Motivation

Digital cash transfers have been demonstrated to be a relatively effective development tool which can be used flexibly managed by implementing partners to create impact in many outcomes of interest. In light of this, CEGA partnered with USAID and GiveDirectly, under the Development Impact Lab (DIL) to develop new methods for comparing traditional development programming to a benchmark of cost-equivalent digital cash transfers, akin to the “standard of care” in medical research. This meant that many programs should be proven to perform as well or better than cash for the target outcomes to be considered effective, not just have more impact than a control arm with no treatment.

Activities

CEGA and USAID collaborated on 6 research projects to pilot two unique benchmarking approaches:

1. Add a cash arm to program RCTs to directly compare cost-equivalent cash transfers to the program and a control group
2. Perform RCT impact evaluations for multiple levels of cash transfers, measuring impact on a  wide range of outcomes of interest to the USAID objectives in the country of interest, for subsequent comparison of program outcomes to cash’s performance

In addition to this research portfolio, CEGA worked with USAID to build a community of practice around the concept of cash benchmarking, and performing related costing and cost effectiveness analysis on present and past USAID programming.

Results

Rwanda: Craig McIntosh (UCSD, Lead PI), Andrew Zeitlin (Georgetown)

This research team led two groundbreaking studies in Rwanda designed as head-to-head comparisons of cash and traditional aid programs. The first of these studies compares an integrated USAID nutrition program, “Gikuriro,” to cost equivalent cash transfers. The second study compares a USAID workforce training program for poor and underemployed youth, “Huguka Dukore,” to cost equivalent cash transfers. The approach demonstrates how a randomized control trial design can be employed to directly compare traditional aid to cash.

Liberia & Malawi: Jon Robinson (UCSC, Lead PI), Jenny Aker (Tufts), Shilpa Aggarwal (Indian School of Business, Hyderabad), Dahyeon Jeong (World Bank), Alan Spearot (UCSC), David Sungho Park (KDI School of Public Policy & Management), Naresh Kumar (UCSC) 

Rather than a head-to-head comparison with a particular USAID program, the research team has run randomized evaluations of standalone cash transfer programs in Liberia and Malawi. The evidence emerging from these studies covers a wide range of socio-economic outcomes for a range of cash transfers.  The resulting estimates of cost-effectiveness help define the “benchmark” of what cash can deliver, against which traditional programming can be measured. The team also used the data to develop and publish a paper on food security and agricultural markets.

Democratic Republic of Congo: Jeremy Magruder (UC Berkeley, Lead PI) 

This research includes a pair of randomized evaluations in the DRC that are measuring the impacts of a cash transfer program as well as a youth employment intervention.  One study compares cash transfers to a control group; the other compares a youth employment intervention to cash transfers and a control. The forthcoming results will help us understand how cash and traditional aid both perform in a setting where mobile money, connectivity, and governance are weaker than in our other cash benchmarking sites of Malawi, Liberia, and Rwanda.