Towards a Competition Enabling Framework in Asia Pacific: Opportunities & Challenges – Japan Chapter

Below, we have provided the full transcript of the panel discussion, Online Advertising, Market Competition, Data and Transparency: Opportunities & Challenges for the Ads Ecosystem in Japan, from the third chapter of our series, Towards a Competition Enabling Framework in Asia Pacific: Opportunities & Challenges.

D. Daniel SOKOL:

Hello, I’m Professor Daniel Sokol at the University of Florida and today I will be moderating the Japan chapter for CPI on online advertising market competition, data and transparency opportunities and challenges for the ad’s ecosystem in Japan. Joining me today is a stellar panel of practitioners and academics at the forefront of these issues. I have Daisuke Korenaga, professor of Tokyo Metropolitan University. Takako Onoki, member of the Competition Policy Committee of the American Chamber of Commerce in Japan, and also Counsel at White and Case. Renato Nazzini, professor of law at King’s College, London, and Kensuke Kubo, associate professor of economics at Keio University and visiting researcher of the CPRC and chief economist of UTokyo Economic Consulting. We’re going to get started with some opening statements and then we’ll have some interaction for some questions. So, let me get started then by asking, and I’ll use first names to make it more intimate, so, Ken, what do we know about the economics of digital ads?

Kensuke KUBO:

Well, Danny, thank you very much for the introductions. So I think I’d like to begin just by noting that there are two issues that are really important, especially in the Japanese context. And those are vertical integration and indirect network effects. So let me start by talking about vertical integration. So I think if you just look back at the traditional advertising industry, they were actually vertically integrated in the sense that ad agencies directly connected advertisers with publishers. And today, when you say that Google is vertically integrated in digital ads, what they’re doing is directly connecting advertisers with publishers. So we’re sort of coming back to square one in the sense that that direct connection has become realized. And the question is rather why it took so long. And I think it’s because the industry was really fragmented in the beginning. The traditional ad agencies were incapable of handling the digital ad business.

And so you had these new businesses that came up to serve various functions along the value chain and Google was one of them. And I think they, some of them, especially Google and Facebook were successful at connecting the two different sides of the market, not only through acquisitions, but also through organic growth. And that organic growth actually created a lot of new technology, a lot of innovations. And of course there was a lot of efficiencies that were generated through this process. And I think that all of us would agree that vertical integration generally creates efficiencies. The most well-known of which is the elimination of a double margin. There are also other effects like facilitation of investments, better sharing of information, et cetera. So why then are we so concerned about vertical integration? It’s because of foreclosure effects, namely customer foreclosure effects and input foreclosure effects.

In the context of digital ads, I think this will probably be mentioned throughout the session, but there are certain issues that have come up in Japan. One of them is alleged self-preferencing by the digital platforms. I think there are several different forms of that practice that are being discussed in Japan. I think there are open questions about whether self-preferencing actually occurs or not. And there are also other questions about whether if there are certain advantages enjoyed by vertically integrated firms, whether that’s because of efficiency or if it’s because of any alleged anti-competitive conduct.

I’d like to move on to the other issue, the other economic issue, which is indirect network effects. It’s actually very closely related to vertical integration because these indirect network effects have been generated mostly because the two sides of the market, namely advertisers and publishers, they’ve become connected through some of these digital platforms like Google.

And so when you have these two different sets of players that are being connected through a single platform, that’s where you have indirect network effects. And I think the most important thing about these network effects is that they do generate a lot of efficiencies. In fact, if you look at the UK merger assessment guidelines, they explicitly state that indirect network effects can lead to higher consumer welfare, all else equal. Which is saying that if the prices were the same, consumers would be better off if there were such network effects relative to the situation where there were no network effects. So it’s just wrong to assume that network effects have anti-competitive impacts when the primary effects are probably pro-competitive. But at the same time, the same guidelines, namely the UK merger guidelines, they say that network effects can lead to, for example, market tipping effects where you have one single operator, which sort of becomes dominant because of the network effects.

And also it becomes harder for new entrants to enter the market once you have a single operator, which has achieved a certain scale. And there are also issues about certain pricing peculiarities when you have indirect network effects. For example, in Japan, some ad tech companies have noted that Google’s ad server is actually provided for free to publishers and that’s giving sort of an unfair competitive advantage to Google. But if you think about how Google operates, if it’s a vertically integrated firm, which enjoys indirect network effects, it actually makes a lot of economic sense to provide free service to one side of the market so that, for example, they can collect a lot of publishers, which makes their platform more attractive to advertisers. So a lot of these issues can be explained by conventional economics. But there are some issues that are more difficult.

