Google’s Academic Influence Campaign: It's Complicated

Update: Google's Transparency Project has posted an addendum to its Academics Inc. report.

Earlier this week the Wall Street Journal published a detailed investigation showing that Google has been systematically paying academics to publish research favorable to the company’s policy and business positions—often without disclosure of the financial relationship. Concurrently, an organization called the Campaign for Accountability published a report from its Google Transparency Project showing the same thing, and naming many more as the recipients of direct or indirect Google dollars.

It’s the kind of thing that would seem to contravene Google’s one-time exhortation “don’t be evil,” and the Journal article seems to have the company nailed. But in the case of the Google Transparency Project, things aren’t that simple. Several academics named in GTP’s database say that they’ve never received money from Google—that they’re innocent dolphins caught up in what was, best case, a methodologically sloppy tuna net.

In at least one instance, the GTP database captured a scholar with no financial ties to Google. Annemarie Bridy, a professor of law at the University of Idaho, got named because of her status as an affiliate scholar at Stanford’s Center for Internet and Society. Stanford CIS does list Google as one of its funders, but the center’s director, law professor Barbara van Schewick, confirms that it has no financial relationship with its affiliate scholars. “The idea that because I have those affiliations I’m somehow tainted by a relationship to the donors to those centers, I think is ludicrous,” Bridy says. “They said I receive indirect funding. That is a verifiably false factual claim harmful to my reputation, which is pretty much the definition of libel.”

Bridy isn't the only one. Van Schewick herself shows up in the database, owing to the Google contribution to CIS. Like Bridy, she has requested the Google Transparency Project remove her name—CIS doesn’t pay her salary. Stanford Law does.

Aaron Perzanowski, a law professor at Case Western Reserve University, seems to have been slurped into the database only by (dubiously-interpreted) association. Perzanowski says he has never taken money from Google, but his frequent co-author Jason Schultz once worked for the Electronic Frontier Foundation and then for the Samuelson Law, Technology, and Public Policy Clinic at UC Berkeley—both of which have indeed gotten Google dollars. But it was in the form of cy pres payments, basically a judge taking money from a class action settlement and directing it to some other recipient. “I suppose in some sense you could say that Google funded that center, perhaps unwillingly, right?” says Perzanowski. “That’s where this idea of indirect funding gets really messy. I’m not suggesting there’s an easy answer one way or the other, but this is saying individuals who at one point two jobs ago worked for an organization like the University of California that received funding from Google are forever tainted.”

It’s not clear how far back and how deep such connections should be to register as compromising. What warrants disclosure? Casey Fiesler, of the Department of Information Science at the University of Colorado Boulder, is in the database because of a Google Policy Fellowship she had in 2011, which paid some of her living expenses while she was working for Creative Commons (another recipient of Google funds) after law school. Fiesler says she doesn’t even remember if Google or CC determined she’d get the money. “I never worked for Google, and as far as I know they had nothing to do with the work I was doing at Creative Commons,” Fiesler says. “And the work I did that summer had nothing to do with the research referenced in the database.”

Here’s where “indirect” and the lack of clarity on what constitutes a disclosure-worthy association become complicated. Professionals are terrible at judging what kind of influence even a small gift or exchange can have on later behavior. And disclosure is a poor metric for potentially unethical behavior.

Still, though, a grad school fellowship is a long way from prospective pay-for-play. “The only way I can think of that would in any way relate to me is that Google is finding impressionable graduate students and giving them money so that maybe when they do policy work later they’ll favor Google,” Fiesler says. It does sound unlikely.

The Campaign for Accountability buys some but not all of that. “A lot of the emails I’ve received sort of don’t change my mind. I think we’ll add what they say about it or their defense,” says Dan Stevens, the Campaign’s executive director. “We’re working on a post or something we might add to allow these folks to voice what they say.” Stevens says that should happen in the next few days.

Bridy, he says, received what might be described as an indirect benefit through her association with the Stanford Center for Internet and Society, but “given that she’s not receiving any payments, it seems like something we can update in the database.” Perzanowski’s connection to Schultz, he says, is harder to figure out. “I think this is one where we’re going to have to think about how we disclose accurately.” And as for Fiesler? “Yeah, I think that’s fine,” Stevens says. “There’s a specific relationship between her and Google. She received direct funding from them.”

The Wall Street Journal article didn’t include Bridy, van Schewick, Perzanowski, or Fiesler. A person familiar with the article’s reporting (who wasn’t authorized to speak for the paper and so asked to go unnamed) says the writers, Brody Mullins and Jack Nicas, went through GTP’s list and retained entries, adding their own academics and throwing out many of GTP’s. Mullins and Nicas did their own investigation and analysis out of due diligence, but also because they were aware that Campaign for Accountability had a history of going after Google specifically.

Stevens says as much. “Our point is to say, look at Google. They’re funding people to extract this result,” he says. “Building this database to show their influence campaign, that’s the point. We don’t want to impugn anybody specifically. The larger issue is Google’s attempt to influence all these academics.”

Even the people who don’t think they should have been in the database agree that the pay-for-play the Journal uncovered and GTP wants to talk about is a pernicious, serious problem. “I think the issue is so important that I don’t take any money and I wouldn’t take any money that could be seen as compromising my independence,” Bridy says.

Perzanowski says he had the same reaction to the initial reports as any casual reader. “I saw the headlines and thought, ‘This is bullshit, people shouldn’t be out there taking money from Google and not disclosing. Who are these bad colleagues of mine?’” he says. “And I pull up the database and, like, that’s my name.” He adds: “Because of the sloppy way this was done, it’s distracting from a really important issue.”

“Sloppy” is, in fact, a word that came up again and again in my reporting. But so did a question: Who funds the Campaign for Accountability? I should have checked before I wrote my initial story, and didn’t—which I'm embarrassed about, because it turns out the organization, dedicated to transparency in financial affiliations, does not disclose its financial affiliations. “It’s just always been our policy since the beginning not to disclose our funding sources,” Stevens says.

Isn’t it … ironic? “Of course I’ve heard that before. I would just say that we’re not a major company trying to extract things from government and policymakers. We have a different purpose,” he says. “Our statement from the beginning has been, let the work speak for itself.”

As a 501(c)(3) nonprofit, the Campaign for Accountability should have an Internal Revenue Service form 990 detailing its funding. It doesn’t. “We started in May of 2015 as a project of the New Venture Fund. They spun off some of their projects into the Hopewell Fund, which we were part of through 2016,” Stevens says. “We’ve been a standalone project through 2017, so because we are still new and getting spun up, we don’t have a 990 yet.”

Prior reporting has said that at least one funder of the organization is Oracle, which—perhaps unrelatedly—has been locked in a bitter legal battle with Google, the outcome of which could bring one side or the other billions of dollars.

That all makes it hard to draw conclusions about whether the Campaign for Accountability made the best-possible decisions with messy data, made bad decisions with messy data, or had some other motive. “Maybe they don’t care about the academics. I think what they probably care about is making Google look bad by whatever means they can,” Perzanowski says. “We’re not the target. We’re just sort of the ammunition.”