(FORTUNE Magazine) – The average price of a Manhattan apartment south of Harlem has hit more than $850,000--at a time when two-fifths of New York City's residents make $20,000 or less a year. In Silicon Valley teachers struggle with the rent while dot-com-rich parents wonder how to cope with "affluenza"--the perils of new and great wealth. (Hint: Just don't buy that helicopter.) In leafy suburbs nurses and cops commute from 50 miles away: They cannot afford to live near their work.

This dichotomy--between new wealth and the not-so-wealthy--has lately become something of an academic and political obsession. Economists and social scientists have turned the study of income inequality into a thriving cottage industry. And while the rich-poor gap has not cropped up explicitly in the presidential campaign, it is the subtext for a number of front-burner issues like tax cuts, educational reform, and the "digital divide." When a politician uses the word "fairness" in an economic debate, that's often shorthand for "inequality."

Why the concern about inequality? Basically, because there's more of it. From 1977 on, the cash earnings of the poorest fifth of the U.S. population fell about 9%, estimates the Center on Budget and Policy Priorities; middle-class earnings rose 8%; and upper-income earnings, 43%. The exact numbers are hotly contested, but it is clear that the distance between the top and the bottom tiers of the income distribution has grown strikingly since the 1970s. By some measures, Americans' earnings are more unequal today than at any time in the past 60 years; at best, even after the past several years, when income has grown throughout the income distribution, the gap has plateaued at or near record levels.

Of course, no serious person would argue that everyone should get the same-sized piece of the economic pie. That would be unfair to those who work hard, as opposed to those who watch reruns of Gilligan's Island all day. And if spectators want to pay more to watch a baseball game than, say, a badminton match, there is no reason both sets of athletes must be paid alike. At the same time, no serious person would deny that inequality can hit such levels (think medieval societies) that it comprises both an ethical problem and a threat to social peace (the peasants revolt). Finally, there is little disagreement about whether inequality has increased. It has. But there is also massive mud-wrestling about how much it has grown, why, and what it all means.

FORTUNE will spare you the arcane details--for now, anyway. But the fundamental argument about inequality is simple. The pessimists contend that income distribution has grown so lopsided that all society is worse off. Richard Freeman of Harvard speculates that there is a link between inequality and crime. He notes that high school dropouts fill the nation's jails--and that these men have lost the most ground economically. Edward Wolff of New York University contends that if young men had a better shot at earning a stable living they might be more willing to marry and stop having children on a freelance basis. Robert Greenstein of the Center on Budget and Policy Priorities argues that earnings disparities are one of the reasons that almost one in five children lives in poverty. America's lowest-paid workers make less, as a percentage of the median wage (the point at which 50% are above and 50% below), than their counterparts in any other country (38%, compared with 46% in Britain and Japan and more than 50% in France and Germany). This means that many low-skilled parents just cannot earn enough to escape poverty. "If there were somewhat less inequality," Greenstein concludes, "more would have a better standard of living."

There is also considerable (but contentious) literature that more-equal societies are healthier. And there is the inchoate but deeply felt belief that inequality at current levels is simply un-American. It gives the rich too loud a voice. It makes it too hard for those at the bottom to rise to prosperity. And it allows the wealthy to separate themselves from society through private clubs, private schools, and gated communities.

The optimists respond to that critique with a polite yawn. Or perhaps a rude word along the lines of "Rubbish!" Sure, inequality has grown, but so what? As long as people at the bottom have not become absolutely worse off, goes this set of arguments, it doesn't matter that the rich got richer faster. And no, the poor are not worse off. Though men's earnings seem to have fallen since 1973 (and maybe they haven't), women's have clearly risen. That trend and smaller households mean that family income and income per head have increased all along the income distribution. Housing quality and access to medical care have improved markedly for the poor since 1973. Besides, people don't necessarily stay in the same position. They move up and down the income ladder: Horatio Alger was not just making stuff up. Today's income distribution is the result of long-standing economic forces and social trends. Nothing is broke, so don't fix it.

Those are the broad outlines of a debate in which the devil is most definitely in the details. What follows is a primer of the arguments, followed by a suggestion about how to get out of this thicket.

What are people so concerned about? Students of inequality use several tools in their trade. One is the Gini coefficient; a 0 coefficient is perfect equality (everyone has exactly the same share of the economic pie). A coefficient of 1 is perfect inequality (Bill Gates gets it all). In America the coefficient has risen from 0.323 in 1974 to 0.375 in 1997, according to the Luxembourg Income Study, higher than in any other rich country. Britain's is 0.346, Germany's 0.300, Canada's 0.286, and Sweden's 0.222.