And I think one of them is, and I think it’s not an issue that’s solely seen in Japan, something that’s probably more serious in Australia and some European countries, which is fair compensation. A few Japanese news publishers have claimed their share of ad revenues is actually too low relative to what they would expect in a competitive market. There are a few numbers that are sort of being thrown around, like less than 30% of their ad revenue is going to the publishers, but nobody’s really sure what that number is. But unlike in Australia and other places, I think the question is still about how to share the ad revenues that are directly attributable to news publishers rather than ad revenue in general, which is I think what’s happening in Australia. So the question then is who decides what’s fair?  And also whether there are remedies that are based on competition law.  For example, there’s some discussion about requiring digital platforms to disclose the revenue from each advertiser for each advertisement so that the publishers can enter into negotiations with the digital platforms with more information than they currently have. But it’s not clear how that’s going to affect the negotiations or how it’s going to affect the competitive environment. So those are the three points that I wanted to discuss from the economics point of view.

SOKOL:

That’s very helpful. I wonder since now we’ve talked about the economics, Daisuke, I wonder if you can work us through, what is the state of competition law in digital ads? I think you’re still on mute.

Daisuke KORENAGA:

Thank you, Daniel. Can I speak Japanese?

SOKOL:

(Nonverbal affirmative)

KORENAGA:

Okay.

To deal with the problem of digital advertisement, a new law was enacted in Japan. I would like to talk about this law. It is a law on improving transparency and fairness on digital platforms which, on 27th of May, 2020, was passed. The enforcement of this law is planned for next spring.

The basic view of the regulation on digital platforms should be to maintain and promote competition and innovation. It doesn`t have to be based on ex-ante regulation, comprehensively and intervention. It is established based on the thought of the issue that overregulation is problematic and it isn’t appropriate.

Because of this reason, the Anti-monopoly Act and Competition Law are positively operated and parts which has to be made up for supported by this law. There are 3 difficulties of the Anti-Monopoly Law to operate. As you know, the enforcement of competition law has faced strict judicial judgments. It needs strict investigation. And it needs details of determination of facts and collecting proof, so it is not easy.

First is the difficulty in defining the relevant market. For example, it has only 10 years since iPhone came out. It has already given an impact on computer market. The result smart phones have begun to spread, now they are forming one market. Second, it is difficult to evaluate of competition in the relevant market. For digital platform, the competition is often promoted and restricted. It has the precipitative aspect both effects have to be investigated strictly and impact of competition has to be valued. It takes time to relate different to facts. It cannot be taken measure officially, it takes time for the investigation. The Japan Fair Trade Commission is one of the fastest among advanced countries, but it takes a lot of time for the investigation. Then, it may not be possible to take effective legal action while the competitors remain sufficiently competitive.

As I mentioned 3 points of difficulties on antitrust enforcement, having a legal system that overcomes the difficulties of operating the Antimonopoly Act is beneficial for building a competitive market environment by complementing the Antimonopoly Act. From this point of view, we are considering the introduction of regulations that contribute to the maintenance and promotion of competition and innovation, focusing on the following 3 points. First, acts that have the effect that Anti-Monopoly Act can be protected in advance. Second, acts that can be expected the effect to promote the users’ rational choice. Third, acts that can lower down switching cost. Beyond that, in terms of announcing algorism and enforcing big data sharing, there is the issue of the effect against competition and the operation of competition law. Compare to EU, it can be parallel regulation with this. It is important of the difference from EU.

While all online intermediate service and online searching engine are targeted in EU, Japanese law on improving transparency and fairness on digital platforms is restricted a lot. For now, only large online mall and App store are targeted. It is about 5 companies. Both EU and Japan prepare parallel procedure about disclosure obligation and code of conduct. Another difference is the system of monitoring review. The system of monitoring review is based on self-review by the targeted operators, and reports have to be submitted based on its review. The report will be reviewed from certain standard in Ministry of Economy, Trade and Industry. It is not a one-side approach, they discuss each other in the review process. Low review can be put on public pressure and high review can have advanced treatment. To promote this movement, it works on other digital platform operators.

Regulation on Digital advertisement is under consideration so far. I join in this though, this is under consideration including the use of this law. I mentioned this new law on digital platforms because digital advertisement is likely to be subject to the law.

SOKOL:

Thank you so much.

Takako, I’m going to move to you to ask, how do you counsel clients about digital ads related competition law?

Takako ONOKI:

Thank you, Danny. I will make my opening remarks in Japanese.

Danny asked me, how do I counsel clients. I will try to respond to his question, but from a bit different angle. We have to have some kind of rules to deal with it. On April 28th of this year, the Japan Fair Trade Commission published an interim report about digital advertisement. On June 16th, the Digital Headquarters, which was set in the Cabinet Secretariat, issued an interim report about the digital advertising market. The Digital Headquarters said they would aim to announce the final report within this winter, but it is already December and it’s cold. I guess it will be released before spring when the cherry blossoms open up when they deliver the final report.  The JFTC will apparently be issuing a final report around the same timing.

In regards to this, as an answer to Danny’s question, it is useful to perceive the developments of Japan and the world. While various questions are going on, it probably is unrealistic to perceive all of the developments, but just watching the direction of relevant discussions would still be helpful. By the way, I am a member of the Competition Policy Committee in the American Chamber of Commerce in Japan (ACCJ). We communicate member companies about competition law and policies, including to coordinate public comments and submit them to the Japan Fair Trade Commission.

