Here's How Algorithms and Data Can Help You Find the Best Charities to Donate to  

Illustration by Lincoln Agnew

Like everyone in college, David Winegar wondered what to do with his life. Politics seemed like a good fit. At Duke University, he served as co-president of the Duke Dems. He fought North Carolina’s initiative against same-sex marriage. He read deep into Paul Krugman and income inequality. But Winegar became more interested in math than in retail politics. He eventually found his way to computer science, graduated in 2015, and moved to Silicon Valley to write software code.

He did it, he admits, for the paycheck. “When I realized I wouldn’t go into politics, it made me think of other ways I could help the world,” says Winegar, a tall, pale, earnest 26-year-old with brushcut red hair. “And computer programming is one of the things you can do to make a lot of money.” The more he made, he thought, the more he could give away.

Today, Winegar works, in part, to help others. He gives more than 10 percent of his income, about $1,000 a month, to organizations like the Against Malaria Foundation, which distributes mosquito netting and is considered one of the world’s most effective charities. The cost per net—to buy them and hand them out—is about $4. The foundation estimates Winegar’s annual donation to the effort saves four lives each year.

“For a lot of people in this generation, and particularly in tech, their goal is to give back,” says Winegar. “And they want to go about it the same way they go about their jobs, through an analytic lens. They want to measure their effectiveness on other people’s lives.”

Philanthropists have always cared, to a greater or lesser degree, about how their donations are spent. Groups like Charity Navigator have drilled down into nonprofit balance sheets, studying overhead and expense ratios and looking to increase financial accountability.

But that’s not the same as impact. Today, the most transformative innovation in philanthropy is what’s known as “effective altruism”—EA to its followers. This young and growing movement is based on the notion of maximizing the good that each individual can do in the world, regardless of wealth. It is popular among tech billionaires, like Facebook co-founder Dustin Moskovitz and his wife, Cari Tuna, who created Good Ventures, a foundation tasked with giving away their wealth.

Like many EA adherents, they don’t let personal passions drive their giving. “We see ourselves as stewards of the capital we have, and we feel that we have an obligation to use our money to help the world as much as we can,” Moskovitz said by email. “We take that role seriously and try to approach our philanthropy in an open-minded way, rather than being motivated primarily by our personal experiences or relationships.”

Moskovitz and Tuna, just like Winegar, found their way to EA through the writings of Peter Singer, an Australian moral philosopher and professor of bioethics at Princeton. His 2009 book, The Life You Can Save, argues that people in affluent countries have a moral obligation to save lives and end poverty in developing nations. In an earlier essay, he stated, “It makes no moral difference whether the person I can help is a neighbor’s child ten yards from me or a Bengali whose name I shall never know, ten thousand miles away.... The moral point of view requires us to look beyond the interests of our own society.”

EA brings that same moral pragmatism, along with a Silicon Valley mindset, to giving. It emphasizes logic and results, and encourages giving where your dollar will do the most proven good, regardless of any emotional attachment to a given cause. Its sexiness is that it takes the sexiness out of philanthropy, which it argues is the very problem of philanthropy. The classic example of wrongheaded giving, according to this model, is donating to your alma mater or to an arts institution to get your name plastered on a building. Or sending money to the Red Cross, which already has a hefty name-brand war chest, after a disaster.

EA’s moral pragmatism appeals to Luke Ding, a former fund manager based in London, who once donated $30,000 a year to build a school for autistic students in North London. Today that same money, roughly the cost of one year’s tuition, can deworm an estimated 100,000 children in Burundi. “You can support one child and benefit their life outcomes for 70 cents,” says Ding. For him, helping 100,000 kids get well enough to attend classes—for their entire academic careers—outstrips the need of one child to attend one year of school.

Ding doesn’t crunch those numbers on his own. He and people like him are getting spreadsheet help from a new generation of charity analysts. Among them are Australian philosopher Toby Ord and his British colleague Will MacAskill, who co-founded the EA organization Giving What We Can. (Ding served on its advisory board in the past.) Perhaps the most high-profile proponents in the techie world are Elie Hassenfeld and Holden Karnofsky. Former investment associates at the mega hedge fund Bridgewater Associates, they started GiveWell, a nonprofit that analyzes charities and helps people figure out where to give. While traditional charity assessment has focused solely on fraud and swollen budgets, GiveWell is willing to look at how your money is put to work and calculate the lives you’ll save.

“I approach charitable giving as a math problem,” says Benjamin Clark, a 24-year-old Google software engineer who gives his money ($54,000 last year) mostly to GiveWell’s stress-tested charities. “If you get the inputs wrong, it leads to very wrong conclusions.” Clark, who lives frugally, has a roommate, and rides his bike to work, says the same ethos he holds as an engineer—efficiency-driven optimization and getting the details right so a thing performs well—is what he values about GiveWell’s guidance. Also, he says, “everyone likes to get a deal, to get value for your money. That’s one way to view this doctrine.”

It’s a cool, clear Tuesday in late April.

About 40 young people, nearly all of them from the tech industry, have crowded into a conference room off the lobby of a downtown San Francisco office tower. Bike helmets and backpacks, black tees and New Balance sneakers predominate. The back wall is lined with folding tables full of veggie wraps.

“Our version of a fund-raising gala,” jokes Hassenfeld.

