A year after the nasty secret got out of how race-norming works on the nation’s most widely used job test, the establishment news herd suddenly discovered the story. There were spots on NBC Nightly News and the Today Show, a front-page story in the Washington Post, an editorial in the New York Times, and a piece by me in these very pages. It was front-page copy throughout the month of May 1991. House Democrats quickly inserted language into the then-proposed civil rights bill purporting to ban this kind of manipulation of test results (although actually doing no such thing). Whereupon a silence then descended upon the Big Media. For all most Americans knew, race-norming was a quirky practice that, once discovered, was quickly terminated.
Pssst! Race-norming just continues to roll along with no restraint in sight. Having published on July 24, 1990, an official notice of intent to suspend its race-normed General Aptitude Test Battery within ninety days, the U.S. Labor Department fifteen months later seemed to have placed the GATE in a state of permanently proposed suspension—a limbo allowing state employment services and thousands of major corporations to continue to hand out phony scores to millions of hapless job seekers. And even with the recently enacted civil rights bill, statements of legislative intent that congressional staffs have salted away in the fine print of committee documents make it highly likely that regulators and judges will find race-norming to be an imperative for hiring by the racial numbers. The legislative history states that if a job test has “disparate impact” on protected minorities (that is, if passing rates are unequal), it will fail to meet the “business necessity” criterion of the civil rights bill. In such circumstances, employers must either trash the test or resort to race-norming (or similar subterfuge). Actually, they would rely on Big Brother to continue to do the dirty work for them, because the bogus ban places no restraint on government employment services.
Paradoxically, President Bush has railed against the quotas the civil rights bill would foster, but he has ignored this smoking gun of quota-mongering in his own bureaucracy. Perhaps he is hoarding this issue to spring during the 1992 presidential campaign on the assumption that voters, not without reason, will blame the party of Ted Kennedy for the scam.
Why is race-norming the subject of such little reporting or critical examination? The reason may be that so many large newsrooms are also becoming the products of strict quota hiring. Or perhaps the self-censoring influence of political correctness has captured news reporting. Whatever the reason, it is clear that not all dirty secrets about employment discrimination are equally titillating to the gods of the liberal media or to the Beltway cognoscenti; nor do those presumed opinion leaders deem all such revelations worthy of exposure to the disinfecting sunshine of full disclosure.
Meanwhile, segregated scoring of job tests according to an illiberal presumption of settled racial inferiority is not an unsubstantiated secret. The GATB race-norming also happens to be about ten years old; it traces to the final days of the Carter administration when the focus of affirmative action was changed from equal opportunity to equal results. Moreover, race-norming does not affect just a few individuals. Indeed, Scripps Howard political reporter Peter Brown reports in his book Minority Party that roughly sixteen million unsuspecting Americans received race-normed GATB scores during the 1980’s. An even larger number were handed racially rigged marks on tests administered by corporate Carterites.
The way the scam works sounds like something out of South Africa’s playbook of separate development: test takers are scored against norm groups composed solely of members of their own racial or ethnic designations, the purpose being to compensate for the tendency of blacks and Hispanics as a whole to score lower than Asians and whites on such tests. This paternalistic scheme assumes that individual members of designated victim groups cannot hack it on their own, and by using norming data that are 20 years old it ignores the gradual narrowing of the black-white performance gap on standardized tests. The effect of what Department of Labor bureaucrats euphemistically call “within-group” scoring is to skew the final scores wildly and preferentially. For example, if a white, an Asian, a black, and a Hispanic all apply for a mechanic’s job, and all make a real score of 300, the white and the Asian will each receive an adjusted score of 44, while the Hispanic will get a 67 and the black will go to the head of the line with an 83. Those scores reflect the test taker’s percentile ranking within his own racial or ethnic group, but the marks are reported as a set of grades on a college exam would be. And those are the scores, never the real marks, that the employment services report to employees and to unsuspecting test takers, many of whom are unemployed. This is a fraud, and a racist fraud at that, but evidently nothing more than unpleasant trifles to affirmative action fans in the news media and government.
The secret charts used to convert raw scores to within-group percentiles were leaked to the Richmond Times-Dispatch in May 1990. At the time, the Department of Labor had had its race-normed version of the GATB under seemingly permanent review since the Reagan Justice Department had lodged a reverse discrimination complaint in 1986. In 1989, the DOL obtained a friendly analysis from a politically attuned National Academy of Sciences panel, which concocted a pseudoscientific rationale for race-norming. But the outcry that erupted in Virginia upon publication of the conversion charts reached the DOL’s ears, following which then-Labor Secretary Elizabeth Dole published the July 24, 1990, proposal in the Federal Register to suspend the GATB for a two-year retooling.
As is customary in federalia, the DOL solicited “public comment” to this proposal. From my efforts to review the responses, I have learned two lessons: (1) federal bureaucrats do not really consider these comments to be “public,” and (2) bureaucracies have no more scruples about manipulating public comments to their own advantage than they do about falsifying test results.