Matters naturally are not quite that straightforward. Alan Greenspan has pointed out that while the Gini coefficient is comparatively high for income, when applied to consumption it is about 25% lower. In other words, poorer people are spending more like the rich; they are, for example, almost as likely to own such things as dryers and microwave ovens. So the economic distance between the top and the bottom may be narrower than the income numbers suggest. And Europe's greater equality may simply reflect the widely accepted premise that while America has adapted to economic change by allowing inequality to rise, Europe has adjusted by allowing higher unemployment. Which is better?

Another favored analytical tool for measuring inequality is to divide the population into fifths, or quintiles, and see what share of the nation's earnings each fifth took home. According to the Census Bureau, in 1998 the bottom 20% earned only 3.6% of total income (4.2% in 1973), compared with more than 49% for the top 20% (44% in 1973).

But wait a minute. The Heritage Foundation points out that the Census defines quintiles in terms of households--and households in the bottom quintile are much smaller than those at the top. Therefore, while there are 64 million people in the richest quintile, there are fewer than 40 million in the poorest one. Adjust for population, and the share of the bottom fifth grows. Also, many Americans have income that is not in the form of wages or cash transfers--food stamps and housing subsidies for the poor, realized capital gains for the better-off. Adjust for that, and the distribution narrows again, as it does after accounting for taxes. Should the adjustment include Medicaid and Medicare? If so (and that is debatable), the gap shrinks further still; put it all together, and Heritage figures that the bottom quintile takes in 9.4% of national income, and the top 39.6%.

There is, then, no consensus on how to measure inequality. There is, however, broad agreement that it has indeed grown. Since the early 1970s the cash incomes of the rich have indeed risen faster than those of the poor, with the middle class hanging in there; the higher up the income ladder, the faster the growth. That may help explain why the poverty rate, now 12.7%, has still not dipped to 1973 levels (11.1%). Median household income (the point at which 50% are above and 50% below) has grown grudgingly, rising about 9% in real terms from 1973 to 1998 and passing its 1989 peak only in 1998.

Men have had a particularly dismal time. The median income of men is significantly lower than in 1973 ($27,394 then vs. $25,212 in 1997, in 1997 dollars). Men under 45 are making less now, in real terms, than they did in 1967, and blue-collar workers have taken the biggest hit. Blacks and women, however, have seen their earnings rise.

Why is inequality increasing? Income inequality is increasing because wage inequality is. The U.S. economy has evolved to reward highly educated people even more than in the past--a trend that social scientists, in a flight of whimsy, call "skill-biased technological change." This means that demand for labor has shifted toward the skilled and away from the unskilled. Brains beat brawn--hands down.

That explains the rise in the college premium--the extra income college graduates can expect to earn compared with those who finish only high school. The premium rose much faster in the U.S. than in Europe because the supply of graduates in the U.S. did not rise as fast in the 1980s and 1990s as the demand for them; Europe came closer to matching demand and supply. It sounds like a tautology, and perhaps it is: Income shifted toward the more highly skilled because employers would pay more for their services. But it really is that simple.

Of course, that by itself doesn't explain the income gap. Another significant factor has been family structure. Weighing on the downscale side of income distribution has been the burgeoning number of single-parent families, particularly those headed by never-married mothers; overall, single-parent families earn about half as much as two-parent households. On the upscale side, there has been an increase in families in which both spouses make lots of money. To put it another way, there are almost 2 1/2 times as many people working in the richest fifth of households as in the poorest fifth. Less than a third of the people in the bottom quintile live in households headed by a married couple; the rest are single (55%) or in single-parent families. In the top quintile some 90% live in married-couple families.

Changes in family structure account for more than a third of the increase in income inequality since 1979, figures Gary Burtless of the Brookings Institution, making it a slightly more important factor than the widening wage gap. Lynn Karoly of the Rand Institute in California calculates that the wage gap is a bigger deal, but no matter: No one disputes that both factors are crucial.

Other suspects in the inequality lineup are the declining minimum wage (lower in real terms than in 1973), declining unionization among men (accounting for as much as 20% of the gap, estimates Freeman), deregulation (protected industries kept wages high), immigration (which can depress wages), and trade (that giant sucking sound). Higher levels of entrepreneurship may also be associated with higher inequality.

All those things probably count, but to a minor degree compared with the changes in earnings patterns and family structure. Immigrants, for example, can drive down wages in local labor markets, particularly among the low-skilled, but that effect is muted across the country as a whole. When it comes to trade, the effect is even more difficult to identify. While some companies have certainly shipped jobs to cheaper climes, most U.S. trade is with other rich countries, and most low-paid jobs are domestic, such as cleaning or food service. Remember, too, that to critique immigration and trade strictly in terms of their impact on inequality is to look through a cracked mirror: Doing so ignores the contributions immigrants make to America and the opportunities wrought by freer trade.

