The Innovation Economics Conference for Antitrust Lawyers

On 21 April, economists Soledad Pereiras and Justin Coombs spoke at the Concurrences Innovation Economics conference for antitrust lawyers at King’s College London. Soledad focused on market definition in digital markets, while Justin discussed potential exclusionary behaviour by companies and their effects on the market and consumers.

Panel 1 – Market definition in innovation markets: Time to rethink?

Soledad Pereiras focused on market definition in digital markets. According to her, the market definition is still a valid exercise. Indeed, it is difficult to think of a case where market shares are not calculated and to calculate market shares, a market must be defined. However, we should recognise the limitations of market definition in the context of digital markets.

In digital markets, we typically observe a company that is acting as a platform and sells different products to different groups of customers. In this context, the first question is the number of markets that will need to be defined. In general, the authorities have defined different markets on either side of the platform. However, considering separate markets does not provide a full view of how the markets work. Indeed, when an assessment of competition is made it is necessary to consider that the two sides of the market are closely linked. Platforms in this type of markets face different competition in each side of the market and, in fact, they may face competition from companies that are only in one side of the market. However, there is a strong interdependence between the demand in both sides of the market because the platform needs both sides on board and, therefore, competes for customers in both sides. Companies make pricing decisions considering both sides of the market, not just one side, so the competitive constrains that a platform faces in its pricing strategies can only be assessed considering both sides of the market.

A second challenge to market definition is how these markets will be defined. The most rigorous conceptual tool that we have is the SSNIP test, where we identify the smallest set of substitute products on which a monopolist would find it profitable to increase prices by a small but significant amount. But in this type of markets it may be perfectly normal for a platform to subsidize one side of the market when the presence of consumers in this side is important in the other. The application of this test is even more complicated in multi-sided markets as platforms always re-optimise the price structure across the platform. The SSNIP test could still be used as a conceptual guide or adapted, for example, to assess to what extent a hypothetical monopolist can profitable impose a “small but significant non-transitory decrease in quality”.

The third challenge concerns how market shares will be calculated. Again, it is necessary to consider whether these market shares will be calculated on one side of the market or the whole platform. One should also carefully select what unit will be used to calculate these shares. Revenue from sales may no longer be relevant for the assessment of market power (ultimate purpose of market definition) and there may be additional complications because the monetization model may be different across platforms. Some alternatives measures may be number of users, number of views, number of downloads, number of subscribers or the time a person has spent on the platform.

What do we do in practice to define markets? The starting point for market definition is generally based on qualitative evidence and in many cases this will lead to an uncontroversial market definition. Quantitative evidence can also provide useful insights for market definition. An analysis that she will expect to gain more relevance in this respect will be natural experiments where one analyses the reaction to unexpected events (supply shortages, advertising campaigns, market entry…). Consumer surveys will also be increasingly used to infer consumer preferences.

Panel 3 – Essential facilities: Back through the window?

Justin Coombs pointed out that in the context of potential exclusionary behaviour by companies three questions are raised: whether the company has the ability to exclude competitors, whether it has the incentive to do so, and the effect that such behaviour would have on the market and consumers.

First, under the ability question, an undertaking would have the ability to foreclose a downstream undertaking from the market by refusing to supply if three conditions are established. The first condition is that the upstream firm supplies the downstream firms with an essential input (downstream firms cannot supply their own customers without access to this input). Second, there must be no viable alternative supplier of the input. Third, someone can't enter the market within a reasonable time and supply this input, so no one will suddenly enter the market and customers will not be able to start supplying themselves or sponsor new entry.

The next issue is incentives. In general, this kind of case involves a situation in which the upstream company not only supplies the downstream market but is itself present in the downstream market. Therefore, this company has a kind of conflict of interest in the sense that its customers are also its competitors. For this reason, it might have an incentive to try to foreclose these downstream firms to monopolise the downstream market and earn monopoly profits. In this context, the Chicago critique says that in practice there is only one monopoly profit to be made in this industry, so the upstream firm can already make that monopoly profit by charging a monopoly price for the essential input. However, while there are some very limited circumstances where it is true that there is only one monopoly profit, in most industries this is not the case. There may also be strategic reasons why a company might want to do this. The final question concerns the effects of the obligation to supply.

In this context, there are two effects that we have to balance. First, we have the short-term effect, which is why we might impose an obligation to supply, i.e. it will increase competition in the downstream market in the short term. There is also a potential long-term adverse effect of such a supply obligation, which is that it could affect incentives to invest and innovate. These two effects must therefore be balanced. This means that we need to think very carefully about the circumstances in which it would be beneficial for consumers and society to impose a supply obligation. This will be the situation where the effect of increasing competition is greater than the effect on investment incentives. There are situations in which we can expect this to be the case. Three questions are relevant. The first question is whether there will be a significant increase in competition as a result of the obligation to supply. The second question is the nature of the products that are going to be supplied by downstream competitors: are they true copies or differentiated products? If they are differentiated products the benefits to consumers from an obligation to supply will be higher. It may also mean that there is less of a negative impact on investment incentives because if they provide a differentiated product they are less likely to cannibalise the dominant firm's customers and therefore have less of an impact on the investment incentives of the dominant firm and other firms. Lastly, it may be that a supply obligation leads not only to differentiated  products but also to completely new products. In this case, the supply obligation is even more likely to have benefits and even less likely to harm investment incentives.

Recent legislative proposals introduce a system of ex-ante regulation imposing obligations on  companies that have operated in the private sector and where the entire investment has been privately financed. The question now is what effect this will have on investment incentives. On the one hand, you could say that it may remove the incentives for these companies to invest and create new products because these regulations are imposed on them. On the other hand, ex-ante  regulation could reduce uncertainty relative to ex-post competition law enforcement, which might enhance investment incentives. So, these new regulatory frameworks need to be developed in such a way as to provide the benefit of greater certainty for business rather than a situation where the business has less certainty.

Read the full synthesis here (subscription required)

This synthesis was originally published by Concurrences and is available on their website (subscription required). The views expressed are those of the authors only and do not necessarily represent the views of Compass Lexecon, its management, its subsidiaries, its affiliates, its employees, or clients.