A Multifaceted Program Causes Lasting Progress for the Very Poor: Evidence from Six Countries

A multidimensional intervention program improves the lives of the extreme poor with effects lasting up to a year later, including increasing women’s empowerment in the short-term.


The question of how to best fight global poverty has been a dilemma of our generation. Various initiatives, including the United Nation's Millennium Development Goals, have targeted efforts toward eradicating extreme poverty and hunger, and yet specific international development programs have had limited success and mixed results. This study tests the effect of combining several different interventions to see if an integrated approach could lead to lasting changes for the poorest of the poor - particularly increasing consumption and achieving stable improvements in their well-being. The "Graduation" program, originally designed and implemented by BRAC (a large Bangladeshi NGO), provides a holistic set of services to the poorest households in a village. Based on a theory that says the extreme poor need a “big push” to escape from the poverty trap, the Graduation program gave participants an initial grant of a productive asset and  support with a combination of other services. The researchers sought to test whether or not this multidimensional approach is effective and cost-effective, and if it could be scalable and applicable to different contexts and cultures. Between 2007 and 2014, the extreme poor in six countries were randomly assigned to either the treatment group, participation in the Graduation program, or the control group, no intervention, to evaluate the effects of this multidimensional approach.


Participation in the Graduation program showed significant results for the poorest of the poor; compared to the control group, the program participants reported positive effects lasting at least a year after the implementation ended. The results pooled across all countries and sites are as follows:

  • At the end of the program, participants demonstrated positive improvements on all ten of the primary outcome measures: consumption, food security, productive and household assets, financial inclusion, time use, income and revenues, physical health, mental health, political involvement, and women's empowerment.
  • The per capita consumption increased by roughly 5% of the control group mean, from $78.80 to $83.35.
  • One year after the conclusion of the program, the effects on economic gains remained: productive asset value increased by 15% measured at the end of the program, and 14% measured one year after compared to control group means; household savings increased by 156% of the control mean at the end of the program and 96% of the control mean  one year after.
  • After a year, the gains in physical health and women's empowerment declined and were no longer statistically significant.
  • Although there were variations between the countries, the results were not driven by any one country and the program appeared to be effective in most places.

In short, the study found that both at the conclusion of the program and lasting through the next year, the Graduation program participants were earning more income and achieving lasting improvements in their well-being. The study concludes that while more research can be done to optimize the design and implementation of such programs, a multifaceted “big push” to helping the ultra-poor escape the poverty trap led to sustained increases in consumption and income, and should be further considered for creating sustainable and cost-effective impact.


This study evaluated the effects of a multidimensional Graduation program designed to help the extreme poor improve their well-being by establishing sustainable self-employment activities. The program was conducted in six countries - Ethiopia, Ghana, Honduras, India, Pakistan, and Peru - with a total of 10,495 households (21,063 adults). In three countries, randomization occurred at the household level so that half of the eligible participants in each village were randomly assigned to the program with the other half serving as the control group. In the other three countries, first villages were selected to be either treatment or control and then within the treatment villages eligible households were randomly selected for treatment. In all cases, researchers identified eligible households – the poorest of the poor – through a participatory wealth-ranking process. The researchers conducted a baseline survey with all eligible participants prior to the start of the program. Two end-line surveys, one at the end of the intervention and one a year after that, captured the longer-term effects of the multidimensional Graduation program. The intervention provided the following assistance to its participants to start a productive self-employment activity:

  1. Productive Asset Transfer - one time gift of a productive asset chosen from a list by the participants (e.g. agriculture or livestock)
  2. Consumption Support - temporary transfer of food or cash to stabilize consumption and discourage participants from eating or selling the given asset
  3. Skills Training - technical skills training and support on managing the chosen livelihood
  4. High-frequency Home Visits - regular interaction with field officers throughout the 2-year program for encouragement and accountability
  5. Savings - access to a savings account and/or a deposit collection service to help cope with shocks
  6. Health Services and Life Coaching - provision of, or access to, health education (health, nutrition, and hygiene training), basic health services, and/or life-skills training.

Related GAP Studies