Escaping Poverty: Exploring the Link Between Mental Health and Economic Productivity

How to cite

Barker, Nathan; Bryan, Gharad; Udry, Christopher; Ofori-Atta, Angela; Karlan, Dean. 2022. "Escaping Poverty: Exploring the Link Between Mental Health and Economic Productivity." AEA RCT Registry. AEARCTR-0008915. https://doi.org/10.1257/rct.8915-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Nathan Barker* , Gharad Bryan , Christopher Udry , Angela Ofori-Atta , Dean Karlan

* Primary investigator

Registry ID
AEARCTR-0008915
DOI
https://doi.org/10.1257/rct.8915-1.0
First registered
Last updated
Status
On going
Trial dates
2016-04-21 – 2022-09-30
Countries
Private
JEL codes
I15, I31, O12
Registry record
http://www.socialscienceregistry.org/trials/8915

Abstract

The Escaping Poverty (EP) Project is a multi-arm randomized controlled trial in Ghana that studies the link between mental health and efforts to improve economic productivity for those in extreme poverty. The main arms in the study are as follows: a 12-week cognitive behavioral therapy (CBT) program; the same CBT program followed by a multi-faceted anti-poverty asset transfer and training program; the multi-faceted program without CBT; a one-time cash transfer equivalent to the cost of the asset and training in the multi-faceted treatment arms; and a control group.

Intervention

The EP Program was implemented in twelve districts in five regions of Ghana: Northern, Upper East, Bono, Bono East, and Ashanti. Innovations for Poverty Action, a research NGO, and Heifer International, an NGO focused on livelihood development, are overseeing the implementation and data collection of the program. In total, there are several program components, each of which is administered to only a subset of the households. The EP Program consists of the following components:

Cognitive Behavioral Therapy: individuals enrolled in CBT participated in a twelve week program that focused on productive thinking, dealing with negative thought distortions, setting goals and time management, and related topics. Individuals met once a week for approximately 90 minutes at a time in a group of roughly 10 people from the same community, all of the same gender. They were led by one CBT counselor and one assistant. The counselor in each community was someone with a bachelor’s degree or higher, in about half of the cases with their degree in psychology. These counselors received two weeks of classroom training from an IPA staff who designed the CBT program and one week of practice with individuals who agreed to attend a part of the curriculum.

Productive Asset transfer and training: in each district, IPA and Heifer staff met with government officials in the district assemblies, the ministry of food and agriculture, and business and enterprise development to discuss livelihoods that a low-income individual with no previous technical skills could manage in the area and where there would be a sufficiently large market for their goods. In each district, the staff developed a list of 4-5 assets for individuals to choose from, in each case with the assets worth approximately $280 USD. Common assets include goats, inputs for making shea butter, and seeds and fertilizer for growing crops.

Staff with Heifer or IPA met with chosen individuals in each community, informed them they were being chosen into the program, and gave them a list of the productive assets to choose from. Immediately prior to receiving the asset, they attended a training that lasted on average two weeks, focused on how to properly and profitably manage the asset.

Access to insurance and savings: all individuals who receive the asset transfer also had bank accounts opened for them at Ghana National (GN) Bank. They were also enrolled in the government’s health insurance, the National Health Insurance Scheme. These program components are used to help the households build wealth, and to be better handle any negative shocks that might limit their ability to properly manage their productive assets.

Consumption support: a subset of households also receive consumption support in the form of monthly cash transfers to the GN Bank account opened for them. Each transfer is worth approximately $7, and is supplied in order to ensure that each household is able to meet their most basic needs, thus discouraging the immediate sale of their productive assets.

1-on-1 visits: in some communities, a subset of households received 1-on-1 visits from Heifer or IPA staff. The staff checked in with them about any issues they might be having with their productive asset and offered guidance or answer questions as needed. These staff are also meant to shift aspirations, and help the households overcome any temptation they may have to liquidate their assets or decrease their time worked.

Group visits: in other communities, all individuals who receive the asset transfer were encouraged to attend group trainings, overseen by Heifer staff. These trainings offered a blend of asset-specific and more general training, focusing both on management of the asset, as well as skills like budgeting or managing finances. These group trainings were similarly meant to have an aspirational component.

