 Okay, so I should start with a couple of disclaimers. The first is my usual disclaimer that I'm a non-exec director at the Competition and Markets Authority. So I always need to say that all my views are personal and don't necessarily reflect the position of the CMA, not that I am necessarily gonna say anything they would disagree with. But the second disclaimer for the purposes of today is this is really not a paper like the sorts of paper that are usually given in this lunchtime seminar, which I fairly religiously attend and I know the sorts of papers that are presented and this isn't one of them. So it was Julian Wright's idea to invite me to present this and he claims that we, that actually occasionally there are kind of very policy-orientated kind of translation papers presented and this is basically one of them. So there isn't a huge amount of economics exactly in there but the whole thing is kind of underpinned by economics. So and a lot of the economic thinking that this group are doing links into this policy debate. So I think it should be useful. Some of what I'm just gonna say, many of you will know but some of you may not. Hopefully there will be something new for everyone. And it's the paper is forthcoming in the Oxford Review of Economic Policy early next year. There's an SSRN submitted version on SSRN. So why is my, why it's not going on? There we are. So what I'm gonna cover is very briefly the first three sections I hope. What problem is pro-competition digital regular platform regulation trying to solve? Second is why regulation are not competition law, which obviously we're all more well-versed in competition law. Third, what are the design challenges involved in developing this kind of regulation? The third is what are the pros and cons of divergent regulatory approaches? And the final one is how might these risks be mitigated? So first of all, what problem are we solving? And it is true that there is a decent case for leave well alone. So we know that digital platforms have brought huge benefits for consumers. I'm not gonna go through all these, but for firms, for competition, for innovation. We also know that there has been and can be intense competition to win markets in the first place. And that even if we see monopolized markets, they can still be contestable the kind of Schumpeterian creative destruction idea. And that markets here have been at least very dynamic and very fast moving. So there's an argument that any intervention will be at best irrelevant and at worst create serious harm. And I think it is worth bearing in mind all of those concerns as we enter the world of regulation. However, on the other hand, we do see a number of digital markets and this is just a simple figure that looks at European market shares in search, mobile and desktop operating systems and mobile browsers. So a number of really core digital markets which have tipped to being highly concentrated seemingly over the long-term that exhibit or seem to exhibit very high barriers to entry and expansion where market power has been or at least risks being extended into related markets creating ecosystem-wide issues. And also that platforms have bottleneck power over competition in related markets. And in some environments, I would need to explain what I meant by that but I don't think I need to here. And the issues are global. So this is the market shares in search by continent. And what you can see there is that Google has a very high share everywhere except China where they do as a very, very high share. So it's a similar situation in China but the people, the parties are different. And so what are the economic implications of that? There's a risk of direct harm to end users because actually we know that in two-sided markets even though things are free, zero price may actually be higher than the competitive market price. There may also be if there's very little competition reduced quality, there may be excessive data collection for them end users may also suffer via the impact on business users. So for example, high fees to business users for advertising for example, or for intermediating a product sale are likely to be passed on and also if you are a producer of content so if you're a publisher, then if the advertising chain is taking too large a share of the advertising revenues then there's less advertising revenues to fund content and that's going to be bad for consumers and users as well. And then probably the most important risk is the dynamic risk. And I know that there's a big debate about competition and innovation which I'm not going to get into here. But if you look at how innovation works in tech what's clear is that for an innovation to be successful it needs to have access to data. It needs to have access to users. It needs to have access to kind of the relevant infrastructure that it's going to fit with. So it needs to interoperate and it needs to have access obviously to fair reward if it's successful. And for many potential innovations and many of these aspects the biggest tech platforms essentially control these things. So they control the access to the data, the users the infrastructure and in the end they control a lot of the reward that you might gain as well. They can also, they're also in a uniquely good position to copy any successful innovations and although they've done a huge investment in innovation in the past their incentive to invest themselves is clearly reduced the more that they don't feel assailable by other innovative rivals. There can be Schimpaterian creative destruction clearly but there's a growing feeling that while that's possible and it may occur in certain markets and maybe TikTok is going to be the destruction of Meta we don't know then it can't be relied on to solve these problems. And none of this covers all the stuff around political power, online harms, democracy, resilience risk, et cetera. This is just about