At this moment, I will introduce my experience as a member of the ACCJ. We often hear that competition is important, but at the same time, innovation is also important. We discuss that competition policy should not ignore innovation. I believe that competition authorities around the world including the JFTC understand both innovation and competition are important. When I think about what companies can do to let the Japanese government move to an appropriate direction, for example, it is meaningful to transmit information through trade associations, such as the ACCJ. For example, the ACCJ submitted public comments on the Digital Headquarters’ interim report, which is posted on its website.

One of the various opinions the ACCJ submitted is that we should respect companies’ efforts and self-imposed regulation. Because the environment of digital platforms and digital advertising is changing so fast, and as Mr. Korenaga said earlier, if competition authorities investigate a case and enforce the law, they probably cannot follow such drastic changes of the competitive environment appropriately. Eventually there may be a risk that it will have no effect on the enforcement.

By the way, the JFTC issued “Guidelines Concerning Abuse of a Superior Bargaining Position in Transaction between Digital Platform Operators and Consumers that Provide Personal Information, etc.” It’s a very long name, but in short, the guidelines are about application of an abuse of superior bargaining position between digital platform operators and consumers. There was a public comment opportunity and the ACCJ submitted a public comment. About the regulation of personal information protection, even though digital platform operators abide by its regulation. At the ACCJ, we discussed whether a company who complies with the regulation of personal information protection still would violate an abuse of superior bargaining position?  The JFTC’s answer to this question is yes. However, it is still not clear enough what such companies should do in addition to complying with the personal protection regulation in order to avoid violating an abuse of superior bargaining position. We hope that it a clearer guidance will be provided in the near future.

That’s all of my opening remarks.

Thank you, Danny.

SOKOL:

Thank you so much, Takako. Renato, I’d like to get your comparative perspective on background of what can the European experience teach us about the direction that Japan is now going in with regards to digital advertising.

Renato NAZZINI:

Thank you, Dan. And I take the European perspective as meaning European Union, EU member States and the independent United Kingdom as well. Starting from the independent United Kingdom. Of course, there has been quite a bit of analysis and study of these markets in the UK and in member States of the European Union. And you are aware that the Competition and Markets Authority in the UK published in July, 2020 a major report on digital advertising, which I think it’s a very thorough report as you usually expect of the CMA with very interesting findings. I would say I read the report as containing findings going in different directions. Of course, there is some evidence of harm and market power in these markets, but also interesting findings like the one that Google’s intermediation fees at the end of the day are in line with the fees charged by or other intermediaries, which of course will strongly suggest absence of at least significant market power in intermediation.

So and what’s surprising, I think, in the CMA report is then the very drastic and draconian conclusions and recommendations that are, whereas as I would say the rigor in the analysis in the report is very much what you would expect of the CMA. The conclusions that go in the direction of a mandatory code of conduct administered by a regulatory unit, which now we know will be the digital market’s unit sitting within the CMA with very strong powers, not only to police this code of conduct, which is still to be drafted, but it’s expected will contain obligations of fairness and transparency on these markets. But going as far as mandating structural separation and even ownership separation. So divestiture in order to avoid perceived harm in these markets. So I think something which is very extreme given mixed evidence, I’m not one of those who say that there is absolutely zero market power and zero issues in digital advertising, of course there is.

There may be some problems and by and large, I think competition authorities have the tools to address those. Perhaps they should better equip themselves with digital market experts and engineers alongside economists and lawyers. But I think our competition tools certainly can address most of the harm that is going on if there is evidence of it. But I was certainly surprised by those conclusions. What’s happening in European union is a number of member States conducting studies on digital advertising as well, and we have now a number of European commission initiatives, which are expected, are in consultation phase or in the proposal stage, but are expected at some point to become legislation in one form or another, that will affect the digital markets generally and digital advertising as well. So for example, many will be aware that European Commission has consulted on a digital services act and on a digital markets act. A package which is intended to include essentially regulation of digital platforms, again, based on principles of fairness, level playing field and ensuring equality of opportunities.

The specific content of this legislation is still very unclear and very vague. Again, I was struck by the fact that the evidence is at best mixed. As in UK, I think that the CMA has done at least a better job at evidence gathering and consultation than European Union institutions have done. But again, it is expected to be some quite intrusive regulation potential, we will see. Based on, as I said, the best mixed evidence and most of the harm could be addressed by existing tools. European Commission has also consulted on a new marketing investigation tool, they call it the new tool. It’s essentially marketing investigation as a result of which unlike today, European Commission would be able to impose behavior on structural remedies.

And one of the options is to have a general instrument, which would apply to all sectors of the economy. But the driver for that is clearly the digital economy and there is clearly political intention and determination. Again, we will see the legislation to use it in digital markets. And again, the commission there will have powers to impose behavioral remedies, but also structural remedies including ownership separation.

