Evidence

Research on Cash Transfers

Last updated January 21, 2020

Many people (including us) grew up hearing that “you can’t just give money to poor people.” As it turns out, this view was largely based on anecdote and hearsay; the first rigorous experimental testing of this approach (and many others) did not begin until the early 2000s. Since then we have learned an enormous amount about the impacts that cash transfers have had on the lives of people living in poverty, with well over 150 studies from around the world.

Summarizing the results of these studies is a difficult task –– in part because the whole point of cash transfers is to give different people the freedom to do different things, so there is no one answer to “the impacts” of cash. We do our best below to draw out the main lessons as we understand them, citing independent external reviews of multiple studies whenever possible and framing individual studies as examples when not.

We are also actively engaged in producing new evidence through a series of more than a dozen experimental studies, which you can read about here.

Cash as a percentage of aid

Today, people living in extreme poverty have very little control over how to spend the aid money intended to help them.

Yet independent reviewers agree that cash is an effective way to help people living in poverty.

quotation

Evidence was extracted from 165 studies, covering 56 cash transfer programmes in low- and middle-income countries… There is strong evidence that cash transfers are associated with reductions in monetary poverty. (source)

ODI global think tank logo
quotation

Cash transfers have the strongest track record we’ve seen for a non-health intervention, and are a priority program of ours. (source)

GiveWell logo
quotation

Cash transfers are more cost effective than vouchers which are more cost effective than in‐kind food assistance. (source)

Campbell Collaboration logo
quotation

Cash transfers are one of the more thoroughly researched forms of development intervention. (source)

Department for International Development logo

Recipients of cash do not systematically waste or misuse it.

Among the hundreds of millions of people who have received cash transfers, some have surely bought beer or worked fewer hours. But systematic reviews of the evidence find that on average these are not the impacts of giving money; if anything, they find the opposite.

quotation

This article reviews 19 studies… on average cash transfers have a significant negative effect on total expenditures on temptation goods [alcohol, tobacco, etc.],… This negative result is supported by data from Latin America, Africa, and Asia, for both conditional and unconditional cash transfer programs. (source)

The World Bank logo
quotation

We re-analyze the data from seven randomized controlled trials of government-run cash transfer programs in six developing countries throughout the world, and find no systematic evidence that cash transfer programs discourage work. (source)

MIT logo
quotation

This article brings together evidence from seven experimental and non-experimental impact evaluations of government-run unconditional cash transfer programs… There is no evidence that cash transfers translated into an overall reduction of labor supply or work effort — in fact quite the opposite: the transfers were used to improve household income-generating activities. (source)

Agriculture & Applied Economics Association logo

Recipients of cash typically end up less poor and put cash towards improving different aspects of their lives.

quotation

The evidence consistently showed an increase in total expenditure and food expenditure and a reduction in poverty measures… There is robust evidence that cash transfers increase beneficiaries’ savings, investment in livestock and, to a lesser extent, agricultural assets. (source)

ODI logo
quotation

The existing body of evidence, which is based on several cluster‐RCTs, suggests that UCTs have probably had a large, clinically meaningful, beneficial effect on the likelihood of having had any illness… (source)

Cochrane Library logo
quotation

Using data from 75 reports that cover 35 different studies, the authors find that both conditional cash transfers and unconditional cash transfers improve the odds of being enrolled in and attending school compared to no cash transfer program. (source)

World Bank logo
quotation

We conducted a mixed-method review of the impact of CTs on IPV in LMICs and built a program theory to help understand the mechanisms behind this impact. In total, we identified 14 quantitative and nine qualitative studies that met our inclusion criteria, of which 11 and six respectively showed support that CTs (cash transfers decrease IPV [intimate partner violence]). (source)

UNICEF logo

Of course, people, programs, and priorities vary… so effects do too.

