Latest Findings on Unconditional Cash Transfers: GiveDirectly
A policy brief on the NGO GiveDirectly was just released today. GiveDirectly is an NGO founded by two MIT Economics PhD students to provide unconditional cash transfers (UCTs) to poor households in Kenya, in which they give large lump sums of money (up to two years worth of income) with no strings attached. Unconditional cash transfers have become one of the most exciting opportunities and areas of research in development economics over the past few years. GiveDirectly, one of the foremost UCT programs in the field, has received some press in the past year, and this marks the first thorough study of GiveDirectly. I think the promising results of the study hold exciting potential for a new paradigm of charity-giving for the poor.
In brief, this report demonstrates that UCTs have large potential to improve the lives of the impoverished, especially those who primarily struggle from a lack of income. In a RCT conducted in poor villages in Western Kenya, the study randomized UCTs along three dimensions: whether it is paid to the husband or wife of the household, the payment schedule, and the total amount paid. Haushofer and Shapiro, the authors of the study, are able to show that their UCTs allow households to build assets (particularly in household repairs and accumulation of livestock), increase consumption, reduce hunger, increase business profits, and improve overall happiness and well-being. The money was not spent on alcohol or tobacco — one of the historical concerns with giving poor people money directly, as opposed to in-kind services. Notably, the study was not able to demonstrate any change in education or health during the time period of the study. Of course, there are some minor concerns with the study, as with nearly any study, but it seems like experts are generally happy with the methodology and results.
For an overview of the current literature and state of the field on UCTs, take a look at this excellent article by the Economist, which does a good job of reviewing much of the academic literature at this point and giving a sense of the continuing debate between conditional cash transfers (CCTs) and UCTs. Conditional cash transfers have shown to be effective, and have become a favored tool by policymakers around the world to support impoverished households while encouraging good behavior, such as sending children to school. Progresa and Oportunidades, CCT programs started by the government of Mexico in the late 90s, were some of the earliest large-scale CCT programs, and their successes have engendered more programs like it around the world. For a detailed analysis of the impact of Progresa in Mexico, take a look at this report. Note however that it is a policy report, not a peer-reviewed journal article, and so it may be somewhat less objective. For a more rigorous comparison between CCTs and UCTs, take a look at “Cash or Condition? Evidence From a Cash Transfer Experiment” by Baird et al., a 2011 paper published in the Quarterly Journal of Economics (might be paywalled).
I was impressed by the use of technology in GiveDirectly’s operation, from the use of satelite imaging and Amazon Mechanical Turk to identify poor villages by the material of roofs (thatched vs. iron) to the use of M-Pesa for the cash transfers themselves. I have long seen M-Pesa as a very exciting opportunity for social interventions to provide monetary gifts or incentives in a low-cost and scalable manner.
In the same way that the positive results of Progresa and Oportunidades at the turn of the millenium helped to kick off a wave of CCT programs, I believe that the strength of this study could similarly result in new UCT interventions in poor regions of the world. UCT programs appear to be effective in improving social outcomes, seem to be cheaper to run (GiveDirectly reported that 92.1% of their donations were transferred to recipients), and most of all, respect the autonomy and agency of the poor to make their own decisions about their lives. I am eager to hear more about how the research in this area continues to progress.
Update: David McKenzie on the World Bank Impact Evaluations blog has also posted his thoughts on the report.