(FORTUNE Magazine) – The battles raging in Washington since Newt Gingrich became Speaker of the House have been epic in their ferocity. But what is the war really about?

For some 20 years--ever since the end of the great postwar boom in the American economy-debates over economic policy have been dominated by promises of revived growth. Ronald Reagan promised that his package of tax cuts, targeted mainly to high-income families, would produce a rising tide that would lift all boats. Bill Clinton pledged to "focus like a laser beam" on the economy, making it grow through investments in training and infrastructure. And Newtonian economics, if I understand it, seems to promise that tax cuts will be really effective now that we're all on the Internet.

The fact is, all these promises are silly: Administrations don't cause recession and recoveries--if anyone is in charge of the business cycle, it's the nonpartisan technocrats at the Federal Reserve. That doesn't mean we should ignore the effects of Washington's actions on economic efficiency. Some policies (educating children and respecting property rights) are clearly good for growth; others (running huge budget deficits and taxing at punitive rates) are bad for it. But the overwhelming evidence is that the growth rate of the U.S. economy has remained stolidly unmoved by all the posturing of the past two decades. Any politician who claims he can raise it by as much as three-tenths of a percentage point is naive--or worse.

This statement will, of course, outrage true believers on both sides of the ideological divide. But it's a hard statistical fact. Look at economic recoveries since 1975. Every one has reduced the unemployment rate to around 5% but no further. (It fell to 5.8% in 1979, 5.3% in 1989, and 5.4% today). In other words, no Administration has succeeded in changing the level of unemployment associated with "full employment."

Now look at the relationship between changes in unemployment and economic growth. It turns out there is a stunningly close correlation between the two--an equation that is one of the few things in economics reliable enough to be called, with tongue only slightly in cheek, a "law." And it can be expressed like this: The growth rate of the U.S. economy during any given year equals 2.5% plus twice the fall in unemployment from the beginning to the end of that year--or minus twice the rise in unemployment over that same period.

So if the unemployment rate falls from, say, 6.5% to 5.5%, the economy that year will have grown by 2.5% + (2 2 1%) = 4.5%. An increase in joblessness, signaling economic weakness, would have the reverse effect. If unemployment climbs from 6% to 6.75%, the economy will have grown by 2.5% - (2 2 0.75%) = 1%. Apply this formula to U.S. data since 1975, and you'll find that the correlation, though not exact--this is economics, not quantum physics--is remarkably close (see chart).

The fact that this equation works so well tells us two things. First, since there has been no shift in the "full employment" rate of unemployment, the underlying growth rate of the U.S. economy has been a very stable 2.5% right through the past five Administrations. Second, because almost all the variation in growth tracks with fluctuations in unemployment--that is, with the up and downs of the business cycle, which has been managed by Paul Volcker and Alan Greenspan, not Ronald Reagan, George Bush, or Bill Clinton-no recent President has had any visible effect, positive or negative, on U.S. growth.

Maybe you believe that Phil Gramm, or a miraculously recovered Bill Clinton, can somehow do better (or worse). But that belief is a triumph of hope over experience. The hardheaded, realistic assessment is that whatever is at stake in U.S. politics, it isn't the economy's growth rate.

So what are the real economic issues? The main difference between economic liberals and economic conservatives is over what they propose to do with the welfare state. Conservatives want to lower the welfare state's safety net as close to the floor as possible, if not remove it completely. Liberals want to preserve that net, maybe even raise it a bit higher. This isn't a dispute about how to promote growth--it's a dispute about interests and values.

Given that, let me make my own values clear. I am an economic liberal. That means that, while I believe in free trade and have no sympathy with the sort of liberalism that wants to centralize economic decision-making in Washington (cases in point: Jimmy Carter's energy planners and Bill Clinton's health planners), I also believe in the need to preserve, even strengthen, the social safety net.

