TEACHER quality has become a highly politicized issue. In November, Californians will vote on Proposition 74, which weakens tenure rules for kindergarten through 12th-grade teachers. Not surprisingly, Republicans and Democrats are lined up on opposite sides of the issue, with debate focusing on the impact of the proposition on teacher quality.
But what is teacher quality? How can one measure it reliably? How does it relate to student learning? What can be done to improve it?
Though these issues are endlessly debated, there is little consensus. Generally, good teachers like to teach good students in good schools. Well, sure, but what is cause and what is effect?
A National Bureau of Economic Research working paper by Eric Hanushek, John Kain, Daniel O'Brien and Steven Rivkin called "The Market for Teacher Quality" sheds some light on these contentious issues. (A summary of the paper and a link to the text are at www.nber.org/digest/aug05/w11154.html.)
Their analysis is based on a sample of data from a Texas school district. They argue convincingly that teacher effectiveness should be measured by students' gains on standardized tests: Mr. Smith is presumably a better teacher than Ms. Brown if his students consistently improve their test scores more than hers do.
Though this approach is appealing, there are tricky issues. For example, what if Ms. Brown teaches in a school where students score so high there is little room for improvement?
The authors try to correct for this problem, as well as other sorts of measurement issues, to generate a measurement of teacher effectiveness. The paper is primarily concerned with how this measure is related to other observable teacher characteristics.
The first finding is that there is a large variation in teacher effectiveness: some teachers consistently have a larger impact on their students' achievement than others.
Second, easily observable characteristics like having a master's degree or a passing score on the teacher certification exam are not correlated with teacher effectiveness.
Then what does matter? The most important single influence is experience: first-year teachers are much less effective than others. The second year is significantly better, and by the fourth year, most teachers hit their stride.
It is not entirely clear whether this experience effect is learning by doing (the more you teach, the more effective you become) or survival of the fittest (those who are not good at teaching tend to drop out early).
If learning by doing is the major effect, then principals should emphasize teacher retention, reasoning that new teachers will do better with time. If survival of the fittest is the major effect, then principals should encourage teachers who are faltering after their first year to leave sooner rather than later.
From my reading of the paper, both effects appear important and there is no simple answer. The data do suggest, however, that teacher effectiveness is pretty clear by the end of the second year, so the information to make an informed decision is available at that time.
The authors also investigate the contentious issue of racial matching of students and teachers. Here they find strong evidence that minority teachers tend to be more effective with minority students. "Students who have had both a black and a white teacher perform better relative to classmates during the year in which they had a teacher of their own race," they said. Again, it is unclear whether this is because of a role model effect (students respond better to a teacher of their own race) or an empathy effect (teachers empathize better with students of their own race) or something else entirely.
The authors also look at teacher mobility. There is some evidence that teachers who quit teaching or switch schools tend to be below average in effectiveness. This is consistent with the survival-of-the-fittest model.
But, interestingly enough, the quality of those who leave is significantly lower in the year they leave than in the previous year. Once again, two interpretations are possible: Perhaps the teachers have a particularly bad year, which pushes them over the edge in terms of the decision to leave, or once they have decided to leave they may slack off in effort.
Finally, the authors examine whether experienced and effective teachers who move to a different district receive better compensation packages. They find little evidence that they do, though this is a weak conclusion because of the small sample size. Other work suggests that schools may trade quality for lower turnover.
Clearly, teacher turnover is a critical issue in several ways: it absorbs a significant amount of management attention and has a major impact on student learning. Getting teacher retention policies right can make a big difference for student achievement, particularly in schools serving disadvantaged students.