SUCCESS is dangerous, as American economists discovered in the 1970s. The spectacular performance of modern economics in the great boom of the early 1960s left its principles entrenched in habits of mind and political practice long after the boom had faded.

As governments here and in Europe kept anxiously applying the proven rules, they produced, to their rising dismay, more inflation than real growth -- and unemployment soared.

In his latest book, Lester Thurow undertakes to explain what went wrong, not in the engineering of economic policy, but at the level of fundamental theory as the professionals know and use it. He took his PhD from Harvard in 1964, just as the magnitude of the boom was becoming apparent. Now a professor of economics and management at MIT, he has been an active participant in the painful process of academic post-audit.

Although he writes with his customary vigor and clarity, this book -- Dangerous Currents: the State of Economics -- remains in the end incomplete.It gives the impression of being fragments of a larger work that the author has not yet fully thought through. While Thurow is illuminating on several important defects in recent theory, it's never quite clear where that takes us. He concludes that the world is a messy place full of variables unknown to econometric equations. That is a thought to which many readers will have preceded him.

The value of the book is in its explanations, in particular, of two crucial lapses in the structure of economic thinking. One is the lack of any connection between macroeconomics, on which national policy is based, and the microeconomics of the various industries and markets. In the 1950s and 1960s economists embraced the view that what really counts is the mathematical relationships among large abstract aggregates -- income, consumer demand, employment and the rest. But microeconomics is based on sharply different assumptions, and the two are difficult to reconcile.

Most economists resolved the conflict by losing interest in microeconomics, relegating such subjects as the automobile industry and the grain trade to the business schools. Instead, they turned to their computers and increasingly elegant statistical models.

That led to the second lapse, the overreliance on econometric models -- that is, families of equations that were supposed to reflect the operation of the economy as a whole without having to pay much attention to any actual part of it. With time, Thurow says, modeling "took on a life of its own."

"Where does all of this leave us? Economics is in a state of turmoil. The economics of the textbooks and of the graduate schools not only still teach [the] price-auction model but is moving toward narrower and narrower interpretations. The mathematical sophistication intensifies as an understanding of the real world diminishes."

For example, what about the question of unemployment? When people lose their jobs, why don't they get new ones by reducing the wages they ask -- in effect, putting their labor up for auction? But that doesn't happen often. "The downward rigidity of wages is the key problem in labor economics," Thurow observes, and he has some ingenious comments to offer. But like most questions of economic theory, this one is loaded with political implications. In this book Thurow deals with them only occasionally, and in passing.

Is it possible that unemployed people don't reduce their wage demands because they have learned that the federal government can be relied on to push up the demand for labor until they are hired back at their accustomed pay? Do they conclude that it's more profitable to wait out the recession than to take permanent wage cuts in a temporary slump? That's currently a point of great controversy and the evidence appears to be inconclusive. But if there's anything to that suggestion, it leads to a paradox: that successful economic policy creates public responses that make the economy more rigid -- and ultimately causes successful policy to fail.

One central question for economists is the nature of their subject -- not a machine that responds predictably, but a crowd of people who will try to anticipate what's going to happen next. They will act on the basis of their experience and, because people's experience was different in the late 1970s from the early 1960s, their reactions to policy will also differ.

Another omission: Like much American writing on these matters, Thurow gives only brief and passing notice to the international economy. Many of the errors of the 1970s clearly spring from a systematic underestimation of the force on the domestic American economy of events abroad. It's not that different principles apply to the world economy; it seems to be, rather, that most American economists still aren't in the habit of paying much attention to them.

When the authoritative work on the economics of the 1960s and 1970s finally appears, the systematic neglect of international realities will, I suspect, be central to the explanation of what went wrong. But that will be a deeper book -- for which this one may turn out to be an early and partial draft.