely on rather coarse statistical breakdowns -- dividing the population into five ''quintiles,'' each containing 20 percent of families, or at most 10 ''deciles.'' Indeed, Greenspan's speech at Jackson Hole relied mainly on decile data. From there it's a short step to denying that we're really talking about the rich at all. For example, a conservative commentator might concede, grudgingly, that there has been some increase in the share of national income going to the top 10 percent of taxpayers, but then point out that anyone with an income over $81,000 is in that top 10 percent. So we're just talking about shifts within the middle class, right?

Wrong: the top 10 percent contains a lot of people whom we would still consider middle class, but they weren't the big winners. Most of the gains in the share of the top 10 percent of taxpayers over the past 30 years were actually gains to the top 1 percent, rather than the next 9 percent. In 1998 the top 1 percent started at $230,000. In turn, 60 percent of the gains of that top 1 percent went to the top 0.1 percent, those with incomes of more than $790,000. And almost half of those gains went to a mere 13,000 taxpayers, the top 0.01 percent, who had an income of at least $3.6 million and an average income of $17 million.

A stickler for detail might point out that the Piketty-Saez estimates end in 1998 and that the C.B.O. numbers end a year earlier. Have the trends shown in the data reversed? Almost surely not. In fact, all indications are that the explosion of incomes at the top continued through 2000. Since then the plunge in stock prices must have put some crimp in high incomes -- but census data show inequality continuing to increase in 2001, mainly because of the severe effects of the recession on the working poor and near poor. When the recession ends, we can be sure that we will find ourselves a society in which income inequality is even higher than it was in the late 90's.

So claims that we've entered a second Gilded Age aren't exaggerated. In America's middle-class era, the mansion-building, yacht-owning classes had pretty much disappeared. According to Piketty and Saez, in 1970 the top 0.01 percent of taxpayers had 0.7 percent of total income -- that is, they earned ''only'' 70 times as much as the average, not enough to buy or maintain a mega-residence. But in 1998 the top 0.01 percent received more than 3 percent of all income. That meant that the 13,000 richest families in America had almost as much income as the 20 million poorest households; those 13,000 families had incomes 300 times that of average families.

And let me repeat: this transformation has happened very quickly, and it is still going on. You might think that 1987, the year Tom Wolfe published his novel ''The Bonfire of the Vanities'' and Oliver Stone released his movie ''Wall Street,'' marked the high tide of America's new money culture. But in 1987 the top 0.01 percent earned only about 40 percent of what they do today, and top executives less than a fifth as much. The America of ''Wall Street'' and ''The Bonfire of the Vanities'' was positively egalitarian compared with the country we live in today.

III. Undoing the New Deal

In the middle of the 1980's, as economists became aware that something important was happening to the distribution of income in America, they formulated three main hypotheses about its causes.

The ''globalization'' hypothesis tied America's changing income distribution to the growth of world trade, and especially the growing imports of manufactured goods from the third world. Its basic message was that blue-collar workers -- the sort of people who in my youth often made as much money as college-educated middle managers -- were losing ground in the face of competition from low-wage workers in Asia. A result was stagnation or decline in the wages of ordinary people, with a growing share of national income going to the highly educated.

A second hypothesis, ''skill-biased technological change,'' situated the cause of growing inequality not in foreign trade but in domestic innovation. The torrid pace of progress in information technology, so the story went, had increased the demand for the highly skilled and educated. And so the income distribution increasingly favored brains rather than brawn.

Finally, the ''superstar'' hypothesis -- named by the Chicago economist Sherwin Rosen -- offered a variant on the technological story. It argued that modern technologies of communication often turn competition into a tournament in which the winner is richly rewarded, while the runners-up get far less. The classic example -- which gives the theory its name -- is the entertainment business. As Rosen pointed out, in bygone days there were hundreds of comedians making a modest living at live shows in the borscht belt and other places. Now they are mostly gone; what is left is a handful of superstar TV comedians.

The debates among these hypotheses -- particularly the debate between those who attributed growing inequality to globalization and those who attributed it to technology -- were many and bitter. I was a participant in those debates myself. But I won't dwell on them, because in the last few years there has been a growing sense among economists that none of these hypotheses work.

