Giving in the Light of Reason

The Humane League, one of Open Phil’s grantees, stages a protest outside McDonald’s in downtown Chicago as part of its “I’m Not Lovin’ It” campaign against the chain’s chicken menu items. (Photograph courtesy of The Humane League) 

There’s an old saying in philanthropy: If you’ve seen one foundation, you’ve seen one foundation. Each is distinctive, which makes sense: Extremely wealthy people do not get to be that way by following the crowd, so they want their foundations to stand out as well.

Still, of the 86,000 or so grantmaking foundations in the United States, few stand quite so far outside of the mainstream as the Open Philanthropy Project, which guides the charitable giving of Dustin Moskovitz, the cofounder of Facebook, and his wife, Cari Tuna, a former Wall Street Journal reporter.

Open Phil, as it’s known, has a vast fortune to give away. Moskovitz’s net worth was estimated to be about $14.3 billion at the end of 2017, and Moskovitz and Tuna say they intend to disburse nearly all of it before they die. In terms of assets, that puts them ahead of the Ford Foundation and behind only the Bill & Melinda Gates Foundation, the Chan Zuckerberg Initiative, and George Soros’ Open Society Foundations in a ranking of America’s philanthropic giants.

Their giving is shaping up to be a grand experiment in rationalism—the idea that it’s possible to think through nearly all of the messy questions at the heart of philanthropy. Should grants go to education, science, or the arts? To a nearby community or to poor people overseas? To cure disease or influence public policy? As Open Phil grapples with such questions, it is guided by the principles of effective altruism, a philosophy and a movement that seeks to use reason and evidence to determine the best ways to do good. “They are unabashed technocratic engineers of good outcomes,” says one insider.

As befits its name, Open Phil is also radically transparent—more so, arguably, than any other big foundation. Staff members publish long, analytical blog posts explaining major decisions, both to test their reasoning against outside critics and as part of a deliberate effort to influence other philanthropists. They have posted about hiring decisions, too. About one new program officer, they wrote, “We believe he will have a steep learning curve in order to get up to speed on philanthropy.” They are even open when they decide to be less open, posting a 2,000-word blog post to explain why.

Open Phil is open in another sense as well: It got going with no devotion to any particular cause. Open to many possibilities, it funds an eclectic and seemingly disconnected set of causes, organizations, and projects: Global poverty alleviation. Criminal justice reform. Scientific research. Farm animal welfare. Existential risks to humanity. The Federal Reserve’s monetary policy. Solar geoengineering research. The history of philanthropy. Efforts to improve human decision making. Anything, really, that is judged to be an important and neglected problem that the organization can help solve.

Their approach raises questions for other philanthropists and, for that matter, anyone who gives money away. Are you doing as much good as you can? How do you know? How do you choose your causes? Do you share what you learn?

These questions don’t get as much attention as they should. Holden Karnofsky, the executive director of Open Phil, says, “There’s a lot of debate in our society about what the government should do. There’s debate about what corporations should do. But there’s very little debate about what foundations should do.” It’s time for philanthropists to engage in that debate, he says.

A Deep Dive

Last winter, I flew to San Francisco to meet with the leadership of Open Phil. Cari Tuna wasn’t feeling well, alas, and didn’t make it into the office, so we talked over a video connection. She is friendly, thoughtful, and, well, young. I couldn’t help but think that the woman on the screen, who at 32 holds the keys to a $14 billion fortune, is younger than my oldest daughter, who has been a grantmaker for about a decade.

Not long into our conversation, I asked Tuna if she identifies as an effective altruist. “Yes,” she replied. “Do you?” (I do, more or less, and told her so.) Effective altruism is inspired by the ideas of Peter Singer, a Princeton philosopher, which he first expressed in a 1972 essay titled “Famine, Affluence, and Morality.” “If it is in our power to prevent something very bad from happening, without thereby sacrificing anything morally significant, we ought, morally, to do it,” Singer wrote. In the late 2000s, Toby Ord and Will MacAskill, who teach philosophy at Oxford and jokingly call themselves “super hardcore do-gooders,” began looking for ways to bring their brand of committed, rational, outcome-oriented altruism to the mainstream. Effective altruism has since become a small but growing movement, with more than 300 chapters around the world, many on college campuses. It has inspired a half-dozen books and at least as many charities and meta-charities, including The Life You Can Save, Giving What We Can, 80,000 Hours, and Animal Charity Evaluators. Nearly all are grantees of Open Phil, which has become the single biggest funder of the movement.