What is more important than any of these individual factors, Karoly notes, is how all of them have reinforced one another. At the same time, there have been few countervailing forces. The U.S. could have tried to slow these trends, as Europe has done, through high minimum wages or centralized wage bargaining or protective trade barriers or high taxes. It chose not to.

What can be done? The primary rule of economic policy should be like that of medicine: First, do no harm. And the problem with many of the knee-jerk policy responses to inequality is that they cannot pass that test. Looking at the list of culprits responsible for the run-up in inequality, for instance, one could argue for less technological change, less trade, more regulation, and less entrepreneurship. Would America really be better off with such an economic blueprint? To ask the question is to answer it.

Even the more plausible approaches carry side effects worth thinking about. Take unions. Unions are an essential part of a free society, and they do an excellent job of raising wages for members. But they can also be associated with not-so-good things, such as protecting their workers at the expense of those trying to get into the labor market--an important factor in the high level of European unemployment. In July, Alan Greenspan contended that it was America's greater labor-market flexibility that had allowed it to take advantage of information technologies faster and more fully than Europe; tech-led productivity has been the bedrock of America's recent wage and productivity surge. In this context, the case for actively encouraging more unionization begins to weaken.

What about raising the minimum wage? That's plausible too, and the increased minimum wage probably played a role in steadying inequality in the past few years. Moreover, countries like France, which has a high minimum wage, have seen inequality grow much less. America may be robust enough to swallow the proposed minimum-wage increase to $6.15. But there is clearly a point where a minimum wage can become burdensome, killing job opportunities, as has happened in Europe. And raising the minimum wage is an awkward way to lessen inequality. Most minimum-wage workers do not live in low-income households (think of suburban teens), and many poor households have no workers at all. So most of the gain from a higher minimum wage goes to families that are not poor. Worse, the Organization for Economic Cooperation and Development has documented a connection between the minimum wage and youth unemployment: the higher the wage, the more idle youngsters. That has to be a large part of the reason a quarter of France's under-25-year-olds are out of work.

Is all this simply an argument for complacency? Not quite. It is really an argument for looking at the issue from a different perspective. Let's face it: Normal Americans do not fret about rising Gini coefficients or quintile displacements. They do, however, worry if hard-working people, even professionals, cannot find a home of their own that fits their means. They don't want children suffering, even if their parents made bad choices. They believe that opportunity is available to all and that government should not hinder people's ability to take care of themselves. Americans, in short, are hapless at class warfare (perhaps because they are so absorbed in racial and ethnic issues). If they were better at it, they would be howling, say, at the proposed death of the death tax, which applies to only a tiny share of estates. Instead, most people want it killed. The attitude seems to be, "Hey, that might be my estate someday."

Given such attitudes, a plausible list of goals for government might go something like this:

--Enhance the prospects of poor children. --Improve living conditions. --Reward work. --Bolster family responsibility. --Keep taxes from impoverishing people. --Ensure mobility.

And surprise, surprise: American social policy in the 1980s and '90s has done almost precisely that. The Reagan Administration can take credit for the 1986 tax reform, which released many lower-income Americans from federal income-tax liability. The earned-income-tax credit (EITC), also a Reagan-era initiative, supplements the pay of low-wage workers with children through a refundable tax credit of up to 40% of earnings. The Bush and Clinton Administrations expanded the EITC (the latter in the teeth of strong Republican opposition). Both also expanded the provision of support services for poor children outside the home--child care, foster care, Head Start, and so on. Child-support enforcement expanded under all three (with, it has to be said, spotty results), and health insurance and child-care subsidies for poor children expanded under Bush and Clinton. The welfare reform of 1996 (in the teeth of strong Democratic opposition) explicitly connected working to the receipt of benefits. Overall, these policies make up a broadly consistent approach that Americans are in tune with--and that has delivered real improvements.

Perhaps, then, the way to remedy inequality is not so much to try to lessen the Gini coefficient--through redistributive taxation, for example--but to ameliorate the problems of those snagged at the bottom. One such problem is clearly housing. There is a gap between the growing numbers of low-income renters (10.5 million in 1995) and the shrinking numbers of low-cost rental units (6.1 million). A record 5.4 million households spend more than half of their income on rent or live in substandard housing. The feds can and should do more in this regard by boosting the number of housing vouchers. (Congress eliminated new housing vouchers for four years in the 1990s; the 2000 budget envisions expansion.)

