Explore J-PAL's current work in education and the policy lessons that have emerged.  
Key Image: 

Low student attendance, absent teachers, and underperforming schools are persistent problems across the developing world. Randomized evaluations have shown that simple interventions, such as providing deworming treatment and informing parents about the returns to education, can dramatically improve attendance at a low cost. To increase student learning, matching instruction to students’ ability levels has proven to be very effective. Incentives, such as merit scholarships for students, and performance- or attendance-based pay for teachers, have also improved learning outcomes in some contexts. There is little evidence that simply providing more inputs (such as textbooks) has much impact on learning.

Education remains one of the most promising ways for children to attain a fuller, more productive life. While many developing countries have improved access to education, the quality of education has remained low, and many children leave school having learned few real skills. If universal education is to be attained in the truer sense of enabling these children, their families, and their communities to realize the promise of education, developing countries must improve the quality of the education they offer.

J-PAL's Education Program promotes research aimed at providing policymakers and practitioners with the knowledge needed to design and implement more effective education policies and programs, including:

  • Surveying current research to provide descriptive information on the status of education, and the challenges and opportunities that students, parents, and communities, as well as teachers, schools, and governments, face
  • Conducting randomized impact evaluations to provide rigorous scientific evidence on what policies and programs work
  • Cost-effectiveness analysis to provide information on the relative cost effectiveness of potential solutions

The primary open questions in education can be divided into the following seven areas:  

Pedagogy. The provision of additional resources unaccompanied by changes in teaching behavior often has little or no effect on learning. What teaching methods best meet the learning needs of children within the large and heterogeneous classes that are common in most developing countries? How much emphasis should be placed on teaching competencies rather than teaching specific content? Given the heterogeneous achievement level in these classes, should the grade structures be more flexible so that some children can take different subjects with different peer groups, taking math with one group and reading with another?

Technology. While most attempts to use information technology to promote more effective teaching in wealthier countries have not been very successful, we see a more positive picture from the few available studies in developing countries. However, it seems as if the details of the program—whether or not computer classes are taught during or after school hours, or by existing teachers or external instructors—matter a great deal.

Teacher training. What is the best way to train teachers who may themselves not be very well educated? What can be done to provide current teachers with the skills they need to better target children at different levels within a large classroom? How structured should the teaching be: should teachers be asked to follow a very specific plan or should they be given discretion?  

Teacher incentives and school governance. Evidence suggests that teacher incentives help improve student test scores when they are implemented effectively. But evidence on programs with decentralized incentives and those that aim to empower communities to control local schools has been much more mixed.

Many questions remain: Is it better to give teachers incentives based on attendance or test scores? Is the answer different if we care about end-of-year test scores or long-term learning? Why do we find many instances of parents being reluctant to exercise control over schools even when they are legally entitled to do so? Is it because they do not believe that they can affect the learning outcomes of their children? Would more information on school performance induce greater participation from parents? What is the effect of offering school choice on parental choices, teacher effort, and child outcomes?

Incentives for children. What can be done to get children and parents from disadvantaged families to invest more effort in schooling? Evidence suggests that providing incentives for children based on educational outcomes leads to increased effort and improved performance. Evidence also suggests that a significant fraction of children and parents seriously underestimate the economic benefits of schooling for children of backgrounds such as theirs, and react positively to accurate information about the returns to schooling.

Preschool and early childhood development. Institutionalized early childhood education, such as preschooling and initiatives that provide extra support to children already in lower primary school, could give underprivileged children a head start in the learning process—in particular first-generation learners whose parents never went to school or are illiterate. These programs could also enable the delivery of proper nutrition to children at an age when it is vital for brain development and learning potential.

Post-primary education. Evidence is much stronger on primary education than on post-primary education. As many more children finish primary school, the need to provide them with quality post-primary education looms large. While many of the same issues raised above—pedagogy, technology, training, and incentives—may be important here too, there remain a number of issues specific to post-primary education about which little is known. J-PAL's Education Program aims to further examine these specific issues through the Post-Primary Initiative (PPE).

Policy Lessons

Impersonally administered and direct incentives for attendance are the most effective at increasing service provider attendance. However, in settings where supervisors were given discretion over administering incentives, these programs became ineffective.

There are many ways of increasing school enrollment and attendance, but there is huge variation in the costs of different programs.

Numerous strategies to improve student learning have been evaluated, and found to have widely different impacts. These different strategies also incur drastically different costs, and some programs therefore achieve learning gains with much greater cost-effectiveness than others.