Facebook cofounder Dustin Moskovitz and his wife, Cari Tuna, launched Open Phil to guide their charitable giving. (Photograph courtesy of Good Ventures) 

When Tuna and Moskovitz became the youngest billionaires to sign the Giving Pledge in 2010, the term “effective altruism” hadn’t been coined yet. He was a 26-year-old Harvard dropout who made his fortune during four years at Facebook, as its first chief technology officer; she was a 25-year-old Yale graduate who as a Wall Street Journal reporter had written about the California economy, gay marriage, and Turkish food in San Francisco.

They approach their newfound wealth with humility. “Cari and I are stewards of this capital,” Moskovitz said later. “It’s pooled up around us right now, but it belongs to the world. We are not perfect in applying this attitude, but we try very hard.”

With Moskovitz devoted to Asana, his second technology startup, it fell to Tuna to manage their giving. She quit her reporting job to investigate philanthropy, making the rounds of most of the big foundations. Some advised her to “think about the causes that really touch your heart,” but she didn’t find that advice helpful; her heart was telling her that what she wanted to do was the best she could to help humanity thrive.

An “aha” moment came when she read The Life You Can Save, a 2009 book by Singer. It argues that rich people have moral obligations not only to share their wealth, but to do so effectively. The book features the work of Karnofsky and Elie Hassenfeld, who founded GiveWell, a nonprofit that seeks through rigorous research to identify the world’s most effective charities. They did so because, as investment analysts at Bridgewater Associates, the world’s largest hedge fund, they had struggled to figure out the best ways to give away their money.

Tuna has a similar problem—just on a bigger scale.

“It introduced me to the idea of not just trying to do some good with your giving, but doing as much good as you can,” Tuna told me. A friend connected her to Karnofsky, and they met for Sunday brunch when he visited the Bay Area. “I was immediately impressed by the kinds of questions he was asking, the kinds of causes that GiveWell was interested in,” Tuna said. They talked about how fortifying salt with iodine can increase cognitive development in children with mild to moderate iodine deficiency at a low cost. “It’s not a sexy topic,” Tuna says. “It just made me wonder how many other causes out there are like that—promising and neglected.”

She plunged in. Tuna joined the board of GiveWell in 2011, formalizing a partnership that laid the foundation for what became the Open Philanthropy Project. Good Ventures, a foundation formed by Moskovitz and Tuna, made its first grant, for $50,000, to GiveWell, and decided to give another $1.1 million to charities recommended by GiveWell, all in 2011. Its biggest grants were to the Against Malaria Foundation and the Schistosomiasis Control Initiative, a deworming charity; typical of nonprofits favored by GiveWell, these deliver short-term, low-cost proven interventions to help the world’s poorest people. Since then, Good Ventures has been by far the biggest donor to GiveWell, accounting for about 60 percent of the estimated $375 million that has been moved by GiveWell to its favored organizations.

In 2012, GiveWell decided to relocate its office and staff of five people from New York to San Francisco to be closer to Tuna and other technology-industry donors. By then, Good Ventures and GiveWell had begun a research project, then called GiveWell Labs, to look for giving opportunities beyond direct-aid charities. GiveWell Labs eventually morphed into the Open Philanthropy Project, which was spun off from GiveWell in 2017, although their boards and some staff overlap, and they share office space in downtown San Francisco.

All told, Tuna plays a key role in no fewer than seven entities, as a director or major donor. There’s the Open Philanthropy Project, a limited liability corporation (LLC) that recommends grants, tracks results, and publishes its findings; the Open Philanthropy Action Fund, a small 501(c)(4) social welfare fund that supports nonpartisan advocacy; the Open Philanthropy Project fund, a donor-advised fund housed at the Silicon Valley Community Foundation; the Good Ventures Foundation, a private foundation that makes grants and investments; Good Ventures, a supporting organization, also at the community foundation; and Good Ventures LLC, a for-profit impact investment firm. Finally, there’s GiveWell, the meta-charity, which is a conventional 501(c)(3) nonprofit. Henceforth, for simplicity’s sake, we’ll say that “Open Phil” makes grants or investments when, in point of fact, money could flow out of any of three grantmaking entities.

