Abstract

The Graduation Approach, a model for holistic livelihoods programs, has been proven to have lasting impacts on poor families’ income, assets, food security, and mental health, but these programs can be expensive to implement. Innovations for Poverty Action is working with researchers in Ghana to evaluate the impact of different variations of the Graduation program—without coaching visits, with cognitive behavioral therapy, without additional services, and a cash transfer only—on household income, consumption, assets, psychological health, and other measures of well-being.

Policy Issue

Nearly 900 million people live on less than US$1.90 per day. Many of these families depend on insecure and fragile livelihoods. Recent research has shown that holistic livelihoods programs can have a wide range of benefits for these poor families, from increasing household consumption and income to improving food security and mental health. The successful Graduation Approach provides families with a range of services, including income-generating assets, training, access to savings accounts, consumption support, and coaching visits, making the model expensive for many developing country governments to implement. However, it may be possible to retain positive impacts of these programs at lower cost by providing a smaller set of services. This research investigates ways to make the graduation program both more scalable and more beneficial.

Context of the Evaluation

The study is spread across four administrative regions of Ghana: Upper East, Northern, Brong-Ahafo, and Ashanti. Researchers selected peri-urban communities from three districts in each region. These communities are largely within 30 minutes of the district capitals, but they remain reliant on agriculture and have significant populations of ultra-poor households.  IPA partnered with the international NGO Heifer International to implement the program, following the same model IPA and Presbyterian Agricultural Services used in 2011-2013.

Details of the Intervention

Innovations for Poverty Action is working with researchers to evaluate and compare the impacts of different versions of a livelihoods program on the lives of the ultra-poor in Ghana. To select the poorest members of participating communities, the project team will conduct a census of all individuals in the poorest neighborhoods in the selected communities and use a short set of questions (a proxy means test)  to identify the poorest households. Researchers will then randomly assign 7,696 households to six groups:

Full Graduation program: Households will be offered the full Graduation program, which includes six core components: productive asset transfer (worth approximately US$270 or enough to purchase five goats); consumption support (typically a cash transfer) to  reduce the likelihood that households immediately sell the asset; training on management of the asset; weekly coaching visits over a two-year period; access to a savings account; and access to health services by enrolling them in health insurance.

Graduation program without coaching visits: Households will be offered a modified version of the Graduation program that that was identical to the full program, except there were no coaching visits. This reduced the total program cost by approximately 30 percent.

Cognitive behavioral therapy (CBT): Households will be offered 12 weeks of group-based cognitive behavioral therapy followed by the full Graduation program. The sessions will be led by Psych Corps, a program of the Government of Ghana. During the sessions, counselors introduce participants to a systematic way of listening to thoughts, differentiating between helpful and unhelpful patterns. They will also teach behavioral methods of coping with anxious and sad thoughts, goal setting, and other skills to increase participants’ ability to take advantage of opportunities and to respond to adverse events.

Heifer model: Households will be offered a productive asset and training on how to manage the asset. This treatment arm was similar to the programs traditionally offered by Heifer International.

Heifer and CBT: Households will be offered the full Heifer model as well as the CBT sessions.

Cash only: Households will receive lump-sum, unconditional cash transfers equivalent to the value of the asset transfer (approximately US$270). Households are free to decide how to use the money.

Comparison group: Households will not receive any intervention.

Researchers will measure the impacts of the interventions on economic outcomes (including household consumption, income, and assets) and mental health outcomes (including mental distress, subjective well-being, and working memory). Researchers will measure mental health outcomes immediately after the CBT intervention (before the start of the full Graduation program), and they will measure economic and mental health outcomes two and three years after the asset transfer. By measuring both categories of outcomes, the research design will allow researchers to determine whether receiving cognitive behavioral therapy is critical to reducing depression or if improving incomes through cash transfers or the Graduation model is equally effective at improving mental health. Researchers will then be able to test if this improved psychological well-being allows individuals to more productively utilize their new assets.

Results and Policy Lessons

Project ongoing; results forthcoming.