At the end of World War II, income inequality was lower in the United States than at any time since the 1920s. During the ensuing three decades, incomes grew briskly and at about the same rate - almost 3 percent per year - for households up and down the income ladder.

That pattern began to change in the 1970s. Since 1979, for example, the incomes of families in the bottom 80 percent of the income distribution have grown by less than 1 percent each year, and only households in the top 20 percent have enjoyed income growth comparable to that in the earlier period. For a small group at the very top of the economic ladder, however, incomes have been growing explosively.

For more than 25 years, Business Week has conducted an annual survey of the earnings of chief executive officers of the largest U.S. corporations. In 1980, those executives earned 42 times as much as the average American worker, a ratio larger than the corresponding ratios for such countries as Japan and Germany even today. By 2000, however, American CEOs were earning 531 times the average worker's salary. The gains have been even larger for those above CEOs on the income ladder.

With corporate malfeasance much in the news, we know that at least some of the spectacular corporate pay packages were not won on merit. Most of them, however, are a simple consequence of market forces. As local markets have given way to regional, then national, and now global markets, even a few slightly improved executive decisions can now add hundreds of millions of dollars to the bottom line.

More generally, rapid pay growth at the top owes much to the spread of reward structures once confined largely to markets for sports and entertainment. In these "winner-take-all markets," small differences in performance often translate into enormous differences in economic reward. Now that we listen mostly to recorded music, the world's best musicians can literally be everywhere at once. The electronic news wire has allowed a small number of syndicated columnists to displace a host of local journalists. And the proliferation of personal computers has enabled a handful of software developers to replace thousands of local tax accountants. Each change has benefited consumers but has also led to greater inequality.

Around the globe, income inequality has been growing for essentially similar reasons. In most countries, public policy has attempted to counter this trend. Not in the United States. With the market's push toward greater inequality already apparent, for example, Congress reduced the top marginal income tax rate from 50 percent to 28 percent during the 1980s.

These tax cuts have increased inequality not only through their direct effects on after-tax incomes, but also through indirect effects on federal spending policies. Although supply-side economists predicted that the cuts would increase tax revenues by stimulating more than enough income growth to offset the lower rates, this did not happen, and hence the large budget deficits of the 1980s.

Those deficits were eliminated during the Clinton years but have reappeared, larger than ever, under President Bush, who has reduced tax rates on earnings, dividends and large inheritances. Once the enabling legislation is fully phased in, more than half of the resulting cuts - 52.5 percent, according to one recent estimate - will go to the top 5 percent of earners. The nonpartisan Congressional Budget Office now forecasts deficits larger than $300 billion for each of the next six years.

Many proponents of smaller government applaud these deficits, arguing that they will force legislators to cut wasteful spending. As always, however, budget cuts focus not on wasteful programs but on those whose beneficiaries are least able to resist them. Recent proposals by House Republicans would eliminate free school lunches for 40,000 children and food stamps for 225,000 people in working households with children. House Republicans also propose $12 billion in cuts for Medicaid, a program on which 25 percent of American children now rely heavily for access to medical care.

The combined effects of market forces and changes in public policy have clearly made life more difficult for middle- and low-income people. They are working longer hours, saving less, borrowing more, commuting longer distances, and doing without things once considered essential. Personal bankruptcy filings have set new records in each of the last several years. The personal savings rate, always low by international standards, has fallen sharply since the 1980s. It has hovered close to zero since the late 1990s, and in recent months has actually been negative. About 45 million Americans now have no health insurance, 5 million more than in the early 1990s.

Although income inequality has increased sharply in recent decades, it has always been greater here than in other industrial democracies. Can a case be made for it? Many have described inequality as the price we must pay to achieve high rates of economic growth.

The evidence, however, suggests otherwise. As economists Alberto Alesina and Dani Rodrik have found, for example, growth rates across countries are negatively related to the share of national income going to top earners.

Others have portrayed inequality as a necessary condition for socioeconomic mobility, arguing that people who are willing to work hard and play by the rules face a better chance of making it to the top here than in any other country. But here, too, the evidence suggests otherwise. Even as economic inequality has been rising, social mobility has been declining. According to sociologist David Wright, the probability that a child born to parents in the third quartile of the income distribution would move up into the top quartile was only half as large in 1998 as in 1973. Economist Thomas Hertz has found that children whose parents are in the bottom fifth of the income distribution have only a 7.3 percent chance of making it into the top fifth. In contrast, children born in the top fifth have a 42.3 percent chance of remaining there. Contrary to popular impressions, socioeconomic mobility is now lower in the United Stated than in most other industrialized countries.

Although the market forces that have been producing higher inequality show no signs of abating, Republicans in Congress are now calling for an additional $70 billion in tax cuts aimed largely at high-income families, arguing that because the most prosperous Americans have worked hard, they are entitled to keep a greater portion of their pretax incomes. But tens of millions of less prosperous Americans have worked hard, too. And in winner-take-all markets, examples abound in which some earn thousands of times more than others just as talented and hardworking.

The economist Herbert Stein once said, if something cannot go on forever, it won't. History has repeatedly demonstrated that societies can tolerate income inequality only up to a point, beyond which they rapidly disintegrate. Numerous governments in Latin America have been overthrown largely because of social unrest rooted in income inequality. And in a survey of more than a quarter of a million randomly selected individuals worldwide, economist Robert MacCulloch found that people in countries with high income inequality were much more likely to voice support for violent revolution.

Major social upheavals are sometimes preceded by years or even decades of rising levels of social unrest. If such unrest is currently building in the United States, it remains well-hidden. But as recent experience has made clear, social upheavals often occur with virtually no warning. Almost no one predicted the fall of the Eastern European governments in 1989. Because revolutions almost always entail important elements of social contagion, even small changes can launch political prairie fires once a tipping point is reached.

As Plutarch wrote almost 2000 years ago, "An imbalance between rich and poor is the oldest and most fatal ailment of all republics." Before the United States succumbs to that ailment, we might want to reconsider the wisdom of policies that widen that already large gap.

Robert H. Frank (rhf3@cornell.edu) is an economist at the Johnson School of Management at Cornell University and the coauthor, with Philip J. Cook, of "The Winner-Take-All Society."