the competition side. Another key point is that many of the key drivers of what we're seeing here are inherent in the economics of these markets. And again, this group have discussed lots of these but the huge global economies of scale and scope the network effects both within platform and ecosystem wide the data feedback loops. So you collect data by providing these services but the data is also a key input to the production of the services and the more data you have generally the more or often the higher quality the products you can provide which again, then wins you more customers more data, et cetera. And then also I think important in this space is the cognitive limitations of end users and the ability to really be quite granular in manipulating the choice architecture that end users see you can do constant AB testing to make sure that basically the choice architecture is playing on consumer biases in the way that most it may be doing it in good ways in some ways and you're designing your user interface to make consumers have a nice easy consumer journey but that can also be used in more exploitative ways and even making an easier consumer journey can kind of inadvertently create harm as well. And these factors all apply across markets as well as within them. So that really drives ecosystem growth and that's even before any strategic firm behavior so and deliberately anti-competitive behavior deliberately anti-competitive merges and there has been that too. So the key question then is can ex post competition law deal effectively with all of this on the plus side, there has been a huge amount as many of you will be aware of activity in the competition law space. These are a whole set of cases internationally that have got to fruition. There's lots and lots of cases ongoing, many more than this but this is the ones that have actually got somewhere and so they've been successfully used to change conduct in the digital sector. On the negative side, there are concerns that and I think, you know, these concerns but they are I think these are very serious concerns that ex post competition law enforcement it takes an awfully long time is very backward looking. So you're always looking, you can look a little bit prospectively but inherently you're looking pretty retrospectively in these relatively dynamic markets in these tippy markets where you need to act quickly sometimes if you're gonna stop them from tipping and once they've tipped it's too late or it makes it a lot harder. Ex post competition law enforcement is very narrowly focused typically so you need to make things narrow to create a decent story that's gonna get through the courts but then if you end up with these kind of narrow apparently somewhat esoteric cases that are true on the fact that work on the facts that's not so great at creating deterrents given the complexity of these markets and the patchwork of different issues arising and the interlinkages between those issues and there's a risk and we've seen some of this already that if you end up with remedies coming out of your ex post competition law enforcement you end up with a kind of panoply of requirements of these firms which require ongoing monitoring anyway sanctions for non-compliance anyway and that's effectively a form of regulation anyway but in a slightly ill thought out patchwork way. So that's one set of reasons why competition law enforcement is not great but the other set of reasons and I think they're really important is going back to the slide before many of the key drivers that we see in these markets are kind of inherent in the economics of these markets. So in economic terms entry or expansion in these markets may be blockaded in the Bain sense rather than deliberately strategically deterred and if you've got that sort of situation and you're concerned about the lack of competition that arises then you may need to be more proactive to promote competition than using competition law which is most typically used where firms are deliberately behaving anti-competitively there's a legal debate about whether it can be used more widely but it typically isn't because it is people are cautious and then there's also been concerns about anti-competitive mergers that is somewhere something where potentially the authorities could do more and they are starting to do more but over, this is a set of the this is just an image of the M&A activity by the five largest tech companies over the last 11, 12 years and what you see is an average of nine acquisitions per company per year clearly some of them are gonna be fine the vast majority of these should probably have passed even if they had been investigated which most of them weren't but even if they had been investigated they should have been allowed but some of them should not we could all debate which one should not it's a fun game over dinner but in practice, none have been a few were remedied usually the remedies weren't very effective none were blocked until the CMA blocked Metagiffy this year so merger law and there's a debate about whether merger law needs strengthening or whether the authorities just need to be braver and think about theories of harm in a more effective and brave and thoughtful way but that's obviously not gonna help deal with retrospective issues so the outcome of all of that is there's a growing consensus at least amongst quite a lot of people that we need pro-competition regulation Germany actually went first with this with their amendments to competition law in January 2021 and what that is it's formally competition law but it allows the regulator to designate a set of platforms that are considered undertakings with paramount significance for competition across markets which is the very catchy up scam digital platforms and once you've been designated that reverses the burden on a whole set of different rules such that as long as basically you have to show that doing