Fair Merit Pay Schemes, Part VI

Some neat new evidence from the Indian state of Andhra Pradesh, where they’ve just carried out a randomised field experiment on teacher merit pay.

Teacher Incentives in Developing Countries: Experimental Evidence from India
Karthik Muralidharan & Venkatesh Sundararaman
Performance pay for teachers is frequently suggested as a way of improving educational outcomes in schools, but the empirical evidence to date on its effectiveness is limited and mixed. We present results from a randomized evaluation of a teacher incentive program implemented across a representative sample of government-run rural primary schools in the Indian state of Andhra Pradesh. The program provided bonus payments to teachers based on the average improvement of their students’ test scores in independently administered learning assessments (with a mean bonus of 3% of annual pay). Students in incentive schools performed significantly better than those in control schools by 0.19 and 0.12 standard deviations in math and language tests respectively. They scored significantly higher on “conceptual” as well as “mechanical” components of the tests suggesting that the gains in test scores represented an actual increase in learning outcomes. Incentive schools also performed better on subjects for which there were no incentives. We find no significant difference in the effectiveness of group versus individual teacher incentives. Incentive schools performed significantly better than other randomly-chosen schools that received additional schooling inputs of a similar value.

You can’t get much simpler than the formula used to determine the merit payment:

Bonus = Rs. 500 * (% Gain in average test scores – 5%) if Gain > 5%
= 0 otherwise

This entry was posted in Economics of Education. Bookmark the permalink.

4 Responses to Fair Merit Pay Schemes, Part VI

  1. Kevin Cox says:

    The way I read it it did not matter whether the incentive was on the school or on the individual teacher. Taking away the individual teacher evaluation and making it a school evaluation will remove many of the objections to performance based pay while still giving the same results.

  2. Andrew Leigh says:

    Yes, I noticed that too. It’s different from Victor Lavy’s finding that Israeli school-based bonuses were less cost-effective than teacher-based incentives. But I guess we shouldn’t be surprised that the effects of incentives can differ across countries.

  3. Kevin Cox says:

    It could be something to do with the size of the schools. These schools were small and so the group is more likely to see itself as a team whether they were evaluated individually or as a group. My guess is that it works best when there are groups where people can identify with the group and they know that helping each other will lead to the best results. Maybe in looking at the evidence some measure could be made of “tribalisation” or group coherence or ?

  4. Yes, the small school size was probably the main reason. One of our hypotheses was that the interaction of school size and group incentives would show an inverted U relationship with outcomes – as in, there would be gains to cooperation at first (while peer monitoring would still be able to solve most of the free-riding problems), but that as the schools got larger, the free riding would dominate. The problem is that the typical rural government school is quite small and so our (representative) sample does not give us too much variation in school size (with 92% of schools having between 2 and 5 teachers), and so these interactions were not significant.

    But the fact that group and individual incentives show similar results for groups of 3-4 teachers might mean that even in larger schools, you could create groups of say teachers in the same grade, and treat them as the unit to be rewarded. We will have the data from the second year of the program soon and the differences might become more pronounced then.

Comments are closed.