Adapted from Bastagli, Zanker, et al (2016)

These results don’t mean people living in poverty will always be the best at deciding how money should be used, but it raises the question why we’ve been trusting the vast majority of our giving to an aid industry with a pretty patchy track record of allocating capital.

quotation

It is hard to find a skills training program that passes a simple cost-benefit test. After repeated studies of technical, vocational, and business skills training programs, most programs do not have positive impacts … Those that do are often so expensive that costs far outweigh benefits. And most poor people turn these programs down or drop out. This is incredible given that the World Bank alone invests nearly a billion dollars a year in skills training programs of some form. (source)

U Chicago
quotation

In its heyday, microcredit was the basis for the 2006 Nobel Peace Prize and embraced by policymakers, donors, and funders worldwide as an effective policy tool. … Summarizing and interpreting results across [six] studies, we note a consistent pattern of modestly positive, but not transformative, effects. (source)

MIT
quotation

… FFS [Farmer Field Schools] has been used to train 12 million farmers in over 90 countries across Asia, Africa and Latin America, … Drawing on a systematic review of over 500 documents, this study finds that although FFSs have changed agricultural practices and raised yields in pilot projects, they have not been effective when taken to scale. (source)

International Initiative for Impact Evaluation

Frequently asked questions

+Conditions: why give “no strings attached”?

Conditions can be costly to enforce and can exclude the most vulnerable recipients.

“Without targeting or monitoring of conditions, the estimated long-run CTR [cash transfer ratio declines] about 40 per cent, underscoring that targeting and conditioning require substantial resources.”
Caldés & Maluccio 2005

“By denying noncompliant [(excluded)] adolescent girls and young women cash transfers at precisely the moment when they are most likely to start childbearing, a myriad of potential benefits might be missed under CCT programs.”
Ozler, 2019

+Training: shouldn’t you “teach a man to fish” instead?

Training can be expensive and hard to do well.

“Skills training and microfinance have shown little impact on poverty or stability, especially relative to program cost.”
Blattman & Ralston 2017

“Drawing on a systematic review of over 500 documents, this study finds that although FFS [Farmer Field Schools] have changed agricultural practices and raised yields in pilot projects, they have not been effective when taken to”
Waddington & White 2014

Also, the people we serve often know a lot about fishing and other ways to make a living.

+What’s the evidence on GiveDirectly’s programs specifically?

We review the research studying cash transfers delivered by GiveDirectly here. We are constantly testing different ways of delivering cash transfers to different types of people to find which approaches best achieve different outcomes.

+What about the long-term effects of cash?

There are a lot of examples of the effects of cash transfer programs lasting a long time. But long-lasting does not mean forever, and there are also many examples where effects fade over time. How long effects persist may also vary by how long the cash is provided to the recipient.

Ultimately, what matters is whether total effects last long enough and are large enough to be worthwhile relative to the cash delivered. That’s difficult methodologically and morally because it requires adding up different types of outcomes, but there are some positive examples from more simplistic, single-outcome approaches — e.g., where recipients’ earnings gains alone exceeded cash transferred.

Here are some examples of these dynamics.

“Four years on, an experimental evaluation found grants raised earnings by 38%… We return after 9 years to find these start-up grants raised earnings and consumption temporarily only. Grantees’ investment leveled off; controls eventually increased their incomes through business and casual labor; and so both groups converged in employment, earnings, and consumption… We estimate a discounted present value of total earnings gains of roughly $665 — almost twice the size of the grant…
Blattman et al 2019

“[W]e collected… records of applicants to the Mothers’ Pension program—the first government-sponsored welfare program in the United States (1911–1935)… Male children of accepted applicants lived one year longer than those of rejected mothers. They also obtained one-third more years of schooling, were less likely to be underweight, and had higher income in adulthood than children of rejected mothers.”
Aizer et al 2016

“We examine the medium-term effects of a two-year cash transfer program targeted to adolescent girls and young women. Significant declines in HIV prevalence, teen pregnancy, and early marriage among recipients of unconditional cash transfers (UCTs) during the program evaporated quickly two years after the cessation of transfers. However, children born to UCT beneficiaries during the program had significantly higher height-for-age z-scores at follow-up.”
Baird et al 2016

+Do women and men spend cash transfers differently?

“While one might expect differences in outcomes depending on the gender of the main recipient, based on the four studies included in this review, for most of the indicators in this review there does not appear to be strong support for differences arising from specifically targeting either men or women.”
Bastagli, Hagen-Zanker, et al 2016