Tall, with graying temples and clear green eyes, Hassenfeld is standing by the open door. Few of the attendees have ever met him, or each other. They are not part of a fund-raising circuit. They’re here because they care about people and measuring the impact of their giving.

“It’s still a very small slice of the overall philanthropic community,” says Hassenfeld. “We don’t think what we do is mass market, and the people who find us have pretty similar stories. ‘I got my first job and noticed my savings building up, and I felt fortunate and wanted to give back.’ But they want to be confident it’s effective, and they find us.”

GiveWell conducts its price-shopper methodology using four criteria. It looks at a charity’s cost-effectiveness, such as measuring the “cost per life saved”; whether the charity has room for more funding to save more lives; the transparency of its reporting; and evidence of impact. For that evidence, it seeks out public health data and randomized studies—the gold standard for scientific research. One famous study, which got its start at MIT, found that handing out malaria nets for free was a highly effective way to stop the disease’s spread and prevent premature death. The Against Malaria Foundation does this, and has been a GiveWell toprated charity for years.

Other groups include Evidence Action’s Deworm the World Initiative, which can improve life outcomes at a low cost (53 cents per child in places like Kenya, and far less than that in India) and GiveDirectly, which makes cash transfers to families in developing countries. Randomized controlled trials show that people in those areas know best what they need (a new roof, a new water tank, money for food, transportation to school) and will use the money accordingly.

GiveWell’s impact is impressive. Since 2011, it has directed more than $500 million to charities. The organization is expected to deliver 317 million deworming treatments, to transfer about $90 million directly to some of the poorest people in the world (between $1,000 and 90,000 per family), and will avert more than 75,000 deaths from malaria and other diseases.

But like everything in Silicon Valley, the question is: Can it scale? GiveWell fears its donor rolls will stagnate. And with most donations coming from people who give in the four to five digits, a big chunk of tonight’s presentation is devoted to the challenge of finding more donors.

“We are expecting to be more constrained in the long run by our ability to increase the amount of funding that we can direct to [our] charities,” Catherine Hollander, a research analyst at GiveWell, tells the crowd at one point. She surmises that effective altruism tends to appeal to a very limited demographic, namely young professionals in the tech or finance professions—people “with a quantitative mindset.” To reach them, GiveWell is going where they spend most of their time: online. It’s investing in search engine optimization for its social feeds and advertising on podcasts of interest to the EA crowd, as well as trying to build their communities in real life with events like this one.

One donor in his late forties asks Hassenfeld what he has been doing to expand beyond that cohort. The short answer, according to Hassenfeld, is not much. “We think that [approaching] people who are broadly demographically similar, but who don’t yet know what GiveWell is or does, will more likely be an effective way to raise more money” than reaching beyond them, he says.

Hassenfeld recognizes that EA’s approach is not for everyone. Many in the philanthropy field are critical of what they view as EA’s cold, hyperrational approach to helping other people. A 2013 editorial in the Stanford Social Innovation Review co-written by Charity Navigator’s then president Ken Berger labeled it “defective altruism.” While noting all donors “should be informed,” the authors wrote that EA claims a moral high ground by “weighing causes and beneficiaries against one another.” It’s not moral, he wrote, but “moralistic in the worst sense.”

Other critics are skeptical of the EA community’s efforts to rewire a natural human instinct: empathy for those we know over those we’ve never met. “I would be wary of inverting the natural order of affections,” the New York Times columnist David Brooks wrote. “If you see the world on a strictly intellectual level, then a child in Pakistan or Zambia is just as valuable as your own child. But not many people actually think this way.”

People in EA, who are typically as selfcritical as they expect charities to be, agonize over what they call the proximity question. “The money you give to an American living on the street, in a very rich country, could in some ways have a bigger impact elsewhere,” says Winegar. “At the same time it’s someone in front of you, in need, who actually has this serious issue that you should help solve. If you say no every time, it can lead to coldness.”

GiveWell, like the EA movement at large, hopes to extend its impact.

Hassenfeld tells the donors here tonight that his group will offer its research to government agencies and public health lobbyists in middle-income countries. The goal, he says, is to expand its mission beyond just funding the direct delivery of services (like bed netting) into influencing policy.

Hassenfeld says his organization wants to help change public health regulations by funding groups that seek to pass tobacco-tax legislation, which reduces tobacco use and improves health. “This would not be GiveWell itself advising a government but funds going to organizations that work closely with governments on these types of regulations,” he says.

He notes that last year, government aid agencies around the world, including the U.S. Agency for International Development and its U.K. counterpart, collectively spent $25 billion on their programs. His goal is to provide them with GiveWell’s research. 

He admits that working with these players will be tricky for his group’s mission. “This space as a whole is going to be much less quantifiable and harder to assess” than the highly targeted, quantifiable programs he and his team have recommended so far.

A young woman stands and asks if engaging with another country’s political process was fraught and perhaps could be viewed as overtly paternalistic.

“We have a point of view of what would lead to better outcomes for people, like giving someone a malaria net, so that certainly is paternalistic in a certain sense,” Hassenfeld answers, demonstrating the kind of honest self-assessment that lies at the core of the effective altruist mission. He acknowledges that working with groups trying to change health laws will be a challenge. He says, “We’re just going to have to move cautiously to be confident that the types of groups we’re supporting...that we’re convinced what they’re doing will lead to well-being rather than major problems.”

With fewer metrics to rely on, it may be that the people giving with their heads will just have to follow their gut.