GiveWell began with a simple question: where should I donate?
Elie Hassenfeld and Holden Karnofsky (bios) were 20-something friends who worked at a Westport, CT, hedge fund called Bridgewater Associates. They made a good money.
They wanted to give some away. That was easy.
They wanted their donations to help people—by reducing suffering or increasing opportunity—“as much as possible.” That was much harder.
Hassenfeld and Karnofsky scoured the Internet and found enormous amounts of data on nonprofits. Most of it was useless. They then approached about 60 charities with two simple questions — what they do with donors’ money, and what evidence exists that their activities help people? Most did not or could not answer. Foundations were reluctant to share what they knew, even with a couple of Ivy-educated, well-connected hedge fund guys. Charities, they learned
compete for individual contributions in many ways, from marketing campaigns to personal connections, but not through comparison of their answers to our two basic questions. Public scrutiny, transparency, and competition of charities’ actual abilities to improve the world is thus practically nonexistent.
In 2007, Hassenfeld and Karnofsky left Bridgewater to start GiveWell. They set out to do the kind of thorough, rigorous and time-consuming research about charities that they couldn’t find anywhere else, and to publish it. They are still at it, and by at least one metric they have been successful: In 2014, GiveWell steered about $28 million in donations to a handful of organizations that they have identified Top Charities. (Here’s the current list.)
But GiveWell has not had much impact on the broader world of charitable giving—not yet, anyway. Hassenfeld and Karnofsky remain a long way from achieving their goal of “a world in which donors reward charities for effectiveness in improving lives.”
“Most charities, in my experience, have basically no thoughtful, analytical assessments of how well their programs are working,”Hassenfeld told me, when I visited him at GiveWell’s downtown San Francisco office.
Perhaps surprisingly, many donors don’t particularly care about making a difference. As for those who do, they—we—remain, for the most part, in the dark.
To be sure, what GiveWell is trying to do isn’t easy. “This work is challenging,” Hassenfeld says. “We don’t want to be overconfident.”
There’s lots to be said about GiveWell. Happily, GiveWell has lots to say about itself on its website, which might be the most comprehensive website of any nonprofit in the world. (I can’t say that I have verified that claim.) Consider, for example, a section of the website called Mistakes, which chronicles in painstaking detail both major missteps—the GiveWell founders, in the early days, posted comments on other blogs and websites without identifying themselves, for which they were reprimanded by their board and financially penalized—-as well as such minor flubs as a website crash that put GiveWell offline for a few hours. The goal is not self-flagellation, but to set an example for other nonprofits. GiveWell also explains, in sometimes excruciating detail, how it selects its top charities, and publishes notes of dozens, if not hundreds, of Conversations it has with charity representatives, funders and experts. GiveWell staff members disclose their personal donations. This is radical transparency, and it’s not easy, as Karnofsky explains in an illuminating blog post on The Challenges of Transparency. But it’s a practice other nonprofits and foundations should strive to emulate.
GiveWell’s top charities share a number of qualities. They serve the poorest of the poor. They are focused. They try to measure their impact. And, not surprisingly, they are committed to openness.
“They’re extremely transparent organizations,” Hassenfeld told me. “They want to be open-books to us and via us to the public. They want to share the good and bad. When things go wrong, they’re not afraid to talk about why they go wrong and what they could do differently. Most organizations don’t have that attitude.”
Where possible, Hassenfeld and Karnofsky try to bring rigor and discipline to their analysis. “We put a heavy weight on what we call independent evidence — evidence that’s not coming from the charities themselves,” Hassenfeld says.
Consider the Against Malaria Foundation, a GiveWell top charity since 2011. Studies by development and health experts have found that distributing bednets treated with insecticide, which is the focus of the foundation, reduces child mortality and malaria. The charity, which operates through partners in Africa, goes to great lengths to make sure the bednets are in use and in good repair, following up six, 12 and 24 months later. “No one else that I know of, except maybe the other charities that we’ve looked at, comes close to that level of monitoring,” Hassenfeld says.
