The role of coaching within bundled ultra-poor graduation programmes: Experimental evidence from Bangladesh

Last registered on July 08, 2022
View Trial History

Pre-Trial

Trial Information

General Information

Title
The role of coaching within bundled ultra-poor graduation programmes: Experimental evidence from Bangladesh
RCT ID
AEARCTR-0009618
Initial registration date
July 06, 2022

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
July 08, 2022, 10:22 AM EDT

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

Locations

Country
Bangladesh
Region

Primary Investigator

Name
Rocco Zizzamia
Affiliation
University of Oxford
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Stefan Dercon
PI Affiliation
University of Oxford
Contact Investigator
PI Name
Ashley Pople
PI Affiliation
University of Oxford
Contact Investigator
PI Name
Munshi Sulaiman
PI Affiliation
BRAC Institute for Governance and Development
Contact Investigator
PI Name
Imran Matin
PI Affiliation
BRAC Institute for Governance and Development
Contact Investigator
PI Name
Narayan Das
PI Affiliation
BRAC Institute for Governance and Development
Contact Investigator

Additional Trial Information

Status
On going
Start date
2016-04-01
End date
2023-03-01
Keywords
Behavior, Welfare
Additional Keywords
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
We study the effects of a "big-push" program providing large asset transfers, coaching and other support to very poor women in rural Bangladesh. We focus in particular on the role of coaching within the programme, using an RCT implemented in 2016 to test whether changing the size and nature of the coaching element has lasting impacts on programme effectiveness.

Big push interventions have received much scholarly attention over the last decade, with most of these studies finding large, positive and enduring impacts of the programme on outcomes ranging from health and education, to asset holdings and political involvement. However, since these interventions are really a composite bundle consisting of an asset transfer, consumption support, savings facilities, and regular training/coaching, less is known about what is driving impact and through what mechanisms. At the same time, a major draw-back of these programmes, with implications for scaling, is that they are expensive and logistically burdensome to implement. This study seeks to understand the role that the form and intensity of the coaching element of BRAC's Targeting the Ultra-Poor (TUP) programme has in explaining some key outcomes, and the mechanisms through which coaching has an impact on these outcomes.
External Link(s)

Registration Citation

Citation
Das, Narayan et al. 2022. "The role of coaching within bundled ultra-poor graduation programmes: Experimental evidence from Bangladesh." AEA RCT Registry. July 08. https://doi.org/10.1257/rct.9618-1.0
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)
In 2016, BRAC implemented a forth phase of their Targeting the Ultra-Poor programme (TUP). The programme consisted of a one-off transfer of assets to ultra-poor women in Bangladesh, complemented by regular coaching through home visits and/or group meetings, consumption support, access to savings facilities and other social services. In this phase of the TUP, BRAC introduced experimental variation in the intensity and mode of delivery of the coaching element of the graduation model.

The traditional model has been built with substantial coaching of recipients. This coaching aims to support the beneficiaries in their day-to-day actions, providing moral support and simple advice leading to behavioural and attitudinal change. However, this coaching is expensive. For this reason, BRAC experimented with implementing variations of the TUP model in which the intensity and mode of delivery of the coaching element was modified. Specifically, three variations were tested, corresponding to three treatment conditions:

T1: All transfers accompanied with weekly ``home visits” by staff for individualized coaching
T2: All transfers accompanied with fortnightly ``home visits” by staff for individualized coaching
T3: All transfers accompanied with fortnightly ``home visits” by staff for individualized coaching and fortnightly group meetings

Among the three coaching variations the weekly visit (T1) has the highest intensity and operational cost. This variant represents the approach to coaching that was used in BRAC TUP programmes prior to 2016. The second treatment arm (T2) is designed to test whether reducing the coaching intensity by half (from weekly to fortnightly home visits) results in a reduction in impact (on our primary outcomes -- see Section \ref{measurement}) that is higher than the reduction in operational costs. Finally, the third treatment (T3) includes a fortnightly group meeting along with fortnightly household visits, resulting in a variation in coaching intensity with operational costs somewhere between T1 and T2. Besides the cost-effectiveness aspect, the group meeting is also a potential way of creating social networks among the beneficiaries. In addition, group-based coaching may be complementary to one-on-one house visits in other ways, for instance, representing a potentially more interactive and engaging approach to hard- and soft-skill acquisition, a more binding form of commitment device, and allowing recipients to learn from the experiences of other recipients. Importantly, T3 is closest to the approach to the coaching element in the TUP design which was adopted by BRAC from 2017 onward.
Intervention Start Date
2016-08-15
Intervention End Date
2018-12-15

Primary Outcomes

Primary Outcomes (end points)
Household consumption, household asset holdings
Primary Outcomes (explanation)
See pre-analysis plan (attached) for details.

Secondary Outcomes

Secondary Outcomes (end points)
Income, labour supply, social inclusion, locus of control, depression, "hard" skills, health, investments in children's education, savings, access to credit, informal risk sharing, access to government and non-government social services
Secondary Outcomes (explanation)
See pre-analysis plan (attached) for details

Experimental Design

Experimental Design
In its fourth phase of the Ultra-Poor Graduation (UPG) programme (starting in 2016) BRAC implemented an RCT in Bangladesh to test three variants of the programme in which the form and intensity of the coaching element was experimentally varied. A baseline survey was conducted in 2016 prior to rolling out the randomized interventions. Baseline surveys were completed by 8,468 recipients and their households. The baseline survey drew a sample of between 50 and 200 eligible ultra-poor households from each branch. Our analysis of this RCT will be based on an endline survey on the beneficiaries and an additional sample of non-beneficiaries. Data collection on this endline survey began on the 7th of June 2022, and is expected to be completed within two months -- by early-to-mid August 2022.

Figure 1 (see analysis plan) summarizes the RCT design. The study covers 88 branch offices (the smallest unit in BRAC’s administrative structure) from 11 districts. These 88 branches were then randomly assigned into three treatment groups and a control group of 22 branches each (stratified by district) where

a) T1: All transfers accompanied with weekly ``home visits” by staff for individualized coaching (n = 2,101 at BL)
b) T2: All transfers accompanied with fortnightly ``home visits” by staff for individualized coaching (n = 2,358 at BL)
c) T3: All transfers accompanied with fortnightly ``home visits” by staff for individualized coaching and fortnightly group meetings (n = 2,297 at BL)
d) Control (n = 1,712 at BL)
Experimental Design Details
Randomization Method
Randomisation using computer assignment
Randomization Unit
Randomisation was undertaken at the branch level. Of 88 branches, 22 were assigned to each treatment arm/control. The branch is the smallest unit in BRAC’s administrative structure. Each Branch services several villages.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
88
Sample size: planned number of observations
8,468
Sample size (or number of clusters) by treatment arms
T1: n = 2,101 at BL
T2: n = 2,358 at BL
T3: n = 2,297 at BL
Control: n = 1,712 at BL
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
CUREC-DREC
IRB Approval Date
2021-04-01
IRB Approval Number
ECONCIA21-22-12
Analysis Plan

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

Request Information

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