Print
Share
A New Kind of Overseas Charity Is Born
September 30, 2011
I was struck by the advent of a new form of overseas charitable giving. GiveDirectly, which opened this summer, allows anyone on earth to directly wire money to the poorest of the poor. The transfers go to very low-income Africans over mobile phones, which have become suddenly ubiquitous across the continent. The transfers are targeted to people very likely to be extremely poor, and almost all the donor’s money goes right into the recipient’s pocket.
Three things make this effort new. First, these are unconditional transfers. They are targeted to people living in low-quality structures in areas of high poverty, meaning that these people are destitute by the standards of almost anyone reading this blog. But the transfers are unconditional in that the people selected to receive the money don’t need to do anything to receive it. Second, this kind of giving only recently became possible because it arises from technological change. The money is wired directly to the recipient’s mobile phone, keeping cost, fraud, and leakage very low. And the targeting depends on easy digital access to granular area-level poverty data. Both of these were unthinkable in Africa just a few years ago. Third, the founding team involves some of the smartest young development economists in the world, people with a commitment to impact and rigorously evaluating that impact—so that others can learn from their experiences. Three cheers for that.
This mechanism is different from related mechanisms like DonorsChoose, GlobalGiving, and Kiva. GiveDirectly donors don’t pick and choose individual recipients from among the recipients who meet the targeting criteria (though with Kiva, that’s complicated). GiveDirectly transparently states the targeting criteria—other mechanisms have more discretion about who gets to be in the pool—and donors simply decide whether to donate or not. Also, unlike Kiva, GiveDirectly gives grants, not loans.
Because some of the founders are leading researchers, it’s not surprising that the GiveDirectly model is rooted in good research. Conditional Cash Transfer programs have taken the anti-poverty world by storm ever since Mexico created and rigorously evaluated the PROGRESA program. Conditionality on cash grants makes donors feel reassured. But recent research suggests that the on-the-ground consequences of such conditions are complex. A new paper in the Quarterly Journal of Economics points out that unconditional transfers can have many of the same benefits as conditional transfers, and conditions can have unintended consequences that change with children’s age. Michael Carter of the University of Wisconsin and his co-authors show that unconditional cash transfers in South Africa have large health benefits for children, benefits that translate into higher earnings when those kids grow up. Jenny Aker of Tufts University and CGD has co-run an experiment in Niger showing that unconditional cash transfers over mobile phones in particular can be cheap and highly effective.
And this is part of why new unconditional transfer models like GiveDirectly are interesting. Aid agencies must pay to enforce every condition, and those conditions often cause offsetting changes in recipients’ behavior. It remains an open question whether, net of their costs and unintended consequences, conditions on cash transfers improve their impact in some settings relative to unconditional transfers. The UK government has a good review paper on these issues (see especially page 49).
And there’s something undeniably compelling about the instant, person-to-person, very low-leakage channel of a transfer via mobile. In aid as in many other global phenomena, intermediaries are a slowly vanishing species.
Possibly Related Posts
- $100 Billion to End Global Poverty –Déjà vu, Déjà Dismissed?
- Oil to Cash: An Idea to Fight the Resource Curse
8 Responses to “A New Kind of Overseas Charity Is Born”
Post a Comment
We value frank and constructive exchanges and encourage you to use your real name in your comments.
September 30th, 2011 at 2:16 pm
Michael – This is very interesting. I wonder about the incentive effect of random transfers that seem to fall from the sky. My response if I were the potential recipient: get more cell phones! More fundamentally, do I plan for the future differently if I think I might get an uncomditional transfer at any moment (or not)? And how do I feel when my neighbor/brother/wife gets one and I do not? Presumably these are the sorts of things (among many others!) that the researchers will be looking at. For my part, I’m going to include a GiveDirectly transfer on my (modest) annual giving list.
September 30th, 2011 at 2:35 pm
Thanks very much Lawrence. These are good points and it seems to me that any kind of cash giveaway will generate some degree of mistrust and skewed incentives.