“Each has a reason for being,” says Alexander Berger, a cofounder, managing director, and board member of Open Phil. Not surprisingly, tax considerations are key: Donor-advised funds provide more generous benefits to living donors. The Open Philanthropy Project LLC was set up with the idea that it might advise other donors in the future. Outside of the various structures, Moskovitz and Tuna have cash of their own that they devote to favored causes. In 2016, they wrote that they were donating $20 million to help elect Democrats, including Hillary Clinton, and they are investors in Bill Gates’ Breakthrough Energy Ventures, a fund that is researching climate-change solutions.

Perhaps inadvertently, the complex structure means that Open Phil does not always live up to the “open” part of its moniker. While Open Phil says it will list all of its grants online, except in rare circumstances, grants made through donor-advised funds can’t be traced back to the original donors. The various entities disclose some but not all of their investments. What’s more, unlike private foundations, Open Phil does not disclose its annual operating costs or the salaries or benefits of its highest-paid staff members. It’s impossible, as a result, to know how much Open Phil spends on its research and analysis, and whether it might do more good by pushing more of that money out the door.

That said, Open Phil blogs about big decisions, reports on the outcomes of its grants, and publicly tests its own assumptions. “We see transparency as one of the core areas in which we are trying to experiment, innovate, and challenge the status quo,” Karnofsky says. Early on, Tuna posted detailed notes on more than 100 conversations she had with funders, nonprofits, and academics. Karnofsky has written about three key issues he changed his mind about. Other Open Phil researchers have published deep research into such questions as the impact of immigration on US workers and the debate about which animals warrant moral concern, the latter in a report that ran well over 100,000 words.

“The analysis of how they operate is out there for the world to see,” says Rob Reich, a GiveWell board member, Stanford political science professor, and codirector of Stanford’s Center on Philanthropy and Civil Society. “You come to understand what they’re doing by reading their preposterously long and complex blog posts. I find that a virtue.”

The Most Important Decisions

So how does Open Phil decide which causes to support? Through a long, arduous, and ongoing process led by Tuna, Karnofsky, and Alexander Berger, who joined GiveWell right out of Stanford in 2011 and is now a managing director and board member at Open Phil. Unlike GiveWell, which supports time-tested, direct-aid charities, Open Phil set out to identify a broader set of giving opportunities, including high-risk, high-return grants. Selecting causes, Karnofsky says, is “one of the most important decisions you make. Maybe the most important.”

Tuna, Karnofsky, and Berger dug into the history of philanthropy. They read case studies of philanthropic success, seeing how earlier philanthropists had promoted the green revolution, funded research that led to the birth control pill, and helped to reduce the risk of nuclear war. (An Open Phil conference room is named Nunn-Lugar, after a 1991 law shaped by philanthropy that brought about the deactivation or destruction of nuclear weapons.) They commissioned new research from historians, to evaluate the role that philanthropy played in driving passage of the Affordable Care Act and marriage equality in the United States.

Patterns emerged. “A lot of philanthropy’s biggest claimed successes have come from improving policy,” Tuna says. “Many came from supporting breakthrough scientific research. Those were two big categories that we wanted to learn more about.”

They made long lists of causes and ranked them according to three criteria that they describe like this:

• Importance: How many individuals does this issue affect, and how deeply?
• Neglectedness: All else equal, we prefer causes that receive less attention from others, particularly other major philanthropists.
• Tractability: We look for clear ways in which a funder could contribute to progress.

Tuna, Karnofsky, and Berger identify as effective altruists. So it’s no surprise that this process of cause selection led them to a set of causes and programs that align, more or less, with those of the movement.

Take, for example, existential risks—that is, the possibility that future events will devastate or end humanity. Effective altruists worry a lot about this, for better or worse. “In the beginning, EA was mostly about fighting global poverty,” wrote Dylan Matthews of Vox, after attending EA’s global conference in 2015. “Now it’s becoming more and more about funding computer science research to forestall an artificial intelligence-provoked apocalypse. At the risk of overgeneralizing, the computer science majors have convinced each other that the best way to save the world is to do computer science research.” The concern about x-risks, as they’re called, grows out of the belief that even a small reduction in the likelihood of a global catastrophe has a high expected value because billions of lives are at stake.

Existential risks are, arguably, a classic example of a neglected cause. “Governments do not have the incentive, corporations do not have the incentive to worry about really low-likelihood, super-duper worst-case outcomes,” Karnofsky says. The Open Phil team researched doomsday scenarios, ranging from geomagnetic storms to nuclear war to widespread famine, compiled a spreadsheet, and settled on two focus areas. They work on biosecurity, which aims to protect the world against natural pandemics, bioterrorism, and biological weapons, and on efforts designed to head off the dangers posed by advanced artificial intelligence. (See “The Bot You Can Save” below.)

