Self-Preferencing Practices and Their Future After the DMA

The chapter describes the self-preferencing prohibition of Art. 6.5 DMA and the main interpretive problems that are likely to emerge in its enforcement. Certain features of the prohibition and its functionality are relatively clear: for example, the fact that it applies exclusively to rankings, and that it demands disclosures by gatekeepers of their ranking parameters. As the discussion in the chapter shows, however, identifying the types of ranking that will precisely substantiate the prohibited self-preferencing conduct presents several difficulties. The difficulties are linked mainly to the need of distinguishing between user-serving features of a ranking algorithm (e.g. the intention to privilege the quality of the results shown) and platform bias that works to the detriment of consumers, as it reduces choice and quality, and business users, as it deprives them of the benefits of their economic contribution to the platform. Based on these observations, as well as design choices of the DMA with regard to its enforcement (e.g. the importance of compliance reports), the chapter argues that Art. 6.5 is unlikely to really work as a per-se rule for the ex-ante identification of prohibited conduct, which is how the DMA prohibitions are mostly described in the scholarship. Most likely, article 6.5 will work as a form of quality standard for the design and presentation of rankings by vertically integrated platforms, to be enforced collaboratively with the undertakings. This functionality does not undermine the overall impulse of the DMA to facilitate scrutiny over digital platforms through a shift away from effects analysis, which had proven difficult in more traditional antitrust. As the chapter begins to show, the form of quality control over rankings that Art. 6.5 prefigures is different from an economic analysis of competitive effects. It focuses not so much on welfare effects and efficiencies, or the analysis of counterfactuals, but rather on explainability of the ranking parameters, and the design of the consumer interface design, to minimize platform bias and, where it had to emerge, its harmful effects. This interpretation points at a vision of the DMA as offering consumers constant monitoring of the quality of the platform, and specifically its rankings, which in turn ensures opportunities for business users to thrive on the platform and potentially, on the long run, contest its role.