http://www.nber.org/papers/w2098 ... &utm_source=ntw

The long term effect of teachers’ pay for performance is of particular interest, as critics of these schemes claim that they encourage teaching to the test or orchestrated cheating by teachers and schools. In this paper, I address these concerns by examining the effect of teachers’ pay for performance on long term human capital outcomes, in particular attainment and quality of higher education, and labor market outcomes at adulthood, in particular employment and earnings. I base this study on an experiment conducted a decade and a half ago in Israel and present evidence that the pay for performance scheme increased a wide range of long run human capital measures. Treated students are 4.3 percentage points more likely to enroll in a university and to complete an additional 0.17 years of university schooling, a 60 percent increase relative to the control group mean. These gains are mediated by overall improvements in the high school matriculation outcomes due to the teachers’ intervention at 12th grade. The pay scheme led also to a significant 7 percent increase in annual earnings, to a 2 percent reduction in claims for unemployment benefits, and a 1 percent decline in eligibility for the government disability payment.