---

The Bot You Can Save

Two years ago, the Open Philanthropy Project made its biggest grant, a $30 million, three-year donation to OpenAI, a nonprofit group of researchers and engineers dedicated to advancing artificial intelligence (AI) to benefit humanity. It’s a grant about the future—not just the future of AI, but the future of Open Phil.

Signaling the grant’s importance, Holden Karnofsky, Open Phil’s executive director, joined the board of OpenAI. To expand its work on advanced AI, Open Phil is seeking to hire several technical and policy experts. Recently, Karnofsky said that more than half of Open Phil’s grantmaking in the near term will probably be devoted to long-term work like preventing catastrophic risks to humanity, including those posed by advanced AI.

Open Phil is not alone. Stephen Hawking, the theoretical physicist who died last March; billionaire entrepreneur Elon Musk; and Microsoft co-founder Paul Allen, who recently made a three-year, $125 million grant to his Allen Institute for Artificial Intelligence, also have warned of AI’s dangers and supported research to keep AI safe.

Advanced general intelligence “is at least 10 percent likely in the next 20 years,” Karnofsky says, but not enough people are paying attention. Advanced AI is “on a very short list of the most dynamic, unpredictable, and potentially world-changing areas of science,” he says.

Behind Open Phil’s work on AI is the belief that the technology could lead to catastrophe if it is deliberately misused by governments or terrorists, or if not enough care is taken to prevent super-intelligent machines from becoming smarter than people and spiraling out of control.

“If institutions end up ‘racing’ to deploy powerful AI systems, this could create a significant risk of not taking sufficient precautions,” Karnofsky has written. “The result could be a highly intelligent, autonomous, unchecked system or set of systems optimizing for a problematic goal, which could put powerful technologies to problematic purposes and could cause significant harm.”

Yes, it sounds like dystopian science fiction. But Karnofsky takes pains to say that he believes that future progress in AI is likely to do enormous good by, for example, improving the speed and accuracy of medical diagnoses or reducing traffic accidents by making cars safer. He worries that too much focus on the downside could open the door to premature or counterproductive regulation.

But Open Phil AI safety work, like its work on preventing global pandemics, is driven by a worldview that places a high value on the distant future. Nick Beckstead, a philosophy PhD and leading proponent of this view—his PhD thesis was titled On the Overwhelming Importance of Shaping the Far Future—leads much of Open Phil’s work on catastrophic risk. Many effective altruists argue that even very small reductions in catastrophic risks have enormous expected value because they have the potential to save the lives of countless numbers of future beings.

So far, Open Phil has donated about $51 million to about a dozen universities and non-profits that work on AI safety, including OpenAI, the Future of Humanity Institute at Oxford University, and the Machine Intelligence Research Institute in Berkeley, California.

If and when advanced AI is developed, Karnofsky says, “we believe the world would be a lot better off if there’s already a large, robust, excellent field of experts who have spent their careers thinking very deeply about what could go wrong with AI, and what we could do to prevent it.”

---

OpenPhil also applied the criteria of importance, neglectedness, and tractability to US policy issues. Of nearly two dozen listed on a spreadsheet, five rose to the top: criminal justice reform, farm animal welfare, macroeconomic stabilization, immigration policy, and land-use reform. It’s a list that raises eyebrows. Farm animals and not climate change? Criminal justice reform and not inequality?

Climate change is the most glaring omission from the priority list. “While climate change is obviously highly important,” Karnofsky says, “we thought there were other similarly important issues that were more neglected and more tractable.” Other foundations have poured billions into climate change advocacy in the United States, with little to show for it, critics say. Karnofsky told me, “Some degree of emissions reduction is tractable. Getting below 2 degrees over the relevant time period—that looks really tough.”

That helps explain why Open Phil has supported research into geoengineering, a term used to describe large-scale efforts to artificially cool the planet and offset some impacts of climate change. It gave $2.5 million to Harvard, to support research into solar geoengineering (the project of cooling the Earth by reflecting sunlight away from it) led by David Keith, one of the field’s leading experts. (Very little philanthropy and almost no government money supports the study of geoengineering.) OpenPhil also kicked in $3 million to an effort led by the ClimateWorks Foundation to help poor countries replace polluting refrigerants with efficient, climate-friendly cooling. Open Phil expects to do more climate change funding in the future, Karnofsky says.

