Test Guidelines Will Coerce Colleges and Cheat Students By ROGER CLEGG and LENORE OSTROWSKY The U.S. Department of Education's Office for Civil Rights is circulating draft guidelines -- which bureaucrats euphemistically call a "resource guide" -- among college and university admissions offices, ALSO SEE: Colloquy: Join a debate on the issues raised in this article putting them on notice that relying on standardized tests like the SAT and ACT could land them in legal jeopardy. The officials at the O.C.R. claim that, if certain racial or ethnic groups don't score as well as other groups on a standardized test, you can presume that a college that relies on the test in its admissions process has engaged in illegal discrimination. Never mind that the test itself is neutral on its face. Nor that the college had no intention of keeping anyone out because of skin color, ethnicity, or gender. Welcome to the world of what civil-rights wonks call "disparate-impact theory." In this world, the federal government may find a college or university guilty of civil-rights violations if there are racial, ethnic, or gender imbalances in the student group that it selects for admission, and if some bureaucrat determines that the selection device leading to the imbalances is not "educationally necessary." If the O.C.R. guidelines are adopted, any college that runs afoul of the new policy could lose federal funds and face a heightened risk of lawsuits. The guidelines, called "Nondiscrimination in High-Stakes Testing," are being hailed by pro-preference civil-rights activists, who see them as the latest weapon in the fight to insure that quotas for their preferred minority groups can be maintained in higher education. In fact, as Congress held hearings on the guidelines in June, those groups began mounting a campaign to pressure college and university leaders to support the guidelines. They are afraid that educators will criticize the guidelines for their naked animosity toward intellectual achievement and individual merit. Indeed, the guidelines would encourage the sort of discriminatory admissions practices that Education Department officials are supposed to remedy. They also would open the door for the federal government to arbitrate -- quite inappropriately -- admissions policies and processes throughout the nation. And the guidelines would inhibit the use of what the O.C.R. surely knows to be the single most reliable predictor of academic success for students from all racial, ethnic, social, and economic backgrounds. It is no coincidence that the civil-rights office is circulating such guidance now and not decades ago, when the underlying statutes, Title VI of the Civil Rights Act of 1964 and Title IX of the Education Amendments of 1972, were passed. The O.C.R. knows that anti-affirmative-action court decisions and voter referenda are pressuring colleges and universities to abandon the use of racial and ethnic preferences in their admissions decisions. It also knows that, if standardized tests are a significant part of those decisions, the enrollments of some minority groups may well drop. So the time has come for the O.C.R., in desperation, to tell colleges and universities that they need to play down the importance of the tests. It also is typical for the O.C.R. to coerce people through subregulatory guidance -- in other words, without giving formal notice and opportunity for comment to the general public. The office instead proceeds by threats and intimidation tactics that it administers informally and, where possible, secretly. But this time, college officials were sufficiently perturbed by the guidance to leak it to the press. The civil-rights office claims not to understand what all the fuss is about, asserting that its guide is just a restatement of settled law. But the oh-so-politically- correct educational establishment would hardly have complained had this been the case. The fact is that the guidance is extremely strict and goes far beyond settled law, even if we accept the disparate-impact approach. And it remains to be seen whether the O.C.R. -- now under pressure from both colleges and Congress to revise the guidelines -- will actually modify them in any meaningful way. If we are to avoid what is a certain blow to education, we must break open the world of disparate-impact theory and examine both halves of its rotten core. The first half is that, under the guise of getting rid of unintended imbalances in student enrollments, the disparate-impact theory itself requires deliberate discrimination. The theory -- borrowed from employment-discrimination law -- requires educators to choose selection devices with an eye to the racial, ethnic, and gender bottom line that the devices will create. Such a practice is condemned as discriminatory under any other circumstances -- and rightly so. The other half of disparate impact's rotten core is this: Federal bureaucrats will decide the best way for colleges and universities to make admissions decisions. Any selection criteria, and any combination of them, will "cause or contribute" -- the guidelines' words -- a disparate impact on some group or another. In addition, the guidelines state that an institution