If you use social media, can your vote be unbiased?

Do digital media outlets have political bias, and can they manipulate how we vote and see elections because of it? We asked Dipayan Ghosh, Ph.D, the Pozen Fellow at the Shorenstein Center at the Harvard Kennedy School. He was a technology and economic policy advisor in the Obama White House and served as a privacy and public policy advisor at Facebook.


The past few weeks have seen President Donald Trump fuel the already-heated public controversy over supposed political bias on digital media outlets. ”Conservatives have been treated very unfairly” by internet companies like Facebook, Twitter, and Google, he claimed, adding that leading technology firms now present a “very antitrust situation.”

Trump and other leading Republicans have pressed forward with this narrative and have taken it beyond rhetoric to action. The nation’s senior-most law enforcement official is leading an inquiry into the matter, and called last month on state attorneys general to come together and discuss what can be done to the digital media platforms through antitrust enforcement.

As the midterm elections approach, these comments and sentiments should be very alarming. Democracy requires a clean and equitable media ecosystem that does not arbitrarily favor one side over the other. The electorate must have the opportunity to access any and all legitimate political material and content it wishes. Individual voters should have the autonomy to develop political positions and choose their electoral representatives accordingly.

A quick glance at these companies’ services might lead an observer to conclude that there is some political bias behind them; Google News search results, Facebook’s News Feed, and the Twitter feed have at times appeared to favor liberal outlets over conservative ones. And conservatives do not appear to be well-represented among the personnel at these companies, which only adds to their dissatisfaction with Silicon Valley.

But conservatives are not alone in harboring the notion that digital media platforms are going after their politics. Many liberals have similarly argued that Facebook favored the Trump campaign by dispatching staff to support Brad Parscale’s 2016 digital strategy operation in favor of then-candidate Trump.

Are technology platforms really spinning their algorithms such that conservative content to sinks to the bottom of our feeds? No. In fact, the reality is that Trump’s claims – along with the many conservatives and liberals who make similar statements about the existence of political bias on these platforms – couldn’t be further from the truth about Silicon Valley’s intentions. The industry’s decision-making on these matters isn’t anti-conservative, anti-liberal, or politically intentional at all. If the past few years have proved anything about the nature of the algorithms that curate content and route personalized ads over these platforms, it is that the companies behind them are not set up to make value judgements. It is simply not in their interest to do so.

The business models behind these companies’ core services — which eventually define and dictate their algorithmic and editorial policies — are simple: they develop highly compelling social media services that have been described as addictive; collect extensive amounts of personal data on their users through those services; maintain behavioral profiles on their users by drawing inferences from that data; and develop algorithms that curate content and target ads. They want to do these things as efficiently as possible, because the key to their entire business is to engage the individual and keep them on the platform, exposing them to greater amounts of ad space that marketers pay small fortunes to access. Messing around with what a user actually wants to see – for instance, by suppressing liberal or conservative content – would only hurt the core of that “pay-for-attention” business model. In fact, advertisers might start taking their business elsewhere — to those platform services that target advertising more effectively.

Digital platform services like Facebook and Twitter also have no interest in making the kinds of political value judgements the President and Attorney General are alleging. Every time they have to set their foot down on one of these issues, they expose themselves to loud public outcries. The people on the “wrong” side of their decisions resent the company, and those who were pushing for them all along wonder why the industry didn’t act sooner. This is exactly what happened in the case of Alex Jones, the conservative conspiracy theorist whose  accounts were removed from many social media platforms.

The industry’s primary interest is to remove itself from the equation entirely by enlisting third parties who represent the public interest to develop policies: the federal government, a certified fact checker, the National Center for Missing and Exploited Children, the Anti-Defamation League, a pesky journalistic investigator. It doesn’t matter as long as their companies are not held liable – in the courts of public opinion or law – for the policy that enforced the final decision.

That is the industry’s perspective. And without a strong regulatory regime dictating and shaping content policy, it is unlikely that we’ll see changes that can hold the industry accountable.

The tech industry needs these policies. And the Attorney General needs to be shown that these companies don’t have some nefarious anti-conservative bias written into their algorithms. But to effectively accomplish this, we have to look under the industry’s hood and begin to give the public the ability to understand why they see certain commercial or political ads over others. Ultimately, we need develop a “digital social contract” – a means by which the voter, the consumer, the citizen, can know where they stand in digital society and have the power to use technology effectively for their means – rather than feeling they are the industry’s pawns in a game they don’t fully understand.

To assure our evolving new media system remains a functioning part of our democracy, we must create a single movement for regulatory reform that can garner support across both sides of the aisle. In recent months, many – from the Democrats pushing internet policy reform, to the Republicans considering appropriate competition policies for the sector – have presented comprehensive and compelling proposals. Ben Scott and I recently charted one possible way forward for regulatory policy.

But until such reforms are implemented, we as individual voters must remain vigilant, and accept that the content we see online may not be what it seems.