In contrast to climate change, the plight of farm animals—a cause favored by many effective altruists and championed by Peter Singer since the publication of his landmark book, Animal Liberation, in 1975—cleared the hurdles of importance, neglectedness, and tractability. Tuna, Karnofsky, and Berger believe that farm animals can likely feel pain, and that their suffering matters. Spending on farm animal welfare, across the entire animal protection movement, previously amounted to no more than $25 million a year. “Very little [money] is going to the welfare of farm animals, even though there are billions and billions of farm animals being raised around the world in terrible conditions, every year,” Tuna says.

Open Phil stepped up in a big way. Late in 2015, Lewis Bollard, a Yale Law School graduate who worked at the Humane Society of the United States, joined Open Phil as its first program officer for farm animal welfare. He was the hire who faced a “steep learning curve,” but he made a rapid ascent, disbursing $47 million in grants to 50 nonprofits in 24 countries since then. A flurry of donations supported activist groups to push large companies to end the practice of confining egg-laying chickens in small cages. Peter Singer has described these caged hens as “the most closely confined, overcrowded, and generally miserable animals in America,” and there are a lot of them—about 320 million at any given time.

The funding transformed nonprofits like the Humane League, which previously had an annual budget of less than $1 million a year and fewer than a dozen paid staff. “They were amazing people, with amazing ideas, working on a shoestring,” Bollard says. The Humane League was granted a total of $5 million by Open Phil in 2016 and 2017; this year, it expects to spend more than $7 million and employ 75 people. Its hard-hitting campaigns have helped persuade Kroger’s, Subway, and Panera Bread, among others, to improve their animal welfare policies. Altogether, advocates supported by Open Phil have secured promises to eliminate battery cages from about 300 US food companies. “We’ve seen an exponential growth in campaign victories,” says David Coman-Hidy, the Humane League’s executive director. The organization has professionalized, hiring its first human resources director and staff attorney, and it has raised salaries and benefits. “You no longer have to be psychotically committed and take a vow of poverty to work for us,” Coman-Hidy says.

Open Phil’s push for cage-free policies sparked a backlash, not just from chicken farmers but from Direct Action Everywhere, an animal rights group. Direct Action Everywhere argued that getting chickens out of cages might not improve their lives and could instead lead to “positive brand feelings around eggs” that increase consumption. For its part, the chicken industry has argued that hens are better off in cages.

Open Phil then did what few foundations would: It revisited the evidence on the animal welfare impacts of cages on hens. Ajeya Cotra, a research analyst at Open Phil, spent the equivalent of six weeks compiling a 9,500-word report that acknowledged that cage-free housing gets mixed reviews from some scientists, while affirming Open Phil’s support for cage-free campaigns. Cotra wrote, “We continue to believe our grants to accelerate the adoption of cage-free systems were net-beneficial for layer hens, but we feel we made a mistake by not conducting a more thorough review of the research on this topic earlier.”

If nothing else, the farm animal welfare program illustrates the power of Open Phil. With an expenditure of $47 million—a lot of money for the animal protection movement, but a fraction of Open Phil’s resources—Open Phil supercharged a small number of groups in the United States and expanded the animal protection movement in China, India, and Latin America. (Its grants funded about 190 jobs in the global farm-animal movement, Bollard estimates.) Its money also steered the movement toward welfare reforms for chicken and also steered the movement toward welfare reforms for chicken and fish, and away from the abolitionist agenda or competing strategies, such as vegan advocacy or farm sanctuaries that shelter animals and educate visitors about factory farms.

This is not unique to Open Phil—many big foundations play an agenda-setting role—but the stakes are higher for Open Phil because it seeks out causes that are neglected. Almost by default, it becomes a dominant player. So misjudgments or abuse of its power could have serious consequences, at least to those causes.

Understanding that, people at Open Phil say they strive to practice what some effective altruists call “epistemic humility.” “We have a whole bunch of projects going on that question our basic assumptions,” Karnofsky says, citing the cage-free study and the deep research done by David Roodman, a senior advisor, who has studied immigration, deworming, and whether putting more people in prison for more time reduces crime. (It doesn’t.) Open Phil also tries to improve the accuracy of the judgments of its people—by, for example, engaging in what’s called “calibration training,” which involves efforts to help people avoid overconfidence and become better forecasters.