So there are also other proposals such as the EU data governance act, which aims at facilitating the sharing of data. We will talk about it later, I think today. But so what is the lesson that one can learn from Europe? I mean, first of all, nothing much has happened yet in terms of hard law. And so one should probably draw no lessons at all because you may want to wait and see what the UK and European union do and see how this legislation looks like and what impact it has. Secondly, I mean the other lesson that one can learn is that actually, if you look at the economic record of Europe in the digital environment, in the digital marketplace, it’s actually not very good.

All the major platform, of course, there is a very good check search engine and there are all sorts of European platforms as well, but the very large players in these markets come from the United States or indeed from the far East some of them, but certainly not from Europe. Sometimes it is said, well is European Union or European commission always finding or investigating US platforms for abuse of dominance and so on and so forth? Do they have anything against US? No, the sad reality is there are no EU platforms to investigate. So that record in other words of the European economy is not very good in this space. So one would want to think twice or perhaps three times before importing any solution from the European Union, which is clearly not very platform friendly environment.

There is also another lesson to be learned, which is that the European union would ever want things about cases like Google Search, AdSense, for example, which is about online advertising. That is a case where European Commission took issue against certain contractual and softer kind of indirect exclusivity type restrictions that Google was trying to implement, was allegedly implementing in relation to its intermediation services with websites and competition law of course, is very well placed, I think, to tackle issues such as those. Of course if you have exclusivity type or interoperability type of arrangements which prevents multihoming for advertising, in principle, as always you have to check the facts and the evidence, but in principle, this of course could limit competition. As European Commission enforcement record shows, a competition law may very well and can very well deal with these issues and has been used to that effect.

And the final point I would like to make here is that also what the European record shows in digital advertising ending up where I said, I mean, there is still not much except one or two cases. Google Android of course was a case about Android and tying of Android with Google Search or Google Chrome through the Google Play Store as a mass stock product. But it was also about harm happening on the digital advertising markets. But these cases, of course, and their appeal is a bit more controversial. I think that Google Search outsets that exclusivity case for intermediation services. But the other thing, as well as the possibility of taking action under existing competition law, I think the fourth point I wanted to make was one should be very well cautious against and about the risk of regulatory fragmentation worldwide.

We know what the general data protection regulation has meant for many business around the world, operating in Europe, increased compliance costs and increased complexity in actually the being and acting on the digital markets in a competitive way, because the general data protection regulation is in itself a hindrance to competition in many ways. I’m not saying it’s necessarily hundred percent a bad piece of legislation, surely personal data must be protected. But I think what we need to look for now is a better way of doing that, especially in digital because these digital markets bring lots of benefits because they are global because the economies of scale that can be realized are unprecedented and across jurisdictions and across continents. If every jurisdiction around the world, by in large, if I base myself or look at the European experience based on political perceptions and political will, rather than hard evidence start regulating in more or less draconian ways, the digital space, I think who will at the end of the day suffer will be the consumers in these jurisdictions which will have lower quality services.

And actually advertisers and publishers as well, which will reach fewer consumers and have at the end of the day lower services as well. So I would then as my final point, just make a plea for a multilateral and more international approach. I know it’s difficult, but has been done in other area. For instance, the Asia Pacific Economic Corporation Group actually has put forward in the area of data protection a very good multilateral framework I think for the protection of privacy, which has been recognized in the Mexico, Canada, US agreement of 2019. That’s just a small example, but an example of how actually multilateral action by market players themselves are also backed by governments and institutional organizations can happen. And I think it would be a much better solution to the problems that there may be in this sector.

SOKOL:

Thank you. With those opening thoughts, hoping for me to frame the discussion, I want to move to the first substantive question for all of you to answer, which is, if we look at the institutional changes in the Japanese competition policy system, the creation of the digital headquarters and shifts that we’ve seen in recent years in the JFTC’s approach to digital ads, what has this meant in practice? Daisuke you have, perhaps, unique knowledge. Things both that you can talk about and things that you have to be a little more circumspect in talking about. So I’ll let you begin and from there I’ll ask some of our other panelists.

KORENAGA:

The purpose of the headquarter of digital market competition is because of 3 organizations (Minister of economy, trade and industry, general agency). Those 3 organizations thought they had to establish the regulation about digital platform. On the other hand, to think about necessary regulation, a regulation, and they devised a certain policy.

In fact, each ministry and agency takes action, but the Japan Fair Trade Commission is not restricted and they can enforce their authority freely as before. That part didn’t change. On the other hand, as I explained earlier, the Japan Fair Trade Commission is expected to play a role concerning law on improving transparency and fairness on digital platforms. If eventually acts by digital platform companies may go against the Competition Law, the headquarters can let the competition agency deal with the acts. While thinking about which department can play which role and not being restricted on, we have to think about necessary regulation.

SOKOL:

Thank you.

