The marketplace of ideas and its externalities: Who pays the cost of online fake news?

Since the nineteenth century, thanks to J.S. Mill, an argument against limiting the freedom of expression has been gaining ground, namely that the truth of an argument is the result of free competition between different ideas in a “marketplace of ideas” and that this competition can yield an approximate form of the truth or a robust argument. In this paper I will argue that Mill’s model of the marketplace of ideas is an “idealised” model because it assumes that there is a subject who seeks to know in a detached manner, without being influenced by his or her expectations, needs, ideologies, or biases. However, the metaphor of the marketplace of ideas is still relevant because it also harbours a limitation to the marketplace itself, namely the production of “negative externalities”: the spread of fake news (conspiracy theories and the deliberate spread of alarm), the polarisation of the debate, and the concentration of the advertising market in the hands of a few companies. Just as in the case of pollution, these costs need to be taken into account. How can this be done? The available conventional mechanisms seem to be insufficient because they need to pursue two conflicting goals, i.e., reducing the number of disinformation sites on the one hand and protecting users’ freedom of expression on the other. Platforms may not allow competing apps to be downloaded or may impose automated and intrusive regulatory schemes. However, this means that censorship would return in the form of a “private” censor. I will compare two models for possibly solving the conundrum: the one proposed by Paul Romer of regressive taxation of platforms and the mixed model proposed by the European Commission.