Whether this will improve decision making at Open Phil is, well, unpredictable. Much depends on the program officers, who are expected to be well connected, well respected, and experts in their field. They operate under a 50-40-10 rule. They have to convince Open Phil’s board that about 50 percent of their grants, by dollar value, are good ideas. Another 40 percent need be merely okay, meaning that Tuna, Karnofsky, and Berger don’t entirely buy in but see why the grant makes sense. The last 10 percent are discretionary, and get a quick approval unless they raise red flags.

Outside the Mainstream

Foundations are among the least accountable powerful institutions in the United States. Provided they fulfill the IRS’s reporting requirements, they have no obligation to explain what they do or, for that matter, to do any good at all. This lack of accountability can lead to insular thinking and ineffective granting. It is also a license to embrace risk, support unpopular or neglected causes, and tackle problems that will take years or decades to solve.

Capitalizing on that freedom, Open Phil practices what Karnofsky calls “hits-based giving.” Philanthropy’s biggest successes, he says, often come from taking risks and being unafraid of failure. “When philanthropists are funding low-probability, high-upside projects, they’re doing what they do best, relative to other institutions,” he has written. This theme runs through much of Open Phil’s grantmaking, including its funding of scientific research.

Consider David Baker, a PhD biochemist and the director of the Institute for Protein Design at the University of Washington. He has been trying for a quarter century to create proteins not found in nature—“artisanal proteins,” the New York Times called them, in a glowing profile of Baker—and he is making dramatic progress. Using crowdsourced computers, cell phones, and open-source software built by a collective of scientists known as the RosettaCommons, Baker and his colleagues have designed thousands of proteins, with enormous potential, including a nano-scale particle formed from several proteins that millions of Americans would have cheered had it been available last winter: a universal flu vaccine.

You’d think that the government or pharmaceutical industry would be interested in an lifetime immunity with a single dose, but no. Instead, Open Phil delivered an $11.4 million grant to Baker and his collaborators to support research into the vaccine, along with the development of the software tools needed to create it. Influenza causes an estimated 290,000 to 650,000 deaths each year, and it has the potential to cause a severe outbreak, such as the 1918 flu pandemic, which may have killed more than 50 million people.

Chris Somerville and Heather Youngs, who oversee funding for scientific research at Open Phil, say the potential of Baker’s work aligns with the goals of Open Phil. Somerville told me, “This grant, for us, is important in several ways. It addresses an existing problem, which is the ineffectiveness of current seasonal flu vaccines. We would like to get a better one-time childhood vaccine that protects you for life. Secondly, it addresses our pandemic risk concerns. One of the most likely pandemics that we would face in the future is influenza. Third, it addresses our goal to contribute to improving science. There a lot of things we’ll be able to do if we can design novel proteins.”

Open Phil’s scientific research program remains a work in progress, with focus areas to be determined. So far, though, since Somerville and Youngs left their academic jobs at the University of California, Berkeley, to join Open Phil in 2016, they have made nearly 20 grants worth about $56 million. Easily the biggest is a $17.5 million grant to Target Malaria, a Gates Foundation-funded consortium that is seeking to develop and deploy genetically modified mosquitoes to curb malaria in sub-Saharan Africa. Other grants have gone to scientists working on specific problems that are neglected by private or government funders, such as research into the repair of damaged livers or methods for mapping the structure of the brain, while still others support “breakthrough fundamental science”—that is, research intended to provide broad insights without focusing on specific short-term results.

The flu grant spans both categories, by targeting influenza while underwriting basic research into the tools to create proteins. Prominent scientists have urged the federal government to spend up to $1 billion a year to research a universal vaccine, to no avail. The drug companies that make the annual vaccines are also uninterested. “If David’s universal flu vaccine works, the [seasonal flu vaccine] industry would be obliterated,” Somerville says.

Baker has received federal funding over the years, but the National Institutes of Health turned down his proposal about the flu vaccine because it was thought to be “too early a stage and therefore risky,” he says. The government’s reluctance to fund experiments with unclear outcomes creates opportunities for Open Phil, according to Somerville and Youngs. Last year, they invited researchers whose applications were rejected during an NIH competition designed to fund high-risk, high-impact ideas to try again. About 120 researchers resubmitted proposals, and four teams were awarded $10.8 million.

