The General Equilibrium Effects of Graduation Programs: Experimental Evidence from Ethiopia
How to cite
Egger, Dennis; Vinez, Margaux; Loeser, John; Abate, Gashaw. 2024. "The General Equilibrium Effects of Graduation Programs: Experimental Evidence from Ethiopia." AEA RCT Registry. AEARCTR-0013236. https://doi.org/10.1257/rct.13236-1.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This entry covers the evaluation of the impact of Livelihood (LH) interventions in the context of Ethiopia’s rural Productive Safety Net Project (PSNP) using a nationwide clustered randomized controlled trial (RCT) to evaluate. This economic inclusion (EI) program provides livelihood services to a subset of PSNP public work households. It complements the transfers received as remuneration for the participation in public works by a sequenced package of training and coaching as well as a livelihood grant or facilitated access to subsidized credit. This evaluation will estimate the direct impact of the interventions on beneficiaries and will examine who benefits the most from them. It will investigate the value-added of additional government implemented life-skills training. Finally, it will study the indirect (spillover) effects of these interventions on the local economy.
Intervention
We evaluate the impact of Livelihoods (LH) interventions in the context of Ethiopia’s rural Productive Safety Net Project (PSNP). This EI program provides livelihood services to a subset of PSNP public work households. It complements the transfers received as remuneration for participation in public works by a sequenced package of training and coaching as well as a livelihood grant or facilitated access to subsidized credit. The program targets woredas with timely payments in past iterations of the PSNP program, and aims to cover all major regions within Ethiopia. A total of 34 woredas were selected to implement LH interventions, and the project was designed to cover around 40,000 households in the first year of implementation, when the IE will be implemented.
Within participating woredas, we randomize LH interventions across kebele – administrative units of approximately 500 households each. In kebeles selected for LH, public works clients are invited to express interest in participating via a newly introduced targeting or “profiling” exercise. Clients who express interest in participating go through a targeting exercise which assigns each interested client household a wealth ranking, credit score, and commitment score. Households with the lowest wealth rank (20% of the livelihood caseload) are selected for the grant track (300USD$ in cash), while households with the highest credit scores are selected for the credit track (80% of the livelihood caseload). Households are encouraged to diversify their activities and can choose between an on-farm and off-farm pathway, which determines the type of technical training they will receive. Development Agents (DAs) and Community Facilitators (CF) then form groups of beneficiaries and deliver training in financial literacy, technical training (depending on the pathway chosen), business skills, business planning over a four to five months period. In selected kebeles, they also receive life-skills training. After developing a business plan, clients are linked with microfinance institutions to apply for credit or receive the grant. Finally, they receive weekly coaching sessions organized by CFs over the 12 months following the reception of the grant/application for credit.
Experimental design
348 Kebele in our survey sample, randomnly selected among participating kebele, stratified at the Woreda level (30 Woreda), into:
Control: 116 Kebele
Livelihoods without Life Skills Training: 116 Kebele
Livelihoods with Life Skills Training: 116 Kebele
In each Kebele, households in the "Missing Middle" stratum (described below) were randomly assigned to LH-Credit and Control. We sampled 5 households in each of these groups, and as a consequence in our survey sample:
Missing Middle LH-Credit: 5 households/Kebele
Missing Middle Control: 5 households/Kebele
Design details
Cited by
No citation data available for this registration yet.