But inequality begins at home. It is not coincidental that two cities with massive affordability problems--New York and San Francisco--may also have the most tortured housing markets in the country. Byzantine regulations suppress new construction and raise its cost. Insiders--those who have scored a price-controlled apartment--benefit at the expense of outsiders, who pay prices exaggerated by the artificially induced constraint in supply. So while rent decontrol rarely makes the egalitarian to-do list, it deserves to be on it. And Silicon Valley and other wealthy communities should take a hard look at regulations--two-acre zoning and the like--that put up a KEEP OUT sign for the unrich.

Expanding the EITC further--by increasing the credit (particularly to families with three or more children) and extending it to childless full-time workers--would also help. The EITC is first-rate social policy. Essentially it promises parents that if they work, their income will exceed the poverty line. In 1998, EITC supplements lifted almost five million people out of poverty, and that money has proved an important carrot to get former welfare recipients into the job market. A further expansion would put more dollars in low earners' pockets and reduce the ranks of the working poor, without the scattershot effect of the minimum wage. It also makes perfect equity sense in the context of the tax cuts both parties are fiddling with. Don't believe the fluff: Tax cuts would benefit the better-off most, for the very good reason that they pay the lion's share of taxes. The top 1% of earners, for example, pays almost a fifth of all individual federal income taxes, according to the Congressional Budget Office, and the top fifth almost 60%. The bottom two quintiles contribute 8%. An expanded EITC, in combination with tax cuts, would spread tax largesse all the way up and down the income distribution. Along the same lines, states that are considering cutting taxes would do well to cut sales taxes, which hit the poor hardest, rather than income taxes. Or they could start or expand their own versions of the EITC, as more than a dozen states have already done.

Third, surely a country as rich and talented as America can figure out some way to ensure reasonable, regular health care at a level of access that, say, Ireland provided in the 1960s. There has been expansion of guaranteed medical provision for poor children, but about 15% still slip between the cracks. A system with fewer gaps could also promote mobility; it is scary for low-income people in a job with health coverage to try to improve their position by moving to a new job without it.

Fourth, let's remember that not every problem comes with a ready solution, from government or anywhere else. For example, it would be an unambiguously good thing for America as a whole if families formed more readily and stayed together more reliably. This would also narrow wage inequality and boost family income. It's just far from obvious how to get there from here.

Social policy is not a field of dreams; miracles are rare. Across the rich world, estimates Ignazio Visco of the OECD, the long-term poor are some 2% to 4% of the population. But at any given time, these families make up half of the population living in poverty--everyone else moves up and out. The major problem in such homes is not lack of money but disorganization, illness, lack of social skills, and general cluelessness. In her book What Money Can't Buy, Susan Mayer of the University of Chicago argues that after basic needs are met, additional income has little effect on children's prospects. Using a form of regression analysis that only a social scientist could love (or indeed understand), Mayer estimates that doubling the income of the poor would reduce high school dropout rates by one percentage point, increase education by a few months, have no effect on teen pregnancy, and possibly worsen male idleness. "Any realistic redistribution strategy," she concludes, "is likely to have a relatively small impact on the overall incidence of social problems." Enhancing living standards to provide dignity and reasonable comfort is a social good in itself. But humility is warranted in terms of the long-range benefits of doing so.

In the long run, because so much of inequality is connected with the higher returns on skills, it is crucial that Americans learn the things they need to know in order to succeed. Which brings us to education, the most important component of the mobility that is the bedrock of the American dream. Poor people in poor communities are educationally shortchanged, and the problems begin early. That Americans of almost any intellectual level can find a college to accept them does not excuse the lack of basic skills too many high school graduates demonstrate. Money may be part of the answer, but only part. Cash can be spent wisely or stupidly; there is, at best, an ambiguous correlation between spending and achievement. But evidence indicates that increased attention to education in early childhood brings enduring and positive results. It's clear that there has to be more emphasis on accountability and outcomes--what children actually learn--as opposed to how much is being spent. That's beginning to happen. And it's hard to believe that competition--vouchers, charter schools, and the like--would not be a goad to improvement.

Finally, let's remember that nothing good is going to happen if the economy goes into the tank. Tight labor markets have done more to make welfare reform work than any aspect of its design; productivity has driven up wages since 1993 faster than any transfer program could have done. Remedies to inequality that hurt the economy as a whole will hurt the poor first and worst.

Laura D'Andrea Tyson, former head of the Council of Economic Advisors under President Clinton, offered a striking way of looking at these issues at a Federal Reserve conference in 1998. Imagine the income distribution, she suggested, as an apartment building in which the penthouse is more and more luxurious, and the basement, in which a number of dwellers (and their children) are stuck year after year, is rat infested. What to do? Well, some social critics, offended by the presence of wealth amid such distress, would like to pillage the penthouse. Tyson simply notes, "We need to do something about that rat-infested basement." Taking care of the rats and making sure people can climb out of the cellar: That seems about right.

FEEDBACK: cmurphy@fortunemail.com