What do I mean by the welfare state? In the U.S., this national web of protections against economic hardship consists mainly of six programs: two that benefit the old (Social Security and Medicare); one aimed at workers (unemployment insurance); and three aimed at the poor--Medicaid, food stamps, and last, and very much least in terms of actual spending, Aid to Families With Dependent Children, the part most people think of as "welfare." Taking all levels of government, spending on these programs--and the taxing required to pay for them--amounts to somewhat less than 15% of GDP. (In most other advanced countries, the total is about twice as large.) Almost two-thirds of that money goes to the elderly, via Social Security and Medicare. Only about one-fifth, or less than 3% of GDP, is in the form of public-aid programs aimed at the poor. (And contrary to popular impressions, a majority of these recipients are white, not black, Americans.)

Who wins and who loses from this redistribution? The great myth of modern American politics is that the middle class is groaning under the burden of huge taxes deployed to support the undeserving poor. In fact, what the welfare state really does is to take from the well-off (a little) and give to the poor (also a little, but because they are so poor it matters a lot). Families in the middle are not much affected either way.

The Statistical Abstract of the United States tells the story clearly enough (see chart). In 1992 the richest fifth of families received slightly more than half of all private, pretax income: 51%. By the time government taxes and transfer programs were through with them, their share was down to 43.3%. At the other end, families in the bottom quintile received a mere 0.9% of private income; government programs raised their income share (including noncash benefits) to a still less than princely 4.9%. For families near the middle, benefits were somewhat larger than the taxes used to pay for social welfare. The pretax share of the middle quintile was 15.4%; its post-tax-and-transfer share 16.7%.

Conservatives will argue, with some justification, that the true cost of the welfare state is higher than this simple calculation indicates. After all, taxes surely discourage some people from earning as much as they might, and benefits encourage some people to work less than they would otherwise. So it is possible that families in the middle quintile actually end up with less income than they'd have if the welfare state were abolished--but at worst, no more than a few percent less.

Even so, why should middle-income families support a system that gives them roughly the same amount in benefits that they pay in taxes?

One answer that, I believe, still has more force with the public than the 1994 election might suggest is simple compassion. It does bother most Americans to see deep poverty in the midst of general affluence. A second answer, which reinforces the first, is that the benefits of the non-Social Security portion of the welfare state flow disproportionately to children. Even those who feel little sympathy for the "undeserving" adult poor generally believe society owes the very young decent food, housing, and clothing.

But there is a final argument one rarely hears in this country, which is surprising because in other advanced nations it's the key reason a majority of the populace strongly supports the welfare state. It's this: social welfare spending is mainly a form of insurance.

America is, after all, a dynamic market economy --a society that not only accepts but even welcomes constant change. Most of us believe that dynamism is part of what makes our country great. But with change comes risk. Today's well-paid worker may discover that the industry he works in has suddenly been downsized, that the skills he has developed over the years are suddenly made obsolete by a new technology. The middle manager of 1985 may find that the new, flat organization of 1995 has no need for his services. And, of course, alongside the hazards of the market are the hazards of the human condition: illness, widowhood, divorce.

For all these reasons America in this era of just 2.5% average growth--an era that neither liberals nor conservatives know how to end--is marked by just as much downward as upward social mobility. Indeed, since Americans tend to think of themselves as middle class even when their incomes put them well into the top 5%, it's worth bearing in mind that there are more than twice as many families below the poverty line as there are families with incomes of more than $100,000.

Now consider what effect the welfare state has on the prospects for a middle-income American family in this inevitably risky economy. If it stays in the middle, that family can expect to receive about as much in benefits as it pays in taxes. The family or its children may, of course, rise into the upper-middle class, or even become rich; if they do, they will pay considerably more in taxes than they receive in benefits, and their income will be something like 20% lower than it would have been without a welfare state. But it is equally likely that the family will fall into the bottom fifth of the income distribution, perhaps even below the poverty line; in that case, their income will be double or triple what it would have been in a society with no public safety net. I believe there is no question that the overwhelming majority of people would, if they understood it, regard this as a good bargain--that the security that comes from knowing food, shelter, and health care will be there in the worst case is well worth paying an easily affordable tax bill in the best case.

So by all means, let's have a vigorous national debate about reforming Social Security; it can't be sustained in its present form. Let's also debate how best to care for our poorest citizens; the current system is too easily abused and contains various perverse incentives. But dismantle the welfare state? Some middle-class Americans may think that's what they want. But if they get their wish, they'll be amazed how much they miss it.

Paul Krugman is a professor of economics at Stanford University.