must meet the difficult standard of proving educational necessity, and demonstrate that alternative selection criteria -- which would "substantially" serve its interest with a lesser disparate impact -- do not exist. Obviously, the guidelines point a litigation revolver at a college's head. Which leaves the following question for those institutions that want to avoid litigation: What process will not be challenged by the O.C.R.? Only the bureaucrats can determine that. More is going on in this controversy than just disagreement over a particular set of draft guidelines. For years, the education and civil-rights establishments -- both of which are left-of-center -- have each been happy with the system of racial and ethnic double standards that has increasingly been used to determine college admissions. To put it simply: One standard has existed for black and Hispanic students, and another for white and Asian students. But now, the educational and civil-rights establishments find themselves at loggerheads because the law is increasingly inhospitable to double standards -- and they can't agree on a new single standard. The pro-preference civil-rights groups advocate a low standard because that is what's needed to guarantee that colleges and universities will continue to meet racial and ethnic "targets." But the leaders at selective institutions can't abide that standard, since their livelihoods hinge on their claim to be, well, selective. In fact, 82 per cent of four-year undergraduate institutions require admissions tests for one major reason: Without them, the college or university would probably enroll a high number of students with low academic ability, and the institution and its reputation would suffer. Remove that fear, and the SAT and ACT would be long gone. Not surprisingly, the O.C.R. has weighed in on behalf of the civil-rights establishment. One cannot help but indulge for a moment in some Schadenfreude at the academic elite's misfortune. But just for a moment, because too much is at stake for the students whom our colleges and universities are supposed to be serving. Liberal civil-rights groups are attacking testing for a whole series of reasons -- not only when it has a disparate impact on racial minorities and women, but also whenever it fails to make a reasonable accommodation for the mentally impaired. The latter challenges rely on the increasingly controversial Americans With Disabilities Act. As Anthony Carnevale, vice-president for public leadership at the Educational Testing Service, which designs the SAT, recently told a reporter: "In a sense, tests are on trial." Yet how a student does on standardized tests such as the SAT and ACT remains the single best means of assessing future aptitude for learning. The SAT was introduced in the 1920s to help colleges and universities identify bright students with high academic skills and abilities who might otherwise be overlooked because of their poor economic and educational backgrounds. It and the ACT measure reasoning ability and bear some similarity to I.Q. tests, from which they were drawn. They have been remarkably effective in predicting future college success -- regardless of the wide variations in students' backgrounds and the rigors of their previous academic training. Indeed, the tests are also now widely used to identify gifted children under the age of 13. Because they are valid cognitive tests, the SAT and ACT measure something tangible and real: the ability to master the complex material taught at any institution of higher education that has traditional academic standards. In the final analysis, we need to distinguish between more and less qualified students for their own benefit, not that of the educational establishment. Nearly as many individuals drop out of colleges and universities as complete their degrees. At some less-selective institutions, fewer than 30 per cent of students graduate within five years. If college-admission tests were not valid, they would not have predicted -- as they unerringly have -- the way dropout rates have increased as average cognitive abilities among students at less-selective colleges have decreased. Pretending that differences in ability, irrespective of race and ethnicity, do not exist may suit the posturing of some ardent civil-rights advocates. But, in the end, it will help no one -- and will result in even more dropouts. The fact, revealed by standardized tests, is that students vary widely in their academic abilities. That stubborn datum has provoked increasingly illogical educational "solutions." Colleges and students deserve better. The O.C.R. should withdraw its guidelines. Congress should pass legislation banning the use of the disparate-impact theory under Title VI and Title IX. And teachers and educational administrators should be allowed to do their jobs: bringing out the most in each student, to the best of every student's ability. Roger Clegg is general counsel of the Center for Equal Opportunity, in Washington. Lenore Ostrowsky is a Washington lawyer whose article on dropouts and standardized testing appears in the spring 1999 issue of Academic Questions. http://chronicle.com Section: Opinion & Arts Page: B8