Individual-Level Longitudinal Data on School Competition and Test-Score Inequality

What the evidence says

The available evidence on this topic covers teacher mobility in relation to school-level student achievement, not school competition or test-score inequality directly. Using panel data from 1993-1996, one analysis finds that teachers in schools in the bottom quartile of average student achievement leave their school at a rate above 25 percent per year, compared with less than 20 percent in top-quartile schools — a gap of roughly 5 percentage points. A regression-based measure of the same relationship, using standard-deviation units of school average achievement, finds that a one standard deviation decrease in achievement increases the probability a teacher exits the district by 1 to 2 percentage points, against a baseline exit rate of 13.7 percent. Both measures point the same direction: lower average student achievement at a school is associated with higher teacher exit, whether measured as a quartile comparison or as a continuous regression coefficient. Confidence in the direction of this relationship is high across both specifications; confidence in the exact magnitude is high for the quartile comparison and medium for the regression estimate, since the regression coefficient's stability across model specifications is flagged as an open question.

A separate finding describes where teachers go when they do move: on average, teachers who switch schools move to a school where district average achievement is about 0.08 standard deviations higher, equivalent to roughly 3 percentile points on the state performance distribution. This is a descriptive, not causal, measure of the destination schools of movers, and it is consistent with — though not identical to — the exit-rate findings, since it describes the endpoint of mobility rather than its trigger.

The sources say nothing about school competition (for example, choice, charter entry, or open enrollment) or about test-score inequality as a distributional outcome. All three findings concern teacher labor-market behavior conditional on school achievement levels, drawn from a single source using panel data from the 1990s. No claim here measures effects on students' test scores, on inequality in test scores across schools or groups, or on any competitive mechanism between schools.

Where the evidence splits

No contradictions in the current sources.

Thresholds and dose-response

No threshold established in the current sources.

Open empirical questions

• Whether school competition (choice, charter entry, open enrollment, or similar mechanisms) has any measured relationship to teacher mobility or to test-score outcomes — not addressed by any claim here.
• Whether the achievement-exit relationship observed in 1993-1996 panel data holds in more recent periods or other districts.
• Whether teacher mobility toward higher-achieving schools changes test-score inequality between the schools teachers leave and the schools they join.
• What mechanism connects lower school achievement to higher teacher exit rates — the source establishes the association but not the cause.