The DMA and default positions in search

Picking up on our latest post on the direction of the Council with respect to the DMA, we would like to touch on an important issue that seems to have been overlooked by EU lawmakers: the role of default positions in search. Rather surprisingly, both the original DMA proposal of the European Commission and subsequent drafts contain little, if anything, with respect to defaults in search. As we argue in this post, there is a good case for EU lawmakers to include in the DMA specific provisions tackling defaults in search.

Defaults matter

The power of defaults is an area of behavioral economics that has been well-researched and evidenced across a wide range of settings, including food consumption, marketing, and even organ donation. A recent Wall Street Journal article on the ongoing Facebook drama suggests that the solution to fix social media is to change default settings which benefit companies and their bottom lines instead of users.

At the same time, defaults may entrench dominant positions and prevent expansion of rivals. In particular, defaults seem to be a key barrier protecting Google’s position in general search and preventing rivals from gaining scale. Google pays device manufacturers and browser vendors substantial sums of money (up to $15 billion a year in the case of Apple, according to analysts) to have its eponymous search engine set as the default across key search access points (e.g., the pre-installed browser). As the UK CMA found after reviewing a wealth of evidence in the context of its online platforms market study (including internal research and documents of Google), “default positions have a significant impact on consumer behaviour in search and, in turn, on the search volumes (and search revenues) of search engines.” The CMA held that Google’s extensive default positions “act as a significant barrier to expansion for rival search engines and lead to weaker competition to Google in general search.” In addition, Google benefits from a positive feedback loop “between its position as the largest and most revenue-generating search engine and its ability to acquire extensive default positions that further reinforce this position.”

Meanwhile, the so-called “status quo bias” caused by default positions and pre-installation is at the core of the European Commission’s Google Android decision condemning the pre-installation conditions in Google’s MADA agreements (check also our post on the recent hearing in Luxembourg on Google Android, which includes a discussion of nudges in the digital economy). As part of complying with the Google Android decision, Google has rolled out a choice screen for new Android phones and tablets in the EEA where the Google Search app is pre-installed, that is a screen appearing during device set up that allows users to select an alternative search engine as their default.

Against this background, one may think that the issue of default positions in search has been effectively tackled, hence there is no need for further (antitrust or regulatory) action. Yet this would be a misguided approach in our view. First, the Google Android decision is limited to Android devices; it does not include iPhones, which in certain regions (e.g., the Nordics) account for the majority of smartphones. Second, even with respect to Android devices, the choice screen rolled out by Google suffers from important limitations (e.g., it does not cover all search access points and can be accessed only during device set up) even after Google’s decision to drop the previous controversial auction format.

Considering how the state of competition in search leaves a lot to be desired (save for Google, we are not sure anyone will argue that competition is just “one click away”), EU lawmakers should seriously consider addressing the key issue of default positions within the context of the DMA. Relying on the Google Android decision and its implementation by Google simply won’t do the trick.

The DMA on defaults

Rather surprisingly, the original DMA proposal is entirely silent on the issue of default positions. Subsequent drafts we had the chance to review are also rather underwhelming in this respect. Specifically, the third compromise text circulated by the Slovenian presidency provides that a gatekeeper shall allow end users to change default settings on an operating system that direct or steer end users to products or services offered by the gatekeeper. In itself, this obligation does not add much. In fact, in many instances users are technically able to change default settings on an operating system; the trouble is that they often do not, as defaults are sticky and changing them may require several steps (for instance, the WSJ article mentioned above refers to a Facebook setting as being “still more buried than tulip bulbs”). Moreover, this obligation would presumably not capture Apple’s deal with Google for Safari, as in this case the default setting on Safari steers the end user to products or services offered by another gatekeeper.

Some proposals from the European Parliament (and in particular the ECON committee, where MEP Yon-Courtin is the Rapporteur) are more ambitious but they could be improved. In particular, according to the ECON proposal a gatekeeper shall “prompt” end users to change the default settings for any pre-installed core platform service on an operating system, as soon as the end user uses such core platform service. While it is not clear what this “prompt” would be in practice, it could arguably take the form of a choice screen/preference menu. Depending on the services that finally make it into the list of “core platform services,” this obligation could thus capture default positions on pre-installed browsers (e.g., Safari on iPhones, Chrome on Android) and perhaps voice assistants. On the other hand, this obligation could be overly broad as it is not limited to search and could thus mean that users would be flooded with “prompts” to change all sorts of defaults, which could negatively affect user experience. At the same time, this obligation would presumably not cover all search access points (which go beyond pre-installed core platform services on an operating system – think of the home search bar on smartphones).

As a result, it might be preferable to have a more targeted obligation that would be limited to search. For instance, there could be an obligation on undertakings designated as gatekeepers with respect to their search engine not to secure default positions for their search engine across any or all search access points of the operating system by the same or another gatekeeper. Such an obligation could address the issue of default positions in search in a comprehensive manner, without at the same time being overly broad or affecting user experience.

Conclusion

The issue of default positions is key in promoting contestability in search. Even so, defaults have received surprisingly little attention from EU lawmakers, perhaps because the issue is considered as settled by the Google Android decision. Yet this is wrong; promoting contestability in search would at the very least require measures covering iOS devices. Lawmakers should seriously consider amending the DMA to reflect that.

Leave a Reply