By Richard P. Phelps
Two weeks ago, The Honorable Lamar Alexander of Tennessee, the chair of the Senate Committee on Health, Education, Labor, and Pensions (HELP) invited three allegedly independent education researchers to discuss possible revisions to the Elementary and Secondary Education Act, currently known as No Child Left Behind. All other testimony came from practicing educators, politicians, or interest group representatives.
The three could have, and should have, broadly represented the research literature relevant to the issues at hand. Instead, they did the opposite. They promoted their own research and that of a tiny group of their friends and, when not ignoring the vast abundance of useful and relevant evidence available, openly declared it to be nonexistent.
For example, by far most independent education research, and most of the research pertinent to the issues the Congress is now considering, has been conducted by psychologists. Yet, research psychologists were nowhere represented in the HELP Committee hearings, and I do not just mean in person. I’ve combed the six pages of reference sources listed in the testimonies. Here’s a breakdown: about 175 of the authors listed are economists; a couple of dozen are political scientists; another couple of dozen are education professors; and at least one sociologist is represented. The number of research psychologists? Zero.
Moreover, most of the economists and political scientists are members of a particular tiny group of researchers that has assumed for itself the role of Republican Party education policy advisors and exclusive spokespersons for education reform. Though “tiny” may exaggerate the group’s size. Former students—and students of former students—of Harvard Political Science professor Paul Peterson comprise a substantial proportion of the group. Others come from just a few other universities, such as Stanford and the Universities of Washington (State), Michigan, and Chicago, and their associated think tanks and research centers, such as the Hoover and Brookings Institutions and the National Bureau of Economic Research (which is based in Cambridge, Massachusetts and more Harvard than national). A well-known venue managed by the group is Education Next, where members publish, lavishly praise each other’s work, and, just as often, dismiss the existence of competing work done by others. Call them, inbred and unread.
Congresspersons may have expected their invited guests to present a panorama of the research evidence to consider. They gave them a pinhole to peep through.
The researchers recommended—and claimed that all the “best research” supported the recommendation—that the federal government continue to require annual testing in reading and mathematics, largely because it gives them great data sets to analyze, but also because annual testing is necessary to continue value-added measurement of teachers.
What of the main alternative—“grade span” testing—testing students at the end of the main levels of education, such as the end of the primary grades, the end of middle school, and the end of high school? Not even mentioned.
Completely ignored then were obvious points in favor of grade-span testing, such as: every one of the dozens of countries beating us on international tests requires grade-span testing and not annual testing; grade-span testing typically includes a full-battery of core subjects, whereas our annual testing includes only reading and math; and it is far easier to apply stakes to student test performance with grade-span tests than it would be with annual testing, and students learn far more when the tests they take “count” for them.
Indeed, one drawback to annual student tests, with no stakes for the students, for measuring teacher performance is that student effort does not just vary, it varies differently by demographic subgroup. The economists who like value-added measurement (VAM) just assume away these and other critical flaws.
The researchers testifying before the HELP Committee offered as their primary piece of evidence supporting the use of VAM two economists’ study of the District of Columbia Public Schools’ (DCPS) program. They asserted that teachers who received low VAM scores tended to leave DCPS and, on average, were then replaced by teachers who earned higher VAM scores. So, teacher quality improved, you see.
Intelligent people conducted this study, so one would assume that they recognized both the tautology and the regression toward the mean. That they proffer the study as valid evidence anyway seems cynical. Instead of ranking teachers by VAM scores, DCPS could have ranked them by height, weight, age, or their favorite colors. If height were the criteria for evaluating teacher performance, one would expect shorter teachers to leave DCPS and, if replaced at random, on average their replacements would be taller. That doesn’t mean that students learn more from taller teachers.
Republican lawmakers typically vouch for the value of competitive markets and eschew the harms of monopolies. Unregulated monopolies, most Republican Congresspersons can probably tell you, will inevitably lower quality; raise prices; reduce output, efficiency, and innovation; and raise barriers to entry from potential competitors. True to form, the Republicans’ education policy advisors exhibit all the classic monopoly behaviors, not only poorly serving the American public, but poorly serving the Republicans, too. When will Republican politicians finally recognize that they are being used?
Researchers who tell you “listen only to us and do not talk to anyone else” are typically ruthlessly ambitious, but they are not shy. Have we witnessed this ethically challenged scholarly behavior before in education policy-making discussions? Why, yes, in 2001 when the No Child Left Behind Act was first written.