I don't mean to say that there was nothing to these stories. Yet as more evidence has accumulated, each of the hypotheses has seemed increasingly inadequate. Globalization can explain part of the relative decline in blue-collar wages, but it can't explain the 2,500 percent rise in C.E.O. incomes. Technology may explain why the salary premium associated with a college education has risen, but it's hard to match up with the huge increase in inequality among the college-educated, with little progress for many but gigantic gains at the top. The superstar theory works for Jay Leno, but not for the thousands of people who have become awesomely rich without going on TV.

The Great Compression -- the substantial reduction in inequality during the New Deal and the Second World War -- also seems hard to understand in terms of the usual theories. During World War II Franklin Roosevelt used government control over wages to compress wage gaps. But if the middle-class society that emerged from the war was an artificial creation, why did it persist for another 30 years?

Some -- by no means all -- economists trying to understand growing inequality have begun to take seriously a hypothesis that would have been considered irredeemably fuzzy-minded not long ago. This view stresses the role of social norms in setting limits to inequality. According to this view, the New Deal had a more profound impact on American society than even its most ardent admirers have suggested: it imposed norms of relative equality in pay that persisted for more than 30 years, creating the broadly middle-class society we came to take for granted. But those norms began to unravel in the 1970's and have done so at an accelerating pace.

Exhibit A for this view is the story of executive compensation. In the 1960's, America's great corporations behaved more like socialist republics than like cutthroat capitalist enterprises, and top executives behaved more like public-spirited bureaucrats than like captains of industry. I'm not exaggerating. Consider the description of executive behavior offered by John Kenneth Galbraith in his 1967 book, ''The New Industrial State'': ''Management does not go out ruthlessly to reward itself -- a sound management is expected to exercise restraint.'' Managerial self-dealing was a thing of the past: ''With the power of decision goes opportunity for making money. . . . Were everyone to seek to do so . . . the corporation would be a chaos of competitive avarice. But these are not the sort of thing that a good company man does; a remarkably effective code bans such behavior. Group decision-making insures, moreover, that almost everyone's actions and even thoughts are known to others. This acts to enforce the code and, more than incidentally, a high standard of personal honesty as well.''

Thirty-five years on, a cover article in Fortune is titled ''You Bought. They Sold.'' ''All over corporate America,'' reads the blurb, ''top execs were cashing in stocks even as their companies were tanking. Who was left holding the bag? You.'' As I said, we've become a different country.

Let's leave actual malfeasance on one side for a moment, and ask how the relatively modest salaries of top executives 30 years ago became the gigantic pay packages of today. There are two main stories, both of which emphasize changing norms rather than pure economics. The more optimistic story draws an analogy between the explosion of C.E.O. pay and the explosion of baseball salaries with the introduction of free agency. According to this story, highly paid C.E.O.'s really are worth it, because having the right man in that job makes a huge difference. The more pessimistic view -- which I find more plausible -- is that competition for talent is a minor factor. Yes, a great executive can make a big difference -- but those huge pay packages have been going as often as not to executives whose performance is mediocre at best. The key reason executives are paid so much now is that they appoint the members of the corporate board that determines their compensation and control many of the perks that board members count on. So it's not the invisible hand of the market that leads to those monumental executive incomes; it's the invisible handshake in the boardroom.

But then why weren't executives paid lavishly 30 years ago? Again, it's a matter of corporate culture. For a generation after World War II, fear of outrage kept executive salaries in check. Now the outrage is gone. That is, the explosion of executive pay represents a social change rather than the purely economic forces of supply and demand. We should think of it not as a market trend like the rising value of waterfront property, but as something more like the sexual revolution of the 1960's -- a relaxation of old strictures, a new permissiveness, but in this case the permissiveness is financial rather than sexual. Sure enough, John Kenneth Galbraith described the honest executive of 1967 as being one who ''eschews the lovely, available and even naked woman by whom he is intimately surrounded.'' By the end of the 1990's, the executive motto might as well have been ''If it feels good, do it.''

