The 'squeeze' play.

IVORY-TOWER economists have a mathematical model for just about any occasion. They even have one--the "median voter" model--that predicts how successful politicians will compete to craft tax plans pandering to the middle class.

This pandering tendency does, however, present something of a challenge to politicians. If previous governments enacted policies that were designed to bring bliss to homo medius, then politically attractive changes to current policy become more difficult to devise, for there are fewer gifts left to give. The income tax today is a nice case in point. According to the latest numbers from the Congressional Budget Office, the top 20 percent of the population pays a whopping 82 percent of all income taxes, while the middle 20 percent pays only 5.2 percent. Lowering rates at the top might make everyone better off, but no politician has stepped forward and argued that 5.2 percent is too small a burden on the middle class.

Nevertheless, the latest rhetorical twist on the campaign trail is that there is a "middle-class squeeze." Sure, the economy has turned around, the story goes, but the poor guy in the middle is not seeing the benefit. Amillion new jobs have been created this year, but the jobs are "bad jobs with low wages." And what's worse, health-care costs are so high that folks in the middle just can't make ends meet. Other necessary expenditures are being postponed. What follows is that new non-tax policies, such as the $2 trillion in new social spending offered by Senator Kerry, are needed to rescue the middle class from all that trauma.

If there were a middle-class squeeze, it would be most helpful to political challengers, who could claim that incumbents paid too little attention to the middle class. But the notion of the squeeze is a stretch. As measured by GDP, the economy has now grown every year since 1991. It even grew three-quarters of a percentage point during 2001, the recession year. Before 1991, one has to go all the way back to 1982 to find another year when the economy shrank. If the economy on average is getting richer and richer, how exactly can the middle class be squeezed at the same time?

The latest squeezologists support their claims by alluding to a narrow set of facts. Most cite recent work by Jared Bernstein of the Economic Policy Institute that reports that the real (adjusted for inflation) income of the median family declined in the two most recent years for which we have data. Bernstein has also shown that real family incomes have increased over time, but much of the increase is attributable to second earners. The implication is that individuals, like runners on a treadmill, are working harder just to stay in place.

But these observations provide an incomplete view of the current welfare of the middle class. Bernstein's most recent numbers are for 2002. The economy has, however, accelerated significantly since then. For example, it grew 4.8 percentage points over the past twelve months, a growth rate that was not exceeded in any one-year period during President Clinton's two terms in office. Presumably, this healthy growth has led to improved welfare for the typical denizen of our economy.

More important, incomes vary across individuals for many reasons. Many retirees have high wealth but not a lot of current earnings. The welfare of such people is poorly approximated by their income. There are others who can expect to have high lifetime wages but experience transitory shocks in a recession. If they plan ahead, their standard of living might hold up nicely despite bad news.

Because of these and other factors, academic economists have moved toward a consensus that it is best to rely on consumption rather than income when evaluating changes in welfare. If a retiree is wealthy, her income might be low but her consumption could be high. If a shock has temporarily made a high-income fellow look poor, his consumption might be more revealing of his true economic standing.

The government's Consumer Expenditure Survey provides the best picture of the recent evolution of middle-class consumption. The data tell a strikingly different story from the income data. Real consumption expenditures by the middle 20 percent of Americans increased about 2 percentage points from 2001 to 2002, the last two years for which we have data. Over a longer period they did even better. From 1984 to 2002, consumption of those in the middle increased by about 10 percent. That is hardly a squeeze.

What about the claim that health-care costs are crowding out other expenditures? That notion is also quite dubious. Excluding healthcare spending, consumption grew from 1984 to 2002 by about 8 percent, only a fraction smaller than the overall gain.

The claims that homo medius is not sharing in the latest boom are also almost certainly false. The statistics are striking. Let's compare numbers for 2004 with those from 1996, the last year a president was seeking reelection. In 1996, about 5 million individuals reported that they were working part time for economic reasons. In 2004, the number is down to 4.3 million. In 1996, 65 percent of Americans owned a home. Today, 69 percent do. In 1996, 41 percent of Americans owned stock. Today, half do. The growth rate of real average hourly earnings over the past four years has been 3.3 percent. During Clinton's first term, the growth rate was negative 0.5 percent.

What about those high education costs? In 1996, average financial aid per student (in constant 2002 dollars) was $5,851. In 2004, it was $9,099. And that money is having an effect. In 1995, 62 percent of high-school graduates went on to college. In 2003, 64 percent did.

So the claims that things are getting worse for the middle class seem weak. Which leaves only one part of the squeeze theory unscathed: It may well be that much of the improved welfare of the middle class is attributable to the fact that married females have increased their participation in the labor force. It is difficult to support the view that providing women with employment opportunity worsens the welfare of the middle class. If individuals voluntarily depart the home and enter the labor force, they must be doing so because they believe their choice to be a satisfying one. Moreover, that choice provides many significant benefits. Chief among these is the underappreciated diversification of a household's labor income. If a family has two workers bringing home paychecks, it is much less likely that a recession will cause family income to drop perilously low. The Cleavers might have lost their home had Ward ever been fired. Today's two-earner couples are much better prepared for calamity. Today, June's income would help make ends meet until Ward found another job. Accordingly, households feel less compelled to stockpile resources as insurance against layoffs. This likely explains why consumption has held up so well--even in the bad years.

The consumption data and the topline GDP numbers suggest that the middle class has been doing fine. This view is reinforced by studies of voting behavior. The leader in this field, Ray Field of Yale University, has found that broad measures of income--not distributional analysis--are the best predictors of election outcomes. If GDP is growing, voters are happy. In 1996, for instance, real GDP grew 3.7 percent, which was more than enough to overpower all of the negative middle-class statistics for President Clinton. Income has been booming lately, so votes should go to incumbents this election cycle. If income were shrinking, there might be a middle-class squeeze. It isn't--so there's no squeeze.

Mr. Hassett is director of economic policy at the American Enterprise Institute.