Mike Salem is a Senior Political Adviser for the Kava Coalition and a political and economic analyst.
On Friday, I went with the curate of my local Anglican parish church to watch The Social Reckoning, the sequel to The Social Network. Moving beyond the controversy surrounding Facebook’s creation, the film explores its algorithms and the investigative work of Jeff Horwitz. Although it draws on real reporting, I could not help feeling that a particular agenda was being advanced, especially regarding social media bans.
The film also draws parallels between Big Tobacco and Big Tech, an analogy often used to justify intervention. Its suggestion that Facebook changed its name to Meta following the Facebook Files revelations was particularly unconvincing: a rebrand of that scale could hardly have been achieved in the four weeks suggested. Meta’s creation also reflected Facebook’s ambitions for the metaverse and its desire for an umbrella identity covering its various businesses.
Nevertheless, the film could not have been more timely. At their party conference, the Conservatives claimed credit for the Government’s reported U-turn in favour of a social media ban for under-16s. I oppose such bans. Parents, not the state, should retain responsibility for raising their children, and restrictions risk providing a rationale for further interventions, including intrusive digital identification measures.
Yet the film left me with a different concern. While I remain sceptical of state intervention, I am increasingly unconvinced that the market for ideas on social media is as free as we assume.
The problem with meaningful social interactions
One of the central issues in The Social Reckoning is Facebook’s introduction of Meaningful Social Interactions (MSI) into its News Feed ranking system.
In 2018, Facebook announced changes intended to prioritise interactions between friends and family. But reporting by The Wall Street Journal on internal Facebook documents raised questions about the incentives this created. Its reporting described a scoring system in which different interactions carried different weights, with reshares and certain reactions valued more highly than likes.
The economic significance is straightforward. A like registers a reaction; a reshare can expose a message to an entirely new audience. Content that provokes arguments, anger or outrage can therefore generate more engagement and potentially greater distribution.
But engagement is not the same as value. A thoughtful explanation of an economic issue, such as the excellent work produced by the Institute of Economic Affairs, may inform thousands of readers without prompting them to share it. An inflammatory post may be circulated thousands of times, including by people sharing it to condemn it. A misleading claim can gain traction precisely because it provokes a strong reaction.
The platform need not intend to promote misinformation for its incentives to favour content that generates engagement. The Facebook Files raised concerns about these dynamics, although they do not prove that algorithms alone cause political polarisation or that every reshare is harmful. The question is whether a system that rewards particular behaviours can distort the distribution of attention, irrespective of what users would prefer under different incentives.
Is this really a free market?
A defender of Meta could reasonably argue that Facebook is a private company entitled to design its own product. Users are free to leave, and businesses routinely influence consumer behaviour. Supermarkets choose what to display prominently; newspapers decide which stories make the headlines. Influence alone does not constitute market failure.
But what happens when users cannot properly understand the product they are receiving? Facebook users see a feed of posts, photographs and videos. They may know an algorithm determines its contents, but they cannot readily observe the full ranking process or understand precisely why one post reaches millions while another remains unseen.
Meta, by contrast, possesses extensive data about user behaviour and the effects of its ranking decisions. This creates asymmetric information: the company knows considerably more about how its product operates than its customers do. That imbalance does not automatically justify intervention. Businesses will inevitably know more about their products than consumers. The more important questions are whether users can make informed choices, whether competitors offer meaningful alternatives, and whether the company bears the costs of the incentives it creates.
Network effects complicate matters further. Leaving Facebook can mean losing access to friends, communities and audiences. Even dissatisfied users may struggle to move elsewhere without sacrificing the benefits of the network they have built. The market for attention is not necessarily competitive simply because participation is voluntary.
From hosting speech to allocating attention
We do not ban knives because some people use them to commit assaults. Nor should Meta automatically be held responsible for every harmful statement published by its users. Doing so would threaten freedom of expression and encourage platforms to remove lawful content defensively.
But the knife analogy has limits. A knife does not decide which actions receive an audience. Facebook’s ranking system actively determines how content is distributed, predicting what users will engage with and ordering posts accordingly. Meta is therefore more than a passive host. It designs the mechanism through which speakers compete for attention.
Responsibility for what people say is different from responsibility for the system that determines how widely their words travel. Recognising that distinction does not make Meta the author of every post, nor does it mean the company should be compelled to suppress controversial opinions. It simply means we should scrutinise the incentives built into the distribution mechanism, just as we would examine the incentives created by any other market design.
When private incentives create public costs
Advertising-funded platforms benefit commercially from attracting attention. Yet the costs associated with the content their systems amplify may fall on people other than the company or the users generating engagement.
Misleading or inflammatory material can affect public debate and communities well beyond the original audience. If the platform captures the commercial benefits of engagement while others bear some of its costs, private incentives and social welfare may diverge.
This is a familiar economic problem: an externality. It does not follow that every controversial post is harmful, or that government officials can reliably determine which opinions deserve prominence. Political disagreement is not itself a market failure. The narrower argument is that where a platform systematically rewards certain forms of engagement, while some resulting costs fall on third parties, the market may not produce the most socially beneficial outcome. That is not an argument against free markets. It is an argument for taking their conditions seriously.
The danger of an overconfident state
Recognising a potential market failure does not mean government intervention is necessarily the answer. Regulators have their own information problems and incentives. They may misunderstand technology, become dependent on the companies they oversee or use powers intended to prevent harm to suppress politically inconvenient speech.
For classical liberals, freedom of expression is an essential safeguard against political power. We should be wary of giving politicians discretion over which lawful opinions people may encounter.
Instead, we should ask whether users can choose how their feeds are ranked, whether chronological alternatives are readily available, and whether independent researchers can assess algorithmic effects while protecting privacy. X, formerly Twitter, offers users a chronological feed option, illustrating one way to provide greater choice.
Competition matters too. Making it easier for users to move between services and retain their networks could strengthen the pressure on platforms to respond to consumer preferences. Such measures have costs and limitations, but they focus on transparency, choice and competition rather than government control of speech.
Defending free markets means asking difficult questions
The lesson of The Social Reckoning should not be that technology companies are inherently untrustworthy, or that the state must take control of social media. Nor should we assume that every private product is beyond criticism simply because consumers choose to use it.
A free market requires more than the absence of government commands. It depends on meaningful choice, competition and incentives that do not systematically conceal costs or distort consumer decisions.
The MSI controversy raises questions about all three. Opaque ranking rules, network effects and engagement incentives may weaken the market discipline that would otherwise constrain a platform. That does not prove every algorithmic decision is wrong, or that regulation is the solution. It means the economic case deserves serious examination.
We should defend lawful expression, resist state control of political debate and distinguish those who create harmful content from the platforms that host it. But we should also recognise that platforms make consequential choices about how that content is distributed.
A free market in speech is not automatically a free market in attention. Defending the former should not prevent us from scrutinising the latter. If we believe in competition, informed consumer choice and the discipline of markets, we should want to know whether those principles genuinely operate in the digital public square. That is not a retreat from free-market thinking. It is precisely the kind of question free-market thinking ought to ask.
I'd highly recommend reading Cory Doctorow's 'The Internet Con' on this topic. I think he makes a pretty convincing case that the product itself is not the issue, but rather the legal strangehold they have over their so-called intellectual property. If the market were free from these legal shackles, then competitors which serve users better would certainly rise to the top.
I'd highly recommend reading Cory Doctorow's 'The Internet Con' on this topic. I think he makes a pretty convincing case that the product itself is not the issue, but rather the legal strangehold they have over their so-called intellectual property. If the market were free from these legal shackles, then competitors which serve users better would certainly rise to the top.