My understanding is that GiveDirectly tries to address questions of “why not me” by having a very simple targeting rule that can be easily explained to locals: They give to people in high-poverty, low-quality housing areas, and within those areas they give to people living in homes built of mud, wood, or grass. That’s dramatically simpler and more transparent than any of the targeting rules used by other NGOs I know of, thus dramatically easier to explain to people.
And they try to address incentive effects by arriving in communities unannounced and working quickly, helping to prevent weird outcomes like someone building a grass hut in order to affect targeting.
But these are issues that, to me, don’t seem possible to eliminate completely.
October 2nd, 2011 at 4:00 pm
@Lawrence:
Is the incentive to own a cell phone necessarily perverse? If anything, it seems like a positive externality.
@Michael:
Thanks for the Cliffs Notes version of their methods. It seems a bit simplistic and arbitrary, but I suppose that is better than complicated and arbitrary!
October 3rd, 2011 at 4:32 pm
Michael, this is cool. I’m going to give. I don’t know why but i guess i feel like, after years of giving aid the ‘traditional’ way – through NGOs, governments, and living with the reality that this traditional approach may reinforce some unhelpful power structures at worst, and limit the flexibility of use of funds by poor people, at least – that it is time to give something new a try.
Do you see any reason why a government (like the US government) couldn’t use this instead of the approach that they typically use? For example, perhaps we could take 25% of our current aid to Pakistan and try giving it directly to poor Pakistanis instead of the Pakistani government? Maybe a randomized trial of the two approaches?
I appreciated the GiveDirectly website’s bit on the use of funds for ‘temptation’ spending for things like alcohol. Anyone (and i’m among this group) who has been to an NGO fundraising with drinks, wine, and a nice meal has got to find that at least a bit ironic!
Thanks for bringing it to our attention and i look forward to following its progress.
October 5th, 2011 at 1:03 pm
@Orin: Thanks so much for this thoughtful comment.
I don’t see any reason at all why direct transfers to people could or should not be part of aid efforts. The fact that they would inevitably be imperfectly targeted must be considered in light of the alternatives. First, aid to governments has had titanic problems of targeting in the past (e.g. the half billion dollars that the Reagan administration gave to Samuel Doe). Moreover, no U.S. program of family-level subsidies is perfectly targeted. The better question is what delivery method is most effective among real alternatives.
CGD has a lot of research on this question: most recently the Oil to Cash initiative, suggesting reserving some portion of natural resource rents for direct transfers to individual citizens in developing countries. Many of the same questions of effectiveness and feasibility apply. And there is also one of CGD’s first working papers, in which Bill Easterly discussed the idea of reserving some portion of aid flows as family-level vouchers for services, which flexibly aligns incentives to provide services that people genuinely need.
October 12th, 2011 at 4:49 am
It would require a local organisation (e.g Church) to do the assessment at grassroots level to ensure the grants go to inviduals that meet the targeting critieria.
Those making the donations would need a entire network of local organisaitons that they have assessed in the developing countries…….and this sort of stakeholder network takes years to build.
October 12th, 2011 at 4:54 am
The direct funds transfer mechanism is built on the model of mirco-financing popular in East Africa.
This requires a robust network of local organisations to do the assessments to ensure grants are properly targeted.
In my experience, creating this type of stakeholder network is a process that takes time, experience and resource. Organisations like Oxfam have spent years developing these structures, and they are necessary to do the targeting.
October 12th, 2011 at 3:27 pm
Joe: Thanks for your comment. The targeting criteria are so simple that no such further targeting is needed. That is one of the innovations in this model. If people live in a high-poverty area and live in an unimproved structure (mud, wood, or grass), then they qualify, period. So the targeting is actually quite easy to implement without local partners. What GiveDirectly believes is that such people are so likely to be extremely poor by the standards of potential donors that donors will give anyway, and I think that’s plausible.