Recipient purchase of a bicycle.

Recipient purchase of a bicycle.

Recipient purchase of livestock

This is a recipient purchase of a school uniform

This is a Give Directly recipient with the motorbike they purchased to start a taxi service

Google Earth screenshot from Kenya, where each pin represents a recipient

Google Earth screenshot from Kenya, where each pin represents a recipient

Giving And Relieving: How GiveDirectly Cuts Out The Middleman

Four economists launched a company around a single idea: The best way to help the world’s poor is to put a stack of money in their hands.

In one of the many feature-less buildings that comprise Google's sprawling Mountain View campus, Michael Faye and Paul Niehaus, two directors of the charity GiveDirectly, are being told to dream bigger. "Think of yourselves as a billion-dollar organization influencing the big funders like USAID and the World Bank," Jacquelline Fuller, Google Giving's director, tells the pair. "This is a way to transform how people think about aid. There is no reason why a lot of development dollars couldn't be shifted to a model like this."

The model in question is the unconditional cash transfer program that GiveDirectly has launched in Kenya: $1,000 given to qualifying households in two lump sums, no strings attached. No grand development agendas, no middlemen, no foreigners rolling up in lumbering four-wheel drives. In other words, no resemblance to the majority of Kenyan nongovernmental relief organizations, which a 2007 University of Nairobi Institute for Development Studies estimate says grew in number from 836 in 1997 to 4,099 in 2005. Just the cash, sent directly to the poor via mobile money service M-PESA for them to use however they see fit.

In 2011, the United Kingdom's government aid agency, the Department for International Development, published a review indicating that the cash transfer model "directly reduces poverty, hunger, and inequality" and notes that in the past 15 years, a "quiet revolution" has seen the idea move from "the margins of development policy toward the mainstream." This transformation has coincided with the rise of technology such as M-PESA, letting people across the globe link wallets in ways that previously were impossible. The collision of evidence and technology is behind the creation of GiveDirectly, which seeks to empower the world's poorest while promoting a sea change in the development industry.

According to the Organization for Economic Co-operation and Development, an estimated $51 billion in official (read: government-to-government) development assistance went to Africa in 2011 alone. The NGO industry has blossomed alongside these funds—yet accountability has rarely kept pace. "The system rewards storytellers and brands, not the most effective organizations," says Ken Stern, author of With Charity for All: Why Charities Are Failing and a Better Way to Give. "There is far too little money committed by donors and organizations into researching the effectiveness of what they are doing."

It was this hunt for transparency that produced GiveDirectly. In 2008, Faye, Niehaus, and two other Harvard and MIT economists were surprised to learn that economics literature pointed to unconditional cash transfers as one of the most potent poverty fighters. But no organization existed to facilitate such transactions. Pooling funds from friends and family, they launched GiveDirectly. The charity selects recipients—most of whom live on 65 cents a day—based on their village's per-capita income and whether their home has a thatched roof, seen as a near-certain poverty indicator. Kenyan contractors and GiveDirectly then vet the recipient, provide him or her with a SIM card if needed, and transfer the money.

Since 2011, GiveDirectly has enrolled 1,669 Kenyan households and completed 1,503 transfers of as little as $200. The results have been impressive: 33% fewer homes reported having children who went for whole days without food, and investment in land, farm supplies, livestock, and housing—including tin roofs—jumped by more than 100%. That performance, and the organization's relentless data gathering, has helped GiveDirectly raise more than $5 million, including $130,000 from Facebook cofounder Chris Hughes and $2.4 million via Google's inaugural Global Impact Awards, which support "organizations using technology and innovative approaches to tackle some of the world's toughest human challenges."

"Donors typically give to large international nonprofits, which use some funds for management and fundraising, and then work with partner organizations abroad to implement programs," wrote Niehaus, an assistant professor at the University of California, San Diego, in GiveDirectly's annual letter. "These partners have their own cost structures, which typically are not reported. Every organization that asks for money on behalf of the poor should make a clear and compelling case that they can do more good with it than the poor could do for themselves."

In its own fundraising, GiveDirectly chooses to emphasize data. Yet in today's charity landscape, it's unclear whether statistics can supplant photos of a suffering child or an inspiring story of poverty overcome. "Donors respond to emotional stories," Faye says. "We are trying to take a step back and focus on the product and the impact. The data is less sexy than the anecdote, but it is also largely right."

Critics of the idea behind GiveDirectly argue that efficiency is not everything. In India, a push to replace government subsidies with direct cash transfers has met with public resistance and fears of both "massive social exclusion," especially in rural areas, and misuse of aid that comes as cash rather than food. The latter concern is one of the reasons donors cite most often for their reluctance to fund direct grants to the poor, as opposed to high-touch aid interventions. According to Niehaus and Faye, one of the first questions asked by potential givers is, "Won't people waste it on alcohol or cigarettes?"

