Dr Liza Lovdahl Gormsensupplementary written evidence (FEO0102)

 

House of Lords Communications and Digital Committee inquiry into Freedom of Expression Online

 

Answers provided to Questions 4 and 5 after the evidence session on 2 March due to connection problems during the meeting

 

Question 4

 

(a) How well is the social media market functioning? What are the barriers to new entrants and is there a risk that the Online Safety Bill will create new barriers?

 

Social media platforms retain control over the flow of information in society. Dominant platforms use content curation systems that aim to maximise engagement and this has a very negative impact on users' exposure to diversity. 

 

Dominant platforms create barriers to entry for third parties by hosting as well as curating the content. Thus, the DMU should increase competition by being bold and decisive. It should call for the mandatory unbundling of hosting and content curation activities, together with the obligation, social media platforms should provide fair and non-discriminatory access to third-party players. 

 

I see this unbundling as a highly pro-competition remedy as it opens the market for content curation and relies on competition among players to deliver more choices and better-quality services to users. It is also capable of addressing various market failures such as high concentration, high barriers to entry and the users' under-exposure to diversity of content. This regulatory solution is not a novelty in the history of economic regulation. On the contrary, it has often been used in network industries, and especially in the telecom sector, in order to enhance competition and stimulate market entry.

 

Moreover, unbundling is less invasive or paternalistic than other instruments to address challenges related to content curation, such as imposing specific curation policies or establishing ‘must carry’ obligations. It interferes with digital platforms' freedom of economic activities only in a limited way and it supports long term market driven sustainable outcomes rather than involving top down requirements by a regulator, which is often problematic when it comes to freedom of expression and media diversity objectives. 

 

Finally, this form of unbundling empowers users to make their own choices, rather than imposing strict standards on the market.

 

I suggest a form of functional separation, as opposed to a structural one. In addition, the platform that provides the hosting should remain free to offer content curation too. What changes is that it should keep the two services separate and provide competitors the possibility to offer the curation service on its platform as well as allow users to freely choose among a variety of content curation providers.

 

The unbundling rules should be designed and implemented to address the contractual layer (contractual agreements between the platforms with significant market power and the alternative players that provide content curation services to the platforms' users) and the technical layer (how to make this technically possible while ensuring data protection, consumer protection and security).

 

 

 

(b) How should the Digital Markets Unit increase competition in the social media market? Are there any international examples from which it should learn?

 

The DMU should increase competition by being bold and decisive. There have already been numerous studies and evidence-gathering exercises (Furman report, the CMA online platforms market study, the Stigler report, the ACCC’s report, the EC’s three experts' report). No more reports! DMU needs to take action to avoid becoming a Think-tank like the CMA. 

 

The big online platforms have tried for some time to encourage government bodies in the UK to take it easy in the name of ‘chilling competition’. They have argued that regulators should take a cautious approach to avoid making the UK unattractive for inward investment and thereby damage to the UK economy, which is already facing the double challenges of Brexit and Covid-19. However, there is nothing inconsistent with acting boldly and decisively whilst also being fair and proportionate (observing due process). Obviously, the DMU would need to balance its approach, but it should not be intimidated into inaction. Actually, it is the actions of big online platforms, which are currently chilling competition; and decisive regulatory action is needed to unlock the potential of the UK economy.  Bold and balanced regulatory action can and should be a driver for, not an impediment to, inward investment for the UK, and should provide an open and fair market environment for UK firms to flourish in the digital space.  

 

An international examples is of course the ACCC’s recent actions vs. Google and Facebook, but I think what is at least as important as emulating good practice elsewhere is creating networks among regulators across the world.  It is trite to observe that transnational challenges require coordinated international action, but it still bears repeating.  The DMU should therefore be empowered and encouraged to share information with, and to work closely with, other ‘digital regulators’ across the world.  If (and that is a 'big if') the DMU gets it right, it has the potential to be – a ‘world leader’ in developing a regulatory regime for the digital sector. 