your behavior is good and efficient the German authority doesn't have to show that it's anti-competitive obviously if you can show it's good then they have to deal with that argument back but it reverses the burden so that is the first thing that happened Europe I think could see that all of the member states might well follow Germany so they wanted to develop a more formal legislation and you'll all know about this the digital markets app which was legislated in September and the obligations were bites from early 2024 in the UK there was at the UK started looking at this pretty early with the thing called the Furman report which was four economists and one computer science guy who three economists, one lawyer and one computer science guy one of the economists being me and we recommended regulation in 2019 and the UK government immediately said yes we're interested, we will look at doing something in fact they've ended up being much slower than Europe but they did create a shadow digital markets unit so that sits within the competition of markets authority and that is already thinking about what the regulation will look like and what it will need to do and is kind of involved in a whole set of competition law cases and market studies looking at digital markets issues with a view to informing that regulation we thought the legislation might have disappeared with more recent versions of our current government but the latest version of our current government has announced very positively very recently in November that they are going to take this legislation forward in the next year so in the US you'll probably be aware there was a hard hitting report in 2020 which then led to five and in fact in the end six bills proposed for covered platforms which like designated platforms the sixth bill would only apply to app stores that came out in June last year it's some of them have gone by the wayside but the big one, the main one or possibly two are still going through the process because of the most recent elections it's probable I understand that it won't go through but it still might and obviously we don't know what will happen after the next elections interestingly there's some fun wrangling that even goes on even amongst the US Democrats so when the EU was at quite a late stage in the DMA Gina Raimondo who was the Secretary of Commerce wrote a letter to the EU saying this is a profoundly anti-American regulation and it's absolutely terrible because obviously the five biggest firms are American at which point Elizabeth Warren, same party wrote to her to say this is exactly what we're trying to do in the US and indeed our regulation would only cover those five firms too so please basically shut up so interesting that then also very interesting and I'm not enough on top of this but in the paper I do try and say as much as I can about it given I'm not that on top of it but in China the State Administration in 2021 the State Administration for Market Regulation also published, well it's done two things it's published some competition law guidelines that really help target competition law more effectively at platforms and they have been finalized so competition law has been made stronger in China in this space but they also published these draft guidelines which if finalized would create basically regulation not dissimilar to that proposed in the West although their regulation seems to cover both the pro-competition aspects that the DMA covers and some of the contents issues that the Digital Services Act covers in the paper there's links to the draft so we've got this regulation coming and lots of places internationally are looking at this what are the design challenges of creating this sort of regulation remembering that this regulation is intended to complement and not substitute antitrust so we're still going to have antitrust but the key aims are to be quicker and more administrable than antitrust and to proactively open up markets so to deal with that kind of blockaded competition issue and not just to protect against further worsening but that desire to be quick and be quite interventionist does obviously raise some risks I've put Mark Zuckerberg's famous quote here of move fast and break things unless you're breaking stuff you're not moving fast enough he probably didn't mean that he wanted regulation so also move fast and break things so obviously we don't really want to be breaking things but equally there are real risks on the other side if we're too cautious the best could be the enemy of the good and this is just to bring this out I mean many commentators have highlighted concerns have highlighted regulatory risks that regulation risks being insufficiently based on firm economic evidence to one-size-fits-all and that it risks creating unintended harm and clearly that is true those risks are genuine but there are other risks too the more that we try and do things right and get the best answer we have to collect evidence evidence collection takes time if we try and create bespoke rules if we try and allow for proportionality defences all of that creates the kind of legal quagmires and the time that we end up seeing in competition cases also of course in competition cases we usually are keen well not always keen but we are willing to accept efficiency defences so we think the competition law should allow for efficiency defences one of the interesting things about this situation with the blockaded competition set of issues is that efficiencies are actually part of the problem and therefore you don't really want to have efficiency defences if efficiencies are creating the competition harm so on balance there is this kind of trade-off between speed and administability and getting the best answer and what I'm going to say about some of the different jurisdictions is that I think a lot of what you see in the different approaches is different places trying to get to reach