BRAC's Ultra- Poor Graduation Program: An Experimental Evaluation

Last registered on March 31, 2022

Pre-Trial

Trial Information

General Information

Title
BRAC's Ultra- Poor Graduation Program: An Experimental Evaluation
RCT ID
AEARCTR-0008643
Initial registration date
December 08, 2021

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
December 10, 2021, 2:22 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
March 31, 2022, 11:27 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Region

Primary Investigator

Affiliation
Innovations for Poverty Action - Bangladesh

Other Primary Investigator(s)

PI Affiliation
Innovations for Poverty Action - Bangladesh
PI Affiliation
Northwestern University
PI Affiliation
Northwestern University
PI Affiliation
Yale University
PI Affiliation
Yale University
PI Affiliation
University of Maryland
PI Affiliation
BRAC

Additional Trial Information

Status
On going
Start date
2019-04-04
End date
2022-04-10
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
The Ultra Poor Graduation (UPG) Program was initiated by BRAC in 2002 as a grant-based intervention that encompassed various kinds of enterprise development support and subsistence allowance. When the pilot of the program was initiated, the program was designed to provide the ultra-poor with a grant-based support package which included productive assets, a weekly subsistence allowance, enterprise development training, close supervision/coaching through high-frequency home visits, and health supports. The second phase (2007) offered a credit plus grant-based support package, along with the existing grant-based support package while the third phase (2012) also offered support package to households in the climate change prone coastal and haor regions of the country. The latest modification (2017) was dividing the ultra-poor into three groups; group 1 consisting of elderly population receiving government aids and grants from BRAC, and groups 2 and 3 receiving enterprise and other support from BRAC.
With a sample size of 38902, the 2019 cohort, which is the baseline for our impact evaluation study, underwent some modifications in selection criteria of participants (groups 2 and 3 only) and design of interventions (partial asset grant or interest-free loan plus the provision of a savings match or savings messaging) to evaluate the following aspects pertaining to the UPG program:
• Overall impact of the new graduation program
• The marginal impact of matched contributions to a savings account on participants
• The behavioral impact induced by savings messaging
• Impact of the program on eligible non-participants.

With a sample size of 4000 households, the endline survey is expected to be completed by March 2022. However, due to the many challenges posited by Covid-19, the endline survey will deviate from previously intended design due to potential significant bias into our estimates. We will be dropping the treatment arms pertaining to savings messaging (T2 and T3) and the spillover control group (SC) and focus on the overall impact of the graduation program, especially in the context of an exogenous shock like Covid-19.
External Link(s)

Registration Citation

Citation
Alidaee, Hossein et al. 2022. "BRAC's Ultra- Poor Graduation Program: An Experimental Evaluation." AEA RCT Registry. March 31. https://doi.org/10.1257/rct.8643-4.0
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
The interventions are:

1. Enterprise Development Training
Prior to receiving enterprise development training, the finally selected beneficiaries select their enterprise of choice. In order to help the beneficiaries make a suitable selection, programme staff also discuss a number of issues with them including potential income and expenditure from the enterprise, advantages and disadvantages associated with managing the enterprise, loan repayment details (e.g. repayment rate, approximate size of installment, grace period), and other supporting services. Once the enterprise is decided, the beneficiaries are provided with specific enterprise development training. The training duration is three days, and the sessions are organized separately for members of Groups 2 and 3, with no more than 25 beneficiaries in each batch. Later, during the programme cycle, the beneficiaries receive another seven days of subject specific training.

2. Asset Transfer
Upon receiving the enterprise development training, Group 2 beneficiaries receive the assets from the programme (i.e. purchased by the programme staff). Depending on the types and amounts of assets provided, 30%-50% of the asset values (at the time of transfer) are recovered from the beneficiaries. Hence, in a sense Group 2 receives interest-free loans plus grants. Total values of the assets provided to group 2 members ranges from BDT 12,000 to BDT 18,000, and are expected to repay 50% of the final value. Group 3 beneficiaries are provided with conditional cash for purchasing the assets themselves. They repay 80% of the cash provided. The assets are transferred within a month of the training, unless the situation restricts it (in which case, the delayed transfer of asset has to be approved by the Head Office with an explanation for the delay).

3. Hands-on training through group and home visits
The beneficiaries are also provided with hands-on training bi-weekly throughout the programme duration (about two years), through group meetings and one-to-one home visits.

4. Savings Matching
Each beneficiary has a savings account opened with the programme. The programme staff encourage them to develop the habit of savings.
In case of emergencies or for increasing asset, beneficiaries can withdraw their savings (as well as the amount provided by the programme as matched savings) as per their needs with the Regional Manager’s approval.

5. Healthcare
The programme will continue providing the health care facility to TUP beneficiaries and their household members. For severe morbidity, BRAC provides with financial assistance, and for mild illness, the patients are referred to nearby health centers.