Cash transfer: a subset of households received a one-time cash transfer to a GN Bank account opened for them. The asset transfer was worth $340 USD, the value of the asset transfer and the consumption support to graduation program households. They were given no rules or guidance about how to use this cash transfer, leaving them free to use the money as they see fit.

Control group: no goods or services were provided to the control group.

Experimental design

Our Experimental Design Entailed the following Steps

(1) Administered census to all households in potential communities given by District Assemblies. Communities were chosen on the basis of )(a) being sufficiently large, (b) having high levels of poverty, (c) being accessible by road year-round, (d) not having a "graduation" program present.
(2) Identified 258 communities as eligible if they had at least 45 compounds. Randomly chosen households from the 40-poorest compounds to be potentially in our sample.
(3) randomized each of 258 communities into: (a) pure control, (b) pure CBT, or (c) full program communities
(4) Randomly selected 17 of 40 households to administer a baseline survey to in pure control communities, 20 in pure CBT communities, and 40 in full program communities.
(5) Administered Baseline Survey
(6) Randomly assigned active communities (b and c, in 1) to have CBT for men or women
(7) Randomly assigned CBT treatment status to households
(8) Implemented CBT
(9) Randomly assigned (but did not announce to anyone in Ghana office households to full treatment status, conducted below
(10) Administered CBT endline survey
(11) rolled out full economic program
(12) targeted endline survey: 2-3 years after end of full programs

In total, we have the following treatment cells:
PC1: Pure control (no intervention, no programs in community
CBT1: Control household in community in which only CBT (ie no other program activities) took place
CBT2: Offered CBT in community in which only CBT (ie no other program activities) took place
H1: Control household in community where Heifer program took place
H2: CBT only in Heifer community
H3: Received Cash in Heifer community
H4: Received Asset + CBT + Training, consumption support in Heifer community
H5: Received Asset + CBT + Training, no consumption support in Heifer community
H6: Received Asset + no CBT + Training, consumption support in Heifer community
H7: Received Asset + no CBT + Training, no consumption support in Heifer community
G1: Control household in community where graduation program took place
G2: CBT only in graduation community
G3: Received Cash in graduation community
G4: Received Asset + CBT + 1-on-1 household visits, consumption support in graduation community
G5: Received Asset + CBT + 1-on-1 household visits, no consumption support in graduation community
G6: Received Asset + no CBT + 1-on-1 household visits, consumption support in graduation community
G7: Received Asset + no CBT + no 1-on-1 household visits, consumption support in graduation community
G8: Received Asset + CBT + 1-on-1 household visits, no consumption support in graduation community
G9: Received Asset + CBT + no 1-on-1 household visits, no consumption support in graduation community

Design details

Randomization method
Done on computer, for each randomization we picked a list of covariates we wanted to achieve balance on, re-randomized 10,000 times, compared values on the covariates across treatment units, and picked the randomization with the largest minimum p-value.
Randomization unit
We have multiple units of randomization: (1) community-level (into active, CBT only, pure control) (2) individual-level (for observations to omit from our sample in pure control and CBT only communities, to maximize power per dollar spent) (3) community-level CBT assignment to male or female-only (4) individual-level CBT assignment (4) community-level assignment of asset transfer program to "graduation" or "Heifer" (5) randomization into-specific treatment arm
Clusters
258 communities
Observations
In the CBT analysis, 8,000 individuals In the main economic analysis, approximately 11,000 individuals in 7,700 households
Arms
Our sample size for CBT Status, at the individual level: Received CBT: 1,290 individuals Control: 5,937 individuals Our sample size by group (defined above), at the household level: PC1: 1649 CBT1: 200 CBT2: 200 H1: 568 H2: 173 H3: 464 H4: 205 H5: 205 H6: 205 H7: 500 G1: 371 G2: 425 G3: 304 G4: 205 G5: 205 G6: 700 G7: 700 G8: 205 G9: 205
Primary outcomes
In our initial data analysis of the outcomes on CBT, our primary outcomes consisted of five indices: mental health, perceived physical health and effects on labor, socioemotional skills, cognition, and perceived economic status. [We will update our PAP upon completion of our planned analysis for the subsequent post-CBT study]. For each, we will construct a z-score index, using the procedure in Kling, J.R., Liebman, J.B. and Katz, L.F., 2007. Experimental analysis of neighborhood effects. Econometrica, 75(1), pp.83-119.

Cited by

No citation data available for this registration yet.

Keywords