Education writer RiShawn Biddle has offered a spirited response to my blog post yesterday about the failure of the No Child Left Behind act. In it, he asserts that NCLB has advanced school choice, and links to an earlier essay that ostensibly presented his case. Summarizing it, Biddle writes that:

The impact of No Child on advancing choice… starts with the law’s Adequate Yearly Progress requirements. Thanks to the data culled, the low quality of education in traditional district schools was exposed for all to see, providing parents and school choice activists with the information they needed to push for the advancement of choice.

No thanks. The poor performance of U.S. schooling has been evident to a great many people for a very long time. The bestseller Why Johnny Can’t Read was first published in 1955. Over the past 40 years, the NAEP’s Long Term Trends (LTT) tests have revealed stagnation in math and reading and decline in science toward the end of high school. In contrast to the consistent and nationally representative results of the NAEP LTTs, the NCLB is tied to state-administered tests that are so often corrupted by tinkering with their content and cut scores that they are largely worthless for measuring achievement at a single point in time let alone for measuring trends.

Biddle also claims that NCLB

exposed the long-running gamesmanship by states looking to define proficiency downward (a fact that Cato has used to its own advantage in arguing against expanding federal education policy); this, in turn, has rallied more reformers to move toward advancing school choice.

In reality, NCLB exacerbated the gamesmanship of state-level tests by giving state officials incentives to show the appearance of progress rather than actual progress. Moreover, it was not NCLB that exposed this fraud that was partially of its own making. For that we can thank… the NAEP. It was by comparing unreliable state test scores to far more reliable NAEP scores that it was discovered just how badly public schools in many states have been lying to families about their children’s performance. Even Secretary of Education Arne Duncan has noted this fact, saying in 2009 that:

When states lower [their own academic] standards, they are lying to children and they are lying to parents. Those standards don’t prepare our students for the world of college or the world of work. When we match NAEP scores and state tests, we see the difference. Some states, like Massachusetts compare very well. Unfortunately, the disparities between most state tests and NAEP results are staggeringly large.

[Ironically, Duncan seems to have benefited from the absence of such a comparison while he was head of Chicago Public Schools, riding into his current position on the wings of a supposed “Chicago Miracle” that appears, based on NAEP scores, to have been a mirage induced by fanciful state tests.]

Biddle then goes on to praise NCLB’s “focus on graduation rates,” which he claims “forced states to present realistic numbers.” While it is true that many states had been reporting meaningless graduation statistics prior to NCLB, it is not at all clear that the law has improved matters. On the contrary, Nobel Prize winning economist James Heckman concluded from his exhaustive statistical study of the subject that NCLB appears to have fostered further cheating with graduation rates—what he calls “strategic behavior” by states and districts to present inflated graduation rate figures in order to avoid NCLB penalties. So, once again, it appears that NCLB is obfuscating rather than illuminating educational performance in America.

Finally, a note about Biddle’s characterization of my and my colleagues’ work at Cato’s Center for Educational Freedom. Apparently discomfited by my criticism of NCLB, Biddle dubs us dogmatic ideological purists, unthinking and blindered, and claims that we praise or attack policies based on our “worldview,” etc. etc. While I can understand becoming exercised as a result of a policy debate, I cannot understand why someone who wants to be taken seriously would stoop to such obviously fatuous ad hominem attacks. My last paper was a regression study of the link between the performance of charter school networks and the grant funding they receive. It has multiple technical appendices, several of them added in response to peer reviews. Anyone who doubts its findings is welcome to repeat it and see if they obtain different results. The paper I wrote before that was a regression study of the regulatory burdens imposed by voucher and tax credit programs. It, too, can be repeated by other researchers if they wish to verify its findings. The term for this kind of testable, repeatable work is science, not “dogma” or “ideology” or “world view.” My colleagues are likewise engaged in empirical research and we derive our policy recommendations from that research. So our conclusions are indeed very narrowly constrained, but not by ideology. They are constrained by what works, and what does not work, in the real world.