PREFERENTIAL POLICIES An International Perspective By Thomas Sowell Morrow. 221 pp. $17.95

IN A 1987 book, Compassion Versus Guilt, Thomas Sowell expressed almost contemptuous disdain for those he characterized as "deep thinkers." In Preferential Policies, his most recent book, Sowell practices what he preaches. He ranges across time and the globe, culling illustrations to support an argument based on dubious premises, naive assumptions and crackerbarrel prejudices treated as natural law. The result is a monument to shallowness of thought.

Sowell defines preferential policies as those "which legally mandate that individuals not all be judged by the same criteria or subjected to the same procedures when they originate in groups differentiated by government into preferred and non-preferred groups." His central claim is that five patterns join such policies, apparently in all contexts. First, they tend to persist -- even when initiated explicitly as temporary measures -- and to expand in scope, covering ever more groups and areas. Second, benefits tend to cluster among "already more fortunate" members of designated groups. Third, they tend to intensify group polarization and to produce backlash. Fourth, they generate fraudulent claims. And, fifth, discussion of them tends to focus on their "rationales, mechanics and resource inputs" with little regard to their outcomes. He sets out to demonstrate this argument with examples drawn from postcolonial Nigeria, India, Sri Lanka, Malaysia, South Africa, the United States and elsewhere.

The book is not a comprehensive study, however, and Sowell gives no reason for discussing the particular policies that he does except that usually they seem to conform to the points that he wants to make. Often, though, the relation between his examples and his underlying argument -- that preferential policies undermine their own objectives and cause more problems than they resolve -- is tortured and unconvincing even on his own terms. He acknowledges, for example, almost in passing that Afrikaners and residents of the Indian state of Maharashtra have benefited from policies favoring them. But, he cautions, Afrikaners probably would have gotten better off anyway, and he dismisses Maharashtrians' success by sneering that it reflects the Congress Party's irresponsible disregard for some unspecified "longer-run economic consequences for the country".

Nor is Preferential Policies genuinely comparative. Sowell skips from one case to another like a flat stone ricocheting across the surface of a placid lake. The fact that he discusses a large number of cases across a broad range of societies does not compensate for his superficial, largely anecdotal treatment of them. Indeed, the book's "international perspective" seems almost a ruse for its superficiality; a book-length manuscript less broadly cast would require some of the careful "deep thinking" that Sowell finds so objectionable. Instead, he simply approaches policies favoring the Hausa-Fulani in Nigeria or untouchables in India as proxies for his fundamental target, affirmative action for minorities and women in the United States.

Sowell's strategy of skipping from one context to another provides an escape route from confronting troublesome implications and complexities. For instance, he treats his postcolonial cases as if the meter of history started only at independence. In several contexts he describes certain groups matter-of-factly as having had "historic head starts." But how did they get those head starts? Might some of those groups have been beneficiaries of colonial policies that were at least substantively preferential? In fact, Sowell treats colonialism as part of the natural ecological history of the third-world societies from which he draws. Some groups did well under colonialism because they possessed traits -- and here he comes distressingly close to endorsing hoary Victorian notions of innate racial temperament -- which helped them adapt to the colonial setting or to acquire "Western values." Sowell indicates no sense of colonialism as a system of governance imposed on those societies for the benefit of the colonial powers; as a result, he is not at all moved to consider that privileged status attained under colonial rule is no freer than contemporary set-asides from the taint of political policy.

At its root, Sowell's vision is limited by the two-dimensional ideology of neoclassical economics. He sees markets and individuals (laden with the characteristics of their ethnic groups, to be sure) as constituting a perfectly harmonized, natural world; other institutions, particularly government, are artificial and therefore dangerous. His main standard of justice or propriety is the play of market forces; whatever they produce is by definition natural and proper. But markets are not neutral or natural. They, no less than any other human constructs, exist within ideologically shaded social contexts. Markets do not simply produce winners and losers; winners use public policy to shape market rules to support their interests. Sowell can see politics operating in markets when disadvantaged groups attempt to use public policy for compensatory purposes. When politics reinforces existing privilege, though, he blindly accepts it -- excepting cases, like South Africa, Nazi Germany and the segregated American South, that are too blatant to overlook -- as good common sense, social utility or the immutable product of human nature.

SOWELL skirts this contradiction by his ahistorical and narrow definition of preferential policies. In the American context, for example, one could just as easily define as preferential federal tax, housing and transportation policies which for nearly half a century have favored white middle-class suburbs over nonwhite central cities. His narrowness obscures non-governmental mechanisms for reproducing privilege as well. He exalts SAT scores (despite his opposition to arbitrary licensing requirements in other areas) as exclusive criteria of academic merit; in the process he glosses over both the considerable debate over what and how they might reasonably be held to measure or predict and the context of a class-stratified, two-tiered education system built around preparing upper-status children to perform on the SAT.

Sowell is a true believer and a pure ideologue. All his arguments are instrumentalized to his rightist political agenda, and they are predictably muddled and inconsistent. At one moment he argues that fraud is inevitable in compensatory programs because human nature is driven by "the desire for self-aggrandizement at the expense of others"; at another he asserts that employers themselves are best situated to determine whether their employment practices are unfairly discriminatory. He sanctimoniously excoriates proponents of compensatory policies for not basing their stances on facts; then, preposterously, he proposes as the tough-minded alternative that we focus on the incentives that those policies hypothetically would create. But those incentives are not facts; they are logical projections of likely behavior based on ideal assumptions. Such is the intensity of his faith in market ideology that he naively presumes its abstract logic to be interchangeable with real life.

Preferential Policies has more in common with the ultraright crankiness of H.L. Hunt or Jesse Helms than with any serious intellectual enterprise. Much of Sowell's polemic slides toward sophistry, relying too heavily on debaters' tricks, analogies and shibboleths where evidence and thorough analysis are needed. Even when his points occasionally hit the mark, he extrapolates from them inappropriately. Evidence of fraud is neither unique to Sowell's "preferential" policies, for example, nor is it necessarily invalidating. Think of defense contracting. Calls for nuanced analysis hold no appeal for Sowell, however; the fearless opponent of deep thought tolerates no such squeamish complexity.

Is it fair to characterize Thomas Sowell, the leading black intellectual at the Hoover Institution, as a reactionary crackpot? I leave the reader with the following mot, again from Compassion Versus Guilt: "At one time, people who didn't work were called 'bums'. Today, they have been sanctified as 'the homeless'." Decide for yourself.

Adolph L. Reed Jr. is a professor of political science and a fellow in the Institution for Social and Policy Studies at Yale University.