Google CEO Sundar Pichai answered questions from lawmakers at a Congressional hearing Tuesday, focused on Google’s alleged bias against conservative viewpoints. There was no resolution of the ongoing accusations, but the lines of questioning were revealing.
Facebook and Twitter faced similar accusations at a congressional hearing earlier this year, and the allegations have become increasingly common from the right, and enjoyed a high-profile boost from President Donald Trump’s notorious Twitter account in August.
Google has denied the accusations of bias, and continued to do so at the hearing. Pichai said Tuesday:
“I lead this company without political bias and work to ensure that our products continue to operate that way. To do otherwise would go against our core principles and our business interests. We are a company that provides platforms for diverse perspectives and opinions — and we have no shortage of them among our own employees.”
But Google’s critics on the right are obfuscating the nature of search engine algorithms, and more crucially, Trump’s uneven base of support in general.
In one exchange, Pichai was questioned specifically over the search results for the word “Idiot,” which are dominated by photos of Trump.
“Right now, if you google the word ‘idiot’ under images, a picture of Donald Trump comes up. I just did that,” said California Representative Zoe Lofren, a Democrat. “How would that happen?”
The line of questioning came after extensive accusations of bias from Texas Representative Lamar Smith who said he had “irrefutable” evidence, in a controversial study from Robert Epstein, as well as a blog post that was described by its author as specifically “not scientific.”
Florida Republican Matt Gaetz focused on evidence that Google employees are themselves biased against Trump, citing reports that they had discussed “resisting” the president in a chat room conversation, and a video in which employees say they were upset by Trump’s 2016 victory.
There’s quite a bit to unpack in these accusations, but first and foremost, Google employees do not personally assemble search results. Google, a California based company, may very well employee more liberals than conservatives, as with many other Silicon Valley companies.
Instead Google’s search uses an algorithm that takes into account “things like relevance, freshness, popularity, how other people are using it. And based on that, you know, at any given time, we try to find the best results for that query,” according to Pichai.
In an example of how search algorithms actually work, Google’s system seems to have been manipulated by a deliberate third-party effort to link searches for the word “idiot” to Donald Trump. Reddit users started a campaign to simply upvote a post that contained Trump’s photo and the word idiot.
But the fact is that if enough Trump supporters wanted Google to generate more favorable results for Trump, they have the same power to manipulate results, or to simply overpower negative coverage. But there may simply not be as much positive coverage, or as much activity surrounding that coverage, to outweigh negative content. An analysis by the Brookings Institution showed that over the course of 2017, internet search rates on Donald Trump dropped to 20 percent of their peak during his inauguration week.
In the same period, searches on impeachment spiked several times.
Ostensibly “unbiased” search results that generate half negative and half positive coverage might not be an accurate representation of what’s happening on the internet. The notion that it would plays into a broader false narrative that conservatives have been promoting for years, with considerable success.
Functions like the electoral college, Republican gerrymandering, and Trump’s billions of dollars worth of free media coverage in the 2016 election gave Republicans 2 years of an iron grip on power in Washington DC. Yet, a closer look at the numbers suggests that American politics may not be the fifty/fifty split between Trump’s supporters and detractors that this narrative suggests.
It bears repeating that Clinton won the 2016 popular vote by roughly 3 million. Only around 25 percent of voting age Americans actually voted for Trump in that election. In the midterms last month, Democrats lost seats in the Senate, yet received a whopping 12 million more votes than Republicans in Senate races overall. Both the electoral college and the nature of the Senate give more power to rural states, and less power to urban coastal areas where most Americans actually live.
In many ways, the internet is much more straightforwardly democratic. It isn’t constrained by national borders either, and outside the US, Trump has earned even less support. According to a recent survey of 25 nations from Pew Research Center, only about 27 percent of respondents had confidence in Trump.
In countries like Germany, France, and Spain, his support was even lower, at ten percent or less. If most discussions on the internet take a dim view of Trump, this doesn’t mean there’s a bias. Much more likely, it means most internet users take a dim view of Trump.
Google’s search algorithm is likely less than perfect. But accusations that anti-Trump sentiment among employees has driven a conscious effort to suppress positive coverage are completely unfounded.
“If you want positive search results,” California Democrat Ted Lieu said at the hearing, “do positive things. If you don’t want negative search results, don’t do negative things.”
Trump’s polarizing political strategies may have worked well in shoring up majority support in the key battlegrounds necessary to win elections and hold power in the US. But that doesn’t mean that his policies or rhetoric are actually appealing to anything close to a majority of people, which would be necessary for the kind of “balanced” online coverage conservatives say they wish to see.