Incentives, Selection, and Teacher Performance: Evidence from IMPACT
IMPACT—the District of Columbia’s controversial teacher-evaluation system, ushered in by former D.C. Schools Chancellor Michelle Rhee—offers robust incentives and sanctions for teachers and links them to multiple measures of performance, such as test scores, classroom observations, and collaboration with colleagues. And according to this new study, it is working. Analysts studied teacher-level administrative and demographic data for DCPS’s general-education teachers in grades K–12, and their students, over the first three years of IMPACT (2009–10 through 2011–12), including their scores on IMPACT, which placed them into four categories of effectiveness ranging from highly effective to ineffective. Teachers in the latter category were immediately dismissed, while those in the next lowest category (called minimally effective) were subject to a dismissal threat—meaning they could be fired the following year if their rating did not improve. On the other hand, those who were rated as highly effective were eligible for a one-time bonus of up to $25,000, as well as increases in their base pay if they scored at the highest rating for at least two consecutive years. When comparing these two “threshold” groups, the low-performers (whose ratings placed them near the threshold of a dismissal threat) and the high performers (whose rating placed them near the threshold for the big pay bump), the analysts found that the threat of dismissal increased the voluntary attrition of low-performing teachers by 11 percentage points (or more than 50 percent) and improved the performance of teachers who remained by 0.27 of a standard deviation. They also found that the financial incentives further improved the performance of high-performing teachers; specifically, teachers who originally scored at or above the highest rating in the first year scored roughly 10 percentage points higher in the second year. Commentary on these results abounds in the blogosphere, essentially boiling down to one point: IMPACT is a unique program with a lot of factors—political, geographical, and so on—that converged to make it successful, meaning it is not easily replicated.
SOURCE: Thomas Dee and James Wyckoff, “Incentives, Selection, and Teacher Performance: Evidence from IMPACT,” NBER Working Paper No. 19529 (Cambridge, MA: National Bureau of Economic Research, October 2013).