The bill gates problem, p.33
The Bill Gates Problem, page 33
“What becomes problematic is when these numbers are imbued with authority. When those numbers actually … change the way institutions perceive health problems in certain countries, then it becomes a question of … will this country get funding to fight HIV depending on what the estimate of prevalence looks like?” noted Marlee Tichenor, an anthropologist at Durham University, in an interview. “In a lot of ways, these estimates shape what can and cannot be done.” Tichenor sees a fundamental conflict of interest between the Gates Foundation being key “financiers of global health initiatives” while also controlling the “means by which we judge whether they succeed or not.” Much of the “lives-saved” mongering that goes into Gates’s public relations, for example, is based on numbers produced by the Gates-funded IHME.
Indeed, if Murray’s criticism—or condemnation—of the WHO was that it was vulnerable to pressure from member nations, don’t we also have to accept that the IHME is itself extremely vulnerable to outside pressure—from the Gates Foundation, which has its own interests in what the numbers show? Why does it make more sense for the IHME to exist in the private fiefdom of Bill Gates than in a democratically run institution like the WHO? Or, bigger picture, why should any institution have a monopoly? Why not create a vibrant scientific discourse with many competing bodies furnishing estimates?
Bill Gates believes the IHME “democratizes information” by bringing together 281,586 data sources from national health ministries, private insurers, and the scientific literature to a public-facing academic institute. The IHME then runs this vast data through complex analyses to present detailed portraits of the state of health, along with a growing body of other metrics, in virtually every corner of the globe. The institute’s website offers interactive maps that allow users to drill down to virtually any village in sub-Saharan Africa, for example, to find out how many years of education people have; how incidences of malaria, HIV, and lower respiratory infections are changing over time; who has access to piped water; even how many men are circumcised.
Again, the numbers found in these maps aren’t hard data but, rather, estimates—educated guesses, really—based on whatever data is available. The Gates Foundation, instead of focusing its money and energy on building up health records and infrastructure in poor nations to collect actual data about death and disease (the way rich nations do), has created a high-tech apparatus in Seattle to churn out good-enough estimates that flatten the Global South into best guesses. This has raised criticism that the IHME’s work effectively amounts to a kind of “data imperialism.”
“It creates an illusion of knowledge. It tells people in a lot of [poor nations] that they don’t know what they know about themselves. That what you think you know, you don’t know,” Seye Abimbola, a senior lecturer at the University of Sydney, said. “That is the colonial experience.”
A perhaps even more fundamental question concerns the quality of the IHME’s work. Scholars widely describe the IHME as a “black box,” a Wizard of Oz–like production that is carefully organized to disallow anyone from seeing behind the curtain. “It’s quite impossible to criticize or, indeed, comment on their methods, since they are completely opaque,” Max Parkin, of the International Network for Cancer Treatment and Research, told me.
Peter Byass, now-deceased professor of global health at Sweden’s Umeå University, offered a similar critique. “From a scientific point of view, that makes it impossible for anyone to replicate or verify the estimates,” he told me.
Ruth Etzioni, a professor of public health sciences at the Fred Hutchinson Cancer Research Center, echoed these criticisms. “It’s impossible to do what they’re trying to do rigorously.… The data is just not there to really quantify the impact of some of these diseases,” she told me. “Instead of saying, ‘You know what? That’s not possible,’ [the IHME says,] ‘Here are some numbers.’ You’ve naturally got yourself in an overpromising situation.”
The IHME counters that “no estimate of a problem is interpreted as an estimate of no problem.” And in an e-mail, it defended its estimates as transparent and published with statistical confidence intervals that inform users about the limitations of its work. Etzioni sees a pattern in its pushing its findings into the limelight while relegating “key caveats and uncertainties” to the fine print. She pointed out that even when the institute made a major mistake in its early Covid-19 projections—it had been using a bad model—it never issued a clear mea culpa.
And it was the high stakes of the pandemic that brought a new level of scrutiny—and competition—to the IHME. A number of researchers began publishing estimates, and began to see, in real time, what they had long suspected—that the IHME’s complex estimates are not always particularly good or accurate. At times, they might even be hurting public health.
In the spring of 2020, U.S. president Donald Trump held a press conference in which his advisers pointed to IHME estimates as evidence that the pandemic would rapidly peak and then wind down in the weeks ahead. “Throughout April, millions of Americans were falsely led to believe that the epidemic would be over by June because of IHME’s projections,” data scientist Youyang Gu told me. “I think that a lot of states reopened [from lockdowns] based on their modeling.”
Gu was one of many modelers who ended up competing with, and outperforming, the IHME during the Covid-19 pandemic, independently producing projections that appeared more accurate than Bill Gates’s half-billion-dollar health metrics enterprise. Again and again during the pandemic, scholars pointed out major mistakes and errors in the IHME’s research, openly lampooning the institute on social media. Yet, no matter how often the IHME’s estimates proved wrong, or how loudly the wider research enterprise screamed, “The emperor has no clothes!” the message never quite got through.