Takako, you’ve, again, had to think about the realities of these institutional changes in your client counseling. Do you have anything that you’d like to add?

ONOKI:

Well, Professor Korenaga, explained in detail, so I have almost nothing to add. But let me add one sentence. My understanding is that the Digital Headquarters and the JFTC cooperate for this purpose, innovation and competition, and the JFTC, of course, focuses from antitrust law or a competition law perspective. And what they do sometimes, or maybe most of them will overlap, but they also compete. And from client’s perspective, it’s important for government agencies to compete to achieve more effective or good competition policy. That’s my observation.

SOKOL:

Renato, as you’ve heard some of these comments then about institutional structures, I’m sure this gives you a lot of opportunity to think about how you’ve seen institutions develop as well. Anything that you want to add?

I think much like the movie, he seems to be frozen, Ken in the meantime.

NAZZINI:

I can hear you.

SOKOL:

Oh, good. You’re back. So I’ll repeat the question, which was, you’ve seen a lot of institutional change as well and have good perspective. What might you add in terms of how should Japan think about the institutional design to better help with value creation in the digital ad space?

NAZZINI:

Yeah. Yes. Well, I think, when it comes to institutional design and especially now, when it comes to dealing with the digital economy, in my experience, in Europe, this has become a hugely politically charged issue. And the question is, should it be? Of course, I mean, I’m not advocating for taking politics out of everything, quite the contrary of course, politics is more to govern us all in a way.

But the fundamental question is, when it comes to dealing with problems in digital ads, and looking at competition problems in the digital economy, should we, and I think this is a kind of a sense or an approach we have got used to, should we follow the science, so to speak? So, should we follow actually the evidence, or is that a matter of politics? And people’s perceptions that may be right or wrong, we don’t know. And certainly, in Europe, there is now this political wave against big tech. Big tech are bad.

Which is, to an extent, I think, irrational and not based on evidence. And we have seen, I mean, even with the pandemic, how important, actually, digital services have been for all of and free services, especially which are funded by online advertising. So any institutional design should get, I think, the balance between politics, and economics, and technical expertise right.

And again, I cannot comment on Japan too much, but I think that’s absolutely important. I mean, it’s significant to me that the headquarters are with the cabinet office, and then you have the independent Japan fair trade commission.

So where should the balance lie between the two in terms of identifying problems and analyzing the evidence? And also, let’s make… Any institutional balance needs to address the point of keeping then politics separate from the evidence. So we have to look at the evidence and then make political choices. We should not pretend to make decisions on the evidence that are political decisions.

And for example, if I look at the… Especially, I find it interesting to read the incident report of the digital headquarters, which I write I read, of course, in a translation, in an unofficial translation, because my Japanese is almost non-existent, I should say. You see in that report quite a bit of evidence of a competitive market. For giving just a couple of examples, one of the finding is that actually intermediation prices for pay per click in Japan are lower than in Europe or in the United States.

Now, that’s very good news for the digital market and a well-functioning market. Now, you can say anything, but that Japan is a country where the overall structural cost is very significantly lower. Everybody who traveled to Japan knows that Japan is not a cheap country. And so, lower pay per click fees are actually quite significant. There is perhaps a duopoly in online advertising between Facebook and Google that have about 60% of the market. But again, 40% of the market is up for grabs for others. And also, the report points to Amazon as a significant player that now is increasing in importance.

So, I was surprised in that report. And I think it’s a question to our Japanese fellow panelists here, whether this has to do with institutional structure or other issues. I was quite struck by the disconnect, a little bit like in the UK report, but I think even more so in the digital reporters report between the evidence with… Okay, suggests there may be some issues around transparency and there are advertisers complaints and publisher complaints, fine. What do you make of those complaints? Are these commercial issues? Are there competition issues? We can discuss about it and certainly look at it further. But also quite a bit of evidence of healthy, competitive markets.

And then the recommendations that are, again, to use the word that I used in relation to the CMA report, quite draconian as if there were major issues in Japan that have to be urgently addressed by means, so quite important in intrusive regulation. So, I stop here, but I think I would be interested to know what, those who actually know about Japan, think about it.

SOKOL:

Ken, that gives you maybe be an opportunity to respond to Renato and his thoughts as well as what Daisuke and Takako have also shared. And any thoughts that you might have on this topic of institutional design?

KUBO:

Well, I would just like to analyze the capabilities of the institutions that are here in Japan with respect to the economic issues that have been identified, especially in the Digital Market Competition Council report, which came out in June. I think they devote a significant portion of it to the issue of vertical integration, which I also mentioned in the beginning. And namely, I think it relates to foreclosure effects that, as Renato mentioned, some ad tech companies, and in some cases, some publishers, have been complaining about.  These are claims of unfair treatment and, perhaps in certain cases, input foreclosure.

