Android and Microsoft: similarities and differences (I) #ourreadersask
Our readers have asked me a few times about the similarities and differences between Android and Microsoft. The question makes a lot of sense. If Microsoft was a resounding victory for the Commission, and Android is also about tying (and about software), should we waste time and paper discussing it? Is the outcome not going to be exactly the same? I believe that there are genuine differences between the two. This first post addresses some of them.
Allow me to start with a hypothetical that captures, I believe, the essence of the Android case.
A 24-hour news channel is immensely successful. It is so successful that all pay TV operators (cable, satellite, IPTV) see it as indispensable for their business. This channel does not carry advertising and is offered for free. However, pay TV operators wishing to carry the channel must also carry two tele-shopping channels, which are used to raise advertising revenues. In addition, pay TV operators are required to place these two tele-shopping channels immediately before and after the news channel, to ensure that the largest possible number of viewers watches them.
Is this behaviour anticompetitive and, assuming dominance, contrary to Article 102 TFEU? I do not think so, even though there is tying involved and even though the firm would insist on giving more prominence to its tele-shopping channels. Why? For two reasons that are also key to make sense of Android.
- In this example, the existence of the tying product (the 24-hour news channel) cannot be understood without the existence of the tied product (the tele-shopping channels). The tying of the two products, in other words, is what allows for the viability of the 24-hour news channel. Again, it is all about the counterfactual. It looks like the practice does not restrict competition that would have existed in its absence. As a result, it is not capable of having exclusionary effects.
- What is more, prohibiting the tie-in in this case would allow rivals to take a free ride on the investments of the firm. If a 24-hour news channel is immensely successful, rivals would probably want to place their channels immediately before, or immediately after it. Because it does not carry advertising, prohibiting the tie-in would allow rivals to make money at the expense of the company running the news channel. This firm would develop the channel, and rivals (alone or together with the firms’ customers) would benefit from these investments by reaping the advertising revenues.
These two features were not present in the Microsoft (Media Player) case. It is not very credible to argue that the tying product (the operating system) would not have been created without the tied product (the media player). The operating system existed well before the media player. The analysis of the counterfactual shows that the tie-in was at least capable of restricting competition that would have existed in its absence (whether it was likely to restrict competition is of course a different question). Similarly, it is difficult to say that an obligation to seek a version of Windows without the media player would allow Microsoft’s rivals to take a free-ride.
Consider now Android. I do not believe anybody seriously disputes that the Android ecosystem exists because Google Search exists. The mobile operating system was only developed because of the prospect of making money via advertising (ie, via Google Search). The same is true of Google Play and Google Chrome (after all, a web browser is also a way of performing searches).
This point is intriguing and important. Essentially, it reveals that the Android case turns the logic of Microsoft on its head. In Microsoft, the tied product owed its existence to the tying product. In this respect (but not in others), it was a plain-vanilla tying case. In Android, it is the other way around. The tying product owes its existence to the tied product.
As I understand Android, the Commission appears to argue that the tying product is Google Play and the tied product is Google Search. Can one argue that this practice is capable of having anticompetitive effects? Sure, but it would be necessary to show (i) that, when Android was launched, it was possible to enter the market with a different business model and (ii) that Google had the incentive to enter the market with a different business model.
In other words, it would be necessary to show, against the counterfactual, that the practice restricts competition that would otherwise have existed. If Google would not have entered the market for operating systems in the absence of the alleged restraints, we know from the case law (STM, Post Danmark II) that the threshold of ‘capability’ is not met. I am curious to see how this issue will be considered.
Next time, I will address the second point I identified above, which has to do with free-riding and the remedy. In this respect, I believe there are similarities between Android and the Microsoft saga. A taster: this second point reveals, in my view, none of these cases are really about tying at all. Stay tuned!
A new paper on Android, by Damien Geradin and Ben Edelman
A while ago, we invited our readers to send our way their thoughts and comments on the Android case. With some delay ;), Damien Geradin has sent us a paper he has written with Ben Edelman, from Harvard Business School. Here it is: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2833476