Imperfect Vision

IT'S A MARK OF OUR PROGRESS AGAINST INFLATION THAT the greatest need now is measuring it. Once in double digits, it's so low that we're not sure what it is. The Boskin Commission--named after its chairman, Michael Boskin--brings us closer to a better reading. The commission of five economists says that the consumer price index (CPI) overstates inflation by 1.1 percentage points a year. The precise size of the overstatement is less important than the fact that it is large and persistent. It burdens the federal budget, distorts our economic record and creates uncertainty about the future.

The reaction to the report has predictably focused on the budget. Social Security and some other programs are indexed to the CPI. If Congress cuts the adjustment (as it should), future benefit increases will be smaller. Similarly, tax brackets and the personal exemption rise automatically with the CPI. This holds down taxes by making less income taxable or taxing it at lower rates. A lower CPI adjustment would mean slightly higher taxes. All told, a cut of 1 percentage point (say, from 3 percent to 2) in the adjustment would shave budget deficits cumulatively by an estimated $628 billion between 1997 and 2006.

But the Boskin report's larger significance is that it demolishes the theory that living standards have stagnated. No longer should it be possible for some economists and politicians to contend that average families have simply run in place in recent decades. The stagnation theory has always been a statistical illusion. If your income rises 3 percent and prices rise 3 percent, then you've got no "real" income gain. But if inflation has actually risen 2 percent, then income is up 1 percent. This is what's happened.

Consider: By the official statistics, the income of the median household--the one exactly in the middle--rose a meager 6.5 percent between 1975 and 1995. That's close to stagnation. But the gain would have been 17 percent if actual inflation was 0.5 percentage points less annually than the official rate; and the gain would have been 29 percent if inflation were 1 percentage point lower a year. How can we know that whacking inflation like that isn't simply a statistical trick? Well, we can examine other surveys that measure what people buy and own. If they own more and buy more, their purchasing power--their real income--has risen.

It has. Compared with 1975, Americans now own bigger homes with more air conditioners and appliances (VCRs, microwave ovens, personal computers). They watch more cable TV and exercise more at health clubs. And they eat out more often. In 1995, about 46 percent of the food dollar was spent in restaurants, up from 42 percent in 1985 and 36 percent in 1975. If people felt so pinched, there'd be more brown-bag lunches. Starbucks symbolizes the growth of discretionary income, and it's absurd to attribute all the gains to more two-earner couples.

The main reason that the CPI overstates inflation is that it inadequately accounts for new and improved products. Some products that rise in price provide more value for money. Appliances (refrigerators, washers, TVs) last longer and require less repair. So do cars. Since 1980, the average age of a car has risen from 6.6 to 8.4 years. New products (VCRs, microwaves) expand consumer convenience but are added into the CPI with a delay--often after big price drops have occurred. And the CPI misses some price cuts--for example, discount air fares before 1982 and frequent-flier miles and grocery coupons today.

The result is that the CPI understates purchasing power. One reason people may not feel the gains is that their wants are rising even faster. In 1987, the median family judged that it needed $20,000 "just to get by," reports one survey; by 1996, that was $30,000--a 50 percent gain that outstripped inflation or real-income gains. The amount needed to "fulfill all your dreams" jumped even more: 80 percent, to $90,000. (Having said this, we need to avoid too much revisionism. Regardless of statistical jiggering, income growth was faster in the 1950s and 1960s--a fact that economists can't explain except as a "catch-up" after the Depression and World War II. Incomes now may not be rising as fast as people expected. And recent income gains still favor the wealthy over the poor.)

The fact that the CPI isn't perfect doesn't mean that it can be instantly improved. No single statistic will ever exactly measure inflation, because there are too many products whose prices change too often and whose quality variations--for worse as well as better--can never be completely counted. The Bureau of Labor Statistics (BLS), which compiles the CPI, disputes some of the commission estimates; there's room for disagreement. The worst response to the report would be for the White House and Congress to pressure the BLS to alter the CPI. The BLS is studying improvements, but if it's stampeded into adopting only those that lower inflation, we could end up with a less useful indicator--one, for example, that masks some price increases.

It's no secret why Democrats and Republicans alike want a quick change. They prefer that the BLS do their dirty work, to relieve them of the responsibility of altering the legislative indexing of Social Security and taxes. Gosh, won't these guys do anything? Indexing is supposed to prevent the automatic erosion of government benefits (through reduced purchasing power) or automatic tax increases (by pushing people into higher brackets). When inflation was high, adopting indexing was right. But it was never supposed to provide automatic benefit increases or tax cuts; and that's what today's overindexing does.

Although we may not know exactly how much the CPI overstates inflation, we know that it does. Ignoring this mortgages our future. Politicians get paid to make judgments based on limited evidence. That's what they should do here. Slashing the adjustment at least 0.5 percentage points (say, from 3 percent to 2.5) is justified for five years until we see how the BLS alters the CPI. This wouldn't impose hardships. The rise in average Social Security benefits in the first year would be only about $36 less; tax increases would also be small. If the White House and Congress can't do this, political cowardice would be the only explanation.