Mark Zuckerberg is (Mostly) Right

Mark Zuckerberg has been under immense and continued pressure to moderate Trump’s political speech on Facebook. He has continued to refuse to do so in the face of extremely negative press and direct pressure from his own employees. I think Zuckerberg’s contrarian position on this issue is (mostly) right, and I don’t think Twitter’s moderation policy will hold up to scrutiny or be effective long-term.

First, to put this in context - Facebook is not interested in open standards or giving users control over their own data. They’re not interested in privacy outside of an ecosystem that they control. They are interested in connecting the world, but only so long as that connecting happens on Facebook or its subsidiaries. Other companies that connect people are a threat to their dominance. The charitable interpretation of this is that they can’t reach their goal if they’re pushed out of business by competition. That said, ego and competitiveness almost certainly play a role too.

This isn’t necessarily wrong, it’s also the reason Zuckerberg is a good CEO - he recognizes what might be a threat to their revenue and in the interest of Facebook’s continued existence, he makes moves to account for that competitive risk. It also doesn’t mean that his ideals or drive to connect everyone is some sort of false narrative (I think it’s something he really believes), but it’s of course in the context where Facebook is always the platform that is doing it.

It also doesn’t mean that his ideas on speech must be motivated entirely by what’s best for Facebook, but it’s always good to be skeptical and aware of potential motivated reasoning when considering any argument. I think even with this disclaimer, Zuckerberg’s position on not moderating political speech from democratically elected leaders is not only right, but the only position that makes sense without setting an impossible standard and problematic precedent.

“I know many people disagree, but, in general, I don’t think it’s right for a private company to censor politicians or the news in a democracy.” - Zuckerberg: Standing For Voice and Free Expression

Watch this exchange between AOC and Zuckerberg, specifically the question about taking down lies in political ads. AOC also attempts to clarify that she doesn’t mean ‘spin,’ but lies. This really gets to the core of the issue: who is determining the difference and how? Do we really want private companies, or Mark Zuckerberg specifically, determining which political speech from an elected democratic leader is a lie and which is not?

Zuckerberg gives a diplomatic answer here, but I think he could have pushed back harder. This responsibility should fall to AOC and the other elected leaders in Congress. If there’s to be policy around political speech and social media, it should not be the responsibility of private companies to determine when to censor or not censor the speech from democratically elected politicians, operating in countries with rule of law and a free press.

There are a lot of conditions on that statement, but it’s because the conditions are relevant and important. The same standard cannot be automatically held for politicians in non-democratic countries, countries without rule of law, countries that suppress speech themselves, or countries without a free press. The moderation standard is also different for the comments of regular people not in office (though this is still not a trivial problem). For politically elected leaders that meet these conditions though, it’s a bad idea to have private companies determine what speech from these elected politicians should and should not be seen by the public on their platforms.

The precedent this would set is both unmanageable and contradictory.

It’s unmanageable because most political speech is inherently partisan, which means it will be a political mess to effectively determine whether something is being removed because people find it misleading (‘spin’), because it’s a lie that’s intended to mislead, or because it’s a political policy people just disagree with. It will also be extremely difficult to get consensus on this given the political nature of the speech. It’s inherently contradictory because if Trump says something outrageous and stupid that gets written about or quoted in the New York Times and then that is subsequently posted and shared on Facebook, should that kind of secondary quotation be removed as well? What about a less reputable paper? or a blog post?

This doesn’t mean private companies do not have a responsibility to be ethical in the absence of policy - they do, but when it comes to the moderation of elected political speech I think the only tenable outcome is to act as a neutral platform. People should be able to see the speech of their democratically elected politicians. A bias towards free speech and anti-censorship is the ethical position. Citizens should be able to see this speech without it being blocked or modified, even if it’s stupid. The right action is for those citizens to then leverage their own speech to protest, act, and vote them out.

Even beyond the logistic hurdles and the congressional responsibility, there’s a deeper ideological issue. Once you’re setting limits on the political speech of a democratically elected politician, the rules that govern speech are now a political problem. Would you be comfortable with members of the party opposite of you deciding what speech should be restricted? If you’d only be comfortable with the members of the party you identify with setting the speech policies, then that itself is a problem. The precedent that restricting this kind of speech sets is itself harmful to a democracy’s future.

