Paul Krugman is a speed-limit man: Rapid growth brings inflation! So he has said and written for years, in columns for

Slate

,

Fortune

, and now the

New York Times

. And those who doubt it are "scofflaws." But in all that time, inflation hasn't obliged him.

Now, beset by critics, tormented by the sophistries of lesser spirits, worried perhaps by the lengthening record of his own mis-predictions, Krugman seizes the scrap of recent inflation news as a sign that -- just perhaps -- the evidence is swinging his way.

In Sunday's

New York Times

he adopted a tone both Delphic and Olympian: "It's starting to look," he writes, "as if the unemployment rate will have to rise a bit -- from 4% to say 4.5% if we're lucky, to 5% if we aren't. And growth will have to slow, from the 4-plus% of the last two years to 3.5%, maybe 3%."

Where do these numbers come from? Is there a model out there from which they fall? Some authority that Krugman himself has come to trust, and whom he commends to us? If so, the oracle declines to name his source.

Instead, Krugman's argument rests on a wisp of recent news, an assertion about the nature of that news, and a filament of textbook economics. The wisp of news is the economy's recent high growth rate. The assertion -- and here Krugman injects the wiggle words "it seems" to protect himself -- is that high growth in total spending is the result of an "exuberant stock market." The thread of textbook economics is that growth fueled by demand, rather than supply, must produce inflation.

But difficulties arise at each level of this thinking. To begin with, the late 1999 growth rate now dominating the news may prove ephemeral. As

Harvard's

James Medoff has argued, there were vast expenditures then to replace Y2K noncompliant computers. These show up in the growth numbers -- and they may have added to recent average wage gains as well as distorting productivity measures -- but they mean nothing enduring about rates of inflation.

Second, there is no great reason to think that high stock prices are the engine rather than the reflection of growth. The

Nasdaq Composite Index

bubble was spectacular, but how much of the economy did it affect? To some extent tech-stock options have been driving consumption among the dot-com elite, and stock wealth may be pushing up housing prices. But housing supply is elastic; surely higher prices will raise construction costs quite soon. And since the Nasdaq has already crashed, that phenomenon hardly requires further policy action now.

Third, the idea that recent consumer price increases were fueled by demand rather than supply factors is doubtful. Was the 50% run-up in gasoline prices that started this blip driven by consumers -- or by producers? This, you can check: which went up first, the pump price of gas or the delivery price of a barrel of crude? And have wages jumped ahead of price increases as the "too-tight labor market" argument requires? Actually, they have not: nominal wages have

not

accelerated and

real

wages have actually fallen in the past few months.

Krugman misses another key point about demand. If a surge in demand did occur -- over and beyond the Y2K computer investments -- what caused it? Surprise, surprise: when people think interest rates are about to go up, they refinance their mortgages and buy new houses and cars to beat the increase. Anyone who hasn't figured out that the

Fed

will be raising interest rates all summer doesn't have a pulse, nowadays. (Disagree? Shoot me an

email!)

That being so, the seeds of destruction have already been sown. Investment moved forward from the second or third quarters will not be made again later on. Ergo; the economy will slow on its own. If that is so, what purpose would a further rise in interest rates now actually serve?

Krugman's argument for higher unemployment could benefit from a review of history. When, exactly, did inflation accelerate under the influence of low unemployment? The answer: 1966 to 1968, when the country was at war (among other things). This was the only time. In the 1970s inflation hit in response to oil shocks at

higher

rates of unemployment than now. You can look these numbers up, if you want. Low unemployment does not cause inflation

per se

.

It is true, of course, that a dose of unemployment would knock inflation down. But at what cost? Suppose we take Krugman's goal of 5% unemployment, reached over a year. Could we get there by cutting just one point off the growth rate? Not likely. In 1989-91, for instance, a drop of 2.2 percentage points in the growth rate pushed up unemployment by only 0.3 percentage points in the first year. There was worse to come, of course.

Dr. Krugman's higher-unemployment Rx can turn into a recession very quickly. And here is another fact: No recession, once started, has ever stopped with a mere 1.1% increase in the unemployment rate.

I have nothing against Paul Krugman. We have a very old acquaintance; my ex-wife introduced him to his ex-wife. He may be right that price increases will continue, for a time. Suppose he is. Does Paul Krugman really favor kicking the economy in the teeth over a bump in oil prices and its ripple effects?

Do you?

James K. Galbraith is author of Created Unequal: The Crisis in American Pay (Free Press, 1998) and director of the University of Texas

Inequality Project. A professor at the University of Texas at Austin and senior scholar at the Levy Economics Institute, he worked for many years on the staff of the House Banking Committee, where he conducted oversight of the Federal Reserve. He welcomes your feedback at

Galbraith@mail.utexas.edu.