Skip to:


Paying for Performance

There is a strong relationship between poor health and the achievement gap.

Gaping rural-urban inequality is a major challenge facing the rapidly developing Chinese society today. Educational inequality is perhaps among the most severe, where alarming disparities in school performance exist between rural children and their urban counterparts.


Recent research has shown that poor health and nutrition, among other issues, are significant contributing factors to this achievement gap.

Reducing iron deficiency anemia (the most common type of anemia among children in rural China) is in principle easily accomplishable with simple, inexpensive nutritional interventions. However, few corrective steps have actually been taken. The misalignment between suppliers’ incentives and the socially desirable outcome of health may be one reason. In light of this, in 2010, we conducted a pay-for-performance pilot project. We found that schools that received financial incentives to reduce anemia showed higher increases in hemoglobin compared to control schools.

Our research also found that incentives based on anemia reduction are more effective when complemented with incentives based on test scores. If principals believe there is a connection between anemia and school performance, incentives based on test scores alone may raise hemoglobin levels and academic achievement.

Recent research from the MIT Poverty Action Lab has suggested that the size of the incentive has little effect on the outcome. If this is true, incentive schemes that reward performance could be considerably cheaper and more financially sustainable.

In this study, we hope to further our previous work and explore these possibilities by testing more cost-effective forms of performance pay.


Our objective in this study is to assess the potential of performance pay, specifically health and test score based incentives to school principals, as a useful 

element of policy efforts aimed at reducing anemia in rural China. In particular, we hope to assess:

  • If incentives or grants are more effective in improving health and educational outcomes
  • The effectiveness of small incentives relative to larger incentives  


Sampling and Randomization

We randomly selected 300 primary schools in China’s northwestern Qinghai, Shaanxi, and Gansu provinces. These areas are known to have relatively high anemia rates. We surveyed 50 4th and 5th graders at each school for the baseline survey. We then randomly assigned some schools to experimental arms. In these experimental arms, some schools received incentives for anemia reduction or improved test scores, and some received grants. We then compared the effectiveness of incentives and grants to our control schools, who received no intervention.

Baseline survey (September 2011)

At home, we surveyed parents on basic household characteristics. In schools, we surveyed students, teachers, principals, and district superintendents on general and health-related questions. Specifically, we collected data on student hemoglobin levels and performance on standardized tests of math and Chinese. These served as our main outcome variables. Our secondary variables included nutritional knowledge and principal/teacher time use.

Endline survey (June 2012)

We used the same forms and tests as in the baseline survey, then compared the change in hemoglobin levels and test scores (as well as other secondary variables) between the control group and the intervention groups. In doing so, we were able to assess the individual and combined effects of anemia reduction and test score based incentives on health and educational outcomes, as well as test whether small incentives are as effective as larger ones.


We found that both large grants and large incentives were effective means to lower anemia rates in schools (though small grants and incentives were not).

When principals received a grant, anemia rates in their schools dropped by 14.8 percent from a starting point of 36 percent. This suggests that even without explicit incentives to do so, principals were motivated to allocate more resources to nutrition programs. Children in these schools reported eating more vegetables and meat both at school and at home, suggesting that principals also encouraged families to improve their children's diets.

Students also benefited when their principals were given incentives specifically tied to anemia reduction. In these schools, anemia rates also dropped by more than one third. Students reported eating more vegetables and meat at home, suggesting that principals successfully encouraged families to improve nutrition in their households. Additionally, incentives were much more cost-effective than grants in achieving anemia reduction.