How did this change in corporate culture happen? Economists and management theorists are only beginning to explore that question, but it's easy to suggest a few factors. One was the changing structure of financial markets. In his new book, ''Searching for a Corporate Savior,'' Rakesh Khurana of Harvard Business School suggests that during the 1980's and 1990's, ''managerial capitalism'' -- the world of the man in the gray flannel suit -- was replaced by ''investor capitalism.'' Institutional investors weren't willing to let a C.E.O. choose his own successor from inside the corporation; they wanted heroic leaders, often outsiders, and were willing to pay immense sums to get them. The subtitle of Khurana's book, by the way, is ''The Irrational Quest for Charismatic C.E.O.'s.''

But fashionable management theorists didn't think it was irrational. Since the 1980's there has been ever more emphasis on the importance of ''leadership'' -- meaning personal, charismatic leadership. When Lee Iacocca of Chrysler became a business celebrity in the early 1980's, he was practically alone: Khurana reports that in 1980 only one issue of Business Week featured a C.E.O. on its cover. By 1999 the number was up to 19. And once it was considered normal, even necessary, for a C.E.O. to be famous, it also became easier to make him rich.

Economists also did their bit to legitimize previously unthinkable levels of executive pay. During the 1980's and 1990's a torrent of academic papers -- popularized in business magazines and incorporated into consultants' recommendations -- argued that Gordon Gekko was right: greed is good; greed works. In order to get the best performance out of executives, these papers argued, it was necessary to align their interests with those of stockholders. And the way to do that was with large grants of stock or stock options.

It's hard to escape the suspicion that these new intellectual justifications for soaring executive pay were as much effect as cause. I'm not suggesting that management theorists and economists were personally corrupt. It would have been a subtle, unconscious process: the ideas that were taken up by business schools, that led to nice speaking and consulting fees, tended to be the ones that ratified an existing trend, and thereby gave it legitimacy.

What economists like Piketty and Saez are now suggesting is that the story of executive compensation is representative of a broader story. Much more than economists and free-market advocates like to imagine, wages -- particularly at the top -- are determined by social norms. What happened during the 1930's and 1940's was that new norms of equality were established, largely through the political process. What happened in the 1980's and 1990's was that those norms unraveled, replaced by an ethos of ''anything goes.'' And a result was an explosion of income at the top of the scale.

IV. The Price of Inequality

It was one of those revealing moments. Responding to an e-mail message from a Canadian viewer, Robert Novak of ''Crossfire'' delivered a little speech: ''Marg, like most Canadians, you're ill informed and wrong. The U.S. has the longest standard of living -- longest life expectancy of any country in the world, including Canada. That's the truth.''

But it was Novak who had his facts wrong. Canadians can expect to live about two years longer than Americans. In fact, life expectancy in the U.S. is well below that in Canada, Japan and every major nation in Western Europe. On average, we can expect lives a bit shorter than those of Greeks, a bit longer than those of Portuguese. Male life expectancy is lower in the U.S. than it is in Costa Rica.

Still, you can understand why Novak assumed that we were No. 1. After all, we really are the richest major nation, with real G.D.P. per capita about 20 percent higher than Canada's. And it has been an article of faith in this country that a rising tide lifts all boats. Doesn't our high and rising national wealth translate into a high standard of living -- including good medical care -- for all Americans?

Well, no. Although America has higher per capita income than other advanced countries, it turns out that that's mainly because our rich are much richer. And here's a radical thought: if the rich get more, that leaves less for everyone else.

That statement -- which is simply a matter of arithmetic -- is guaranteed to bring accusations of ''class warfare.'' If the accuser gets more specific, he'll probably offer two reasons that it's foolish to make a fuss over the high incomes of a few people at the top of the income distribution. First, he'll tell you that what the elite get may look like a lot of money, but it's still a small share of the total -- that is, when all is said and done the rich aren't getting that big a piece of the pie. Second, he'll tell you that trying to do anything to reduce incomes at the top will hurt, not help, people further down the distribution, because attempts to redistribute income damage incentives.

These arguments for lack of concern are plausible. And they were entirely correct, once upon a time -- namely, back when we had a middle-class society. But there's a lot less truth to them now.