It’s further evidence that Open Phil does not to defer to expert opinion or conventional wisdom. “Our interest in neglectedness will often point us to issues where social norms, or well-organized groups, are strongly against us,” Karnofsky says. Outside of the mainstream is where Open Phil wants to be.

Technocrats with Hearts

Visiting Open Phil, you can’t help but notice that most people who work there look alike. Chris Somerville, who held tenured professorships at UC Berkeley and Stanford, is old enough for Medicare, and his colleague Heather Youngs has a son in college. But the rest are in their 20s and 30s. Just one is African-American. Most have degrees from elite colleges; of the 21 staff members on the Open Phil website, five, including Karnofsky went to Harvard.

“We’re not where we want to be when it comes to diversity,” Karnofsky said three years ago. That’s still true, he admits.

Many at Open Phil share his devotion to effective altruism. This could be a problem. Jon Behar, a former colleague of Karnofsky’s at Bridgewater Associates who served on the board of GiveWell, says, “The effective altruism community and its leadership disproportionately represent populations who systematically lack humility (‘the best and brightest’), experience (the young), and access to alternative perspectives (women, people of color, people who remember the ’70s, etc.). That’s a lot of red flags.”

Open Phil’s close association with the effective altruism movement, as well as its detached, intellectual approach, might lead it to miss opportunities to make change—and to learn. William Schambra, a senior fellow at the Hudson Institute, says an undue focus on “the grand cosmic project of global giving” runs the risk of ignoring the messy but important work of attacking small-scale problems and building community.

“The gritty, unpleasant, contentious world of local politics and civic association is a magnificent school of citizenship,” Schambra has written. He told me, “At the risk of grossly generalizing, I would say that every foundation, even the largest ones, should have at least 10 percent of its giving devoted to very local philanthropy, in their own backyards, that people visit personally and keep track of in a very anchored way.” Get your hands dirty, he advises: The “cognitive elites” who staff big foundations need to “develop the notions of citizenship and compassion in an immediate way.”

Schambra is a conservative. His critique is echoed by some on the left. Open Phil’s work on the distant future, in particular, irks some. “I’m more concerned about a mother watching her child die of diarrhea than I am about someone who hasn’t been born yet,” says the founder of a nonprofit, who asked not to be named because he hopes to work with Open Phil.

Others, like Leah Hunt-Hendrix, founder of Solidaire, a group of donors who fund progressive causes, say that effective altruism underestimates the power of grassroots movements. “The most pressing crises of our time are products of our political-economic system,” Hunt-Hendrix writes. “They are deeply historical, rooted in capitalism and imperialism, compounded by racism and sexism. In our pursuit of economic and political change, Solidaire’s fundamental commitment is to social movements, which seek to contest, disrupt, and transform these systems.”

“We do fund movements,” replies Berger, citing grants not just to farm animal welfare groups but to organizations fighting for criminal justice reform, against restrictive zoning, and for more just monetary policy. But it’s fair to ask whether a more diverse board and staff at Open Phil might lead to support, say, for Black Lives Matter and its struggles for racial justice.

---

The World’s Most Expensive Internship

Some interns are paid. Others are unpaid. A few pay, to gain valuable experience.

Kaitlyn Krieger paid, generously.

In April 2015, Krieger and her husband, Mike Krieger, who is the cofounder of Instagram, formed a partnership with the Open Philanthropy Project. They wanted to learn about philanthropy, so she asked to work several days a week at Open Phil, sitting in on meetings and developing charitable projects. As part of the deal, they promised to make $750,000 of grants to organizations recommended by Open Phil.

The experience was invaluable, Krieger says. “I got to learn about biosecurity and farm animal welfare and global catastrophic risks. It was completely fascinating and a little overwhelming.”

She emerged with a commitment to criminal justice reform, which has become the focus of the Future Justice Fund, the grantmaking organization (it’s structured as a donor-advised fund and a legal liability corporation) that the Kriegers founded in December 2015. Criminal justice reform is also a key focus area of Open Phil, led by program officer Chloe Cockburn, former advocacy and policy counsel with the American Civil Liberties Union. A key goal is to reduce the number of people in prisons and jails in the United States.

Future Justice Fund’s grantees include the Alliance for Safety and Justice, the Ella Baker Center for Human Rights, and the Marshall Project. It does not disclose how much it is giving away. Mike Krieger reportedly owned 10 percent of Instagram when was sold to Facebook for $1 billion in 2012.

