The Washington airwaves have been filled with talk about alleged anti-conservative bias on social media. On Thursday, President Trump held a social media summit of right-wing internet name brands at the White House in which he blasted Twitter and company for “terrible bias.” Next week, Sen. Ted Cruz, R-Texas, will be holding a hearing pitting Google against conservative vlogger PragerU, who claims the search engine has downgraded the vlog’s search rankings.

Ideological diversity online is certainly valuable, and major tech companies have a role to play in fostering that diversity. Yet if Trump, Cruz, and others want a politically neutral, unbiased platform, there’s a simple solution: They should make one. Call it “freespeech.gov” and guarantee that all ideologies will be treated equally.

A federal online platform is not far-fetched as a legal or technical matter. More importantly, it helps to highlight why running a private platform that embraces values like tolerance and free expression is actually really hard.

Perhaps counterintuitively, a public social media service actually fits right in to the government’s traditional role. In the real world, traditional public forums include streets, parks, and town halls — all government-run property. Such venues are perfect for political neutrality because the First Amendment demands it. Officials can restrict the time, place, and manner of speech on public grounds, but they cannot discriminate among viewpoints. The village park of Skokie, Illinois, must be as open to neo-Nazis as anyone else.

freespeech.gov platform would be a virtual public square, so the First Amendment itself would prohibit anti-conservative or other bias. Contrast this to Republicans’ currently favored approach of legislating political neutrality on platforms. Private platforms are emphatically not public forums, as the Supreme Court has explained, and worse yet, this neutrality legislation might be an unconstitutional compelled-speech mandate.

Little stands in the way of building freespeech.gov. The government has its digital consultancies 18F and the U.S. Digital Service to produce the technology. It has already created the “We the People” White House petitions website, a virtual public square where hundreds of thousands of users call for everything from deporting Justin Bieber to building a Death Star. A freespeech.gov site would likely be popular as well — after all, what teenager could resist sharing dank memes under a .gov URL?

Free-market thinkers might initially balk at “socialized social media,” but this government-run service would likely do little to undercut private businesses: Most online users have multiplesocial media accounts, and so people would likely use freespeech.gov alongside, not instead of, other platforms.

If constructed, freespeech.gov would be a teachable moment for the president and Congress, because they will be faced with an important problem: how to moderate content online. At a recent hearing on children and the internet, members of Congress seemed to think that content moderation is an easy task — that illicit material is simple to identify and weed out.

Yet even the government itself, which is bound to political neutrality by the First Amendment, has failed to figure out content moderation. Consider registration of trademarks, where the government approves business names and logos to stop wrongful copying. The registration is a sort of public forum subject to the First Amendment, but for decades, Congress thought that it was fine to prohibit registration of scandalous or derogatory names — until a band called The Slants pointed out that this content moderation amounted to discrimination against political viewpoints, and the Supreme Court agreed.

Or consider We the People again. Surprisingly, the White House petition site has a stringent content moderation policy, prohibiting petitions for commercial endorsements, obscenity, profanity, and “degrading slurs.” (Query the legality of that last prohibition in view of the above Supreme Court case.)

How that policy has been applied is even more concerning. In 2017, a white nationalist group alleged that the We the People content moderator had removed the group’s petition on college funding. I have no love for white nationalists, but kicking them off a White House petition page is plainly viewpoint-based censorship. Ironically, the group noted that the Obama administration had allowed the same petition, so it was the Trump White House that was censoring far-right views.

Managing content moderation, whether on a trademark register, a petitions website, or a social media platform, is hard work. It involves a Sophie’s Choice between allowing the most degrading, offensive, harmful content, and moderating in ways that will inevitably offend someone’s political tendencies. If Trump and Congress want unbiased social media, they can try their hand at moderating freespeech.gov. If they decline or fail, perhaps they will have a better appreciation for why the tech industry has found the problem so difficult as well.

Image credit: Wachiwit