Implementing the DMA: The role of behavioural insights

Vice President Zita Vasas recently authored a paper alongside Amelia Fletcher discussing the role of behavioural insights in the implementation of the Digital Markets Act (DMA) and potential pitfalls that could arise during implementation.

Abstract

The Digital Markets Act (‘DMA’) is a key plank of the EU’s digital markets strategy. This article provides an overview of the role of behavioural insights in the DMA’s implementation. This is highly topical, since the initial implementation period is ongoing. The effectiveness of several DMA provisions will in practice depend on their impact on end user choices and the way in which the designated gatekeepers design the choice architecture within their user interfaces will be critical for this. We provide some overarching considerations relating to this design and then discuss the relevance for specific DMA provisions of some key well-documented behavioural effects such as saliency effects, ranking effects, default effects, social cues, choice overload, information overload, choice fatigue, obfuscation and shrouding, complexification, framing effects and timing effects. Our focus is on highlighting the potential issues, including in relation to circumvention risks. We note where empirical testing by the gatekeepers is likely to be useful in order to demonstrate their compliance with the DMA.

Introduction

The Digital Markets Act (‘DMA’) is a key plank of the EU’s digital markets strategy. Its aim is (i) to ensure that a small set of large and critical online platforms (‘gatekeepers’) behave in a fair way towards their business users and (ii) to promote contestability in markets where gatekeepers have entrenched positions, as well as in related markets. To this end, it contains a collection of provisions that set out what gatekeepers must and must not do. Central to the DMA is a focus on effectiveness. First, Article 8 requires that gatekeepers ensure that implementation is effective in achieving the aims of the specific provisions and also the objectives of the DMA – fairness and contestability. Second, Article 13 expressly prohibits any circumvention behaviour that undermines effective compliance. Third, a number of specific provisions also mention the need for effectiveness.

However, effectiveness does not imply a requirement to achieve maximum impact at all cost. Proportionality is also important. This is clear within the DMA itself (see Article 8) but there is also a general principle of proportionality in EU law.4 We thus interpret the DMA’s wording on effectiveness as requiring gatekeepers to implement the provisions in a way that is effective but not disproportionate in achieving the objectives of the DMA.

The aim of this article is to provide an overview of the role of behavioural insights in this context. This is highly topical, since the initial implementation period is ongoing. The effectiveness of several DMA provisions will in practice depend on their impact on end user choices and the way in which the designated gatekeepers design the choice architecture within their user interfaces will be critical for this.

We provide some overarching considerations relating to this design and then discuss the relevance for specific DMA provisions of some key well-documented behavioural effects, such as saliency effects, ranking effects, default effects, social cues, choice overload, information overload, choice fatigue, obfuscation and shrouding, complexification, framing effects and timing effects. These effects are all relevant to the way in which options are presented to end users – the so-called ‘choice architecture’. The design of the choice architecture can have a substantial impact on the choices end users make.

Such behavioural effects can arise in two ways in the context of the implementation of the DMA. First, they can inherently hinder the effectiveness of provisions in delivering contestability and fairness. Gatekeepers therefore need to consider carefully how best to implement the DMA’s provisions with behavioural insights in mind, which in many cases will require empirical testing of different interface designs. It is beyond the scope of this short article to provide guidance on how to best implement such design and testing, but we highlight where it would be useful. Second, there is potential for gatekeepers to exploit behavioural effect sstrategically, weakening the impact of the provisions in driving contestability and fairness. Such conduct would likely constitute circumvention, and thus breach Article 13. It is important that the Commission is alert to the risk of such conduct.

In this article, we first provide some overarching considerations that are relevant to the two ways in which behavioural effects can arise in the context of the DMA (Section 2). We then consider (in Section 3) a set of more specific issues relating to three categories of obligations:

i. Provisions that require the gatekeeper to impose certain end user actions;
ii. Provisions that require the gatekeeper to enable certain end user actions; and
iii. Provisions that require the gatekeeper to allow certain business user actions.

For each of these, we discuss which behavioural insights are relevant and highlight potential pitfalls that could arise during implementation.

Read the full paper here

This paper was originally published on the Social Science Research Network here. The views expressed are those of the authors only and do not necessarily represent the views of Compass Lexecon, its management, its subsidiaries, its affiliates, its employees, or clients.

Fletcher, Amelia and Vasas, Zita, Implementing the DMA: The role of behavioural insights (July 5, 2023). Available at SSRN: https://ssrn.com/abstract=