Still, questions remain. How well do nets work? Why do they fail? What happens as mosquitos develop resistance to insecticide? How many nets have to be given out to save a life? These are extremely hard questions to answer. GiveWell estimated the full cost of distributing each Against Malaria Foundation-funded long-lasting insecticide net — $5.30 in Malawi and $7.50 in the Democratic Republic of the Congo—and then came up with a estimate of the cost per child life saved at about $3,340. Because of claims made by many charities, some donors might see $3,340 as extremely expensive; it’s not.
By contrast, Nothing but Nets, a UN Foundation-led charity whose partners include the NBA and Major League Soccer, tells donors: “Send a net. Save a Life.” The implication is that, for $10, the price of one net at Nothing But Nets, you can save a life. It’s good marketing but bad math.
[I asked Nothing but Nets about this. By email, a spokeswoman replied: “Send a net. Save a life.” is our campaign slogan and meant only to be an aspiration, not to imply that it only costs $10 to save a life. It’s obviously much more complicated than that.” A World Health Organization meta-study estimated the cost per child death averted of long-lasting insecticide-treated bed nets at between $1242 and $2924, depending on the cost and duration of the net.]
GiveDirectly is another GiveWell top charity. A New York-based NGO, GiveDirectly makes cash transfers of about $1,000 each to extremely poor people in Kenya and Uganda and carefully monitors results. Cash transfers, often with conditions—requiring families to keep their kids in school, for example—have become popular government programs in Latin America.
“Surprisingly, cash transfers are one of the most well-studied development interventions,” Hassenfeld told me. People, it turns out, “spend money on productive things, anything from basic consumption like food or clothes, which they really need, to what these programs called investments.” For example, replacing a thatched roof, which leaks and must be replaced often, with a better-built, longer-lasting tin roof turns out to be an investment with a good rate of return, freeing up cash for other purposes.
The idea of cash transfers is appealing to those (like me) who question the value of experts, and trust people to guide their own lives. “Who is better to make a decision for themselves than the people who get the money?” Hassenfeld asks. “When they have the option to improve their lives, they’re going to think about it seriously.” [Disclosure: I donated to GiveDirectly through GiveWell.]
As you’ve gathered, I’m a fan of GiveWell. But I see two obvious drawbacks to its approach. First, the work done by GiveWell is expensive and thus hard to scale. GiveWell has about 17 people on its staff (and they are hiring). In 2014, GiveWell spent about $2 million and moved about $28 million in donations. About half of its spending was devoted to researching charities.
“We think the 10 to 15 percent range of staff costs to money moved is about right,” Hassenfeld says. Maybe so, but applying GiveWell’s methods to hundreds or thousands of NGOs would be prohibitively expensive. Where, then, can donors who care about climate change, or curing Alzheimer’s, or alleviating poverty in their community, turn for reliable third-party information?
The deeper question about GiveWell has to do with politics, advocacy and social change. My daughter Sarah, who works at the Astraea Lesbian Foundation for Justice, which funds grass-roots community organizing, asked me where a metrics-driven donor who wanted to help black children in the late 1950s might have given. To the Boys & Girls Clubs? Or after-school tutoring? It’s unlikely that a by-the-numbers approach would have led donors to the Southern Christian Leadership Conference, the Student Nonviolent Coordinating Committee or the Black Panthers.
Hassenfeld does not disagree.
“It’s a great point,” he says. “Social change — policy change — may be better ways to give than the current GiveWell top charities. That is big part of the impetus for the Open Philanthropy Project. Our broad goal is to find the best ways to use money to accomplish good.”
The Open Philanthropy Project is the next big project for GiveWell. Working with Good Ventures, the foundation of Facebook billionaire Dustin Moskovitz and his wife Cari Tuna, a former Wall Street Journal reporter, GiveWell is trying to identify outstanding giving opportunities across an array of causes, including US-based policy initiatives. More to come on that, another day.
This story was originally published at Nonprofit Chronicles.