First, the share of the rich in total income is no longer trivial. These days 1 percent of families receive about 16 percent of total pretax income, and have about 14 percent of after-tax income. That share has roughly doubled over the past 30 years, and is now about as large as the share of the bottom 40 percent of the population. That's a big shift of income to the top; as a matter of pure arithmetic, it must mean that the incomes of less well off families grew considerably more slowly than average income. And they did. Adjusting for inflation, average family income -- total income divided by the number of families -- grew 28 percent from 1979 to 1997. But median family income -- the income of a family in the middle of the distribution, a better indicator of how typical American families are doing -- grew only 10 percent. And the incomes of the bottom fifth of families actually fell slightly.

Let me belabor this point for a bit. We pride ourselves, with considerable justification, on our record of economic growth. But over the last few decades it's remarkable how little of that growth has trickled down to ordinary families. Median family income has risen only about 0.5 percent per year -- and as far as we can tell from somewhat unreliable data, just about all of that increase was due to wives working longer hours, with little or no gain in real wages. Furthermore, numbers about income don't reflect the growing riskiness of life for ordinary workers. In the days when General Motors was known in-house as Generous Motors, many workers felt that they had considerable job security -- the company wouldn't fire them except in extremis. Many had contracts that guaranteed health insurance, even if they were laid off; they had pension benefits that did not depend on the stock market. Now mass firings from long-established companies are commonplace; losing your job means losing your insurance; and as millions of people have been learning, a 401(k) plan is no guarantee of a comfortable retirement.

Still, many people will say that while the U.S. economic system may generate a lot of inequality, it also generates much higher incomes than any alternative, so that everyone is better off. That was the moral Business Week tried to convey in its recent special issue with ''25 Ideas for a Changing World.'' One of those ideas was ''the rich get richer, and that's O.K.'' High incomes at the top, the conventional wisdom declares, are the result of a free-market system that provides huge incentives for performance. And the system delivers that performance, which means that wealth at the top doesn't come at the expense of the rest of us.

A skeptic might point out that the explosion in executive compensation seems at best loosely related to actual performance. Jack Welch was one of the 10 highest-paid executives in the United States in 2000, and you could argue that he earned it. But did Dennis Kozlowski of Tyco, or Gerald Levin of Time Warner, who were also in the top 10? A skeptic might also point out that even during the economic boom of the late 1990's, U.S. productivity growth was no better than it was during the great postwar expansion, which corresponds to the era when America was truly middle class and C.E.O.'s were modestly paid technocrats.

But can we produce any direct evidence about the effects of inequality? We can't rerun our own history and ask what would have happened if the social norms of middle-class America had continued to limit incomes at the top, and if government policy had leaned against rising inequality instead of reinforcing it, which is what actually happened. But we can compare ourselves with other advanced countries. And the results are somewhat surprising.

Many Americans assume that because we are the richest country in the world, with real G.D.P. per capita higher than that of other major advanced countries, Americans must be better off across the board -- that it's not just our rich who are richer than their counterparts abroad, but that the typical American family is much better off than the typical family elsewhere, and that even our poor are well off by foreign standards.

But it's not true. Let me use the example of Sweden, that great conservative bete noire.

A few months ago the conservative cyberpundit Glenn Reynolds made a splash when he pointed out that Sweden's G.D.P. per capita is roughly comparable with that of Mississippi -- see, those foolish believers in the welfare state have impoverished themselves! Presumably he assumed that this means that the typical Swede is as poor as the typical resident of Mississippi, and therefore much worse off than the typical American.

But life expectancy in Sweden is about three years higher than that of the U.S. Infant mortality is half the U.S. level, and less than a third the rate in Mississippi. Functional illiteracy is much less common than in the U.S.

How is this possible? One answer is that G.D.P. per capita is in some ways a misleading measure. Swedes take longer vacations than Americans, so they work fewer hours per year. That's a choice, not a failure of economic performance. Real G.D.P. per hour worked is 16 percent lower than in the United States, which makes Swedish productivity about the same as Canada's.

