Benchmarking Heterogeneous Programs
How to cite
Magruder, Jeremy; Wiseman, Eleanor. 2022. "Benchmarking Heterogeneous Programs." AEA RCT Registry. AEARCTR-0009051. https://doi.org/10.1257/rct.9051-1.0
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
We evaluate two interventions: a cash transfer program in urban Goma, D.R.C., and a soft skills, savings, and training intervention in rural Walungu, D.R.C. The objectives of the study are to evaluate which program is a more impactful use of aid dollars. However, as the programs are heterogeneous in location and implementation, they are anticipated to have their strongest impacts on different outcome variables.
This generates challenges in making fair comparisons through a single analysis plan. We propose to use a flexible, data-driven analysis plan following Anderson and Magruder (2022) to identify impacts of each program and generate an apples-to-apples comparison which limits bias inherent to evaluation design.
Experimental design
In Goma: 50 avenues were selected at random to receive $500 cash transfers. All youths residing in these avenues and born between 1st Jan 1994 - 30 Sept 1998 were eligible to receive these cash transfers. 50 avenues were selected at random to receive $1000 cash transfers with the same eligibility criterion. 100 avenues were left as control and did not receive cash transfers.
In addition, we select a subset of respondents at random to receive trainings in FINCA's portfolio of financial products, stratified on treatment.
In Walungu,80 villages were randomly selected from a list of 180 to receive a programming from Pathway 0, programming which involves the creation of savings and lending groups, financial education, employability and entrepreneurship trainings, livelihoods skills trainings, and small ($100) cash grants. An additional 20 villages were added to this list after it was determined that there were too few eligible beneficiaries in the first 80 villages. In addition, 50 of these 100 villages were selected at random to receive $750 cash grants from give directly, again for all eligible youth.
Design details
Cited by
No citation data available for this registration yet.