This post summarizes findings from a group of studies in education that were presented on March 2, 2012 at a half-day event hosted by Innovations for Poverty Action (IPA) and the University of California’s Center for Effective Global Action (CEGA). Please forgive any errors or omissions. Cross-posted from Philanthropy Action.
Karthik Muralidharan from UC San Diego opened the event by presenting findings from a large randomized controlled trial in which he tested four different approaches to improving student learning. In the last decade, student enrollment has improved significantly in primary schools in the developing world, but student learning has not seen a corresponding jump: in India, only eight percent of first graders can read at grade level. Muralidharan’s four approaches correspond to existing theories about why poor children do not learn. His studies tested the theories by seeing what happens when:
- Teachers have better information about where students are failing
- Schools have more money to buy supplies
- Schools gain a low-paid, minimally trained contract teacher from the community
- Teachers have better incentive to perform, either through performance pay based on how all students do in the school overall; or through providing individual performance payments based on how their students do
The findings, in brief, show that better information for teachers had no effect on student learning, and providing money for supplies had only a tiny effect, with diminishing impact over time (the decrease in impact corresponded to a decrease in parents investing household funds in school supplies for their children).
The third approach of providing a contract teacher, in contrast, resulted in significant improvements for students. Though Muralidharan’s data was not precise enough to know for sure, he cannot rule out that contract teachers were ultimately more effective than their better-trained, better paid counterparts, due in large part to a 40 percent lower absentee rate (recall that the contract teachers live in the community, so it is easier to get to school, and if they don’t show up their neighbors won’t have to go to far to complain).
Last, providing incentives through performance pay had the greatest effect on student learning, and individual incentives worked better than group incentives. Interestingly, teacher absence did not change at all, but when teachers were there it seems they more actively applied their skills, so that teachers who had the most training produced the best results in student achievement.
Muralidharan concludes that providing additional instruction in the early schooling years particularly to disadvantaged kids can make a huge difference, and that educators with less training can be an effective resource. His findings support earlier IPA/JPAL evidence from India and Kenya that quantify the positive impact of additional teaching resources on student outcomes.
Read more about the event and find papers presented on the event page here.