But the main point is that though Sweden may have lower average income than the United States, that's mainly because our rich are so much richer. The median Swedish family has a standard of living roughly comparable with that of the median U.S. family: wages are if anything higher in Sweden, and a higher tax burden is offset by public provision of health care and generally better public services. And as you move further down the income distribution, Swedish living standards are way ahead of those in the U.S. Swedish families with children that are at the 10th percentile -- poorer than 90 percent of the population -- have incomes 60 percent higher than their U.S. counterparts. And very few people in Sweden experience the deep poverty that is all too common in the United States. One measure: in 1994 only 6 percent of Swedes lived on less than $11 per day, compared with 14 percent in the U.S.

The moral of this comparison is that even if you think that America's high levels of inequality are the price of our high level of national income, it's not at all clear that this price is worth paying. The reason conservatives engage in bouts of Sweden-bashing is that they want to convince us that there is no tradeoff between economic efficiency and equity -- that if you try to take from the rich and give to the poor, you actually make everyone worse off. But the comparison between the U.S. and other advanced countries doesn't support this conclusion at all. Yes, we are the richest major nation. But because so much of our national income is concentrated in relatively few hands, large numbers of Americans are worse off economically than their counterparts in other advanced countries. And we might even offer a challenge from the other side: inequality in the United States has arguably reached levels where it is counterproductive. That is, you can make a case that our society would be richer if its richest members didn't get quite so much.

I could make this argument on historical grounds. The most impressive economic growth in U.S. history coincided with the middle-class interregnum, the post-World War II generation, when incomes were most evenly distributed. But let's focus on a specific case, the extraordinary pay packages of today's top executives. Are these good for the economy?

Until recently it was almost unchallenged conventional wisdom that, whatever else you might say, the new imperial C.E.O.'s had delivered results that dwarfed the expense of their compensation. But now that the stock bubble has burst, it has become increasingly clear that there was a price to those big pay packages, after all. In fact, the price paid by shareholders and society at large may have been many times larger than the amount actually paid to the executives.

It's easy to get boggled by the details of corporate scandal -- insider loans, stock options, special-purpose entities, mark-to-market, round-tripping. But there's a simple reason that the details are so complicated. All of these schemes were designed to benefit corporate insiders -- to inflate the pay of the C.E.O. and his inner circle. That is, they were all about the ''chaos of competitive avarice'' that, according to John Kenneth Galbraith, had been ruled out in the corporation of the 1960's. But while all restraint has vanished within the American corporation, the outside world -- including stockholders -- is still prudish, and open looting by executives is still not acceptable. So the looting has to be camouflaged, taking place through complicated schemes that can be rationalized to outsiders as clever corporate strategies.

Economists who study crime tell us that crime is inefficient -- that is, the costs of crime to the economy are much larger than the amount stolen. Crime, and the fear of crime, divert resources away from productive uses: criminals spend their time stealing rather than producing, and potential victims spend time and money trying to protect their property. Also, the things people do to avoid becoming victims -- like avoiding dangerous districts -- have a cost even if they succeed in averting an actual crime.

The same holds true of corporate malfeasance, whether or not it actually involves breaking the law. Executives who devote their time to creating innovative ways to divert shareholder money into their own pockets probably aren't running the real business very well (think Enron, WorldCom, Tyco, Global Crossing, Adelphia . . . ). Investments chosen because they create the illusion of profitability while insiders cash in their stock options are a waste of scarce resources. And if the supply of funds from lenders and shareholders dries up because of a lack of trust, the economy as a whole suffers. Just ask Indonesia.

The argument for a system in which some people get very rich has always been that the lure of wealth provides powerful incentives. But the question is, incentives to do what? As we learn more about what has actually been going on in corporate America, it's becoming less and less clear whether those incentives have actually made executives work on behalf of the rest of us.

V. Inequality and Politics

In September the Senate debated a proposed measure that would impose a one-time capital gains tax on Americans who renounce their citizenship in order to avoid paying U.S. taxes. Senator Phil Gramm was not pleased, declaring that the proposal was ''right out of Nazi Germany.'' Pretty strong language, but no stronger than the metaphor Daniel Mitchell of the Heritage Foundation used, in an op-ed article in The Washington Times, to describe a bill designed to prevent corporations from rechartering abroad for tax purposes: Mitchell described this legislation as the ''Dred Scott tax bill,'' referring to the infamous 1857 Supreme Court ruling that required free states to return escaped slaves.