“Many people do not understand how modeling works,” Chris Murray wrote in a Los Angeles Times op-ed, brushing aside his critics before plowing ahead with more highly questionable, headline-grabbing projections. The IHME, for example, began charting the course for the pandemic many months in advance, while competing modelers more conservatively made projections only a few weeks into the future. This put the IHME’s highly contested estimates in a position to guide policy-making ahead of other models, and draw more media attention.
“It seems to be a version of the playbook Trump follows,” the demographer Sam Clark told me in 2020. “Absolutely nothing negative sticks, and the more exposure you get, the better—no matter what. It’s really stunning, and I don’t know any other scientific personality or organization that is able to pull it off quite like IHME.”
When I first reached out to the institute in 2019 and asked about its controversial reputation in the academic community, an IHME spokesperson shot back, “Who is making such criticism, and where has the criticism been published or stated publicly?” Internally, however, the IHME was well aware of this criticism. In correspondence with the Gates Foundation years earlier, released through a public records request, the institute reported receiving criticism as a “black box,” which it acknowledged was a potential “risk” to its future success. Similarly, the institute publicly pushes back on allegations that it has too much power, telling me that “for nearly all outcomes that we publish, there are alternative sources of estimates.” Yet elsewhere, it has called itself the “gold standard in population health metrics” and “arguably the de facto source for global health accounting.”
According to multiple sources, many Gates Foundation staff understand that there are serious problems, if not liabilities, with the IHME. But because Bill Gates personally likes the institute, Murray’s project has become too big to fail—one more illustration of Bill Gates’s top-down style of leadership.
Peter Byass noted in an interview that if the IHME were publicly funded, it would have to operate in a far more open and accountable manner. “If you’ve got enough billions, you can set up a foundation, and you can make the rules entirely as you wish,” he said. The Gates Foundation “is both the rule maker and the rule keeper, in terms of how they choose to scrutinize grant holders. That’s their privilege, because that’s where they are in the marketplace.”
However, the IHME is, technically, a public institution. It’s part of the University of Washington and, theoretically, subject to its oversight. In practice, however, many scholars regard it as a private arm of the Gates Foundation. “IHME, by design, exists in this sort of this gray area,” noted Andrew Noymer. “It’s part of UW, but it’s its own institution. It doesn’t fully answer to them. It’s public when it’s fashionable to be public, but private when it suits them.”
For most of its existence, the IHME was headquartered a few blocks from the Gates Foundation’s offices in Seattle, not on the University of Washington campus. The institute’s first temporary offices were actually located in the foundation’s former headquarters. One former IHME employee told me the Gates Foundation freely calls in bespoke charts and graphs for Bill Gates’s presentations, prompting entire teams of IHME researchers to drop everything else in service to their benefactor. “It really did feel like we were consultants for the Gates Foundation, and the scientific methods we used were often in service of getting the results we wanted … or the story he [Murray] thought the Gates Foundation wanted,” the source told me. “There are thousands of hours cumulatively spent each year just on one-off requests from Bill Gates that trickle through from the Gates Foundation.”
A public records request appeared to confirm this. The IHME, at one point, solicited an additional $1.5 million from the Gates Foundation to address “time-sensitive requests from BMGF leadership [that] often require repurposing IHME staff on the fly from other endeavors to meet analytic requests. Each request has had to be satisfied in addition to normal responsibilities, creating a ripple effect across projects.”
Public records also show the IHME creating a dedicated team to service the foundation. The IHME’s Foundation Response and Engagement Team, according to one grant proposal to the Gates Foundation, was led by Tamer Farag, whose LinkedIn résumé reports that he worked at the Gates Foundation before joining IHME and continued to serve as a “consultant advisor” to Gates while employed at the IHME. (At the same time, notably, Farag also reports serving as an adviser to Mali’s Ministry of Health.)
Most revealing of all, Gates’s original grant agreement with the IHME in 2007, released through a public records request, gave the foundation sweeping authority over the institute: approval rights over new hires for the institute’s executive leadership, approval rights over institute board members, and approval rights over who does external evaluations of IHME and what criteria are used. (Such external evaluations are required by University of Washington bylaws.) Gates also requested “an opportunity to review and approve” press releases and reports related to the work it funds at IHME. The University of Washington signed off on this agreement.
When I first reported this in 2020, sources reached out to me with concern that UW would give such far-reaching influence to a private donor. The American Association of University Professors recommends that schools take steps to preserve “academic autonomy” from funders “by maintaining … exclusive academic control over core academic functions,” including research evaluations and hiring. Some schools have found themselves in hot water for making the kind of concessions that UW has with Gates. After student activists at George Mason University, a public school in Virginia, uncovered that the Charles Koch Foundation had gained influence over university hiring through its charitable donations, an international scandal ensued. Headlines appeared everywhere from the New York Times to the Guardian, decrying billionaire industrialist Charles Koch’s infringement on academic freedom.
