National Tests: A Good Idea Going Wrong

August 26, 1997

Six months ago I was an enthusiastic supporter of the president’s proposal for national testing. I wrote articles in this paper and elsewhere endorsing the idea of testing fourth-grade reading and eighth-grade mathematics. In March I had the privilege of meeting with the president, the first lady, and the secretary of education to discuss the president’s education agenda. At that time, I urged the president to place the national testing program under the supervision of an independent, bipartisan agency, not the federal Department of Education.

Unfortunately, this did not happen, and now the Clinton administration is well on its way to destroying the credibility of its good idea. Instead of turning the testing program over to the National Assessment Governing Board, which is independent and bipartisan, the administration has kept the test development process under its own tight control. It chose a contractor and selected committees to design the tests who may or may not be bipartisan and who may or may not represent the diverse spectrum of views in the fields of reading and mathematics. A few days ago it awarded a $13 million contract to a consortium of testing companies.

Now the Internet is humming with charges that the national tests will be stacked to favor “whole language” theories of reading and against phonics and to promote “fuzzy math” (where the process of problem-solving, not the right answer, matters most) and against computation. I do not know whether any of these fears are valid, but the administration’s partisan control of the test development processing has inflamed such feverish concerns.

The National Assessment Governing Board (NAGB) should be in charge of the national tests. Appointed by the secretary of education, NAGB is required by statute to be bipartisan and broadly representative; its members include governors, legislators, business leaders, teachers, school board members, public members and testing experts. NAGB was created by Congress 10 years ago to set policy for the National Assessment of Educational Progress (NAEP). This is the federal program that regularly tests national samples of American students in such subjects as reading, mathematics and science. NAEP has earned great credibility because it is supervised by an independent governing board. However, the Clinton administration plans to use the NAEP tests while bypassing NAEP’s governing board.

The issue of who should control the national tests is not just a technical quarrel. The decision to have a national test is of historic proportions; the federal government has never done anything like this before. This is certainly not the kind of initiative that should be undertaken without full public discussion and bipartisan support.

The launch of national testing is precisely the kind of issue that ought to have congressional authorization, yet the administration is pushing ahead without public hearings or explicit authorization by Congress. I believe that a public debate would be healthy and that Congress would ultimately agree to endorse national tests. Every opinion poll shows broad public support for the idea.

To date the administration has refused to seek authorization, apparently fearing that Congress would prohibit the tests. It was my experience, when I worked at the Department of Education, that any new program must be authorized by Congress. No matter how terrific an idea the White House might have, the nation’s Founders did not believe that the executive branch should be able to impose new programs unilaterally.

Spokesmen for the administration claim that they cannot ask NAGB to oversee the national tests because it is not authorized to oversee individual testing of this kind. This is sophistry, however, because the Department of Education certainly has never been authorized to launch a national testing program for individual students. If NAGB is prohibited from managing the new tests because it lacks specific authorization, then surely the department is also prohibited.

Worse, the administration is establishing a dangerous precedent. If the party in power gets to control the national test, to pick the contractor without an open competition, to select the committees that will write the tests and to control the reporting of the results, imagine what will happen when the other party comes to power. The next Republican secretary of education (Lynne Cheney? Phyllis Schlafly?) will revise the national tests, choose congenial experts to write the reading and math tests, and make the changes that seem right to the party in power. And all of this can happen without congressional authorization, because the Clinton administration demonstrated that it was not necessary!

It isn’t surprising that only six states have signed up to participate in the first national tests in 1999, even though the administration has offered to pay the cost for the first time only. So long as the testing program is directly controlled by political appointees, it will not gain the broad bipartisan support that it needs to be credible.

Yes, national testing is still a good idea, because our current patchwork of tests does not provide a common yardstick by which parents and educators can compare results. Some critics say that testing won’t help children learn more, but testing based on high standards will give us valuable information about how we are doing.

If we are to have national tests managed by the federal government, it is important to do it right rather than hurriedly. The administration’s insistence on partisan control will destroy the future of national testing.

The writer, a senior fellow at the Brookings Institution, was assistant secretary of education in the Bush administration.