Holden Karnofsky, Open Phil’s executive director, says that partnerships like the one with Krieger could help Open Phil build relations with other funders and “thus could account for a good deal of our influence.”

Krieger now runs Future Justice Fund from her home, without paid staff. “We are definitely lean and mean,” she says. “Having Chloe as a trusted partner, I honestly don’t feel like I need to hire a program officer.”

---

Questions about Open Phil’s priorities will surely arise more often if, as seems likely, less of its grantmaking goes to prevent suffering today and more seeks to protect future generations. In a long update on cause prioritization published in January, Karnofsky wrote that he and the other directors of Open Phil—Berger, Hassenfeld, Moskovitz, and Tuna—continue to share “a desire to allocate a significant (though not majority) amount of capital to ‘straightforward charity’: giving that is clearly and unambiguously driven by a desire to help the less fortunate in a serious, rational, reasonably optimized manner.” But such charity, guided by GiveWell, will probably amount to no more than about 10 percent of total giving at Open Phil, he said. Last year, it was about 37 percent, which amounted to $90 million. (In absolute terms, the dollar amount set aside for GiveWell and its charities is expected to stay roughly constant. GiveWell executive director Elie Hassenfeld declined to comment.) Funding for farm animal welfare, criminal justice reform, and scientific research is expected to be maintained, at a level of at least $50 million, for the next few years.

The major growth in giving, it seems, will flow to work on global catastrophic risk reduction, including the work on AI and pandemics. “It is reasonably likely that we will recommend allocating >50% of all available capital to giving directly aimed at improving the odds of favorable long-term outcomes for civilization,” Karnofsky wrote. This reflects the worldview, popular among effective altruists, that ascribes very high value to the long-term future.

“I’ve come to believe that there is highly important, neglected, tractable work to do that is suited to improving long-run outcomes for large numbers of generations,” he wrote—even if that means condemning people who live today to misery or early death.

Some will view this as proof that Open Phil is dominated by bloodless technocrats. That would be unfair. Tuna is, by all accounts, conscientious and caring. Karnofsky left a lucrative hedge fund job to start GiveWell, with no assurance of success. Berger donated one of his kidneys to save someone else’s life. Their hearts, as well as their heads, shape their work. More important, the $225 million or so that Open Phil has given through GiveWell has saved many millions of lives and improved the lot of the world’s poorest people. How many other foundations can make a similar claim and back it up with evidence?

 

As this story was being written, Marts & Lundy, a fundraising consultancy, released a survey that found that three out of four gifts or pledges of $10 million or more in 2017 went to colleges and universities. (Taner Halicioglu, an early Facebook employee and colleague of Moskovitz, gave $75 million to his alma mater, the University of San Diego.) The next most popular categories for large gifts—this covered individuals, not foundations—were gifts for the arts, culture, and the environment. Who knows how much good they will do, if any?

And while the Gates Foundation’s work on global health has had a huge impact, its long-running campaign to reform American high schools has been disappointing. “What do you have to show for the billions you’ve spent on US education?” the Gateses asked themselves in their latest annual letter. “A lot, but not as much as either of us would like,” Bill Gates replied. As education scholars Jack Schneider and David Menefee-Libey recently noted in The Conversation, the foundation spent at least $700 million merely to improve teacher evaluation systems between 2008 and 2013 before quietly dropping the program.

“It’s a familiar storyline,” they wrote. “Again and again, policy-makers and philanthropists have teamed up to reform public education, only to find that their bold projects have fallen short.” Indeed, in a grant that was not vetted by Open Phil, Moskovitz and Tuna contributed $5 million to Facebook founder Mark Zuckerberg’s ill-fated effort to support school reform in Newark. Zuckerberg and Moskovitz were roommates at Harvard when they started Facebook.

Of course, it may be that some philanthropic spending on education has produced the desired results, at a cost that justifies the expense. But no one really knows.

In the long run, this may be the most important way that Open Phil stands apart from the crowd. There’s plenty of blather in the social sector about “learning organizations” and “risk taking” and “embracing failure,” but very few foundations publicly share what they learned, take big risks, or account honestly for their failures. Open Phil does all of that, and more. Its devotion to reason and evidence; its commitment to do as much good as possible; and its willingness to report publicly on what works, what doesn’t, and why, are, arguably, unequaled in philanthropy.

Tuna and Moskovitz and their colleagues have started an important conversation. Other funders would do well to join in.

Read more stories by Marc Gunther.