| Practice area: | Behavioural Economics | Competition & Antitrust |
|---|---|
| Client: | N/A |
| Published: | 30 September, 2024 |
| Keywords: | behavioural economics competition |
In August 2024, the UK Competition and Markets Authority rejected commitments proposed by Google to resolve concerns around the Google Play billing system for in-app purchases. The CMA’s decision is one of a growing series of rulings in high-profile competition investigations and cases that have explicitly or implicitly drawn upon behavioural economics, including a number of European Commission cases[1] and, most recently, a US federal judge’s ruling that Google violated antitrust law with its default distribution agreements.[2]
James Suter, Divisional Director in the Behavioural Economics practice at London Economics, looks at the CMA’s decision.
The CMA’s investigation into suspected anti-competitive conduct by Google
The CMA’s investigation concerned whether Google’s Play Store policies, which required developers to use Google’s proprietary billing system for in-app purchases, were an abuse of dominance. In 2023, Google offered commitments that would allow developers to use an alternative billing system, either alongside or instead of Google’s proprietary system, with a corresponding reduction in Google’s commission.
While the CMA provisionally accepted these commitments, subject to a consultation, it has now rejected them. As well as the reduction in commission being insufficient, the CMA cited concerns that proposed interstitial (or ‘pop-up’) screens may “create unnecessary friction” and/or “present information in an unduly negative/misleading manner”, thus reducing user uptake.
The CMA noted that although Google’s billing system would no longer be mandatory “in principle”, the conditions and requirements could affect whether they are effective “in practice”. This relates to the fundamental idea within behavioural economics that just because an action can be taken and a remedy solves a problem ‘on paper’, due to behavioural biases and irrationalities this may not be the case upon implementation in the real world. It also implies that the CMA is focused on what will work (and can be shown to work) in practice and may be sceptical of remedies that fix problems in theory without evidence of real-world impact.
This is consistent with what the CMA has said previously: it suggested that using choice screens to ensure consumers make an active choice may be used as a remedy but noted that the effectiveness of active choice is “hugely dependent on the nature of the choice and its design”.[3]
Simultaneously, the CMA announced the closure of its antitrust investigation believing it will be better able to address its concerns in a “more timely, holistic, and flexible manner” using its new powers under the new digital markets competition regime.[4]
Applying behavioural economics to this case
It is the interstitial (or ‘pop-up’) screens that are relevant from a behavioural perspective. Remedies of this type are an area of antitrust where behavioural economics has had a significant impact, while choice screens of the sort rejected by the CMA have been introduced by big tech companies to ensure compliance with EU law, such as in the Microsoft browser[5] and Google Android[6] cases. Here we briefly consider what existing evidence says regarding the issues raised by the CMA, some details of Google’s proposals, and what further might be done.
Existing evidence – the effects of friction and manner of presentation in online consumer processes
The literature is clear, the degree of friction in online consumer processes and the manner of presentation of information and options does indeed affect choices. Friction is often discussed in the context of commercial practices – sometimes labelled ‘sludge’ and ‘dark patterns’ – that use friction to either prevent or induce certain actions. While the literature tends to focus on examples where friction is manipulated to benefit a service provider, against the interests of consumers, there are also examples of it being used positively. For instance, Sunstein gives examples of studies that have shown friction to affect behaviour in negative and positive ways.
For example, a study we conducted for the European Commission on the digitalisation of financial services showed, via a simulated online experiment, that merely altering the number of clicks in the purchasing process can change consumer behaviour. The same study also found that the way in which options are framed, in either a more positive or negative way, has a substantial effect on choices. Moreover, a study we conducted for the FCA found that when a carefully designed extra screen is added to the purchasing process, this can have a significant impact on choices. This is, of course, just a small slice of the literature on ‘online choice architecture’, which the CMA itself has discussed extensively.
Google’s proposals – a behavioural analysis
The ‘interstitial screens’ in Google’s proposed commitments include an ‘Information Screen’ (to be shown when users first make a purchase) explaining the “implications” of using an alternative billing system, and a ‘Billing Choice Screen’ informing users of their alternative options.[7] The proposed text of the Information Screen said:[8]
“Changes to your checkout options
Each time you make a purchase, you’ll now choose either Google Play or the developer as the seller. Your choice will determine:
-
- Who secures your purchase, processes your payment, and stores any payment information
- Who provides customer support for the purchase
- Who is responsible for applicable consumer rights
Only purchases through Google Play are secured by Google. Play features such as Play gift cards, Play Points, purchase controls, and subscription management are only available when you choose Google Play at checkout.”