Twenty years ago, would a prominent senator have likened those who want wealthy people to pay taxes to Nazis? Would a member of a think tank with close ties to the administration have drawn a parallel between corporate taxation and slavery? I don't think so. The remarks by Gramm and Mitchell, while stronger than usual, were indicators of two huge changes in American politics. One is the growing polarization of our politics -- our politicians are less and less inclined to offer even the appearance of moderation. The other is the growing tendency of policy and policy makers to cater to the interests of the wealthy. And I mean the wealthy, not the merely well-off: only someone with a net worth of at least several million dollars is likely to find it worthwhile to become a tax exile.

You don't need a political scientist to tell you that modern American politics is bitterly polarized. But wasn't it always thus? No, it wasn't. From World War II until the 1970's -- the same era during which income inequality was historically low -- political partisanship was much more muted than it is today. That's not just a subjective assessment. My Princeton political science colleagues Nolan McCarty and Howard Rosenthal, together with Keith Poole at the University of Houston, have done a statistical analysis showing that the voting behavior of a congressman is much better predicted by his party affiliation today than it was 25 years ago. In fact, the division between the parties is sharper now than it has been since the 1920's.

What are the parties divided about? The answer is simple: economics. McCarty, Rosenthal and Poole write that ''voting in Congress is highly ideological -- one-dimensional left/right, liberal versus conservative.'' It may sound simplistic to describe Democrats as the party that wants to tax the rich and help the poor, and Republicans as the party that wants to keep taxes and social spending as low as possible. And during the era of middle-class America that would indeed have been simplistic: politics wasn't defined by economic issues. But that was a different country; as McCarty, Rosenthal and Poole put it, ''If income and wealth are distributed in a fairly equitable way, little is to be gained for politicians to organize politics around nonexistent conflicts.'' Now the conflicts are real, and our politics is organized around them. In other words, the growing inequality of our incomes probably lies behind the growing divisiveness of our politics.

But the politics of rich and poor hasn't played out the way you might think. Since the incomes of America's wealthy have soared while ordinary families have seen at best small gains, you might have expected politicians to seek votes by proposing to soak the rich. In fact, however, the polarization of politics has occurred because the Republicans have moved to the right, not because the Democrats have moved to the left. And actual economic policy has moved steadily in favor of the wealthy. The major tax cuts of the past 25 years, the Reagan cuts in the 1980's and the recent Bush cuts, were both heavily tilted toward the very well off. (Despite obfuscations, it remains true that more than half the Bush tax cut will eventually go to the top 1 percent of families.) The major tax increase over that period, the increase in payroll taxes in the 1980's, fell most heavily on working-class families.

The most remarkable example of how politics has shifted in favor of the wealthy -- an example that helps us understand why economic policy has reinforced, not countered, the movement toward greater inequality -- is the drive to repeal the estate tax. The estate tax is, overwhelmingly, a tax on the wealthy. In 1999, only the top 2 percent of estates paid any tax at all, and half the estate tax was paid by only 3,300 estates, 0.16 percent of the total, with a minimum value of $5 million and an average value of $17 million. A quarter of the tax was paid by just 467 estates worth more than $20 million. Tales of family farms and businesses broken up to pay the estate tax are basically rural legends; hardly any real examples have been found, despite diligent searching.

You might have thought that a tax that falls on so few people yet yields a significant amount of revenue would be politically popular; you certainly wouldn't expect widespread opposition. Moreover, there has long been an argument that the estate tax promotes democratic values, precisely because it limits the ability of the wealthy to form dynasties. So why has there been a powerful political drive to repeal the estate tax, and why was such a repeal a centerpiece of the Bush tax cut?

