THE DRUMBEAT on raising educational standards has been one of this administration's better contributions to public policy. In recent years, more school districts have bucked inertia and moved toward addressing persistent achievement gaps between white and minority students and between rich and poor: They have adopted tough tests that, because they have real consequences for students' prospects, force underperforming students and teachers alike to make extraordinary efforts to do better. Yet earlier this summer, the Department of Education took a step that could undercut that useful pressure.
The department's Civil Rights Office circulated draft legal guidelines on "nondiscrimination in high-stakes testing." The draft warned schools that educational testing of this kind may constitute a civil rights violation if it denies a benefit (such as promotion or graduation) disproportionately to female or minority students. A test that exposes an achievement gap between black and white students, in other words, could be open to legal challenge on that basis alone.
The appearance of the draft caused a predictable outcry. Many feared it could speed challenges to admissions tests such as the SAT, encourage lawsuits and even discourage schools from using standardized testing. The department protested that the document -- now undergoing revision -- breaks no new legal ground and was intended merely to help schools keep themselves out of a danger zone as they adopt tests with real consequences. A school or system may go on using a standardized test that shows lower performance on average by girls or minorities -- the draft paper reassures -- if the school can show the test is "educationally valid," and if an outside party cannot prove that another type of measurement or test would meet the educational purposes of the first while having less of a "disparate impact."
This may not be so much reassuring as intimidating to a school contemplating the prospect of a challenge. A revised guide could, and should, offer much more on what testing practices are not liable to such challenges. But the draft's approach also misses the extent to which the wider battle over race and education has moved past the question of whether achievement gaps are an artifact of racist testing. The forced dismantling of college affirmative action plans in California and Texas -- plans that sought to modify just such "disparate impact" by using factors other than grades and test scores in college admissions -- has focused attention on the problem that produced the disparities -- that of unequal K-12 preparation. It can be fixed only by long-running efforts to make poor schools more rigorous.
The civil rights issue here is the need to ensure that minority kids in poor schools get the help they need to pass the tests. They need an equal chance to learn and be taught the material, time to prepare for "high-stakes" tests and, in some places, the chance to take them more than once. What they don't need is quicker recourse to legal action that takes away the pressure to improve.