The DOL’s Employment Service waited six weeks to reply to my March 8, 1991, Freedom of Information Act request for the public comments (even though federal law requires a response within ten days). U.S. Employment Service Director Robert A. Schaerfl then refused to comply with my request because, he said, a “pre-decisional” release “would be detrimental to our decision-making process.” He cited a section of federal law “intended essentially to protect the full and frank exchange, in writing, of ideas, views, and opinions necessary for the effective functioning of the government and the making of informed decisions by its officers.” In other words, those public comments are the property of your incredibly efficient government, so beat it.
After the Times-Dispatch twice appealed the FOIA rejection, Schaerfl relented on June 18 (“pre-decisional” having taken on the look of a permanent condition by then). But now there was a catch: the DOL wanted $465 for copies of these precious documents. That is not a chunk of money that a newspaper gladly parts with during an advertising crunch, as the recession-proof bureaucracy surely understands. My newspaper filed yet another appeal, which cited explicit provisions in the FOIA for waiving or reducing such fees when disclosure would contribute to public understanding of government activities. Schaerfl’s response was a beauty: “Because this system has been publicized extensively by the news media including your newspaper” there is nothing more to be learned. So, again, beat it, pest. Finally, my bosses, bless ’em, forked over the money, and a 25-pound box of public documents arrived at my office seven months after the information had been requested.
Schaerfl was wrong. The 1,700 responses from business, labor organizations, and other parties teach volumes about how government operates. The first point that becomes obvious from a reading of the record is that the “public comments” are hardly vox populi. Anxious to preserve their seam, job services in several states produced letters opposing the suspension of race-norming that corporate personnel (or “human resource”) officers had only to sign. One union apprenticeship program in Rochester, New York, cranked out 83 identical letters on its own. Two dozen Oklahoma businesses separately signed a letter making this mangled statement: “The decision to suspend its use will adversely impact any employer’s ability to hire the best employees they can.” Hundreds of letters were the obvious product of internal campaigns to make the testing scam seem indispensable, and yet they all counted in the DOL’s tabulations of “public comment.”
There were a few dissenting voices. For example, the Anti-Defamation League of B’nai B’rith wrote: “Segregating the GATB test scores merely extends racial and ethnic divisions without targeting or remedying the source of the discrimination, and without providing any meaningful or longterm assistance to minority applicants.” And Evan J. Kemp Jr., Clarence Thomas’ successor at the EEOC, checked in with strong criticism of race-norming. Such responses were buried, however, in an avalanche of solicited letters. “Apparently the guys and gals at the Job Service would like to save their paper-shuffling jobs,” a manager of a small North Dakota business commented sardonically.
Not many corporate officials were brave enough to sass the federal establishment. Indeed, the response to race-norming says something about how government is corrupting the soul of corporate America and diluting its commitment to merit hiring and quality of product. The DOL did not give its clients the option of continuing the GATB (a test that is considered a rather reliable predictor of job success) and scoring all test takers the same way. The choice was between no testing or a race-normed test.
To be sure, the quotacrats have made it exceedingly difficult for major employers to resist race-norming. The Supreme Court’s 1971 Criggs decision turned up the heat on businesses by buying into the civil rights lobby’s argument that it is only necessary to show that a practice has “adverse impact” on minorities, even absent any discriminatory intent, to make a prima facie case of employment discrimination. The 1989 Wards Cove decision restored to employees some discretion to use merit-based hiring practices and shifted to plaintiffs the burden of proving discrimination had actually occurred. But that is one of the Reagan Court decisions congressional liberals sought to reverse with a new civil rights bill.
Then there is all that the bureaucrats have done to kill merit hiring. During the Nixon administration, the Office of Federal Contract Compliance announced that employee selection procedure must result in proportional representation for all protected minority groups; if not, a company had to validate each test it used. Since that can cost anywhere from fifty to two hundred thousand dollars or more per test without providing any warranty against costly discrimination litigation, many companies, “not wishing to lose the strong economic advantages of testing, found the sure bet of race-norming the more practical alternative,” Eliot Long, vice-president of Wonderlie Personnel Test, Inc., has noted. Wonderlie offers businesses a GATB-like “ethnic conversion table” with its widely used personnel test. The Illinois-based testing company opposes any move to prohibit race-norming.
Pragmatism is one thing, but corporate personnel offices also have adopted the doubletalk of the bureaucracy. Panning any use of test results that have not been race-normed, an official of Campbell Soup Company of Omaha opposed suspension of race-norming because he did not believe the Labor Department “should dictate social policy at the expense of workplace productivity.” Huh? A packaging company hailed the test-rigging for providing “a more representative sampling with regard to race.” And the manager of a Ford assembly plant near Atlanta marveled because race-norming provided him applicant pool test score percentages exactly as follows: white male, 31; black male, 42; white female, 5; black female, 24; Asian, 9; and Hispanic, 2. Egads. That adds up to 113 percent, and the Asians and Hispanics appear to have lost their sexual identities.
The larger question race-norming raises is whether any shred of honesty remains in Big Government, Big Business, or Big Media.