The Dismal Science
Illustration by Robert Neubecker The CPI and the Rat Race
New evidence on the old question of whether money buys happiness.

By Paul Krugman
(1,422 words; posted Saturday, Dec. 21; to be composted Saturday, Jan. 4)

       Let's talk about inflation indexing and the meaning of life.
       Early this month a panel of economists, led by Stanford's Michael Boskin, made semiofficial what most experts have been saying for some time: The Consumer Price Index overstates inflation. Nobody really knows by how much, but Boskin and company made a guesstimate of 1.1 percent annually. Compounded over decades, this is a huge error.
       This conclusion is controversial. Some people are upset because any reduction of inflation estimates will reduce Social Security benefits, which are indexed to the CPI. Others are upset because a revision of recent price history would mean abandoning a worldview on which they have staked their reputations. Quite a few people have committed themselves to the story line that productivity is up but real wages are down. If inflation has been lower than was previously assumed, that means the real value of wages may have gone up after all. And some economists with no particular ax to grind simply have doubts about the methodology.

Boskin may be right or wrong, but one argument by his critics is clearly wrong. They say: Suppose it's true that inflation has been less than the official increase in the CPI over the past few decades. If you assume a lower inflation rate and recalculate real incomes back to--say, 1950--you reach what seems to be a crazy conclusion: that in the early 1950s, the era of postwar affluence, most Americans were living below what we now regard as the poverty line. Some critics of the Boskin report regard this as a decisive blow to its credibility.
       The idea that most Americans were poor in 1950 is indeed absurd, but not because of Boskin's numbers. After all, even if you use an unadjusted CPI, the standard of living of the median family (50th percentile) in 1950 America appears startlingly low by current standards. In that year, median-family income in 1994 dollars was only about $18,000. That's about the 20th percentile today. Families at the 20th percentile--that is, poorer than 80 percent of the population--may not be legally poor (only about 12 percent of families are officially below the poverty line), but they are likely to regard themselves as very disadvantaged and unsuccessful. So even using the old numbers, most families in 1950 had a material standard of living no better than that of today's poor or near-poor.

Illustration by Robert Neubecker We can confirm this with more direct measures of the way people lived. In 1950 some 35 percent of dwellings lacked full indoor plumbing. Many families still did not have telephones or cars. And of course very few people had televisions. A modern American family at the 12th percentile (that is, right at the poverty line) surely has a flushing toilet, a working shower, and a telephone with direct-dial long-distance service; probably has a color television; and may well even have a car. Take into account improvements in the quality of many other products, and it does not seem at all absurd to say that the material standard of living of that poverty-level family in 1996 is as good as or better than that of the median family in 1950.
       What do we mean by this? We mean that if you could choose between the two material standards of living, other things being the same, you might well prefer the 12th percentile standard of 1996 to the 50th percentile standard of 1950. But does that mean that most people were poor in 1950? No--because man does not live by bread, cars, televisions, or even plumbing alone.

Imagine that a mad scientist went back to 1950 and offered to transport the median family to the wondrous world of the 1990s, and to place them at, say, the 25th percentile level. The 25th percentile of 1996 is a clear material improvement over the median of 1950. Would they accept his offer? Almost surely not--because in 1950 they were middle class, while in 1996 they would be poor, even if they lived better in material terms. People don't just care about their absolute material level--they care about their level compared with others'.
       I know quite a few academics who have nice houses, two cars, and enviable working conditions, yet are disappointed and bitter men--because they have never received an offer from Harvard and will probably not get a Nobel Prize. They live very well in material terms, but they judge themselves relative to their reference group, and so they feel deprived. And on the other hand, it is an open secret that the chief payoff from being really rich is, as Tom Wolfe once put it, the pleasure of "seeing 'em jump." Privilege is not merely a means to other ends, it is an end in itself.

My fellow SLATE columnist Robert Wright would undoubtedly emphasize that our concern over status exists for good evolutionary reasons. In the ancestral environment a man would be likely to have more offspring if he got his pick of the most fertile-seeming women. That, in turn, would depend on his status, not his absolute standard of living. So males with a predisposition to status-seeking left more offspring than those without, and the end result is Bill G-g-g---I mean, Ronald Perelman.
       Is my license as a practicing economist about to be revoked? Aren't we supposed to believe in Economic Man? And doesn't admitting that people care about fuzzy things like status undermine the whole economic method? Not really: Homo economicus is not a central pillar of my faith--he is merely a working assumption, albeit one that is extremely useful in many circumstances.
       But admitting that people's happiness depends on their relative economic level as well as their absolute economic resources has some subversive implications. For example: Many conservatives have seized on the Boskin report as a club with which to beat all those liberals who have been whining about declining incomes and increasing poverty in America. It was all, they insist, a statistical hoax. But you could very well make the opposite argument. America in the 1950s was a middle-class society in a way that America in the 1990s is not. That is, it had a much flatter income distribution, so that people had much more sense of sharing a common national lifestyle. And people in that relatively equal America felt good about their lives, even though by modern standards, they were poor--poorer, if Boskin is correct, than we previously thought. Doesn't this mean, then, that having a more or less equal distribution of income makes for a happier society, even if it does not raise anyone's material standard of living? That is, you can use the fact that people did not feel poor in the 1950s as an argument for a more radical egalitarianism than even most leftists would be willing to espouse.

You could even argue that American society in the 1990s is an engine that maximizes consumption yet minimizes satisfaction. In a society with a very flat distribution of income and status, nobody feels left out. In a society with rigid ranks, people do not expect to rise above their station and therefore do not feel that they have failed if they do not rise. (Aristocrats are not part of the peasants' reference group.) Modern America, however, is a hugely unequal society in which anyone can achieve awesome success, but not many actually do. The result is that many--perhaps even most--people feel that they have failed to make the cut, no matter how comfortable their lives. (In a land where anyone can become president, anyone who doesn't become president is a failure.) My European friends always marvel at how hard Americans work, even those who already have plenty of money. Why don't we take more time to enjoy what we have? The answer, of course, is that we work so hard because we are determined to get ahead--an effort that (for Americans as a society) is doomed to failure, because competition for status is a zero-sum game. We can't all "get ahead." No matter how fast we all run, someone must be behind.
       If one follows this line of thought one might well be led to some extremely radical ideas about economic policy, ideas that are completely at odds with all current orthodoxies. But I won't try to come to grips with such ideas in this column. Frankly, I don't have the time. I have to get back to my research--otherwise, somebody else might get that Nobel.


At the center of the controversy that may turn American assumptions about the economy on their heads is the Boskin Commission Report. The PoliticsNow Web site has the full text, and the Bureau of Labor Statistics has a whole page devoted to the Consumer Price Index and how it is calculated. "Trouble In Paradise: Eroding Wages and Growing Income Inequality" is a Competitiveness Policy Council report that uses the current calculation of the CPI to come to the common (but possibly misguided) conclusion that real wages are declining. The Urban Institute revisits the question of economic inequality in several reports: "Are Justice and Inequality Compatible?" "How Much Do Americans Move Up and Down the Economic Ladder?" and "American Dreams and Discontents: Beyond the Level Playing Field." If it is not standard of living but inequality that causes dissatisfaction in America, then the responsibility for much unhappiness falls on the shoulders of the Forbes 400. This list is designed in a way that can cause even multibillionaire Warren Buffet financial angst: He is only the second-richest man in America.

Paul Krugman is a professor of economics at MIT whose books include The Age of Diminished Expectations and Peddling Prosperity.

Illustrations by Robert Neubecker