the right balance between speed and administability on the one hand and at least a better answer on the other hand recognising that the best is impossible so some of the key questions in regulatory design the first is around the objectives and I'm not going to talk about the objectives in any detail there's a fair amount of discussion in the paper really I think that we would all agree that the most important aspect of that this regulation should be about is innovation it's about trying to make sure that we have an environment in which the kind of fast and effective innovation that we've seen in the past is preserved the disruptive competitors with great new ideas can actually make some headway and win customers the trouble is is that a good objective for regulation and the problem is it's really very hard to ever evidence we know that from trying to run dynamic effects cases in mergers it's really tricky so actually what the various regulations have sought to do is try and get proxies for what they think will provide a framework in which regulation will occur and so you know in Brussels there's contestability and fairness and in particular the contestability aspect of that is really about trying to create a framework within which firms with great ideas can be successful then the question of scope which is which firms to regulate and I will come back to that and then the question of rules how bespoke to be with your rules how much flexibility to retain and what conduct to address and how and I'm going to talk about this incredibly quickly because I want to get onto the the risks and the mitigations of fragmented approach but just to highlight in terms of the scope the challenge in terms in terms of setting the scope is I mean it's really good to ensure quality so that clarity so that firms that are regulated know that they're regulated and they know which bits of their activity are regulated and that is going to be achieved through pre-designation within all of the proposals and that's very different to antitrust where in antitrust you know dominant firms are covered but you never really know if you're a dominant firm unless you end up in a case and you have dominant proved against you so here you're going to know whether you're subject to the rules then it's good to ensure a narrow scope because we know this is pretty interventionist law and we know that it was motivated by really quite a small number of firms and so actually you know we want to keep it pretty narrow but how many how small is narrow how few and then how flexible the was did we want to be on that do we more flexibility clearly facilitates future proofing it can enhance proportionality but it can reduce administrability again and recognizing that administrability is important and just to highlight that different places I'm going to focus on the US, UK and EU have taken different approaches so the US approach at least in the current proposals is purely quantitative and it would currently capture the top five US firms only there is no flexibility to expand that or to change it over time I think the number there may be some flexibility change precise numbers over time but not anything else the UK approach by contrast is a mostly qualitative approach there is intended now to be a minimum revenue threshold but there's a focus on finding substantial and entrenched market power and a strategic position and so actually the requirement to find market power is quite a big deal it's not unlike having to find dominance and so that raises the threshold makes it a lot harder for the UK but I think that arguably it's more proportionate it's also fully flexible so if new firms come up or if new issues are arised you can be flexible with that framework it could capture more than the top five it actually could capture less than the top five so we will see and then the EU approach is I actually have a lot of admiration for the EU approach I like it I think that so it's like it gets the administability by having a quantitative rebuttable presumption so it automatically brings in a whole set of firms and activities automatically but then it's semi-flexible because that presumption can be rebutted on qualitative grounds and you can also designate firms that fall outside the quantitative thresholds via qualitative analysis so I like that approach what I'm less what I like less about that approach is that currently it's thought lightly to capture around 15 firms and that seems far too many to me it seems much more many more than the firms that have provided the impetus for this regulation I think that some of those have come in kind of by accident and the designation process is ongoing really with the commission now and I suspect that what they're doing is trying to actually accept rebuttals find ways to accept rebuttals for some of those firms so that they get in mostly the firms that they first thought of but we will have to see where that ends up and then in terms of conduct rules what are we trying to address here I think that for all of the regulations there's basically two areas of two core areas of focus the first is interventions that are designed to open up currently monopolized or duopoly duopolized markets to new competition and that's interventions along the lines of interoperability data access and portability measures to enhance multi-homing and that is very aligned with the contestability objective of the EU regular of the DMA and then there's a second area of focus which is interventions that are designed to stop existing strong market positions from being either leveraged into new markets or just exploited and that's where you get into things like prohibitions on self-preferencing rules relating to tying and discrimination more generally you can't draw a really clear line between these two because many of the provisions will do a bit of both but you can they kind of do more of one or the other and