While this seems objective and sensible, from a behavioural perspective what are some potential issues? Firstly, the text sets out benefits of Google Play only (e.g. purchases are secured by Google Play and Play gift cards and Points are available only if the user chooses Google Play). By presenting these benefits in a pop-up, immediately prior to the choice, they become more salient compared to any benefits of other payment options. Second, consumers may be unfamiliar with the developer, which, combined with their risk aversion, may affect how they react to the statement that their choice will determine who secures and processes their payment and stores information. Third, consumers may have an implicit perception that Google Play is the expected or default option, due to the focus of the text on Google Play and since the consumer is using Google Play at the time.
Moreover, although the information screen is only shown to users when they first make a purchase, its influence is likely to be enduring. This is because the choice a consumer makes initially becomes their status quo, which due to status quo bias will likely continue to be their choice in future.
Hence, several known and powerful consumer behavioural biases may be at play, including framing effects, risk aversion, default bias and status quo bias.[9] However, this is not a complete behavioural analysis of the screens and choice architecture of Google’s proposed commitments, and ideally the effects would be tested empirically.
What further could be done
A lesson is that even if a remedy makes alternatives technically feasible, inexpensive, and simple, this does not mean that consumers always will exercise that choice.[10] By providing an understanding of how consumers respond to online choice architecture, insights from behavioural economics can be particularly valuable in designing effective remedies. Remedies need to foresee and account for consumers’ biases, and work with how consumers behave.
Perhaps most importantly, empirical evidence is required. If this is not available from existing studies that are directly relevant to a proposed remedy, then remedies could be tested empirically prior to implementation to ensure they are effective. This could take various forms, such as simulated behavioural experiments, field trials or A/B testing. The CMA has recently advocated such testing[11] and the Digital Markets, Competition and Consumers Act has given it powers which may allow it to ensure such testing is done.[12]
[1] Including the Commission’s Microsoft Media Player, Microsoft Browser, Google Android and Google Shopping cases. For a summary of these cases see: Bennett, M. (2022), Integrating Consumer Behaviour Insights in Competition Enforcement – OECD Competition Policy Roundtable Background Note.
[2] For discussion of relevant cases and the so-called ‘behavioral antitrust’ literature, see, for example: Fletcher, A. (2019), The EU Google Decisions: Extreme Enforcement or the Tip of The Behavioral Iceberg?, Competition Policy International; Huffman, M. (2019), A look at Behavioral Antitrust from 2018, CPI Antitrust Chronicle; Tor, A. (2014), Understanding Behavioral Antitrust, 92 Tex. L. Rev. 573 (2013-2014); and Stucke, M. E. (2012), Behavioral Antitrust and Monopolization, Journal of Competition Law & Economics, Volume 8, Issue 3, 545–574.
[3] See paragraphs 4.32 and 4.33 of the CMA’s 2022 discussion paper on online choice architecture.
[4] The Digital Markets, Competition and Consumers Act gives the CMA power to designate firms as having Strategic Market Status if they have “substantial and entrenched market power” and “a position of strategic significance” with respect to digital activities linked to the UK.
[5] In 2009, the European Commission accepted commitments from Microsoft in light of concerns that the company’s tying of Internet Explorer with Windows may be harming competition and reducing consumer choice. These commitments included introducing a choice screen to allow users to select the web browser they wanted to install, implemented via a pop-up box which prompted users to make an active choice.
[6] Following the finding in 2018 that Google acted anti-competitively (including by tying the pre-installation of the Google Search app with the Play Store, and tying Google Chrome with the Play Store and the Google Search app), Google introduced a choice screen on all new Android phones and tablets in the European Economic Area (the UK included) allowing users to select a search engine default.
[7] Further explanation can be found in paragraphs 6.31 to 6.42 of the CMA’s notice of intention to accept the commitments. The text of Google’s commitments offer is here.
[8] See in paragraphs 6.32 and 6.34 of the CMA’s consultation document. The Information Screen text presented here corresponds to the scenario where the app developer chooses to use ‘User Choice Billing’, meaning that any alternative payment system is offered alongside Google Play’s system.
[9] Each of these is well established in the behavioural economics literature. See, for example, DellaVigna, S. (2009), Psychology and Economics: Evidence from the Field, Journal of Economic Literature, 47 (2): 315–72.
[10] Stucke, M. (2012), Hearing on Competition and Behavioural Economics – Note by Maurice Stucke, OECD Hearing on Competition and Behavioural Economics.
[11] See the CMA’s 2022 discussion paper on online choice architecture, paragraph 4.34.
[12] Section 69 of the Act on power to require information includes the power to require the performance of “a specified demonstration or test”.