Twitter without trump it is much less about freedom of speech than you think
by Detlef Eckert
1 February 2021
Twitter, Youtube, Facebook, Snapchat and others suspended the accounts of Donald Trump. Google and Apple removed ‘Parler’ from their stores. At the time of writing:
Whilst many commentators felt relieved, several politicians voiced concerns:
The German chancellor Angela Merkel called it ‘problematic’.
The French Finance Minister Bruno Le Maire expressed disagreement with Twitter’s ability to remove the American president from the platform. The digital oligarchy cannot do regulation.
The Chairman of the EPP at the European Parliament, Manfred Weber, requests that the EU should not let Facebook and Twitter decide what is acceptable on their platforms.
Back in July, I addressed the issue of ‘warnings’ attached to Twitter messages from Trump. In November I felt social media provide a challenge for democracy. They reinforce the emergence of opinion bubbles (echo chambers), in particular through their recommender algorithms.
Social media give the impression that everybody has the right that his or her voice is published and shared. They provide a powerful basis for demagogues and conspiracy theories. Against this background, does regulation provide the answer as the politicians claim? Let’s break the answer to this question into three parts: First, what is social media? Second, what is free speech? Third, what could regulation say?
Social Media
Social media are private, profit-making companies like newspapers and TV networks. They serve a variety of purposes: commerce, media, and enable people to exchange photos and videos, personal experience, communicate with friends and express opinions. Thus platform regulation must differentiate. The ‘Trump-Twitter’ debate concerns platforms that allow interpersonal communication and distribution of information, audio-visual material, but not relevant for platforms like Amazon or Microsoft Azure.
Platforms define a policy informing their users before they sign up what to expect and how to use their network, i.e. terms and conditions. Thus, if a platform only wants contributions about nature or science, they might not allow commercials for fertilizers or religious content. Platforms are free to set these rules as long as they are in line with general law. As most platforms are US based, they base their policies and user rules on US law, including moral considerations. Hitherto regulation does not question the right of platforms to set their own house rules.
Both in the US and in the EU regulation exempts platforms from liability as regards user generated content as long as they follow a ‘notice and take down’ of illegal content procedure. This is stipulated in so-called section 230 of the Communications Decency Act in the US and the eCommerce Directive in the EU. However, with Trump the issue is not illegal content but violation of the ‘terms and conditions’ of the platforms, i.e. their user policy not to tolerate incitement of violence. Politicians think they through regulation should take such decisions and not ‘digital oligarchs’ (see above). They link their claim to ‘freedom of speech’.
Freedom of Speech
Freedom of speech is a constitutional right in democratic countries, within legal boundaries, which in turn must be in line with the constitution. In some countries these boundaries are wide like in the US, in others they are more restricted. Many people seem to consider freedom of speech as their right to be heard, published and being taken seriously. However, freedom of speech is a defence right to protect citizens against the government. The government cannot prevent me from expressing my opinion about politics and everything else. Whilst I, as a citizen, can express my opinion, I do not have a right that others share it or media publish it. However, one phenomenon of social media is that many precisely think that, including politicians that are ‘being stunned’ that owners of platforms can exclude one of them. This development has given the impression that platforms are the guarantors or enemies of freedom of speech, respectively. Platforms play a role, but they do not establish new rights for being heard.
Some commentators speaking against banning Trump on Twitter argue he is not everyone but a political figure, and at the time of the Twitter ban, he was US President. Indeed, in the past Twitter used a different yardstick for politicians than for ordinary people. Why was this justified in the first place? Twitter argued that politicians have a public role and people need to know what they think, unfiltered. Yet, does this apply to millions of local politicians? What about celebrities? What about experts in certain fields? And should politicians not meet higher standards?
Even without Twitter, as a US President or a prominent political figure, Trump has access to a plethora of media outlets. He can hand out press briefings and give TV interviews. He even has privileged access to certain media, in particular Fox News. Twitter is not the only online platform, yet without doubt Twitter is a uniquely important platform.
