Working Paper: NBER ID: w7082
Authors: Eric A. Hanushek; John F. Kain; Steven G. Rivkin
Abstract: Important policy decisions rest on the relationship between teacher salaries and the quality of teachers, but the evidence about the strength of any such relationship is thin. This paper relies upon the matched panel data of the UTD Texas School Project to investigate how shifts in salary schedules affect the composition of teachers within a district. The panel data permit separation of shifts in salary schedules from movement along given schedules, and thus the analysis is much more closely related to existing policy proposals. In analyses both of teacher mobility and of student performance, teacher salaries are shown to have a modest impact. Teacher mobility is more affected by characteristics of the students (income, race, and achievement) than by salary schedules. Salaries are also weakly related to performance on teacher certification tests appearing to be relevant only in districts doing high levels of hiring, but preliminary examination shows that the certification tests are not significantly related to student achievement. The only significant relationship between salaries and student achievement holds (implausibly) for existing experienced teachers and not for new hires or for probationary teachers.
Keywords: teacher quality; salaries; public schools; teacher labor market
JEL Codes: I21; J31
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Higher salaries (J31) | Teacher quality (I21) |
| Teacher quality (I21) | Pay differences (J31) |
| Hiring and retention practices (M51) | Teacher quality (I21) |
| External labor market conditions (J29) | Teacher quality (I21) |
| Teacher test scores (C12) | Starting salaries (J31) |
| District salary schedules (J31) | Student achievement (I24) |