that second is closer to the thinking around fairness the fairness objective and the digital markets act another key challenge with the rules is you want to make them standard or do you want to make them bespoke to so the one size fits all across firms or are they bespoke are they fixed or are they flexible what we know is that the sorry are they fixed over time or should they be flexible over time what we know is that the firms have a lot of similarities a lot of the economics that apply to these firms are similar but there are also some really obvious differences so even where they provide similar functions they can have different different business models so Google and Apple obviously both run app stores and both have mobile operating systems and mobile browsers but Google is primarily ad funded whereas Apple relies primarily on device sales although it does get quite a chunk of ad funding does this matter should we should that affect the rules that apply but more generally things like you know Amazon is is doing something very very different again Microsoft is doing something pretty different again and there is a risk that rules that are suitable for one firm or one activity within a firm may be unsuitable for another so that raises this challenge of how bespoke to make the rules and the and then there's also this challenge that are thinking about these issues and our understanding of these markets and these markets themselves all are changing and they can change over really quite a short space of time we need to make sure the regulations are going to be appropriate in three, five, 10 years time and not just today even if today and so how flexible or advisable should they be and again the different jurisdictions have taken different approaches here so the US has basically got a one size fits all very high level rule banning self-preferencing it's actually so high level I'm pretty uncomfortable with myself I don't quite know what it means and then but it's very standard very inflexible wouldn't change the UK as I've described generally is it's true in designation it's also true in the rules it's very bespoke very flexible the legislation would only set very high level objectives it would then be for the regulator to develop the core principles and the legally binding requirements based on those it would be an ever a bespoke evidence-based approach it also would have powers to impose specific company-specific pro-competitive interventions where practicable and proportionate so again much better in some ways but arguably much much less easily admissible and then the EU has got this kind of half a house again where it has 22 what they think are pretty specific rules but at least some of them can be further specified in discussion with the and they're set out in legislation but more than half of them can be further specified by the commission in conversation with them with a view to ensuring that the measures are effective and proportionate the rules can be revised over time and there is also specific anti-avoidance provisions and requirements that they be effective both of which are useful in preventing gaming but also just making sure that if the wording in the law isn't quite right they still have their broad effect summary views that people have about these two areas many economists like the UK DMU approach because it's seen as bespoke, flexible, allows for participative debate allows for proportionality and they dislike the EU approach because it's seen as one size fits all there's insufficient potential for engagement although I would say that has actually improved in the final digital markets at least to some extent and there's little role for proportionality although there is some I think in practice these are going to be more similar than they seem on the surface they're basically addressing similar issues so I think the rule book may end up fairly aligned and the big question is going to be administability that said even there I think they may end up more similar the EU claims that its laws are going to be self executing I think for the most part they're going to need quite a lot of engagement with the firms and I think it's going to be an ongoing process of working out exactly what they mean so I don't think it's self executing as is suggested likewise the UK is going to have a whole system of deadlines and reverse burdens of proof which will hope for well they are intended to have the effect of enabling a more bespoke approach while retaining effective administability which in turn is going to if you're going to do things on a quick basis it's going to limit how much evidence you can really put in so they may not end up being so very different so I want to get on because I've spent too long on that so so I've said already that there are even though these jurisdictions are trying to do the same thing they're already differences that we've seen there between the EU the UK the US and the other jurisdictions I haven't talked about have also taken are also to the extent that we know taking slightly different approaches and there's a number of factors pushing towards this kind of divergence the one is actually there's no single right answer I hope that what I've conveyed so far is that there's a problem and there are different solutions to the same problem but it's not obvious that any one of those solutions is the right solution then jurisdictions also have different legal, political and cultural perspectives so the US for example is very anti-regulation even the bill proposals that it has are not formally termed regulation in the way that the EU is perfectly happy to call its regulation, regulation there are political factors at play so for example I'm fairly confident that one of the reasons that the EU has ended up with so many firms in scope is because they didn't want to end up with only the top five firms in scope because they didn't want to only