By researching India's poorly staffed schools, this MIT group figured out how to fix them.A teacher-less classroom is an all-too-familiar scene for primary school students in many low-income countries. In 2006, a study by economists from the World Bank and Harvard University turned up disastrously low teacher attendance in six such countries, including India, where on any given day 55 percent of all classrooms either get a teacher who is present but not teaching or no teacher at all. With salaries for teachers accounting for the biggest fraction of governments' education budgets in the developing world, the study highlights an important challenge: how to get a bigger bang from a country's teaching buck.To that end, researchers associated with the Massachusetts Institute of Technology's Abdul Latif Jameel Poverty Action Lab worked with the Indian state of Andhra Pradesh to randomly select 500 government-run primary schools in which to test interventions aimed at boosting teacher performance. The biggest effect came from performance-based pay. Scores on standardized tests went up in schools where teachers were incentivized with a possible bonus of about a third of a month's pay if their students achieved a 10 percent increase in test scores.How big a grain of salt does a study of 500 schools merit? "It's large enough that the results have to be taken seriously," says lead researcher Karthik Muralidharan, an economist at the University of California at San Diego, "but it doesn't give you a roadmap to implementation." He recommends that policymakers begin larger-scale pilot projects. There was no magic to it, he points out; teachers simply started doling out more classroom work and homework. Still, that's a big improvement over sleeping on the job (or not showing up at all).Photo by Karthik MuralidharanGo back to the interactive site