New impact studies on reaching the ultra poor

In a previously posted video, FAI’s Jonathan Morduch talks about providing the “ultra poor” – people who live on $1.25 a day or less – with financial services.

As Jonathan said, BRAC’s program, Challenging the Frontiers of Poverty Reduction – Targeting the Ultra-Poor (CFPR-TUP), has been a trailblazer in reaching extremely poor households. At the same time, the question of the program’s impact must be taken up more rigorously. BRAC has made valuable efforts to measure the impact of this program, and published several working papers providing insights into whether and how CFPR-TUP works. A new working paper came out in June 2010, investigating the long-term impacts of the program with panel data.

These evaluations, however, did not rely on a randomized controlled trial methodology (RCT), and are therefore susceptible to selection and other biases. It’s hard to know for sure how the beneficiaries would have fared in the absence of the program, and to deduct the net impact of being a beneficiary of CFPR-TUP. Happily, new CFPR-TUP-inspired programs operating in various parts of the world built in RCT-based impact evaluations from the get-go.

This July, I had the opportunity to visit one such program run by SKS in the Indian state of Andra Pradesh. The SKS program kept the essence of BRAC’s CFPR-TUP idea. As with BRAC, SKS emphasizes social development activities (for example, basic literacy education) and the transfer of productive assets (such as buffaloes, cows, goats, or funds to start a small business). SKS, however, has adapted BRAC’s idea to the local context in several ways. One of them is that there is no village committee, but beneficiaries meet weekly with each other (and an SKS staff member at first) to discuss their problems and provide mutual support. All training materials have also been re-designed for the beneficiaries that SKS targets and enrolls. For example, the clients I visited in Medak district, seen in the picture during the closing “prayer,” maintain a “rice bank” (the tin box in the picture). Members save one handful of rice every time they cook, and can later borrow from the rice bank when needed, repaying in rice with no interest. This simple mechanism helps them avoid buying rice in dire times, often at a very high price. Because being ultra-poor is not only about having low income, but also about having irregular income, this consumption smoothing mechanism is highly valued by project participants.

For the RCT-based impact evaluation, 38 villages were randomly assigned to receive the program or not to receive the program (until after the evaluation). All eligible and willing households in a village assigned to receive the program participate, while all households in villages assigned to the control group must wait until the end of the evaluation period to be able to engage with the program. Villages, not households, were chosen as the level of randomization to avoid possible spillover of the program’s impacts from participating to non-participating households within the same village. Spillovers would undo the advantage of the random assignment and likely reduce the measured impact of the program. Spillovers between villages are less likely since they are more distant from each other and their level of economic and social integration is much lower than that of households within a village.

The first group of beneficiaries of the SKS program graduated in May and June 2009. Shamika Ravi of the Indian School of Business and our own Jonathan Morduch collected data from 1,066 households in all 207 villages. The analysis of impacts immediately after the program is ongoing, and a second survey is scheduled to happen soon in order to measure longer-term impacts. Both will be posted on this website – stay tuned.