Abdul Latif Jameel Poverty Action Lab (J-PAL)

Public policies for development and poverty reduction have often been applied as simple formulas, with no prior assessment of the local environment or of their feasibility in the specific context and circumstances. Such policies may have extremely negative results for society and the local economy. New assessment methods that allow the errors of traditional development strategies to be avoided are helping to change procedures and to distinguish appropriate policies from inappropriate ones.

The rigour of this type of assessment is leading to a significant change in emphasis. The Abdul Latif Jameel Poverty Action Lab (J-PAL), which is located at the Massachusetts Institute of Technology (MIT), is a key actor in this process. The need for more and better evidence for optimising development policies is now recognised by the World Bank and the Inter-American Development Bank (IDB), the French Development Agency (AFD) and the United Kingdom’s Department for International Development (DFID), the Gates, Hewlett and MacArthur Foundations, the Indian government’s Planning Commission and the Moroccan government’s Council of Ministers.

The Poverty Action Lab (J-PAL) funds randomised trials of poverty relief projects and programmes. It does not support any programme in particular. Instead, the objective is to emphasise that international and local organisations are responsible for generating knowledge on which policies work and which do not. This knowledge is an international public good that should be available to international organisations and local development agents.

To achieve this, J-PAL, in collaboration with academics and local NGOs, gathers a large amount of field data that are then subjected to econometric analysis to determine the impact of each of the assessed programmes. The results are then disseminated to national and international organisations. Through teaching, J-PAL helps other organisations to increase their ability to undertake randomised trials.

J-PAL was founded in 2003 by three MIT economists, to contribute to substantially improving the living conditions of the millions of poor people in the world. It provides the international development community with specific, scientifically proven information on poverty reduction strategies that have a real, substantial impact on the lives of the poorest people. To achieve this, J-PAL uses an innovative tool in the field of development: randomised trials for assessing social policies.

Most methods for assessing development policies and programmes are flawed as they do not identify whether an action caused a specific result. In other words, traditional assessment methods have causality problems, because they do not consider all of the variables that influence a specific result. For example, if students are bought textbooks and their academic results have improved a year later, the improvement may be due to the books, but it may also be due to the fact that more highly qualified teachers have been hired. How can we assess the impact of buying books on its own? In this case, to verify the impact of the book purchase, books must be given to a number of randomly selected schools and the academic results compared with those found in another group of schools that did not receive any books. When this method is applied to a large number of schools, idiosyncratic differences such as new, more highly qualified teachers are rendered irrelevant. This is the basis for the randomised trials that J-PAL undertakes to rigorously assess development programmes.

In general, J-PAL undertakes three types of activities:

• It carries out randomised trials to assess the impact of development programmes.
• It disseminates the results of its research to policy planners in order to influence their decisions.
• It promotes the use of the randomised trial method, so that political decisions can be taken on the basis of verified evidence. This increases the capacity of decision-makers to assess their own programmes and to use rigorous evidence in the design and implementation of policies.

Currently, the Poverty Action Lab has three centres: one at MIT in the USA, one at the Paris School of Economics in France and one at the Institute for Financial Management and Research in India. Although the first laboratory was set up at MIT, the expansion of the network into Europe and Asia has increased the range of disciplines and geographic areas that are covered and has heightened the Lab’s visibility and the ability of policy makers and donors worldwide to guide policies on the basis of verified evidence. The three centres combined work in a total of 21 countries with approximately thirty of the best researchers in the sector.

Many trials are undertaken by J-PAL centres. Since 2007, work has been carried out on approximately 20 randomised trials in over 8 countries, including those assessing the following:

• The impact of educating women and men on fertility and infant mortality in Indonesia.
• Teaching innovations that improve state school students’ basic maths and reading skills, via a pilot programme by the NGO India Pratham.
• The conditions for transferring funds to improve attendance in Moroccan primary schools.
• Participation and accountability in Uganda’s healthcare system.
• An educational psychology programme for unemployed young people in France.

Several studies compiled by J-PAL are bringing about changes in development policies. In 2008, for example, it made significant advances in the promotion of its educational results through the campaign “Deworm the World”. An event it organised at a meeting of the World Economic Forum showed the benefits of introducing measures to deworm students in schools as a plan to increase access to education. This campaign has had a clear impact and the World Health Organization and World Bank are using the results provided by J-PAL to encourage national governments to approve policy measures for introducing deworming programmes in schools. National deworming programmes are already being implemented in Madagascar, Yemen and Zambia. Bangladesh and Kenya are likely to begin such programmes in the near future. Furthermore, there is clear evidence that the following studies have brought about development policy changes in several national governments and international institutions: a study of free donations of mosquito nets in Kenya and a study of basic literacy teaching for children who are not in the school system and/or illiterate children in schools in India.

In addition, J-PAL’s efforts to boost the rigour of policy designing and the use of randomised trials are increasing the rigour with which many International organisations, NGOs and governments assess their development programmes. Many randomised trials are being undertaken or supervised by bodies that have been trained by J-PAL. Among others, the governments of France, the United Kingdom, Morocco and India, the Millennium Challenge Corporation, Save the Children, Freedom from Hunger, BRAC, and the International Rescue Committee (IRC) have begun or are about to begin drawing up their first randomised trials for social programmes. All of these bodies are supported and guided by J-PAL. The Swedish and UK development agencies are also reviewing their criteria for assessing programmes in order to increase their rigour. The World Bank is now carrying out more randomised trials than ever before, many of them in collaboration with J-PAL.

J-PAL organises week-long courses on randomised trials every year. These courses are run in Cambridge (USA), Chennai (India), Paris (France) and Bali (Indonesia). The 2008 edition had over 160 attendees from 20 countries, including representatives of organisations such as the Aga Khan Foundation, the Azim Premji Foundation, the Millennium Challenge Corporation, the Pratham India Education Initiative, the Pakistan Poverty Alleviation Fund, Save the Children, UNESCO, the US Department of Labor and the World Bank, and ministerial representatives from Indonesia, Jamaica, Madagascar, Niger, Peru and Sri Lanka.

2003; ongoing