And also, I think there was one case where ad tech companies were concerned that Google had removed YouTube content from the traditional advertising channels so that Google was the only one that was handling the ad inventory on YouTube. I think it probably did have a significant impact on some of the ad tech companies. But then again, you have to wonder, is ad inventory on YouTube, is that an essential facility? In other words, does Google have a duty to deal with some of its competitors?

So I think some of these issues are quite traditional in the sense that you can think about them in terms of conventional foreclosure theories, and apply the methods of analysis that the JFTC has used in cases pertaining to unfair trading practices, as well as private monopolization. And as Daisuke-san has mentioned, I think the current setup under the digital market competition headquarters allows for the JFTC to pursue cases which it deems should be subject to their regulation. And I think that will be the case, certainly for foreclosure issues.

I think what’s going to be more difficult would be the issues pertaining to network effects. The JFTC has not been very good at handling network effects. Let’s put it this way, the new merger guidelines that came out in December last year has a new section on network effects, but it doesn’t say how exactly they’re going to deal with network effects in a merger context. And I think it’s the same here. It’s not a merger context that we’re discussing tonight, but the same issue of tipping effects, and also the efficiency effects due to network effects, are extremely relevant in the digital ad industry. But it’s not clear how one should regulate the industry, if at all, because I think that the efficiency effects would have to be overwhelmed by the anti-competitive effects in order for there to be any justification for regulation. So, I can’t say that the JFTC is well-equipped to assess either the efficiency or the anti-competitive effects of network effects and deciding how to regulate, if at all.

The other issue that also presents difficulties is bargaining issues. And I think the ACCC has been at the forefront of handling the news publishers concerns with respect to bargaining vis-a-vis Google, Facebook, et cetera. I certainly doubt that something of that scale and severity would happen in Japan, but at the same time, the recent Digital Market Competition Council report does mention similar concerns from news publishers in Japan about their share of ad revenues being insufficient, and also about issues of transparency that was mentioned by the other speakers. So, I think, the first step would probably be perhaps to increase transparency, which may or may not require new legislation as Daisuke-san has mentioned.

But again, like I mentioned in the beginning, it’s not clear what transparency is going to bring. It’s not clear how clarifying what advertiser paid to Google for each advertisement, how that’s going to affect the payments that news publishers receive from Google. So depending on the outcome, there may be calls for further regulation. I’m not sure if that would actually happen, but if it did, it will certainly be an issue that the JFTC would find extreme difficulty handling. And I’m not sure if there’s any agency in Japan that’s capable of handling this difficult issue. I’m sure the ACCC is going to have a difficult path forward. I think the world is watching what’s going to happen in Australia, and Japanese regulators is certainly one of the parties that’s closely watching the situation over there.

SOKOL:

Thank you for your impressions. I wonder, since you’ve brought up network effects, if I could get the other panels to give their sense of how network effects either apply or maybe sometimes don’t apply, that the network effects are not necessarily durable, for example. Renato, I wonder if we could start with you, your thoughts on network effects in the ad space in Japan.

NAZZINI:

Maybe, Dan, I talk about network effects in the ad space more generally, because in Japan… Is that all right?

SOKOL:

That’s good too. Given that Ken has brought up Australia, you can bring up whatever jurisdiction you’d like.

NAZZINI:

Let’s look at the market itself, because I think then these, in a sense, they function in a very similar way, of course, with different features, but in terms of network effects, I think… And it’s probably very similar around world. I mean, there is no doubt that there are some network effects on these markets. I mean, the more people are on the social network, the more people use the general search services, the more then that advertise medium becomes a valuable for advertisers and publishers. And so, that is quite clear in that sense.

But network effects in intermediation, I am less sure there are so relevant as such. But one point I wanted to make about network effects more generally is this, actually. I think in Europe, and to an extent elsewhere, it has become normal, almost, to equate network effects to some kind of competition problem. Or well, here there are network effects. So, something needs to be done. Well, I think network effects are just a way of describing a way in which certain markets behave, as a matter of the features of that market and the structure of that market. I mean, the more certain users utilize a service, the more that service becomes valuable, the more users on the other side of the platform can find that market, who can find their services more useful. So direct network effects and indirect network effects as well. I mean, they are a feature of many platforms and they are a feature of online advertising as well.

But whether that translates necessarily and automatically into a competition problem, that’s quite a different question. I mean, European Commission itself, in a couple of merger cases, made that point and said, “Well, network effects are an economic phenomenon.” Now the second stage of the analysis to see whether network effects are a barrier to entry, for example, or are leading or have led to the market tipping in some way, so there is now an entrenched market power. And these are the questions that we need to ask. I mean, the analysis of network effects is just a tool to understand whether there is competitive harm going on, whether that particular company has the ability to harm competition, has durable and substantial market power that is no longer contestable.

