The LA Times unveiled this week the first results of a massive report measuring the “value-added” by each of thousands of teachers in the Los Angeles United School District. I hate to throw cold water on it. For the most part, the report is in an impressive advance. “Value-added” metrics attempt to separate the influence on student performance of schools and teachers from the influence of student background. Though elementary, in today’s climate of opinion, the distinction is subversive. As Richard Buddin, the RAND Corporation scholar from whom the LA Times commissioned the report, writes:
Many teachers feel that student performance is based on student background and preparation factors that they are unable to control. The premise is that inner-city teachers serve an at-risk population that will always have lower performing students than their counterparts in more affluent suburbs. This argument has considerable merit for comparing absolute test score levels across schools . . . .
In other words, the best schools get that way because they have the best students to begin with. You don’t say! Yet if school quality is a function of student quality, then “bad schools” can’t explain the persistence of achievement gaps between whites and (non-asian) minorities. (On the contrary, achievement gaps between whites and minorities can explain the existence of “bad schools.”) For many, this is an unwelcome thought, as it eliminates the most obvious and popular (though by no means the only) rationale for blaming society for racial inequalities.
Nonetheless, the LA Times doesn’t hold back: “Contrary to popular belief,” the Times reports, “the best teachers were not concentrated in schools in the most affluent neighborhoods, nor were the weakest instructors bunched in poor areas.” Duggins himself bluntly concludes that school effects “are small relative to . . . student achievement gaps between groups.” In other words, “bad schools” are a myth. They do not explain differences in education outcomes.
Still, for all the report’s courage, the cold water needs throwing. Careless readers of the LA Times may conclude that, to fix the schools, all we need to do is to recruit better teachers. But Buddin in no way shows that good teachers are a panacea. By design, the Buddin study only measures changes in student performance over one year. A good teacher may very well inspire his students to max out their potential in the short term. That does not mean that he has any ability to change outcomes in the long term.
For example, thanks to my good study habits in high school, I routinely pulled better grades than classmates who were smarter than I was. In the end, however, it was those classmates who landed the Supreme Court clerkships after law school or graduated Alpha Omega from medical school, not I. By working harder than they did, I could temporarily outperform them. Ultimately, I couldn’t keep up.
The gains from having a good teacher are similarly temporary. Sure, you can prod a slacker for a year or two and get him to think a little bit less about girls and little bit more about arithmetic. But you can’t prod him to do calculus if he doesn’t have the brains for it. Moreover, as soon as you stop prodding, he may prefer to return to his (more pleasant) slacker ways. Good teachers may provide no more than a temporary bounce in performance that, over the course of a whole academic career (or lifetime), cannot be sustained.
Tellingly, the Bruddin study says nothing about the marginal returns of having good teachers. Suppose a student has enjoyed “value-adding” teachers four years in a row – does he see the same gains every year? The Bruddin report does not say. If Buddin took another look at the data, he might find that, just as in everything else, marginal returns tend to diminish.
In the meantime, teachers without Patton-esque motivational skills are going to get blamed for failing students. Let’s give them a break. In the long term, they’re probably not doing any worse than their colleagues.
UPDATE: In the comments below, “Ciro” links below to research suggesting that gains from having effective teachers several years in a row are cumulative after all. I note that the literature summary to which he links says that “[N]o one has run a true experiment that involves actually randomly assigning students to high-performing teachers for several consecutive years.” So, there is some evidence to support the view that gains from having high-performing teachers are cumulative, but it is not yet conclusive.