Credit plus Grant-Based Approach to Ultra-Poverty Reduction: An Experimental Evaluation

Initiative to fight ultra-poverty needs customized holistic design and targeting tool; without these features, many traditional development approaches are unlikely to generate a positive impact on the poorest though they have promising impacts on livelihoods of the poor community. Thus, BRAC introduced a holistic package consisting of soft loan and some grants for better-off ultra-poor—selected through a multi-staged targeting process. Experimental evaluation of this package shows that it increases female labour market participation by 32%, leading to a 21% increase in per capita monthly income. Moreover, it increases productive asset ownership—livestock, poultry, and cultivable land—and consumption expenditure. It also improves financial inclusion: substituting informal credits with formal ones to some extent and increasing savings. While exploring the intervention’s efficiency, we found that this approach generates benefits on average 4.59 times higher than its costs.

Researchers: Dr Narayan C. Das; Atiya Rahman; Anindita Bhattacharjee

Partners: BRAC

Timeline: 2016- 2018

Status: Completed

Contact: Atiya Rahman;



With an aim to eradicate extreme poverty, in 2002 BRAC’s Ultra-Poor Graduation (UPG) program originated a comprehensive graduation model including a multi-staged targeting process and a sequenced set of interventions which consisted of enterprise training, asset transfer, consumption stipend, handholding coaching for 18 months, and health mobilization. This grant-based approach has already been proven effective to enable the ultra-poor to achieve sustainable livelihood and socio-economic resilience to come out of poverty, however; it is very costly to scale up. The question is, thus, whether there is any way to reduce the cost of the graduation model without compromising the benefits.

A possible way is whether a hybrid of the two approaches—microfinance and grant-based support—can be an effective tool to address ultra-poverty. In this study, we try to address this knowledge gap by evaluating BRAC’s initiated intervention which replaced the grant-based graduation approach’s asset transfer component by soft loan for better-off ultra-poor households.

This study is relevant to SDG 1 (no poverty), particularly to ending poverty in all its forms everywhere.


From 2007–2016, BRAC’s UPG program implemented two support packages: a grant-based support package and a credit- plus grant-based support package. The idea of implementing different approaches under the graduation program is to address heterogeneity (in terms of livelihood opportunities, demographic characteristics, etc.) among the ultra-poor. Both the support packages target the ultra-poor but the target group of the latter is slightly well-off—called ‘’Other Targeted Ultra-Poor (OTUP)”—compared to the former—called “Specially Targeted Ultra-Poor (STUP).” Evidence on pre-intervention status shows that OTUP households had higher household income, asset accumulation, household welfare in terms of food and non-food expenditure, and food security on average compared to the STUP households though both household-groups belonged to bottom three community wealth ranks. The credit- plus grant-based support package—offered to the better-off ultra-poor—includes enterprise development training, soft-loans (conditional on investing it in the enterprises on which training is provided), weekly consumption allowance, handholding coaching through home visits, health care, community resource mobilization, and savings. Soft loans of a maximum of BDT 15,000 was provided for purchasing the assets themselves with lower interest (20%) compared to traditional microfinance interest rate (27%) and also two-month grace period was allowed for repaying the loan. The loan was repaid within 52 weeks.


Our objective is to explore whether a hybrid of soft loan coupled with some grants is effective for ultra-poverty reduction.


For evaluating the UPG program (both STUP and OTUP packages), we adopted a randomized control trial (RCT) for the 2016 cohort of the program. For implementing the 2016 cohort, BRAC identified 11 districts and selected 8 branch offices from each district. The research team randomly selected two branch offices from each district and assigned them to the control group. The rest 6 branch offices served as treatment areas. BRAC carried out the selection of STUP and OTUP households from these 88 branch offices: 66 treatment and 22 control branches. Baseline survey covered all finally selected STUP and OTUP households, however; this study only focuses on OTUP households. During the baseline, we surveyed a total of 8,973 OTUP households: 7,054 treatment and 1,931 control households. In a follow-up survey conducted in 2018, however, the research team attempted to survey all OTUP households from control branches and 66% of the OTUP households from the treatment branches due to fund constraints.

Findings and Recommendations

Evaluating intent-to-treat (ITT) and treatment-on-the-treated (TOT) estimates, we find that the intervention increases working-aged women’s total labour supply. The effect on total labour supply mainly driven by an increase in time devoted to activities related to livestock rearing. It is consistent with the fact that most of the participants mostly use a soft loan to purchase livestock and poultry. Results also show that working-aged male members increase their time devoted to livestock and poultry rearing. Further results indicate that the intervention increases per capita household income by 21% and the value of productive assets by 119%. The intervention also has positive effects on outstanding loans and savings after two years. The estimated effect on consumption expenditure is found to be positive and statistically significant, suggesting that the intervention increases the welfare of ultra-poor households. The benefit-cost ratio (BCR) of the program is estimated to be 7.11 which is higher than that of the grant-based approach. We compare our results with the existing results on the impact of the fully grant-based approach (Bandiera et al., 2017) and find that the impacts are quite similar for these two groups. Also, the benefit-cost ratio of the credit plus grant (7.11) based approach is higher than that of the grand-based approach (3.21). BRAC selected the ultra-poor households for the credit plus grant-based support package but they are relatively well-off compared to those that were selected for its fully grant-based approach. Hence our results imply that the credit- plus grant-based approach is effective for the ultra-poor that are relatively well-off. But whether this model can be effective for all types of ultra-poor in general needs to be studied further.