Clinton And The Cpi Mess

WE COULD CONCEIVABLY GET PAST THE CONTROVERSY over the consumer price index (CPI) with some presidential leadership, but there's been none. Clinton has been absent, and the whole episode shows him at his worst. The man is a compulsive follower. He's a genius at sniffing the public mood--and then going with it. All politicians do this, but Clinton does little else. He seems incapable of leading public opinion and disguises his timidity with soaring rhetoric and strategic silences.

This is the story of the CPI mess. By now, the basics of the controversy are well known. There's broad (though not universal) agreement among economists that the CPI--the government's main price indicator--overstates inflation. Last fall a commission of economists appointed by the Senate Finance Committee and headed by Michael Boskin of Stanford put the overstatement at 1.1 percent a year (with a range of 0.8 to 1.6 percent). What Congress and the president must decide is whether--and how much--to alter the automatic indexing of Social Security, other government benefits and the income tax to the CPI.

Clinton now seems to have torpedoed an effort to deal with the issue. The idea was to create another panel of experts to advise Congress on exactly how much to alter the CPI for indexing. Senate Majority Leader Trent Lott endorsed the concept; Clinton wouldn't--or at least hasn't. The politics aren't complicated. Some powerful Democrats (notably House leader Richard Gephardt) oppose it. So do some potent interest groups (notably the American Association of Retired Persons). Congressional Republicans, having been attacked repeatedly over Social Security and Medicare, won't touch the issue alone. Thus, an impasse.

I think that the way out is for Congress to cut the CPI 0.5 percent for indexing. If the CPI rose 3.3 percent, everything indexed to it would rise only 2.8 percent. This seems a cautious compromise between what we know (that the CPI now overstates inflation) and what we don't (the exact size of the overstatement). Numbers aside, the issue is not just an accounting dispute.

Flawed indexing feeds budget deficits and silently twists the government's priorities. Indexing makes sense. Congress adopted it for Social Security in 1972 and for the income tax in 1981. The aim was to prevent inflation from eroding the purchasing power of benefits or raising taxes. Increasing benefits for inflation maintained purchasing power, and adjusting tax brackets and other tax provisions insulated people from higher tax rates when their real incomes hadn't risen. But overindexing is no virtue. It automatically overcompensates and undertaxes; Congress loses control over more and more of the budget.

Because no one (starting with the president) has had the courage to demystify it, the whole subject has become needlessly symbolic and charged. People wrongly think that altering the index formula would impose huge hardships. It wouldn't. Suppose Congress actually did reduce the indexing formula by 0.5 percent. What would be the effects?

Consider a new retiree in 1996 with a Social Security monthly benefit of $745. The Social Security Administration has compared benefits under its official projection with those after a 0.5 percent cut in indexing. In the first year the monthly benefit would rise to $765 instead of $768: a $3 difference. After 10 years, the monthly benefit would be $1,020 instead of $1,070: the $50 gap is less than 5 percent. After 17 years (average life expectancy for someone at 65), the benefit would be $1,297 instead of $1,408: $111 or 7.9 percent.

The story is the same for taxes. If the personal exemption (and other tax provisions) were raised 0.5 percent less than the official CPI, taxes would increase. But here, too, the effect is tiny. After a decade, the cumulative tax increase is less than 1 percent, estimates the Congressional Budget Office. But because many people are affected--on both taxes and benefits--the impact on the budget can be significant. The CBO esti- mates that a cut in indexing of 0.5 percent would trim the budget deficit in 2005 by $51 billion, or a fifth of the projected deficit ($254 billion).

If the CPI overstates inflation, no one is entitled--at least as a matter of fairness--to the higher benefits or lower taxes that result from overindexing. We are often told that the CPI is a technical issue and that any change should be left to the Bureau of Labor Statistics. Well, the CPI should be left to the BLS; and the agency should be given the money to improve the CPI. But how the index is used for the budget is inevitably a political matter.

Arguing otherwise is self-serving and hypocritical. Imagine that the Boskin Commission had found that the CPI understated inflation by 1 percent so that Social Security recipients were undercompensated and taxpayers were overtaxed. All the people now advising patience while the BLS fixes the CPI would have clamored for Congress to adjust the indexing formula immediately to boost benefits and cut taxes. And that would have been right.

We can't wait for the BLS to construct a perfect cost-of-living index, because it won't ever happen. The job is too hard. The BLS already regularly checks 95,000 prices in 22,000 businesses. But it can't instantly adapt the CPI to changing buying habits, new products and technological changes. At the end of 1996, Americans had 44 million cellular phones, up from 5 million in 1990. But the CPI won't include prices for cell phones until 1998. It will have missed most of their huge price drop. Economists will argue over the size of biases--both overstating and understating inflation--but in the end, Congress must judge their importance.

One job of the president is to clarify hard issues. This is easier to urge than for presidents to do. All of them strive to preserve their political capital by picking their fights selectively. Still, Clinton has almost completely abandoned the job. He's the most talkative president in living memory, but he rarely talks about things people don't want to hear. He must be preserving his political capital. For what is anyone's guess.