Can 4 Economists Build the Most Economically Efficient Charity Ever?

GiveDirectly, the brainchild of four Harvard and MIT graduate students, is so simple, it's genius. Give poor Kenyan families $1,000 -- and let them do whatever they want with it.

GiveDirectly, the brainchild of four Harvard and MIT graduate students, is so simple, it's genius. Give poor Kenyan families $1,000 -- and let them do whatever they want with it.

615 kenya woman poverty.jpg

Every holiday season, the vast majority of America's charitable dollars are donated to local religious and educational institutions, or to particular causes, like protecting endangered species or fighting malaria. Very few of us would consider giving large sums of money to poor people and letting them do whatever they want with it.

Surprisingly, it is precisely that type of charity--and government aid--that some economists say is most effective.

In 2008, four Harvard and MIT graduate students studying developing-world economics decided to form their own giving circle. The research literature on anti-poverty aid was discouraging. In India, an estimated 50 to 60 cents of every government dollar spent on food or employment aid for the poor is lost due to corruption, and private philanthropy, too, is heavily skimmed as it makes its way into the hands of the poor. Micro-credit efforts like Mohammed Yunus' Grameen Bank have been celebrated in the media and embraced by celebrity donors, but there is mounting evidence that microfinance does not reach the poorest populations, and that many recipients default on their loans, leaving them further in debt. While Latin American countries like Mexico and Brazil have had success with "conditional cash transfers" -- government payments to poor families in exchange for "good behavior," such as enrolling children in school or taking them to the doctor -- that model is difficult for lower-income nations, like those in sub-Saharan Africa, to effectively implement. If there aren't enough schools or doctors to serve an influx of new clients, there is little point in requiring the poor to use such services.

So where did that leave four private donors, anxious to fight global poverty, but too savvy to trust many of the leading models for international aid? Paul Niehaus, Michael Faye, Rohit Wanchoo, and Jeremy Shapiro came up with a radically simple plan shaped by their own academic research. They would give poor families in rural Kenya $1,000 over the course of 10 months, and let them do whatever they wanted with the money. They hoped the recipients would spend it on nutrition, health care, and education. But, theoretically, they could use it to purchase alcohol or drugs. The families would decide on their own.


Three years later, the four economists expanded their private effort into GiveDirectly, a charity that accepts online donations from the public, as well. Ninety-two cents of every dollar donated to GiveDirectly is transferred to poor households through M-PESA, a cell phone banking service with 11,000 agents working in Kenya. GiveDirectly chooses recipients by targeting homes made of mud or thatch, as opposed to more durable materials, such as cement or iron. The typical family participating in the program lives on just 65 nominal cents-per-person-per-day. Four in ten have had a child go at least a full day without food in the last month.

Initial reports from the field are positive. According to Niehaus, GiveDirectly recipients are spending their payments mostly on food and home improvements that can vastly improve quality of life, such as installing a weatherproof tin roof. Some families have invested in profit-bearing businesses, such as chicken-rearing, agriculture, or the vending of clothes, shoes, or charcoal.

More information on GiveDirectly's impact will be available next year, when an NIH-funded evaluation of the organization's work is complete. Yet already, GiveDirectly is receiving rave reviews. In November, it became one of just three charities to earn a coveted recommendation from GiveWell, a web publication that conducts exhaustive, on-site research of philanthropies. This month, Google awarded GiveDirectly a $2.4 million grant to expand beyond Kenya. Facebook cofounder and media wunderkind Chris Hughes joined the group's board in August. Nevertheless, GiveDirectly remains an outlier in the development arena, perhaps the only organization that distributes private donations, made online, directly to the poor with no strings attached--no requirement to launch a business or to immunize one's child; no distribution of bed nets, solar lanterns, or goats.

The economics might be sound. But the politics within the non-profit world are more complicated. Niehaus, now a professor at the University of San Diego, says other development experts who have tested unconditional cash transfers are enthusiastic about the approach. The trouble is convincing NGOs to invest in such programs beyond the pilot stages.

"We had conversations with people [in the non-profit sector] who said there was a lot of internal resistance to unconditional transfers," Niehaus told me. "If this works, what are we all here for? Why do we have jobs? There's an industry that exists that tries to make decisions for poor people and determine what's best for them. In some ways, that's the industry I came from. But the value of that hasn't been proven."


Still, GiveDirectly has a lot of work to do to overcome donor resistance to the idea of free money for the poor, much of it based on stereotypes. In the United States, abject poverty and homelessness are often correlated with mental illness and addiction, but that is not the case in rural Kenya where, as Niehaus says, most people are poor simply "because they were born in Africa." Research shows developing world cash transfers are unlikely to be spent on "temptation goods" like tobacco or beer, and the MIT economist Esther Duflo has demonstrated that in South Africa, poor people save or invest more than half of their social welfare payments. The average American, meanwhile, saves less than five percent of his or her income, no matter what its source.

In its report on GiveDirectly, GiveWell flagged several potential weaknesses of its model, most of which the organization is addressing. Because many, but not all, poor Kenyan households own a cell phone that can be used to access M-PESA payments, GiveDirectly provides a SIM card to which the cash transfers can be wired. Each month, the recipient takes his or her cell phone or SIM card to an M-PESA agent, who then distributes the cash in $200 increments. Recipients are able to make this trip in an average of 42 minutes, spending just 64 cents on transportation.

Initially, GiveDirectly worked with village elders to target potential recipients living in subpar housing. But in at least one case, an elder planted friends or relatives with adequate housing in other people's homes with mud or thatch roofs, in order to funnel payments to preferred individuals. GiveDirectly has since stopped relying on village elders to identify potential recipients. The larger effect of GiveDirectly payments on social tensions and jealousies within villages and families remains unknown, though studies of other cash-transfer schemes show few negative results. The ongoing randomized evaluation of GiveDirectly will consider whether it is more socially beneficial to target the poorest households within one village, or to provide payments to an entire village, including those who may be less poor. Research from Duflo and other economists suggests mothers and grandmothers are more likely to spend assets on children's wellbeing than fathers and grandfathers, so the evaluation will also look at any differences between how GiveDirectly's male and female recipients spend their payments.

Unconditional cash transfers to individuals do little to address the structural factors responsible for poverty, such as government corruption, gender discrimination, and the lack of quality jobs, schools, and health care. Those problems need not just philanthropic, but also political and macroeconomic solutions, pursued by governments, private industry, and the non-profit sector working in tandem. Yet if your concern as an individual donor is making life easier and more fulfilling for the poor, GiveDirectly's model warrants close consideration. I donated $200, in part because, after seven years of reporting on American poverty, I'm familiar with the critique that many aid efforts impose needless bureaucratic burdens and constraints on poor people's lives, when what they really need is financial savings: the type of flexible spending power that allows a family to address a problem, like a leaky roof or broken-down car, in a timely and effective manner, before it spirals into a crisis with dire outcomes for health, education, and employment.

"I find the international philanthropic sector very frustrating as a donor," Niehaus says. "We would like to promote much more transparency, and we would like people to ask [when they are making a donation], does this philanthropy have a convincing case that they can do more good with the money than a person would do for themselves?"