6. Community Mobilization
Community resource is mobilized through forming a committee called village social solidarity committee.
Intervention Start Date
2019-05-01
Intervention End Date
2020-12-31

Primary Outcomes

Primary Outcomes (end points)
Business activities of household members, expectations of the respondent about the future (children & initiating new business activities), financial assets and liabilities of households, empowerment of the respondent, crisis management at the household level, and food security.
Primary Outcomes (explanation)
Business activities of HH members: This includes questions about the business networks and activities of the respondent along with other HH members, seasonality of the respondent's earnings, any business activity that had to be stopped in the past year and any new business initiative that the respondent is interested in starting.

Financial Assets and Liabilities of the HH members: Questions about purchase/sale of land and business assets to pay for everyday expenses, savings motivation and obstacles on the ability to save, outstanding loans and sources of supplementary funds.

Empowerment of the respondent: Ability to influence decisions under various scenarios.



Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The current UPG program divided the ultra-poor into three groups; group 1 consisting of elderly population receiving government aids and grants from BRAC, and groups 2 and 3 receiving enterprise and other support from BRAC. For our impact evaluation study, we are focusing on groups 2 and 3, and have implemented two broad treatment arms; Regular program (RP) and Modified program (MP). RP includes partial asset grant or interest-free loan plus a savings match whereas MP includes partial asset grant or interest-free loan plus messaging on the importance of savings.
The sample is randomized at the branch, village, and household level, and ultimately forms 5 groups. RP includes the pure control group (PC), the group with the savings match (T1) and the spillover control group (SC). MP includes the groups with unmatched savings accounts, and messaging on the importance of savings along with suggested savings amount (T2) and only messaging on the importance of savings (T3).
The study is designed to answer the following research questions:
• What is the overall impact of the new graduation program? (T1 and PC)
• What is the marginal impact of matched contributions to a savings account on participants? Specifically, how much more does it induce them to save? Furthermore, how effectively can behavioral messaging act as a substitute for the savings match? (T2 and T3)
• How does the program improve participants’ access to risk sharing networks through informal borrowing and lending? Findings from research on previous UPG programme cohorts indicate that limited income may prevent households from participating in the risk sharing networks. By increasing income, if the program improves access to informal risk sharing networks, this would improve welfare beyond measures captured in prior evaluations. (T1 and T3)
• What is the impact of the program on eligible non-participants? Specifically, do participants provide extra support to non-participants in comparison to the general economic opportunity provided by other village members? (PC and SC)

BRAC has a total of 295 branches across the country and among them 100 branches were chosen via stratified random sampling. Out of these 100 branches, 50 branches were randomly given the RP and 50 MP. Under RP, 50% of the villages in each branch were in the control group (PC) and 50% in the treatment group (RPV). RPV was further divided into 40% of HHs per village being in the spillover control group (SC) and 60% as program participants (T1). Under MP (applicable only in baseline), villages in each branch were divided between two treatment groups (T2 & T3).
However, for our endline data collection and analysis, the effect size associated with the marginal impact of savings matching, savings messaging, and spillover are likely to be much smaller in the context of Covid-19 which may hamper the statistical power to identify an effect, and thus, we will be eliminating every evaluation except for the first, assessing the overall impact of the graduation program.
Experimental Design Details
Randomization Method
All randomizations were done on office computers.
Randomization Unit
100 BRAC branches were selected via stratified random sampling and divided randomly between RP (50%) and MP (50%).
Under RP, 50% of the villages per branch got allocated to PC, and 50% to T1 (60%) and SC (40%). Under MP, 50% of the villages per branch got allocated to T2 and 50% to T3.
For our endline survey, MP and SC will not be included.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
100 Branches
1032 Villages
Sample size: planned number of observations
38,902 respondents for baseline Approximately 4000 respondents for endline
Sample size (or number of clusters) by treatment arms
Pure Control (PC) = 281 villages (9805 HH)
Regular Treatment (T1) & Spill-over Control Group (SC) = 282 villages (6310 & 4202 HH)
Specific saving-message (T3) = 233 villages (9365 HH)
General saving-message (T4) = 236 villages (9082 HH)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Power: 70%, Effect size: 0.1. Effect size is Standard error (SE): Monthly household income (BDT) - ICC: 0.00000, SE: 0.00121 Total value of business assets owned (BDT) - ICC: 0.01526, SE:0.00194 Total savings of the respondents (average in BDT) - ICC: 0.01302, SE: 0.00182 Average amount of land owned - ICC: 0.0294, SE: 0.00266 Amount of outstanding loans of the respondents and their spouses (in BDT) - ICC: 0.04434, SE: 0.00343 Status of food availability: Always deficit (%) - ICC: 0.05483, SE: 0.00396 Decision-making power index (average of applicable indicators) - ICC: 0.06588, SE: 0.00451 Monthly income of the respondents (BDT) - ICC: 0.07521, SE: 0.00496
IRB

Institutional Review Boards (IRBs)

IRB Name
IRB of the BRAC James P Grant School of Public Health, BRAC University
IRB Approval Date
2021-12-29
IRB Approval Number
IRB - 14 December'21 - 042

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials