Digital platforms, like Google, Apple, Facebook, Amazon and Microsoft, are now amongst the largest and most powerful companies on earth. Their success has been primarily driven by their innovative services. However, some people have started to argue that their market dominance should now be curbed through regulations. Should platforms be regulated? Benoit Reillier asked economic expert Dr Richard Cadman.
Benoit:
A number of people claim that platforms like Apple, Google, Amazon and others are dominant. Some even talk about mandating the separation of these large platforms. Should platforms be regulated?
Richard:
We need to have this debate driven by cool heads and rational economic analysis. Not by knee-jerk populist political reactions. Let’s embed any decision to impose economic regulation using the proper processes and rational thinking. The good news is we already have the tools to do that. In the electronic communication industry, the European Commission has established what they call the three criteria test. It provides a way to find out whether a market is what’s called “susceptible to ex-ante regulation”. It means that you can’t rely on competition law because of structural problems in the market. And so, regulation has to come in to open the market to competition.
Should platforms be regulated? If we’re going to regulate these firms, then we should do so on the basis of rational analysis and not just on which direction the political wind is blowing.
Benoit:
It makes a lot of sense. But in some ways, many of these platform companies are already regulated. Uber has got to abide by its license conditions. We know that they have had problems in London and other cities. For example, the way they share data is regulated. We’ve recently heard about Facebook’s data privacy issues. So you are talking about a different kind of regulation, right?
Richard:
We have to distinguish between different types of regulation. All companies have some regulation to do with consumer protection. For example, companies now have to abide by GDPR (General Data Protection Regulations) on private data. There’s also regulation of competition, also known as economic regulation. I think is where the current pressure on regulation is coming from. We want to make sure these markets have competitive outcomes, not necessarily competition, but competitive outcomes that benefit consumers. We may need to ensure that it’s that competition in the market that is subject to regulation. As a result, these firms, some of which are very large in their sector, shouldn’t exploit their position of dominance. That means ensuring that they can’t exclude competitors from that market, or exploit consumers on either side of the platform by charging prices well above what you would find in a competitive market.
Benoit:
You mentioned competition law, which already applies to platforms. As recently demonstrated by the two substantial fines of €2.6 billion and €4.3 billion imposed upon Google for respectively favouring its shopping service in search and tying Android OS to Apps, competition authorities do seem to have teeth and can intervene when there is a problem. So why would regulation be needed?
Richard:
There are two answers to that question. First of all, let’s look at the Three Criteria Test mentioned earlier. The third of the three criteria is whether competition law is sufficient to deal with the identified problems in the market. Competition authorities and lawmakers recognise that there are times when competition law won’t do the job. It is either because competition law could take a very long time or because the right tools don’t exist. The wheels of justice grind very slowly, and perhaps they should do. But it’s not something that’s quick. And it doesn’t happen as quickly as consumers switch from, or don’t go to a competitor because they go to the dominant supplier.
Also, sometimes competition law doesn’t have the right tools. That could be because competition law is designed for markets where there is already competition, but there is some market failure that means that competition needs to be protected in some way.
Regulation is about markets where there isn’t competition, or where there is insufficient competition. Therefore competition needs to be introduced into that market, and where that has to happen, competition law is inadequate because it doesn’t have the tools and it’s too slow.
Benoit:
Very interesting. You mentioned one of Three Criteria Test. Could you remind us what the two others are?
Richard:
First, the three criteria have to be followed in the order in which the European Commission has set them out. Second, all three have to be fulfilled, so they are cumulative. The first criterion is whether the market is subject to high and enduring barriers to entry -which may be economic or regulatory. You can imagine a firm in the telecoms industry building a telecoms network. It’s a very expensive business and the economic costs may be a very high barrier to entry. Similarly, a platform may have become so ubiquitous that its brand becomes a verb, creating a high barrier for potential competitors. So that’s the first criterion. If there are no high barriers to entry, you don’t need to go on with the second and third criteria.
If there are high barriers to entry, and they’re enduring, you go on to the second criterion, which is whether the market is ‘trending towards effective competition’. Now, you can imagine a market where various rivals compete despite high barriers to entry. In that case, while it would be unlikely that new firms will come into the market, it can still be effectively competitive on the other side of those barriers to entry. In telecoms, we can think of mobile markets.
Benoit:
Could you give us an example?
Richard:
There are four operators in the UK, and in many other European countries, and indeed other countries around the world. Those four operators compete pretty aggressively with each other, so you’ve got a competitive market despite the high barriers to entry. For platforms, it would be interesting to think about competition between platforms that are protected by some form of barrier to entry. But if the market isn’t trending towards competition, then you go to the third criterion, which is whether competition law is sufficient by itself to deal with any competition problems. And those are the three criteria. My view is, that if we’re going to think about regulating competition in digital platform markets, if we’re going to think about regulating Google, or regulating Amazon, or regulating Uber, we should be applying those three criteria to the markets to judge whether ex-ante regulation is needed and justified.
Benoit:
It’s very interesting, and we can see how some of these platforms may, based on the judgements of a competition authority, potentially meet some of these criteria. For example, given that network effects often result in a ‘winner take all’ market for platforms, it could be argued that some platforms have erected barriers to entry. These judgements are however pretty difficult to make. They also require a good understanding of the market dynamics, and some assessment about the future level of competition in these markets. Do you think economists are well equipped to make these assessments?
Richard:
I think economists are well equipped, but not perfectly equipped. A lot of the tools that have been used traditionally for market analysis are all here. I’ve mentioned The Three Criteria Test. There are also methodologies for defining markets within which the competition problem might exist, and methodologies for determining whether a company is dominant or not in that market. But these tools may need, and in some cases do need updating, to make them adapted to platform markets.
Benoit:
Why are existing tools not completely adequate?
Richard:
At the risk of going into too much detail, a platform market is two-sided. It’s known as a two-sided market. There are two value propositions for customers on both sides of the market. Take the newspaper reader and the advertiser, the website and the browser, the book buyer and the book seller, etc. Now, each one of those may be charged a price that is not related to the costs of supplying either of those parties. But across the platform on both sides the prices may be at the appropriate level. That’s because it may be free for one side of the market, and the other side of the market pays for it. It’s just an efficient way in which the market can operate.
Now we have to think about the prices that are charged on both sides of those marketplaces. We also have to think about something called multi-homing. So, I have on my smartphone more than one ride-hailing app. If I decide to use Uber, I can, but I’m not obliged to. Simply having it on my phone doesn’t mean I’m going to use it. There are other ride-hailing apps that I could use, or indeed, I could just lift my hand out and stop a cab in the street.
So we have to be very careful of how we define the market, and how we observe behaviour in the market. We need to make sure we understand multi-homing, pricing on both sides of the market, and all these kind of interlaced issues. Economists have the tools, but I think maybe those tools need to be upgraded and updated to deal with some of these platform issues that are there right now.
About Dr Richard Cadman
Dr Cadman is an affiliate of Launchworks & Co and Director of SPC Network. He is an economic expert in network industries and recently published an article on why should platforms be regulated titled Ex ante regulation of digital platforms: Challenges for the three criteria test.