None of this is a defense of Trump, and articles that pretend otherwise are setting up a tired strawman just to knock it down. Much of what Trump says is false, intentionally inflammatory, divisive, and self-interested. His policies go out of the way to be cruel (separating children from their mothers at the border as a deterrent) and anti-democratic (restricting and lying about mail-in voting, denying election results). He is likely the worst person to ever have held the office, and those that continue to enable him have continued to fail both in their responsibility to the Constitution, and to their republic.

It’s possible to both find Trump to be a reprehensible person and think that private companies should not be the unelected moderators of political speech, even if legally they are well within their rights to do just that on their own platforms.

This doesn’t mean there’s nothing Facebook can do, or even that they’re doing nothing wrong.

It doesn’t mean there’s nothing they can do because they are right to focus on the authenticity of the speech. A lot of the political disinformation in the 2016 election came from Russian political interference that intended to manipulate the public, and other foreigners taking advantage of the political polarization by stirring up controversy to generate ad-revenue. Facebook should work to make sure political speech is coming from who it claims to be coming from (and they are focused on this).

It doesn’t mean they’re not doing anything wrong because they’ve long ago abandoned the chronological news feed in favor of an algorithmically sorted news feed focused on engagement. This is decidedly not neutral. Abuse of these engagement algorithms are what allowed spammers to leverage the viral nature of misinformation to enrich themselves. It’s also what gave credibility and reach to the Russian political interference. If Facebook wants to act as a neutral platform for speech, then they should be neutral. If they are elevating certain content algorithmically then they are acting as a publisher. This weakens their argument about being a neutral platform and makes them more responsible for what they choose to elevate on Facebook.

The other issue is the degree to which Facebook enables targeted political advertising.

Politicians have always tailored their message toward the audience they’re talking to, but I’d argue the precision and scale at which they’re able to do this now via targeted Facebook ads is a difference in kind. Most ads exist to persuade rather than inform, but with precise targeting and the selection of only specific messages for certain audiences, Facebook ads can leverage confirmation bias to maximize their ability to manipulate and polarize the public. The Trump campaign did exactly this in 2016: if you were anti-immigrant you’d get the anti-immigrant ad, while if you were Republican but not anti-immigrant you might get the tax cut ad. If you were anti-Trump then you’d get an anti-Hillary ad encouraging you to stay home. I’d argue this is a weaponization of the political process and this kind of targeting should probably be banned.

I think this is particularly problematic because at its best government and politics exist to enable effective human coordination at scale by getting the consensus of a society, and enabling us to do the best that we can. Precise ad targeting at scale that focuses on specific issues for individual voters while hiding information that may contradict their preconceptions only serves to further polarize a country’s citizens by playing into their pre-existing biases and turning everyone into a one issue voter. It’s manipulative, anti-democratic, and the opposite of consensus building - it’s harmful for a democratic society.

There are also a lot of things to take issue with in Facebook’s history that I’m not excusing, most of which revolve around their weak protection of their users’ data and general willingness to give or sell this access to others. I also take issue with their expansion into countries without understanding the language which allowed hoaxes to spread that led to people getting killed and the empowerment of dictators. Steven Levy’s book, Facebook: The Inside Story gives a lot more context and insight into a lot of Facebook’s history, both the good and the bad. I recommend it.

I also reject the idea that this position on speech is the easy or self-interested thing for Zuckerberg to do, or that this is in Facebook’s interest. Political advertising makes up a tiny percentage of Facebook’s revenue. If this wasn’t a principled position they could just ban it entirely and save themselves the headache. Twitter’s limited moves to add a link and hide Trump’s tweet behind a warning generated enormous positive press for Twitter without any serious downside. If Zuckerberg wasn’t operating on principle he could just do the easy thing alongside (likely mild) cheers from his employees and the press, but the easy thing to do isn’t always the right thing. I think he’s trying to do what he thinks is right.

The kind of person who refuses one billion dollars at nineteen in order to pursue what they think is right is not easily pressured into doing something they think is wrong. For better or worse, Mark Zuckerberg thinks he is right.

I think he’s right too.

Post 1/6 Update

I think this essay holds up well after the insurrection on 1/6. I think what changed was not Facebook’s policy, but the way Facebook viewed the USG. After 1/6 the Trump administration no longer met the rule of law conditions required to get a pass so the ban became justified. I think his continued unwillingness to accept the election results in addition to his actions on 1/6 support this decision. In the end it’s a private company exercising its speech rights against the government.