There is an economic argument for repealing the estate tax, but it's hard to believe that many people take it seriously. More significant for members of Congress, surely, is the question of who would benefit from repeal: while those who will actually benefit from estate tax repeal are few in number, they have a lot of money and control even more (corporate C.E.O.'s can now count on leaving taxable estates behind). That is, they are the sort of people who command the attention of politicians in search of campaign funds.

But it's not just about campaign contributions: much of the general public has been convinced that the estate tax is a bad thing. If you try talking about the tax to a group of moderately prosperous retirees, you get some interesting reactions. They refer to it as the ''death tax''; many of them believe that their estates will face punitive taxation, even though most of them will pay little or nothing; they are convinced that small businesses and family farms bear the brunt of the tax.

These misconceptions don't arise by accident. They have, instead, been deliberately promoted. For example, a Heritage Foundation document titled ''Time to Repeal Federal Death Taxes: The Nightmare of the American Dream'' emphasizes stories that rarely, if ever, happen in real life: ''Small-business owners, particularly minority owners, suffer anxious moments wondering whether the businesses they hope to hand down to their children will be destroyed by the death tax bill, . . . Women whose children are grown struggle to find ways to re-enter the work force without upsetting the family's estate tax avoidance plan.'' And who finances the Heritage Foundation? Why, foundations created by wealthy families, of course.

The point is that it is no accident that strongly conservative views, views that militate against taxes on the rich, have spread even as the rich get richer compared with the rest of us: in addition to directly buying influence, money can be used to shape public perceptions. The liberal group People for the American Way's report on how conservative foundations have deployed vast sums to support think tanks, friendly media and other institutions that promote right-wing causes is titled ''Buying a Movement.''

Not to put too fine a point on it: as the rich get richer, they can buy a lot of things besides goods and services. Money buys political influence; used cleverly, it also buys intellectual influence. A result is that growing income disparities in the United States, far from leading to demands to soak the rich, have been accompanied by a growing movement to let them keep more of their earnings and to pass their wealth on to their children.

This obviously raises the possibility of a self-reinforcing process. As the gap between the rich and the rest of the population grows, economic policy increasingly caters to the interests of the elite, while public services for the population at large -- above all, public education -- are starved of resources. As policy increasingly favors the interests of the rich and neglects the interests of the general population, income disparities grow even wider.

VI. Plutocracy?

In 1924, the mansions of Long Island's North Shore were still in their full glory, as was the political power of the class that owned them. When Gov. Al Smith of New York proposed building a system of parks on Long Island, the mansion owners were bitterly opposed. One baron -- Horace Havemeyer, the ''sultan of sugar'' -- warned that North Shore towns would be ''overrun with rabble from the city.'' ''Rabble?'' Smith said. ''That's me you're talking about.'' In the end New Yorkers got their parks, but it was close: the interests of a few hundred wealthy families nearly prevailed over those of New York City's middle class.

America in the 1920's wasn't a feudal society. But it was a nation in which vast privilege -- often inherited privilege -- stood in contrast to vast misery. It was also a nation in which the government, more often than not, served the interests of the privileged and ignored the aspirations of ordinary people.

Those days are past -- or are they? Income inequality in America has now returned to the levels of the 1920's. Inherited wealth doesn't yet play a big part in our society, but given time -- and the repeal of the estate tax -- we will grow ourselves a hereditary elite just as set apart from the concerns of ordinary Americans as old Horace Havemeyer. And the new elite, like the old, will have enormous political power.

Kevin Phillips concludes his book ''Wealth and Democracy'' with a grim warning: ''Either democracy must be renewed, with politics brought back to life, or wealth is likely to cement a new and less democratic regime -- plutocracy by some other name.'' It's a pretty extreme line, but we live in extreme times. Even if the forms of democracy remain, they may become meaningless. It's all too easy to see how we may become a country in which the big rewards are reserved for people with the right connections; in which ordinary people see little hope of advancement; in which political involvement seems pointless, because in the end the interests of the elite always get served.

Am I being too pessimistic? Even my liberal friends tell me not to worry, that our system has great resilience, that the center will hold. I hope they're right, but they may be looking in the rearview mirror. Our optimism about America, our belief that in the end our nation always finds its way, comes from the past -- a past in which we were a middle-class society. But that was another country.