We can hold several ideas in our minds, even though they’re in tension, and try to work through to a better solution.

One one hand …

• Any platform for discussion and communication needs rules. It won’t work if it’s wide open.
• A privately owned platform is free to make up its own rules, and even to enforce them at will (except as governed by contracts that it has freely entered). A private actor is not bound to permit speech it dislikes or to use due process to regulate speech. It enjoys freedom of the press.
• Donald Trump was doing great damage on Twitter and Facebook. It’s good that he’s gone.

Yet …

• It is highly problematic that a few companies own vastly influential global platforms for communication without being accountable to any public. The First Amendment is a dead letter if the public sphere is a small set of forums owned by private companies.
• Twitter’s reasons for banning Trump seem pretty arbitrary. The company refers to how Trump’s tweets were “received” by unnamed “followers” and invokes the broad “context” of his comments. But speakers don’t control the reception of their words or the contexts of their speech. A well-designed public forum would have rules, but probably not these rules.
• If a US-based company can ban a political leader in any given country (including any competitive democracy), then democratic governance is threatened.
• Facebook, Twitter, and Google profit from news consumption, denying profits to the companies that provide shoe-leather reporting. Fewer than half as many people are employed as journalists today, compared to 10 years ago. This is at the heart of the current, very interesting battle between the Australian government and the big tech. companies.
• These companies deploy algorithms and other design features to maximize people’s time on their platforms, which encourages addictiveness, outrageous content, and filter bubbles and polarization.

Regulation is certainly one option, but it must overcome these challenges: 1) private communications companies have genuine free speech rights. 2) Forcing a powerful company to make really good choices is hard; externally imposed rules can be ignored or distorted. 3) The fact that there are 193 countries creates major coordination problems. (I wouldn’t mind if a patchwork of inconsistent rules hurt the big companies–I think these firms do more net harm than good. But it’s not clear that the resulting mix of rules would be good for the various countries themselves.) 4) The major companies are very powerful and may be able to defeat attempts to regulate them. For instance, they are simply threatening to withdraw from Australia. 5) There is a high potential for regulatory capture–major incumbent businesses influencing the regulators and even using complicated regulatory regimes to create barriers to entry for new competitors. Imagine, for example, that laws require content-moderation. Who would be able to hire enough moderators to compete with Facebook?

Antitrust is worth considering. If the big companies were broken up, there might be more competitors. But you must believe very strongly in the advantages of a competitive marketplace to assume that the results would be better instead of worse than the status quo. Metcalfe’s Law also tends to concentrate communication networks, whatever we do with antitrust.

Another approach is to try to build new platforms with better rules and designs. The economic challenge–not having enough capital to compete with Google and Facebook–could be addressed. Governments could fund these platforms, on the model of the BBC. I think the bigger problem is that the platforms would have to draw lots of avid users, or else they would be irrelevant. They would have to be attractive without being addictive, compelling without hyping sensational content, trustworthy yet also free and diverse.

Those are tough design challenges–but surely worth trying.

See also: why didn’t the internet save democracy?; the online world looks dark; democracy in the digital age; what sustains free speech?; a civic approach to free speech, etc.