 

 

Question 5

 

(a) How well is the search engine market functioning? What are the barriers to new entrants?

 

Google's current market share in mobile search is 95% in the U.S., 98% in the U.K., and 98% in Australia. This concludes that Google holds a near monopoly in the mobile search market. I don’t think I need to elaborate on why this is bad for competition. 

 

The lack of competition within the search engine market means that Google does not need to prevent harmful ads from appearing on Google’s searches. For a long time, online platforms have been protected by the exemption in the EU e-Commerce Directive where they have been under no obligation to monitor content online, which has led to huge amounts of financial crime in terms of illegal online scams harming thousands of consumers in the UK. 

 

Due to Brexit, the legal immunity from a monitoring obligation for platforms has fallen away. This leaves the UK with a golden opportunity to impose a monitoring obligation for illegal (for example scams and fraud) as well as legal, but harmful conduct (for example high risk investments). Unfortunately, the Online Safety Bill does not impose such an obligation or deal with online fraud or the psychological harm it brings. Without legislation, however, regulators are obliged to negotiate separately with every social media organisation and to persuade each of them to provide gateway protection to consumers. This will produce a patchwork of inconsistent approaches, take considerable time and will create gaps that can be exploited by scam advertisers. Legislation avoids these gaps and ensures there is a level playing field for all social media organisations and consistent, enforceable standards to protect consumers.

 

As for the barriers to entry – they are huge! How many are using Ecosia, Qwant or Ask as a search engine? 

 

Among the barriers to new entrants in the search market, the most visible is probably Google settings itself as default on its own platforms, like Android and Chrome. The CMA market study on online platforms provides an outstanding recollection and demonstration of what that entails: Google leverages the “power of defaults” to lock consumers in, and makes it impossible to easily change all search defaults on a device. 

 

Google often argues that competition is just one click away. This would be correct only if consumers we faced with a ‘search engine’ preference menu when they bought a phone (that is to say a list of the various search engines available on the market such as Bing, Yahoo, DuckDuckGo, Ecosia and Qwant to mention a few). However, such a search engine preference menu is not available when consumers buy a new phone. Similarly, there is no global device search setting on Android. 

 

Google is the default search engine on Android devices. Thus, if you want to change the search engine on your phone, you would have to click at least 15 times to change to an alternative search engine or alternative browser. Thus, competition is NOT just one click away. Google is using Android and Chrome to further its dominance in the search market, which makes the barriers to entry for competing search engines huge.       

 

(b) How should the Digital Markets Unit increase competition in the search engine market? Are there any international examples from which it should learn?

 

The first step would be to make it mandatory give consumers a choice of which search engine to use when they buy a phone. It would be ideal to give consumers a choice from a ‘search engine’ preference menu listing all the various search engines on the market. This would allow consumers to choose what search engine they would like as their default search engine rather than now where Google is the default search engine on many devices. We have seen this choice menu in the Microsoft case with a choice of media players rather than search engines, but the principle is the same. 

 

In Europe, Google created a ‘search engine’ preference auction rather than a menu. This is not ideal as only the highest bidders ended up on the menu. This is flawed and a search engine’ preference menu should not be designed as an auction or by Google.  

 

A large-sample user testing carried out in the US (12,000 users across three jurisdictions: the US, the UK and Australia) shows that 24% of people in the U.S., 24% in the U.K., and 17% in Australia would choose a non-Google search engine when given the chance to select a preference on their mobile phones. This user testing also showed people will scroll to check out their options before making a search engine selection, meaning Google could be placed on the last screen of any search preference menu.

 

Thus, a properly and carefully designed preference menu (search engines are in random order) can increase competition and empowering consumers. The key point being that a bad preference menu could be worse than none (giving the false impression that a solution has already been implemented and thus preventing the implementation of real solutions), and careful and considered preference menu design is therefore paramount.

 

 

March 2021

4