Helping the Ultra-Poor Use Microcredit in Murshidabad, India

Many interventions designed to reduce poverty fail to improve the lives of the poorest of the poor, suggesting a need for targeted programs. Working with Bandhan, a microfinance institution based in West Bengal, we conducted baseline and post-program surveys of approximately 1,000 "ultra-poor" households. To be considered ultra-poor, households must meet a set of criteria including having an able-bodied female member and not being associated with any microfinance organization. In addition, households must meet three out of the five following requirements: primary source of income is informal labor or begging, land holdings are below 20 decimals, household owns no productive asset other than land, no able bodied males in the household, and school-aged children work instead of attending school.

Half of the sampled households were randomly selected to participate in Bandhan's "Targeting the Hard-Core Poor" (THP) program. The theory behind the program is that the poor are trapped in extreme poverty by their lack of assets and inability to use financial intermediation to overcome their lack of assets. Bandhan purchases and distributes assets such as livestock and inventory to program beneficiaries. The grants are also used to finance other inputs, such as fodder and sheds for livestock. Bandhan staff meet with participants to select the livelihood option best suited to the household. In the study sample, 90% of beneficiaries chose livestock, receiving either 2 cows, 4 goats or 1 cow and 2 goats. The value of the asset transferred was approximately US$100, equivalent to 20% of baseline hosuehold expenditure.

Over the next eighteen months, Bandhan staff meet weekly with beneficiaries. Staff provide information and training on topics related to the household's enterprise as well as broader social and health issues. Beneficiaries are also required to save Rs. 10 per week at these meetings. Eighteen months after receipt of the asset, the beneficiaries are "graduated" to microfinance and become eligible for regular microfinance loans provided by Bandhan.

2007-02-01

2011-11-30

Key outcomes include income, consumption, assets, nutrition, health, food security, financial behaviors, and labor supply.

We collected data for the study in two waves of surveying. The initial wave, spanning February 2007 to March 2008, was conducted among approximately 1,000 ultra-poor households. The survey consisted of a household module covering income, consumption, and various other features of the household, as well as an adult module capturing information on labor supply, time use, and health.

Following completion of the baseline survey, households were randomly selected to participate in Bandhan's THP program. The program, consisting of direct asset transfers (mostly livestock), inoculation of savings habits and integration into microfinance groups, is designed to overcome barriers to upward mobility. The training component is intended to address the assumption that the poor either lack the confidence to escape poverty or that they lack sufficient human capital to make optimal use of assets. Regular meetings which provide both encouragement and training related to enterprise development are also part of the program.

Randomization was done remotely by the research team and selection was stratified on hamlet. A total of 991 households were surveyed at baseline, of which 512 were randomly selected for program participation. Households were re-surveyed in a first endline (18 months after the asset transfer) and a second endline (30 months after the asset transfer). For the third endline, we finished the 1st round of fieldwork on December 22nd 2014 with 868 households and 2090 adult interviews. In January 2015 mostly we are targeted the migrated (21 clients), refused (23 clients) and also those interviews captured during second endline (26 clients).

We estimate the causal impact of the THP program on a number of household and individual outcomes by calculating the mean difference in outcomes between those who were offered program participation and those that were not, controlling for hamlet-level fixed effects.

Randomization done in office by a computer.

Household.

No

991 households

Approximately 2,000 adults in 991 households.

429 households treatment, 466 households control.

This study reports the results of a randomized impact evaluation of a program designed to reach the poorest of the poor and elevate them out of extreme poverty. The program, which includes the direct transfer of productive assets (e.g. livestock) and additional training, was initially developed in Bangladesh, where it has reached thousands of beneficiaries, and is being piloted and studied in over seven countries. The results of this study, based on a pilot in India, indicate that this intervention succeeds in elevating the economic situation of the poorest. We find that the program results in a 15% increase in household consumption and has positive impacts on other measures of household wealth and welfare, such as assets and emotional well-being. Our results are consistent with the notion that the wealth transfer, in the form of asset distribution, directly increased consumption among beneficiary households through the liquidation of assets, but other sources of income, notably from small enterprises, appear to have contributed to the overall increase in consumption as well.

http://www.povertyactionlab.org/publication/targeting-hard-core-poor-impact-assessment