On a busy street in Cambridge, Massachusetts, an unassuming brick building houses the North American headquarters of the Abdul Latif Jameel Poverty Action Lab (J-PAL). Inside MIT’s J-PAL Lab, researchers analyze the results of randomly-designed experiments and the generalizability of their findings. Their test subjects? Social programs.
J-PAL and its Research
Anti-poverty work can take many forms: vocational training, credit access, education and the list goes on. Programs of various types can be valuable and worth funding, but because donors and organizations have limited amounts of funding, the relevant question is often not whether to support anti-poverty work, but how to support it. What programs are making the biggest difference? Where will a donation do the most good?
J-PAL co-founder Abhijit Banerjee puts it best in an interview with The Wire: “you have to basically focus on: where is the lever? That’s what we do, try to help find the lever, and then if you push the lever you go fast. But what [policy-makers] do is often… just hit everything and then some things hit the lever, some things hit the wall.” The strategy of “hitting anything” is certainly not useless, but to maximize impact, a person should concentrate on the mechanisms that maximize impact. J-PAL’s work is in locating the policy levers that will make a difference and encouraging policy-makers to use them.
MIT professors Abhijit Banerjee, Esther Duflo and Sendhil Mullainathan founded J-PAL in 2003 and it has grown quickly. As of 2019, J-PAL has a network of 181 affiliated professors around the world. A truly global organization, J-PAL has offices in Capetown, South Africa; Paris, France; Santiago, Chile; New Delhi, India; and Jakarta, India. J-PAL’s work focuses on researching the results of different cases of anti-poverty action to evaluate what is most effective and advise policymakers accordingly, and educating people about how and when to evaluate social programs.
At its base, J-PAL’s approach is simple. Any case of policy-making has results, and by tracking actions and results in a large number of cases, one can determine which are successful and which are not, and therefore which to support in order to produce a maximum impact. Using this data, one can also attempt to predict the policies whose results could potentially be reproduced elsewhere. J-PAL’s methods also emphasize focusing on the mechanisms behind policies and not on specific details of a single scenario. Particular details vary from case to case, but human behavior does show certain patterns; in similar circumstances, policies that have worked before can work again.
J-PAL and its affiliated professors have performed over 900 evaluations and have developed a selection of case studies and other teaching resources to educate people about the analysis of social programs. Every year, J-PAL runs Evaluating Social Programs, a five-day course for policymakers, researchers and nonprofit workers that covers the basics of organizing scientifically sound evaluations and interpreting their results. Close to 50 people attended the 2018 course, which took place at MIT. J-PAL has also made the content of the course available online for free through the online education platform edX, allowing people around the globe to learn from J-PAL’s work. The case studies discussed in the course are also available on J-PAL’s website.
MIT’s J-PAL Lab has encouraged the scaling up of policies ranging from remedial education in India, deworming in Kenya and community block grants in Indonesia. It also continues to study policy and advocate for effective social policies. The training that it provides for policy-makers allows them to maximize their impact in an economically efficient way. As J-PAL continues to grow, its affiliates will continue to find those levers.
– Meredith Charney