“We’re sort of the poster child for ‘Don’t let this happen to your institution,’” Bethany Letiecq, a former associate professor in the College of Education and Human Development at George Mason (now at the University of Maryland), told me. “And a lot of other universities look to us and say, ‘What went wrong? How can we prevent this going forward?’”
I shared with Letiecq my findings about the Gates Foundation, including its purview of hiring, board appointments, evaluations, and press releases. “What we found at [George] Mason [University], similar to what you’re finding with Gates, is they’re given all kinds of benefits or access or oversight based on their funding. We think that’s highly problematic when that comes to academic freedom,” she said. “Once these relationships are established, I do think it is concerning, in the sense that they can change the whole mission of the university—to just be servicing their [private donor] interests. Public institutions of higher education are sort of like the backstop of democracy. They’re so important to the democratic function to critique, to demand transparency, to seek truth and knowledge. I think that these big-money donors, while they’re important to universities … there’s a serious cost, and I think universities are super vulnerable.” Letiecq said that Koch employed a dark-money strategy at George Mason: instead of making donations to the school, which would be subject to public records requests, it made donations to a private foundation adjacent to the university. More than 80 percent of the Gates Foundation’s giving to the University of Washington—$1.5 billion—followed a similar pathway, going to a university-adjacent foundation.
When I asked the University of Washington Foundation about these dark-money concerns, it did not respond. Instead, the university responded on behalf of the foundation. “The same state ethics laws govern, regardless of whether a donation is made to UWF or UW directly,” the school stated. UW also told me that the UW Foundation is currently subject to public records requests, though it did not respond to follow-up questions about whether this was always the case.
It’s worth emphasizing that the Gates Foundation is not a typical donor to UW. The Gates family name is emblazoned across UW’s campus—the William H. Gates Public Service Law Scholarship Program, Mary Gates Research Scholarships, the Bill & Melinda Gates Chairs in computer science, Mary Gates Hall. The Gates family—Bill’s mother, father, and two sisters—has over the decades held a variety of high-level positions at UW, including sitting on its highest governing board, the Board of Regents, and on the University of Washington Foundation’s board.
The university disclaims that Gates has any untoward influence over the school or that the Gates Foundation enjoys special privileges, for example, in its funding of the IHME. “It’s neither in the university nor the Gates Foundation’s best interests to have a relationship that is not based on open science. That is what keeps our reputation as a top research university secure. And quite frankly the Gates Foundation wouldn’t want to be criticized for that either, I don’t think,” said Joe Giffels, senior associate vice provost for research administration and integrity. “The university wants any university activity, including the research that IHME would do, to be free of undue influence and, in particular, bias from any source, quite frankly.”
Giffels was unaware that the IHME had a controversial reputation, telling me, “I’ve not heard of any [ethical concerns]. And I would have heard of them if there were any.” As he described it, the Gates Foundation does little more than write checks. “We would not consider the IHME to be an institute that was, you know, founded by the Gates Foundation. The Gates Foundation has provided a lot of financial support to the IHME—that’s at the IHME’s request. They [IHME] come up with individual projects, research questions they want to have answered and so on, and then they propose to the Gates Foundation—that the Gates Foundation provide funding for those things, as designed by the IHME. And then Gates either says yes or no,” Giffels told me.
I also asked him about the foundation’s role in approving new hires at the IHME. “Do we allow sponsors to approve hiring or possibly firing—that sort of thing? No, we don’t—in the sense that the university is the employer, they are the employer of record, they are responsible for the employment, and they make final decisions over hiring and firing.”
After the interview, I sent Giffels the grant agreement I had uncovered in which the university explicitly agreed to give the Gates Foundation approval rights over new hires for the institute’s executive leadership. The university then appeared to reverse course. UW spokesperson Victor Balta sent me an email saying that this kind of influence was normative and routine for donors to UW. “The level of funder involvement outlined in the 2007 grant agreement is in line with the type of review and approval included in many research grant agreements with government funding agencies, institutes and other nonprofit organizations,” Balta wrote me in an email. When asked for specific examples, he noted that when a university researcher abandons a government-sponsored research project, the sponsor will play a role in approving who takes over the grant. However, this seems categorically different from the broad influence UW has given Gates—not just in deciding who is in charge of its grant (Chris Murray), but also in holding approval rights over new hires across the institute’s executive leadership alongside other rights and privileges.
After many email exchanges, the university began repeating the same answer: “UW would not sign a grant agreement that does not align with our mission and values.” What I see in these rote responses, and in UW’s failure to meaningfully reconcile the contradictions at hand, is an institution deeply committed to protecting its relationship with a valuable funder. It’s a narrative that academics at other institutions know well. “It all sits under that bucket of undue donor influence and the university’s willingness to sell academic freedom to the highest bidder,” said Letiecq. “Whether it’s the Gates Foundation or the Charles Koch Foundation … the threat to the academic freedom is the same.”
