Student Demographics, Teacher Sorting, and Teacher Quality: Evidence from the End of School Desegregation
June 03, 2009
C. Kirabo Jackson
Journal of Labor Economics
We know that schools with lots of poor kids tend to have less effective teachers while schools with wealthier kids tend to have better ones. What we know less about is why. This study by Cornell economist Kirabo Jackson suggests a disturbing possible explanation. He examined teacher movement patterns in the Charlotte-Mecklenburg schools between 2002 and 2003, which was the year this 137,000-student district ended its 30-year-old busing policy. The result was a naturally-occurring and rare situation where the demographic makeup of schools very quickly converged to that of their surrounding neighborhoods. And it provided Jackson with a unique opportunity to observe teachers' reactions to changes in student demographics while other school and neighborhood characteristics (as well as teacher demand) stayed the same. The results are disquieting. Schools that had an influx of black students as a result of the policy change also saw a decrease in the number of high-quality teachers--both black and white--as measured by years of experience, value-added scores, and certification scores. For the average school that was 60 percent black (before the policy change), a 15 percent increase in black students (after the change) translated into a .3 standard-deviation decrease in teacher quality. In the end, this study refutes the idea that the relationship between teacher quality and student demographics is merely a product of residential segregation. Rather, teachers sort themselves according to student demographics, with some apparently preferring to teach minority students (perhaps because they think they are better at it) and others not. You can read it here (for a small fee).