GiveDirectly will keep harnessing its data to battle these assumptions as it grows—the organization is presently scouting other countries for expansion, including Tanzania and Uganda. If Faye and Niehaus are going to live up to Google's lofty expectations and their own, it will be because they don't waver from GiveDirectly's data-oriented approach. As Chris Hughes, who also serves on the organization's board, notes, "I don't think unconditional cash transfers are a silver bullet, but they are a useful baseline. It is a cheap way to know that you are having an impact on families who are barely able to make it." When it comes to big dreams, that's not a bad place to start.

[Illustration by Am I Collective]

Add New Comment


  • AB

    I think several
    assumptions have been made here. 1. That there are markets within the ecosystem
    so that cash can go and be spent on goods and services 2. That there is a
    direct link between money and poverty and 3. That cash transfers reduce corruption.


    We all know it is not
    actually the money that fights poverty - money buys the power to accumulate
    assets, and saying more about the asset accumulation of what the money can
    bring about, the long-term effects and the *who* the money-asset is benefiting
    would have been more useful. 


    What are the different
    assets categories and what impacts do they bring (fixed v variable; consumable v
    non-consumable; immediate payback v long term payback; safe v risky
    investments; production v marketing; etc)? Also, is the money actually
    trickling down to address a poverty problem: only 33% of beneficiaries in the
    study reported that their children now go fewer days hungry - for poverty
    reduction, shouldn't we expect this to be *every* child and *every* household? What
    is offsetting the importance of child nutrition?


    And, how does the cash
    spent spread out over these asset categories and what does it mean for the
    market dynamic? If the majority of money is spent on fuel for a motorbike so
    that the man of the house can go to town and hawk produce, could he not have
    saved money over a period of time to carry out the same activity? Is the cash
    transfer doing anything to help or is it simply distorting existing practices? What
    does the money saved go on to be spent on? How does this overlay with poverty -
    if most money is spent in the fuel category, and very little in the food
    category, yet only a 1/3 of children are being well-nourished, what does that
    mean for the poverty outcomes? 


    And of course, there are
    some important sustainability issues to think about. How is the cash transfer
    allocated? - is there a selection criteria - and does it lock out certain
    social groups? Have certain social groups been well-positioned to get this
    cash? Has is created conflict among those that 'get' and those that don't?


    And what about follow-up
    costs for asset maintenance - if a woman buys a cow with her $1000, does she
    have the access to the right inputs to keep that cow fed and healthy long
    enough to product the kind of milk needed to sell on the open market? With the
    kind of returns so that the venture doesn’t fail? Does she have the training in
    dairy production? Is it important that the cash be used for successful venture –
    or do failed ventures count too?


    And finally, the
    assumption that cash transfers overcome corruption issues is a little far
    fetched. Cash on its own does not create the institutions needed to create rules
    and regulations and monitoring and enforcement practices needed to tackle
    corruption. Also, so far, there is little evidence to show what mechanisms are
    in place to stop capture by elite groups (by gender, ethnic groups, less


    The same challenges
    levied at microfinance are applicable here. Yet the main difference - the
    unconditionality of the cash - is the one that has not been looked at in
    detail. Microfinance has all the financial mechanisms to encourage borrowers to
    spend a loan on a *revenue-generating asset* - yet we still aren't sure if this
    is really happening. How does unconditional money therefore do that better?


  • Paco de Onís

    What's missing in this story is whether the poverty alleviation from the direct cash transfers provides a sustainable solution to poverty, or is just a bandaid that alleviates while it lasts.  What happens when the cash runs out?  Have any assets been built by the recipients?  The direct cash transfers may alleviate income poverty, but if they're not coupled with financial education I wonder if they will alleviate asset poverty - is there any data on that?

  • Ian Turner

    Maybe you did not notice this part of the article: "investment in land, farm supplies, livestock, and housing--including tin roofs--jumped by more than 100%."

  • Paco de Onís

    You're right - and that's intriguing because in Latin America conditional cash transfers to impoverished mothers have spread rapidly, led by Bolsa Familia in Brazil, and the operating assumption there is that it wouldn't work if the transfers were unconditional.  I'm going to stay abreast of how GiveDirectly fares.

  • Ian Turner

    Thanks for this. I appreciate your attempt to find a downside to this approach to aid, but the comparison to India doesn't make much sense to me. People aren't opposed to the idea of a cash transfer system there because the think the recipients will waste the money; they are opposed to it because they think the money will be embezzled before it ever makes its way to the recipients in the first place. And frankly, I think that's a legitimate concern.

    As GiveDirectly grows, it will be interesting to see what the limits are on unconditional cash transfers.