Corruption in Education: It's the Assumptions, Dummy!

Concerns have been circulating for some time that teachers, or possibly even entire schools, have been cheating on behalf of their students on standardized tests. These concerns were widely publicized by, among others, Steven Levitt & Stephen Dubner of Freakonomics fame. The data they presented was pretty compelling - patterns in test answers indicated that cheating was definitely taking place.

Now the story is back in the headlines - this time with an entire school district. The former Superintendent of the Atlanta schools, along with 34 other administrators and teachers, has been indicted on charges of racketeering, theft, conspiracy, influencing witnesses, and making false statements. The superintendent faces 45 years in prison if convicted - punishment that would run beyond what many convicted Wall Street swindlers have faced in recent years. A total of 178 teachers and administrators were implicated district-wide.

That there is cheating and corruption on this scale in American public schools will come as a shock to some - although the "dramatic increases" in test scores seen in the Atlanta schools should have been enough to raise eyebrows rather than garner invitations to the White House. The article linked above should be read widely by anyone who is interested in K-12 education in America today. This isn't to say that every school district in America has this problem - but if you think it's confined to Atlanta, I have a bridge to sell you.

What's more interesting to me is how we got here. This kind of corruption is a direct product of No Child Left Behind and the high-stakes "accountability" testing that it ushered in. NCLB rests on a set of assumptions that, when exposed, are absurd. The entire system relies on a set of sticks and carrots, with standardized test scores designed to trigger either benefits to those who do well or punishment to those who fail to make the grade (with a decided emphasis on the punishment side, embodied in the widespread term "failing school"). And because you can't punish students directly, these carrots and sticks have been directed at districts, schools, and (increasingly, especially here in Ohio) individual teachers.

If you believe that this kind of accountability is going to improve children's education, you have to accept two assumptions. One is that standardized tests actually measure the educational outcomes we want. That belief has been widely debated and criticized, and I won't rehash that debate here.

Even if you believe that standardized tests are reliable and valid measures of student education, believing that carrots and sticks will change those outcomes requires a second assumption: that the fundamental obstacle preventing kids' learning is the motivation of teachers. You are, in essence, arguing that teachers are either lazy or incapable.

If the former is true, then threats and rewards should induce them to do what they could otherwise do, but won't. If the latter is true, then the system should force out teachers and replace them with others capable of doing the job (at the prevailing wage structure for teachers, no less).

Since we don't see mass teacher firings, the primary mechanism here is motivating the existing teacher workforce. People who believe in the NCLB approach think that if you threaten teachers, they'll suddenly start doing a good job - which they could have been doing all along, if only they were properly motivated.

This is, of course, an absurd assumption. Of all of the obstacles preventing kids (especially kids in inner-city districts like Atlanta) from learning, the motivation level of their teachers is pretty low on the list. Yet somehow, introducing "accountability" is supposed to solve everything.

That this kind of absurdity should produce widespread cheating and corruption, as it has in Atlanta, should not surprise us. NCLB has turned school districts into a version of Kafka's The Trial - a world in which truth is irrelevant and real progress is impossible. In such a world, what else should we expect? Teachers and administrators faced a terrible choice: cheat or be fired.

We should all hope that Dr. Hall, the administrator at the center of the whole mess, is sentenced to a goodly long prison term as an example to others. But even for her, the incentives were skewed. Until she was caught, cheating earned her the highest honors - invitations to the White House, plaudits from politicians and the business community, and some $500,000 in "performance bonuses".

Given the obstacles faced by a large number of the 52,000 children in her district, there was no way for her to earn these things honestly - especially without large numbers of additional resources to invest in Atlanta schools, which she did not have. It's unfortunate that she made the choices she did - but if we present people with a system in which cheating is the only way to get ahead, we shouldn't be surprised when some of them do.

It may be too much to ask that this one case, as shocking as it is, will cause us to reexamine the NCLB approach and the absurdist assumptions which underlie it. Maybe we need to indict a dozen more major urban superintendents. I hope that it doesn't take that long, and that we can someday soon rebuild our approach to education on the basis of a more rational set of assumptions.