Twitter is dominant in its platform segment. This is a common problem with platforms. They become dominant because of network effects: The more people use a network, the more value it has; the more value it has, the more users will join, which will determine advertisement and profitability. This is a virtual circle leading to what we see today, a handful of platforms, each of them a de facto monopoly in their segment. Such a cumulation of power is difficult to handle through competition policy as the monopoly emerges through a market process. Yet, one wonders why these platforms still can acquire other platforms and hi-tech start-ups, cementing their market position.
So, here we are. What can regulation do?
Regulation
The European Commission has proposed two regulations of platforms, with particular attention to dominant ones, and one of them is its ‘Digital Services Act’ (DSA). In short, it obliges platforms to be transparent, consistent, and responsive. It adds some valuable provision to protect users against unfair treatments by platforms, for instance, when other user try to silence political opponents. In principle, such a regulation would have given Trump a basis to challenge the decisions of Twitter, but with no guarantee of success. Facebook, for instance, has transferred the decision to ban Trump to its ‘Oversight Board’. However, this is a voluntary decision by Facebook and not requested by law.
Thus, the gist of the discussion comes down to the ‘terms and conditions’, which—see above—platforms are free to set within boundaries of general law, for instance, anti-discrimination law. When politicians claim platforms should not be the ones deciding about what is permissible, then regulation would have to target the ‘terms of conditions’, which the draft DSA does not propose.
Interestingly, a Polish proposal for a platform regulation attempts to do precisely this. The Polish draft regulation only allows platforms to delete content or ban users based on illegality stipulated by law. This looks like a lesson a government has drawn from Trumps’ experience. It is a kind of reverse ‘free speech’. The government protects itself against platforms. Such a law would invite the government to extend the definition of illegal content in a politically motivated direction. Hopefully, this attempt to target ‘freedom of speech’ through the backdoor will not fly.
I have to admit that it would be ‘nice’ if certain ‘shitstorms’ could be stopped and could not poison a civilized societal discourse. Yet, the limits and boundaries are difficult to draw. In many countries, regulation has established councils, oversight boards or ethical committees to ensure that traditional media, TV and Radio, follow certain rules. However, the more public intervention, the more institutions involved, the more it will be difficult to remove content from established politicians or ban them from a platform. Private sector that worries about its reputation may paradoxically better suited to this.
Conclusions
Triggered by Twitter’s decision to ban Trump from its platform, politicians claim that regulation should decide what is permissible, even beyond illegal content, and not the owners. Yet, there is a snag.
Unless regulation submits the ‘terms and conditions’ of platforms, i.e. their rules for users, to authorization and establishes oversight authorities, platforms will maintain discretionary power to decide what content they want to accept. Such a regulation would cause problems, in particular it would be difficult to enforce globally, conflict with business freedom, not justified by ‘freedom of speech’ (on the contrary) and would therefore fail the proportionality test. The DSA wisely refrains from such a move.
Freedom of speech is a defence right against the government, it does not give a right to be heard, and it does not oblige media to convey everyone’s opinion. Social media have certainly changed this perception, as there are no physical limits to publish comments and posts. Thus, when removed from such a platform, people feel deprived of their perceived right to voice their opinion. Whilst participation is in principle a good thing, unfortunately the downsides as shitstorms, conspiracy theories, and misinformation have become clear.
Some have criticized Twitter that other politically violent figures are still on their platform. This argument has value and should motivate Twitter to review its privileges for politicians. Yet, Twitter could argue that Trump is a special case and through its Twitter reach has caused serious damage (namely the 6 January event) and his presence on the platform, therefore, poses a substantial risk of further radicalization.
The DSA proposal of the Commission ticks many boxes that are important. It will not solve the problem of platforms as an instrument for political charlatanry. Politicians should not give the impression as if regulation in the future were to decide what is reachable on platforms-unless it is illegal. Regulation will ensure more transparency, strengthen user rights, and hopefully reduce the multiplier effects of its recommender systems. This will already be a welcome change.