I mean, these are the real questions to which I think network effects are just a necessary step, or if there are network effects in the market, in the analysis, but are not the end of the story. We cannot say, “Oh, there are network effects, hence we regulate.” Or, “Hence, we intervene.” And I think there is, very often, too much emphasis on network effects and too little emphasis on competitive harm in these markets. If you read certain competition decisions around the world or certain reports around the world, there are 100 pages on network effects and 10 pages on competitive harm. Well, I think this probably should be 100 and 100 and 50 and 100, but certainly, the proportion is not quite right, in my view.

SOKOL:

This discussion reminds me of network effects in Microsoft. There’s no question that Microsoft has network effects. There was a Japanese case, some 15 to 20 years ago, but you know what doesn’t really matter anymore. What overtook Microsoft was in spite of the fact of network effects, the competitive impact is much less now. And it’s because other technologies came in that, A, were unanticipated, but B, showed the limits of network effects. With that, Daisuke, do you have any thoughts on network effects?

KORENAGA:

Thank you.

In my opinion, network effects are not evil. On digital platform, downstream market can be operated for free. Zero-price market is because of the effect of network effect. Network can effect on not two-side effect but multi-side, it is the issue of economies of scale and scope. Because of that, I think disruptive innovation can happen. It is the problem that network effect becomes the major axis in the theory of harm, because network Effect will prompt to be competition at the same time restricting competition.

SOKOL:

Takako I wonder if you have anything to add on network effects?

ONOKI:

No, I’m good. Thank you.

SOKOL:

Okay. Then, I’ll ask one additional question in the interest of time. Which is, what role does privacy play for digital advertising? And what is the nature of the interaction between privacy and competition policy? Is this something, Takako, that you might have some thoughts on?

ONOKI:

Thank you. I will make a quick observation. Of course, privacy and competition are both important. And what recently have been discussing in Japan, I understand is, to protect privacy is important. But as mentioned in my opening remarks, that consumers provide some privacy information to digital platform, and consumers receive services. I use Facebook, I use Google, and they are part of my life. But at the same time, I, as a user of those digital platform, I should be aware what privacy information I provide to digital platforms. And I think that’s important. So, I have choice whether or not to provide or I decide not to use that platform if I think it’s not reasonable. So, that’s something we have been discussing in Japan and I think professors have more detailed explanation about that, so I will give over to Daisuke and Kensuke.

SOKOL:

Ken, why don’t we then, with that, shift to you to get your thoughts on this interaction between privacy and competition?

KUBO:

Yes. I probably have sort of a minority view with respect to this issue, especially in the competition community here in Japan. But maybe it isn’t the minority view, depending on, I guess, who you speak to. For example, I totally agree with Takako-San, what she mentioned in the opening statement regarding the fact that we do have a Personal Information Protection Act.

One thing it does lack though, is administrative surcharge for violations of the Act, which, if one takes a cynical view, is perhaps the only reason why the personal information issue had to be put under the Antimonopoly Act which has a surcharge system. So it does have a deterrent effect on firms who may be tempted to abuse the personal information that they collect.

So to me, it doesn’t really make sense that personal information should be considered an abuse of superior bargaining position with respect to the consumers, because in fact, consumers are already protected against abuse of superior bargaining position through the entire AMA. In fact, the price increase, for example, through a merger, can be considered an abuse of superior bargaining position. It’s not called an ASBP, but basically, that’s what it is.  But we don’t really need to phrase it as an ASBP. So it’s sort of puzzling to me why this information issue had to be particularly placed within the ASBP framework. “Is this another abuse of the ASBP?” was my initial reaction to this particular measure, but maybe that’s just an economist’s opinion, which may differ from the mainstream competition view here.

I’d like to add that we would probably be much better off strengthening our Personal Information Protection Act. For example, by introducing administrative surcharges on violations of the Personal Information Protection Act, which has been opposed by businesses very strongly, especially the Keidanren. Certain laws are meant to serve certain purposes, and I think it’s a mistake to use good laws such as the AMA, which has certain functions, for other purposes.  Those functions should be protected by not using the law to serve purposes for which it was not meant. So, that’s my opinion on privacy/personal information and competition law.

SOKOL:

Daisuke?

KORENAGA:

I agree with Kensuke.

As I mentioned, ASBP, in Japan why Competition Law can deal with about privacy is that Abuse of Superior bargaining position, that is exploitative abuse, is contained in Japanese Anti-Monopoly Act. So that is the difference from U.S.

In Germany, the BGH, the supreme court in Germany, decided the important judgment suggesting illegality against Facebook. This is very important to think about ASBP. To use Japanese ASBP, it is necessary that there is no choice of other alternative business partners. ASBP becomes a requirement for it. We can find the big common point inside in both jurisdictions. When SNS users take considerable choice on more privacy, the Facebook lose valuable data of them. Then advertisers lose targeted users as customer through the platform. Looking at competitors, for some of them, the network effect of some of them will be weakened, so it will be a more competitive market for that part. It seems that the goals of competition law includes to maintain open choice of the customers as I said.

