The Strange Death of American Liberalism. (Political Booknotes: left for dead).

THE STRANGE DEATH OF AMERICAN LIBERALISM

by H.W. Brands Yale University Press, $22.50

IT WOULDN'T BE FAIR TO CALL H.W. Brands's new book a casualty of the September 11 attacks just because it posits an essential and deep-seated American anti-governmentalism that doesn't seem to apply at the moment. One of Brands's main points is that the only purpose for which this country believes a national government is needed is defense. Therefore the only time when government has truly expanded has been during war. So if government is growing now, at the dawn of the war on terrorism, that would confirm Brands's theory, not overturn it.

This is a short book, written, as Brands puts it, more as an "argument" than a history (although he is a historian, at Texas A&M). The argument is a simple one: Americans have never accepted the idea of a big central government. Brands himself, though he writes in a voice of scrupulous neutrality, appears to be an old-fashioned anti-government conservative of the Robert Taft variety, suspicious not only of the Democrats but also of both the moralist and super-hawkish tendencies within the Republican Party. Partly because his basic premise is so forceful, Brands has to spend much of the book explaining why every expansion of government since the American Revolution doesn't disprove it. Therefore, The Strange Death of American Liberalism reads like a long string of caveats, though a consistently interesting one.

The biggest caveat concerns the Cold War, which, if you're willing to accept it as a war, as Brands does, then brings under the capacious umbrella of Brands's theory everything government took on domestically during the fifties, sixties, seventies, and eighties, from Medicare to environmental protection to the War on Poverty to wage and price controls. In each case, Brands produces a nugget of evidence that links the domestic program to the global struggle. John F. Kennedy settled a steel strike because steel production was essential to weapons production, Dwight Eisenhower pushed through federal funding of education in the post-Sputnik National Defense Education Act, and so on. Then, when the Cold War ended, we began inexorably returning to the American state of nature, hostility to government.

Brands's book should work as a useful tonic to liberals who underestimate the difficulty of passing new government programs. But I thought he paid insufficient attention to a couple of structural pro-government forces, other than war. One is the rise of the welfare state all over the Western world during the 20th century. Brands is an implicit American exceptionalist and barely refers to any other country, but industrialism led to a bigger government just about everywhere, by generating societies complex enough to require sophisticated management and rich enough to pay for it. Yes, most countries have scaled back their welfare states somewhat, but the essential phenomenon remains and economic rather than military forces drove it. A second force in government's favor is the power of interest, which figures very small in Brands's account. Interest groups, both from business and labor, consistently lobby to have government grow in ways that help them, and if the groups are powerful they often win. Even more important, government programs that directly help that huge majority of the American public that thinks of itself as middle class don't need a military rationale to be popular. Any politician who called for returning responsibility for Social Security and Medicare to local government would quickly become an ex-politician. And if you think of public education and police protection as completing (with pensions and health care) the Big Four government benefits for the middle class, they are both being gradually federalized, for example through Bill Clinton's bill to put 100,000 cops on the street and George W. Bush's pending, but sure-to-pass, school accountability package.

Nothing would be more wonderful than for us to get the chance to test Brands's theory--because that would require a good long peace.

NICHOLAS LEMANN is the Washington correspondent for The New Yorker.