How Does Teacher Retention Affect Schools?

I've been on the road for two weeks now and all the conference presentations are starting to muddle together in my head. As such, I'm just going to talk briefly about the paper I presented yesterday. It's one of the few things that I can keep straight in my head right now.

Here's the premise of the paper:

Lots of people have investigated why teachers leave schools, but I haven't seen much on how schools are affected by teachers leaving. The simplest version of this question would be "does retaining more teachers improve student performance?"

As I've discussed before, the nature of teacher retention is very different in different schools. Since retention rates are notably low in high-poverty urban schools, I chose to focus solely on them. I had some data and was able to compile a dataset of 43 NYC middle schools that had large percentages of poor students (not including magnet, K-8, and some other schools with unreliable statistics). Across these schools, about 40% of teachers were in their first or second year of teaching at their current schools and less than half had a total of 5 or more years of teaching experience.

There was also a moderately strong (r=.44, p<.05) relationship between the average student score on the 8th grade state math test and the percent of teachers who had been teaching for at least two years at that school. In other words, schools with higher rates of teacher retention also had higher student achievement.

Using regression analysis and controlling for race and attendance rate, teacher retention was still significantly related to student achievement. Teacher retention had a fairly sizable effect (effect size of .20) and the model did a pretty good job of explaining the variance in test scores (R-squared of .70).

So, in short, among high-poverty middle schools in NYC, those with higher rates of teacher retention also had higher test scores, even when controlling for other things that influence student test scores. The question that I can't answer with the data is whether better schools make teachers want to stay there more or if more teachers staying in a school improves the school and boosts student achievement (or a little of both). I hope to come closer to answering this question in the future.