have five US firms they were kind of desperate to have at least some EU firms so that they didn't look like they're being protectionist so I think they're basically going to end poor booking is going to end regulated basically because it's EU and they want to include a EU firm likewise the UK is going to have a different approach you know I can't help thinking that the government quite likes the regulation now not because of its substantive virtues but just because it's different to the EU approach and the UK government is desperately looking for different legislation post Brexit and there are also linkages with other policy areas data protection IP which themselves vary across jurisdictions and there are also differences in timing of course so as first mover I think I mean Germany is kind of the first mover but for proper regulation the EU is the first mover and I think it's therefore going to be very influential even if it shouldn't be if you see what I mean just because it's first I think it's going to require things and that that is going to then change the incremental costs of for firms of doing things in other jurisdictions and that is then going to change the regulations that both the firms are arguing for but also that the other jurisdictions feel justified in imposing in their own in their own jurisdictions there's also issues around firm lobbying and that may get more that may get more effective over time I'm told that one of the reasons that the EU was very keen to get the DMDMA done and dusted is that they felt that Big Tech were getting better with their lobbying efforts and if they didn't get it done quickly they might then get to a position where they couldn't get it done at all so is this good or bad this risk of divergence there can be some benefits of a fragmented regulatory approach so if you have a comparison of approaches across different jurisdictions you know economists are happy because we can do all sorts of difference in difference analysis and that can yield lessons that improve regimes over time and it can be possible for regimes that are designed somewhat differently to function well together so merger control is a nice example I mean people who are merger specialists will complain that they're not sufficiently well aligned but broadly they work the different merger regimes work well together moreover finding a coherent global approach would just take a huge amount of time and we know that time is in short supply here so if we're going to do anything it's global is just not realistic but there are still risks the risks partly depend on the decisions that are going to be made by the platforms so the nature and extent of those risks will depend on whether firms decide whether to comply only within a relevant jurisdiction as is legally required or comply on a wider geographic basis kind of voluntarily and that in turn is going to depend on the nature of the regulations some regulations such as the interconnection messaging interconnection requirements in the DMA it's quite hard to even understand how they could only work in the EU given that EU users want to interconnect messaging wise outside the EU so the nature of the regulation might inherently have extra territorial effects there may be technical costs with having a fragmented approach with interoperability you might want to just have one system of interoperability and have the same thing globally which will be what the EU requires if the EU has made requirements there may be reputational risks of taking different approaches in different jurisdictions because users may think that you'll be that you're not using them well if you don't give them as good a deal in their particular jurisdiction and there may be regulatory risks as well if you are clearly doing some things in some jurisdictions but not in others that may raise the likelihood that those latter jurisdictions decide to introduce their own regulation in practice I think that the firms are take a mix and match approach towards these different rules for some things they will do it very locally so app store, pricing or whatever that very locally jurisdiction by jurisdiction not much cost in doing that things like interoperability much more likely to do that on a global or at least a wide geographic basis if we do end up with fragmentation that can lead to big risks it can be one of the biggest things is it can be very unhelpful for rivals who are seeking to compete because they've got to compete with something with firms that are doing different things in different jurisdictions they'll have different access possibilities different payments etc and recall that the regulation is intended to be about promoting competition and innovation so you want to you don't want to make it as unhelpful for those rivals it also could increase the costs for the regulatory the regulated firms both regulatory costs of compliance but also just the technical costs of designing their systems that could in turn impact on their prices quality or innovation it could lead to them to just increase their focus on compliance rather than actually providing a good service this is something that people sometimes say has happened in financial services where firms are very focused on the compliance with the regulatory rules and not necessarily about making their users happy and then there are risks of regulatory inconsistency between jurisdictions or even creating a regulatory edifice where every single regulatory jurisdiction wants to add something of their own and kind of you just create a mountain of regulation so final slide I realize I'm out of time I think there are some mitigations in theory you know it would be wonderful to have global platform regulation I said I don't given these are global firms I've said I don't think these are that's unrealistic I think it's really unrealistic and it may not even be a good idea even if it was practicable because