Opinion
Income Gap Studies Based on Flawed Statistics
One of the most pernicious impacts of academia on society is the vast advocacy statistics industry that generates phony statistics in support of ideologically based claims that American society is a racist, sexist, selfish, unfair or otherwise bigoted. The standard statistical technique is to omit explanatory variables in the study of some racial or sexual difference, then attribute the unexplained residual to race or sex.

A ploy of this kind won headlines across the country this summer from a complicit press. The study appeared in the August issue of American Sociological Review. Authors Eric Grodsky and Devah Pager found that income disparities between blacks and whites are larger in the professions than in blue collar work, and that these gaps remains after controlling for level of education. Their candidate explanations are "social segregation" by race and racial discrimination.

They are probably right, but not in the way that they think. Consider the inevitable effects of that form of racial discrimination called "affirmative action." It is a fact that people admitted to universities under affirmative action tend to graduate towards the bottom of their class. It is also a fact that people towards the bottom of their class tend to make less money than those at the top of their class. The inevitable result is an "income gap" between those favored by affirmative action and others with the same formal level of education.

The higher one climbs the academic ladder, the more powerful this affirmative action effect becomes. Professional schools essentially carry forth and compound the bias of affirmative action, taking students who have already been advanced ahead of merit and advancing them further, using whatever degree of racial preference is required to do it. Law schools, for instance, are known for pursuing affirmative action with almost religious zeal. The expected result is exactly what was found: a widening black-white income gap as one moves from blue collar to white collar to professional, and this will occur even if the only racial discrimination is against whites.

Technically, the omitted variable here is class rank. Another omitted variable is school rank. In the U.C. system, for instance, blacks and Hispanics mostly attend the second tier universities, whose graduates on average won't make as much money as graduates from Berkeley and UCLA. Other relevant omitted variables would be anything that provides a direct measure of job effectiveness or relevant ability.

Grodsky and Pager cite one such study that tested whether wage gaps can be explained by linguistic test scores. They decide the linguistic variable is irrelevant because in that study "a substantial portion of the occupational variation in black-white earnings inequality remains unexplained." Of course, because a host of other variables, like years of job experience, will also determine wages. The relevant question isn't whether an omitted variable explains everything, but whether it adds explanatory power, in which case it might explain some or all of what remains unexplained in the current study.

In sum, the claims of a wage gap that can't be accounted for by differences in merit is a fabrication, derived by faulty reasoning and by turning a blind eye to obvious explanations. Most disingenuous is the failure to account the role of affirmative action, because the obvious if unstated goal of these advocacy studies is to make the case that affirmative action is necessary. To rely on the assumption that all graduates have equal merit, while having a passionate interest in systems that divorce admission from merit, is more than oversight. It demonstrates a positive refusal to think straight.

The press did its part as well, receiving the study with a credulousness that is reserved for the politically correct. When a study that includes actual measures of job performance is eventually done, and the wage gap mostly or completely disappears, it won't be news. Such corrections never are. Like our dishonest academicians, the press thinks it is doing a great social good by depicting black people as victims. They are actually doing great harm.

In Losing the Race, Berkeley linguist John McWhorter argues persuasively that African Americans are held back far more by the cults of victimology and separatism than by any residual discrimination. Looking for ways to claim victimization breeds separatism from mainstream society and from the norms that lead to success.

Possible support for this view can be found in a 1995 study, conducted by the United States Office of Personnel Management, which found that, between black and white federal employees with similar performance reviews, blacks were far more likely to be fired for misconduct, absence without leave, and other causes. A likely explanation is McWhorter's thesis: that a widespread tendency for blacks to interpret slights and setbacks as racially motivated leads to conflict instead of resolution.

When the press got wind of the '95 study on firing it was worse than credulous. Failing to note the study's distinction between "performance" (past performance review ratings) and "conduct" (the reason for firing), the press charged that, since firings were not for performance, they must be due to racial discrimination. No corrections were ever published.

If one is looking for ways to see discrimination there will always be ways to do it, whether as a racially sensitive individual, as a biased researcher, or as a crusading member of the press. All of these efforts feed misperceptions of racism which, by fostering black alienation and separatism, create racial disparity.

When academics like Grodsky and Pager concoct ways to misinterpret the world as racist, they are doing what Shelby Steele condemned: identifying black interests with claims of victimization which can then support claims for redress. But that isn't how success works in this country. Society rewards people for productivity, not claims of victim-hood.

Productivity means creating value for others. Alienation--feeling cheated, feeling owed--undermines the regard for others that allows people to be motivated to create value for others. Promoting alienation through trumped up charges of racism is an excellent ways to keep blacks out and down.

Academics are generally smart enough to know when methods and claims are intellectually dishonest. What is missing is basic moral fiber: a determination to trust in truth, to follow reason and evidence rather than twist them to fit one's ideological preconceptions.

Every failure to follow honest reason and evidence is a turn away from reality and into a fantasy world where one knows not what one does, wreaking any amount of real damage in pursuit of purely imagined value. In contrast, truth conducts value the way copper conducts electricity. Following truth is the only way to do good in the world.

Right now the demagogues have the upper hand. It is a fact of life that the dishonest reasoners control whole disciplines in academia. The America hating left wing trash of the 60's are right now choosing the next generation of tenured professors. How can they be beaten?

In a truth-loving environment, honest reason trumps dishonest reason every time. If students and administrators and other members of the academic community all embrace their moral duty to create that truth-loving environment then the demagogues will always lose. Like vampires, they can't survive the light, so turn yours on.

Alec Rawls' writings on most subjects can be found at www.rawls.org

Page last modified on Thursday, 02-Mar-2006 00:12:51 MST.