Relaxing whilst doing Competition Law is not an Oxymoron

Archive for October 21st, 2019

Comments on Android (II): follow-up questions on market definition (+ thanks for the comments!)

with 6 comments


I received wonderfully interesting comments to my first post on market definition in Android . That was really the blog at its best. I am really grateful

I felt I would not do full justice to the involvement of these readers if I just added a comment of mine to the pile. The insights are so powerful that it makes sense to give them more exposure (and hopefully lead to yet more comments).

Would a merger between Apple’s and Android’s app stores be unproblematic?

Chris was the commentator to break the ice. The inevitable consequence of the definition of a separate market for Android’s app stores emerged from our exchange. Taken to its logical consequences, Chris explained, the Commission’s approach implies that a merger between Apple’s and Android’s app stores would be unproblematic.

This is a counterintuitive outcome. The fact that it is counterintuitive, however, does not mean anything. Rigorous economic analysis often surprises us with outcomes that contradict our instincts. However, this conclusion could also mean that there is something funny going on.

Any thoughts in this regard would be most welcome. If there is indeed something funny going on, what is it? One possibility is that the assessment of competitive constraints cannot focus on switching alone. A second possibility is that there is scope for refining the assessment of switching (which is what Nicolas was suggesting in his comment). I will focus on the first point.

Competitive constraints: is it all about switching?

The idea that Apple’s app store does not exercise a sufficient constraint on Android’s app store (paras 652-673) relies, by and large, on switching (switching from the end-users’ perspective and the app developers’ perspective).

My immediate question here is: is there something beyond switching opportunities that we may need to consider when assessing competitive constraints? Another commentator, Andrew, gave an excellent example on supermarkets that comes across as a wonderful starting point for this discussion.

Suppose there are two supermarkets that do not compete at the retail level: each one operates in a different geographic area. However, they compete on the upstream market where they buy inputs from providers. Suppose, further, that they are the only buyers on the upstream markets, so suppliers have nowhere else to go, and cannot give up either.

Would a merger-to-monopsony between the two supermarkets of the example be unproblematic? My instinct tells me that, in all likelihood, such a merger would be blocked absent appropriate remedies (and Gianni, in his comment, makes it clear that of course it would be blocked). As a telecoms nerd, the Liberty Global/Ziggo merger, the underlying facts of which are not fundamentally different from those sketched by Andrew, came to mind.

Liberty Global/Ziggo appears to confirm that it is not all about switching (it does not seem to be framed in terms of switching anyway). The Commission identified a series of potential negative effects that would arise even when the suppliers lack the ability to switch from one to the other (or have no choice but to deal with the two buyers). The question was whether the transaction would increase the buyers’ bargaining power.

If a merger-to-monopsony could be problematic even absent switching opportunities for suppliers, perhaps competitive constraints faced by powerful buyers are indeed manifested in other ways (that is, it is not all about switching).

If this is the case, my question is: what are these other ways? Is it, for instance, buyers’ ability to refuse to carry certain products from certain providers? Why would a merger-to-monopsony reduce competitive constraints if suppliers lacked switching opportunities prior to the transaction?

Thanks a lot in advance for sharing your thoughts.

And please feel free to comment on Nicolas’ points. I understand his comments as suggesting that it may be tricky to define markets involving goods such as smartphones, which are not acquired everyday.

It could be interesting to compare the analysis in Android with the analysis of retail markets in the network industries (after all, end-users do not change providers every day, and may be faithful to some providers, in particular incumbent ones).

If you have found an embarrassing number of typos, at least you should know why by now.

Written by Pablo Ibanez Colomo

21 October 2019 at 6:02 pm

Posted in Uncategorized