who would run it how would it be governed how big would the risk be of regulatory capture etc but it may be feasible in specific areas so for example we might be able to we might end up and there already is some of this having global standard setting for APIs you might want global international global transparency standards about what is put into the public domain and how and global standards on for example how to audit algorithms a bit like kind of standard international audit standards some of these extra territorial effects may be helpful in creating more harmonized effects around the world you may be able to mitigate these issues if different jurisdictions just copy out legislation from other jurisdictions and that has been broadly the case in competition policy where different jurisdictions have basically just written out the words of either the EU or the US law and so there's a lot of harmonization there is the potential to use competition law in jurisdictions that don't have regulation to attain some benefits of regulation that would so obviously you can more likely to get a firm to agree to a commitment if the incremental cost to them of doing those commitments because they're already doing them somewhere else where they're regulated is lower and that might lead to a degree of consistency internationally via different legal frameworks you might also and we're already seeing a lot of engagement between jurisdictions on the substance what are the issues what are the pros and cons of different regulatory options if we're going to go be doing this and that's there are kind of lots of debates for example within the OECD and then the real goal would be coordinated enforcement that would require at least memoranda of understanding between jurisdictions ideally information gateways so they can share confidential information and even better the ability to bring joint cases as different states can in the US or to delegate cases between them as different member states can in the EU under the European competition network and sorry I went three minutes over time I will stop there but I hope there was something interesting there No it's quite interesting thank you Amelia and we have Paul Bellaflam as our discussant looking forward to that thank you thanks a lot for the presentation and the paper it is very interesting indeed and especially what I found interesting were the two last sections on design challenges and on the risk of the divergences between jurisdictions and you raise many many questions so it's very hard for me to come up with a another set of questions so let me try to rephrase things in the way I've been teaching them for for quite a number of years now when you deal with markets with network effects I'm usually I teach the kind of contrast there is between competition for the market and competition in the market telling my audience that if you have competition for the market you're going to have a lot of competition beforehand in order to become the winner but then competition will phase out or disappear completely whereas if you have compatibility to start with competition will be in the market and you are kind of you have a guarantee of competition remaining maybe at a lower level than if you have competition for the market but at least competition is going to remain now as you said we are probably in a situation where competition for the markets existed and five big winners emerged and they've kind of killed competition nowadays but also as you as you say perhaps more clearly in the presentation than in the paper they've created a lot of value for their users and probably this is the result of very tough competition between them and other companies before that now it's I think it's also fair to say that the regulation which is proposed would be a way to restore some kind of competition in the market or at least to come up with some some mixture between the two so keeping some form of competition for the market for some aspects while having competition in the market for other aspects but then it seems that we are facing this kind of conundrum that on the one hand the current situation is such that there is an inability for other firms to compete but I'm wondering whether there is an interest for other firms to compete if they foresee that later markets will remain contestable so this is a kind of basic question and I have and another question relates to innovation and here I think there is a sense of making a distinction between those big tech firms that control hardware and I think of Microsoft and Apple basically and other firms that operate what I would call pure two-sided platforms so if I think of pure two-sided platforms whose main value is to let people interact I think the biggest innovation is just well not just but achieving or making it possible for people to for users to interact and out of that many ways to do that right so any other innovation would be incremental but organizing interaction between buyers and sellers like Amazon is doing I don't think there are so many ways to do it right same thing for if you think of Uber or Airbnb there are not so many ways to organize the interaction between the participants of course they have come up with innovation but there are incremental innovations whereas innovations in terms of hardware this is where you would find more basic and and yeah I don't find the term not incremental innovation so if this is and I agree with you that this is hard to make innovation and objective it itself but this is still at the core of the reaction we have nowadays seeing that these companies may kill innovation and competition at the same time so basically these are my two questions so what would be the remedy or the kind of result of regulation is it a mixture between competition for and in the market and what type of innovation do we need to have in mind if we have this dynamic point of view in mind thank you