SOKOL:

Great. Well, at this point, maybe I’ll ask each of the panelists for some closing thoughts. There are so many different things going on in the digital advertising space in Japan that may require us to be thinking more deeply about these issues. Ken, I wonder if we might start with you. Sort of what are next steps and how might we get there?

KUBO:

Frankly, anything I say would probably have been talked through already by the other members. In particular, Daisuke would probably have discussed a lot of these issues already in the Digital Market Competition Council’s Working Group.

That said, I would say that the issue that seems to be attracting attention internationally is the sharing of ad revenues. And I think it’s also the most difficult among the issues that we’ve discussed today. I think some of the foreclosure issues are more easily remedied. In fact, some of them probably don’t require a remedy.

Let’s just say that I look forward to what the new report from the Digital Market Competition Council, coming out in December, will conclude about the revenue sharing between news publishers and the vertically integrated digital platforms. And hopefully, further decisions on individual cases, not so much on general policy matters, but rather individual investigations and so on, would be carried out on the basis of rigorous economic analysis.

I would just like to add that, although network effects is a difficult beast to handle, I think there is one way where it is possible to make the right decision in the context of network effects. And that is to use economic models to conduct simulation. That’s probably the only method where decision-making could be done scientifically, where you have these really complex interactions going on. It’s really impossible to make the right decision just on the basis of simple demand estimation, or simple surveys of consumers, et cetera. That’s really not enough when you have so many complicated network effects going on, especially in the digital ad industry. So, that’s where I would like to close my statement. Thank you.

SOKOL:

Thank you. Renato, do you have some closing thoughts for us?

NAZZINI:

Yeah. Yes. I think, maybe I just comment on this interaction between privacy and competition, because this is something that was very much addressed in the interim reports as well. And again, I think, and my understanding is that it is very much on the agenda in Japan. And I think, while I agree with the view that has been expressed. Obviously data protection, privacy, and competition, or digital app regulation, are two different things, and if we have data protection issues, they have to be addressed under the privacy laws. However, based certainly on the EU experience, I don’t think it’s realistic just to keep the two things completely separate because they do interact and we need to be aware of the interactions. And I’d like to mention just two areas. I mean, there are many, it’s a big topic. As I mentioned, two point that perhaps are going to be relevant going forward in Japan.

First of all, from the perspective of the consumer and the user of services, there are issues to be looked at, I think, they are not necessarily problems, but there are certain issues there in that interim report on digital apps by the digital headquarters about targeted advertising and more generally, the quality of ads that consumers get and the degree of control that consumers have on the apps they receive. And very much linked from the perspective of the consumer, how much data the consumers give to platforms that then can be monetized through advertising.

And the Bundeskartellamt, it’s a German competition authority, took a decision against Facebook that then was upheld by the German Supreme Court on abuse of dominance grounds. And so under German competition laws, saying that Facebook was obtaining too much data, essentially, and there is no time to go into that today, but it’s well known. Saying that Facebook was taking too much data from consumers. Now, I think that decision is wrong, as a matter of competition… Well, not German competition law, but wrong is a matter of competition policy, but I think it’s quite a good demonstration, a proof if you want, that we cannot ignore these issues when we talk about competition policy.

And in the second area, I think is the restrictive effect. And that’s an interim report as well. The restrictive effect that actually privacy laws can have on competition. And measures that, therefore, I think do need to be taken. First of all, to ensure that privacy laws are not drafted in such restrictive way. I think that general data protection regulation in Europe could have been a better piece of legislation. And secondly, whether there are measures that actually can be taken to ensure data portability and data sharing, while respecting our privacy, which is, of course, an important value.

And here, the European Commission has now published a proposal on the EU Data Governance Act that aims at facilitating data sharing. Again. I mean, I think, I’m always a bit critical, I’m afraid of European legislation because it’s very bureaucratic, complicated, heavy on regulation. But I think with improvements, this initiative is not bad initiative. Facilitating sharing of databases, sharing of data, anonymized if possible, otherwise with the consent of users by public entities, if public entities have this data on a non-exclusive basis to third parties, by intermediaries, we can create a market of intermediaries that facilitate data sharing and data portability. I think these are good initiatives. Let’s just not kill them in the start by making them too complicated, and too burdensome, and too risky for all involved. But I think, previously in competition policy and in digital regulation are three areas that now, inevitably, interact.

SOKOL:

Daisuke, we’ll get final thoughts from you to close out, please share with us some of your insights?

KORENAGA:

Okay. OK. As Renato mentioned earlier, I would like to emphasize the importance of data sharing in this area. In Japan as well, consideration of this data sharing has finally begun. Data sharing should be promoted now, but there are still aspects that promote cartels in the form of sharing, and there is a problem with information on what to do for misuse of data. For instance, data sharing may be used as cartel signaling, or data enclosure may impede data sharing. Anyway, there are some problems in digital advertising area, but I’m hoping that we can solve them.

SOKOL:

Great. Thank you, everyone, for your participation. On behalf of CPI, I want to thank everyone for their insights on digital advertising in Japan. Thank you.