Ultra-Poor Graduation Pilot in Honduras

How to cite

Karlan, Dean; Thuysbaert, Bram. 2013. "Ultra-Poor Graduation Pilot in Honduras." AEA RCT Registry. AEARCTR-0000058. https://doi.org/10.1257/rct.58-1.0

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

Dean Karlan* , Bram Thuysbaert

* Primary investigator

Registry ID
AEARCTR-0000058
DOI
https://doi.org/10.1257/rct.58-1.0
First registered
Last updated
Status
Completed
Trial dates
2008-10-01 – 2013-05-31
Countries
Honduras (Latin America)
Registry record
http://www.socialscienceregistry.org/trials/58

Abstract

The Ultra Poor Graduation Pilot in Honduras is a part of the CGAP/Ford Foundation global initiative, currently being implemented in eight countries, that combines safety net programs with entrepreneurial activities to develop a graduation model for the poorest households. The model incorporates a comprehensive package of services delivered over two years: a productive asset (such as chickens or goats), consumption support, livelihood trainings, healthcare, and financial services. By investing in this multifaceted approach, the program strives to eliminate the need for long-term safety net services. A baseline survey was conducted for 2,400 households in October 2008, followed by the intervention. Two follow-up surveys were conducted in March-May 2012 and March-May 2013.

Intervention

The Ultra Poor Graduation Pilot in Honduras provides 800 treatment households in 40 treatment villages with a comprehensive support package. The households receive consumption support, in the form of a family garden, and training in one of two income generating activities: raising livestock (chicken or pigs) and crop (bananas or vegetables) production, or operating a pulperia (small grocery store). Participants are monitored throughout the process. Female heads of households are required to open a savings account at ODEF and are randomly assigned to one of two savings treatments. One savings group is incentivized with savings matching biannually equal to 50 percent of the average account balance while the second group receives monthly direct savings transfers. Both of these treatment groups receive savings incentives valued at 400 Lempiras (about $20 US).

Experimental design

IPA's partner organization for the Ultra Poor Graduation Pilot in Honduras, Plan/ODEF, works in 124 communities in the Lempira department. Out of the 124, eighty communities were selected to be a part of the pilot project. Plan/ODEF carried out participatory wealth rankings (PWR) in all 80 selected communities in order to identify the poorest households in each of the communities. During the PWR, villagers rated the economic status of all members of the community. Those households earning less than 600 Lempiras (approx. $30 US) per month were considered to be Ultra Poor.

The randomization of households into treatment and control consisted of two steps. First, half of the 80 villages were randomly selected to be control villages. None of the households in these control villages received any of the pilot project services. In the 40 treatment villages, we further randomized at the household level. Within each treatment village, half of the eligible households were randomly selected as treatment households. Those eligible households not chosen for treatment comprised the control households. Comparing the control households in the treatment villages with those in the pure comparison villages will allow the study to measure spillover effects.

In the treatment households, the female heads of households were required to open a savings account at ODEF and were randomly assigned to one of two savings treatments. One savings group was incentivized with savings matching biannually equal to 50 percent of the average account balance while the second group received monthly direct savings transfers. Both of these treatment groups received savings incentives valued at 400 Lempiras (about $20 US).

Design details

Randomization method
Randomizaton was done in the office. We used a rerandomization procedure to ensure balance on a number of village-level (first step) and household-level variables (second step). The first-step randomization was stratified by municipio.
Randomization unit
villages, then within the treatment villages, control households and treatment households
Clusters
80 villages
Observations
2,400 households
Arms
800 households in control villages 800 control households in treatment villages 800 treatment households in treatment villages
Primary outcomes
Income, assets, consumption, school attendance of children, health and food security

Cited by

No citation data available for this registration yet.

Keywords