Time to Save Federal Education Data
How can anyone trust the numbers if there is opportunity for politicians to massage the data?
The oldest and most solemn responsibility of the federal government in the field of education, dating back to 1867, is to gather and report "statistics and information showing the condition and progress of education in the United States." Of the many activities housed within the U.S. Department of Education, none is more important than telling the American public the facts about the condition of education.
Now, however, that solemn responsibility is in danger as a result of a bill recently passed by the House of Representatives. ("Research Bill Clears House Without Fuss," May 8, 2002, and "Senate May Vote on Overhaul of OERI Before Fall Elections," this issue.) If HR 3801 becomes law, both the National Center for Education Statistics—the department's data-gathering agency—and the National Assessment Governing Board— the body that reports on student achievement—will see their independence compromised and become subject to political control. This would be a terrible shame, because these functions cannot be credible unless they are well- insulated from such control.
While reauthorizing the Education Department's office of educational research and improvement, or OERI, the House inserted some last-minute provisions that endanger both of these vital activities. Under current law, the NCES is headed by a commissioner who is nominated by the president and confirmed by the Senate to a five-year, renewable term. This post is supposed to be filled by an expert in education statistics with sufficient stature to protect the agency's data from outside pressures and political interests, including those of the administration that he serves and the Congress that appropriates his funds.
Under the House bill, the OERI would be converted into a new "Academy of Education Sciences" headed by a director (in place of the current assistant secretary). This director would henceforth appoint the commissioner of education statistics, who would serve at the director's pleasure. This change in the appointment of the commissioner would not only reduce the position's prestige, making it harder to recruit an outstanding person, but also would remove the political independence that has helped to protect past commissioners from those who appointed them.
The changes envisioned for the National Assessment Governing Board, known as NAGB, are no less ominous. This 24-member, bipartisan board (including governors, teachers, school board members, parents, and business leaders) supervises the National Assessment of Educational Progress, the tests that monitor how American students are doing in reading, math, science, history, and other subjects. NAGB has a measure of independence—not enough, in our view. It now has the authority to set most policies for NAEP and to release its own reports. Much of its freedom is not statutory, however, but a revocable "delegation of authority" from the U.S. secretary of education.
Under the House bill, the new Academy of Education Sciences director would supersede NAGB in key areas, including the authority to publish reports and conduct peer reviews of NAEP studies. In addition, the bill requires that reports from the NCES and NAGB be sent to the secretary of education's office at least a month in advance of their release. This means that the White House—or others to whom these data may leak—would have ample opportunity to fiddle with or delay any awkward findings. (Picture a report on 8th grade math and reading, due for release in an election year, that shows marked declines or gains in states of great importance to one party or the other.)
These two small agencies assume particular salience because of the "No Child Left Behind" Act of 2001, which puts a premium on student achievement. States expect to be measured by how they perform on the National Assessment of Educational Progress. But how can anyone trust the numbers if there is opportunity for politicians, now and in the future, to massage the data or determine the timing and manner of their release?
Surely the nation's premier education statistics agency must be kept free of any taint of partisan meddling or outside manipulation. Washington has many statistical agencies—the Census Bureau, the Bureau of Labor Statistics, and others—and various arrangements by which they're insulated from such interference. This insulation sometimes frays; thus, the periodic fuss over whether the U.S. Census will be "adjusted." But, for the most part, these delicate agencies have been trusted to "speak the truth," even when that's inconvenient for Cabinet officers, presidents, and influential members of Congress—not to mention the swarming lobbyists whose clients' interests are affected by these numbers.
These agencies take pains to shield their data from such influences. At the U.S. Department of Labor, for example, the secretary does not even see the monthly unemployment rate until the morning it's released, nor does the White House or Capitol Hill or Wall Street. Journalists are briefed behind closed doors with no access to their cellphones until the moment of its official release.
It is no less important to preserve the integrity of education data. But that's a sizable challenge. The last full-time NCES commissioner, Pascal D. Forgione Jr., lost his shot at a second term because he objected to then-Vice President Al Gore's taking over a statistical briefing that would advance his own presidential prospects. (A career civil-service statistician has since filled the post on an "acting" basis, and Mr. Forgione is now the superintendent of schools in Austin, Texas.)
If created, the proposed Academy of Education Sciences would probably be led by Grover J. "Russ" Whitehurst, the incumbent assistant secretary for OERI. He's a fine researcher and an honorable man, who would no doubt strive to safeguard the integrity of federal education data. But what about his successor? And the one after that? Tomorrow, the academy might be directed by an anti-testing zealot— the education world contains plenty—or one who is eager to please the party in power or appropriators on Capitol Hill.
The House bill would entrust America's education statistics and assessment results to such an uncertain fate. Instead of thickening their insulation—for example, by fully empowering the National Assessment Governing Board to run the national assessment program—it would make them almost entirely dependent upon one political appointee occupying a single office. It's OK to ask that person to shoulder the thankless task of shaping up the research and evaluation side of the OERI's mandate; there, almost anything would be an improvement, and HR 3801 would make worthwhile and promising changes. But the statistics and assessment programs are working decently. They have earned respect for the integrity of their reports to the American public.
The way to improve these programs is to ensure that both remain utterly credible throughout the land, even to those who may not like what the numbers show and to those who may disagree with the president, the secretary of education, or the director of the Academy of Education Sciences. These programs are our nation's truth-telling activities. Ensuring their credibility is Washington's single most basic responsibility in the field of education. The Senate, which is now turning to this matter, needs to clean up the House's well-intended mess.
Diane Ravitch, a senior fellow at the Brookings Institution, was the assistant U.S. secretary of education for research and improvement from 1991 until 1993, during the first Bush administration. Chester E. Finn Jr., a senior fellow at the Manhattan Institute, served in that role from 1985 until 1988, under President Reagan.
Vol. 21, Issue 42, Page 76