Did you think that maybe teachers are about the only people in the world who don’t improve with experience? I had no idea this meme was out there, until I read this article in EdWeek [beware paywall]. Apparently it’s been conventional wisdom among “reform” proponents (like Bill Gates) t that teachers get better over their first few years (roughly the induction phase) and then stall out. This idea can be used to buttress all kinds of changes to teachers’ employment — as Gates wrote, “we need to raise performance without spending a lot more.” As he writes, “It’s reasonable to suppose that teachers who have served longer are more effective, but the evidence says that’s not true. After the first few years, seniority seems to have no effect on student achievement.”
This is convenient — we need to make big improvements but not spend a lot more. Wonder how we can manage that… Hm, “The United States spends $50 billion a year on automatic salary increases based on teacher seniority.” [emphasis in the original] Great! Since we know that teachers don’t improve after the first few years, we can eliminate all those raises, and all those costly old people, and it’s all “research based.”
Now, one thing about having some knowledge of a field is that when your data provides anomalous results, based on your understanding of how the system works, you feel the need to challenge the surprising conclusions. More research might show that you’ve discovered surprising new insights — or not. In any case, you don’t instantly start prescribing remedies based on counterintuitive findings, until they’ve really been tested and understood. (Unless maybe you’re selling something.)
The EdWeek article reports on two studies recently released in working-draft form, which dig into the impact of teacher experience over some significant portion of a career. John Papay and Matthew Kraft’s paper spends a lot of time unpacking various statistical methods with which to address the fiendishly intangible idea of “returns to experience.” Their evidence suggests that there is indeed a faster rate of growth during the first part of a teacher’s career, but that growth continues thereafter, albeit at a slower rate. They also make the very basic point that [a] what you measure affects what you find, and [b] easily measurable “returns,” for example short-term student gains on tests, are not the only kind of value to examine. For example, they say (pg. 36)
Carrell and West (2011) find that in higher education, more experienced professors have less success in promoting student short-term test-score growth than their less experienced colleagues, but they contribute substantially more to their students’ lasting knowledge and academic skills.
Helen Ladd and Lucy Sorensen examine ELA and math teachers in North Carolina. They also find evidence of continued growth (in terms of impact over time), and also stress that their studies make clear that teachers’ effectiveness cannot well be measured by student achievement alone. While there are interesting differences in the patterns of growth over time between ELA and math teachers, in both subjects more experienced teachers have measurably greater effects on factors that are important for student success such as absenteeism, student behavior, and other factors that fall under the cold rubric of “classroom management,” but reflect, in my humble opinion, increasing wisdom of a practical kind — what helps students succeed as people, at least in the school context.
Of course, both these studies are looking at teachers as statistically modeled, and results may vary depending on many considerations, some of them personal to the teacher, and some reflecting the teaching/learning conditions of a particular year and school.
Nevertheless, these two studies are timely in pointing out that teachers are learning professionals, and that means that they keep learning themselves; that there are measures of student success that matter to the individual and to society at least as much as test scores; and that teachers’ impact in these other, harder to evaluate ways have to be considered seriously in evaluating performance, reward, and value to societal well-being. Surprise!