Tuesday, President Trump tweeted that Google “RIGGED” search results, saying “96 percent of news search results on ‘Trump News’ are from National Left-Wing Media, very dangerous.”
The claim was originally made by Paula Bolyard, a writer at PJ Media, and got picked up by Lou Dobbs. Bolyard manually counted Google results for “Trump,” and found that of the top 100, no stories appeared from “National Review, The Weekly Standard, Breitbart News, The Blaze, The Daily Wire, Hot Air, Townhall, Red State,” while there were many stories from CNN, The Washington Post, NBC, The Atlantic, Politico, The New York Times, The Wall Street Journal, and the like.
The methodology PJ Media used to classify publications’ “bias” was questionable—The New York Times was ranked farther left than InfoWars is right—but Bolyard’s point wasn’t entirely off-base: Google does feature work by traditional media organizations more than insurgent conservative outlets.
Of course, Google’s ability to divine “quality” as distinct from “popularity” is limited. Search-ranking technology relies on the implicit votes of readers, with all the human biases that come bundled with them. Google, for its part, categorically rejected the claim that it tinkered with search results for political reasons. “When users type queries into the Google Search bar, our goal is to make sure they receive the most relevant answers in a matter of seconds. Search is not used to set a political agenda and we don’t bias our results toward any political ideology,” a spokesperson told The Washington Post. “Every year, we issue hundreds of improvements to our algorithms to ensure they surface high-quality content in response to users’ queries.”
There you have it: Google says it is optimizing for relevant answers and high-quality content, not “political ideology.” But—and this is the part Google won’t say—what if relevance and quality are not equally distributed across the media?
As a thought experiment, pick some markers of relevance (timeliness, impact, keyword density) and quality (originality, sourcing, depth). Now apply a ranking algorithm to the work of the mainstream media, which is to say the big networks, papers, and digital sites that abide by traditional journalistic values. Then apply the same standard to the conservative publications that PJ Media bemoaned the lack of.
Of course the mainstream organizations—with larger staffs, generally better-trained journalists, and deeper roots in the field—would rank higher. The New York Times and Washington Post have thousands of journalists between them. Last year, The Times’ subscription revenue broke a billion dollars. The institutions of modern conservative journalism are simply not staffed or funded at the same level as what gets called the mainstream media.
Many right-wing outlets are embedded inside advocacy groups, like the Heritage Foundation’s The Daily Signal. Others are tiny blogs without the human resources to do original reporting: According to its staff page, HotAir, which Bolyard cited, has four editors (one of whom is pseudonymous). Even The Blaze, another outlet Bolyard cited, is reportedly down to fewer than 50 employees; the august Weekly Standard looks to have an editorial staff of only 35. Still other right-wing media organizations don’t adhere to the standards of journalism as the mainstream media recognizes them, peddling conspiracy theories or engaging in ethically questionable “reporting” practices or vowing to “break down the barriers between news and opinion, journalism and political participation.” Left-leaning outlets like Salon and DailyKos likewise shouldn’t expect to compete with The New York Times on Google placement.
All media outlets have to reckon with the power of opaque platforms, and there is plenty to critique about Google’s attempts to rank news stories, let alone URLs. The company’s concept of “relevance,” for example, is caught in a strange loop between what people want and what people think it provides: Google sees pages as relevant if people engage with them, but people trust Google to serve up relevant things, so they engage with what Google shows them.
Quality is hard enough to define among journalism experts, and though some people have laid out indicators of credibility, Google has had a notoriously hard time defining what journalistic value means in machine-readable terms. Some things can be measured (How many fresh quotes are in a story from new sources? How many links are in the story to other high-quality sites? How many readers decided to spend a substantial amount of time with a story? How many different types of readers engaged with a story?), but none of them are airtight indicators of quality. The algorithm will rank plenty of great stories poorly. The system can, to a limited extent, be gamed. It’s far from perfect.
But even if the methodology is flawed, Google applies it equally to all the media organizations in its news universe. It might not be a “free” marketplace of ideas, but it is a marketplace with fairly well-known and nonpartisan rules. If right-wing sites aren’t winning there, maybe Google isn’t the problem.
Google’s mission, as a private company, is to organize the world’s information. It has chosen certain computable ways of doing so and applied them across the media, regardless of the ideologies informing the reporters, editors, publishers, or funders of a media outlet.
There is a reason that Microsoft’s Bing News or Apple News have nearly the same mix of news sources as Google News: By reasonable, measurable standards, those organizations are the ones reporting the state of the world best. Is it any surprise that the main right-leaning media organizations that do make the cut at Google are the Wall Street Journal and Fox News? Those two organizations have real reporters doing journalism that conforms to the industry’s standards.
It would be great to have 10 Wall Street Journals out there doing the hard work of reporting on America. But that’s not what’s happening, and Google shouldn’t be forced to pretend that it is.