1
Select Committee on Communications
Corrected oral evidence:
The internet: to regulate or not to regulate?
Tuesday 1 May 2018.
3.25 pm
Members present: Lord Gilbert of Panteg (Chairman); Baroness Bertin; Baroness Bonham-Carter of Yarnbury; The Lord Bishop of Chelmsford; Viscount Colville of Culross; Lord Goodlad; Lord Gordon of Strathblane; Baroness Kidron; Baroness McIntosh of Hudnall.
Evidence Session No. 2 Heard in Public Questions 12 - 20
Witnesses
I: Dr Damian Tambini, Associate Professor, Department of Media and Communications, London School of Economics; Mark Bunting, Partner, Communications Chambers.
USE OF THE TRANSCRIPT
This is a corrected transcript of evidence taken in public and webcast on www.parliamentlive.tv.
Dr Damian Tambini and Mark Bunting.
Q12 The Chairman: I welcome the witnesses who are giving evidence to our inquiry on internet regulation.
Our Committee has published a number of recent reports containing recommendations on regulation of the internet. The subject generates much public debate and policy-making. We are now looking at whether there needs to be a comprehensive and strategic regulatory framework, whereby we balance the need for regulation with freedom of expression. We are going to review how the internet has come to be regulated, both in the UK and internationally; assess calls for further regulation; and make recommendations about how the internet should be regulated in the future.
I advise the witnesses that the session is being recorded and transmitted and a transcript will be taken. Our witnesses are policy experts, Dr Damian Tambini and Mark Bunting. In introducing yourselves, can you tell us a bit about who you are and your background? So that we can understand where you are coming from on these issues, could you tell us whether in your view there is a need for a new regulatory framework for the internet and, if so, what form you favour? Is it one of self-regulation, more directed co-regulation or direct regulation?
Dr Damian Tambini: Good afternoon, and thank you very much for inviting me. I am research director of the department of media and communications at the London School of Economics.
I agree that there is a need for a comprehensive new set of principles and institutions to deal with the kind of issues that the Committee has identified. If we step back a bit, to understand what has been happening in this space, we see that the Government’s internet safety strategy is only part of the picture. Across the piece, we have been delegating censorship functions or regulatory functions to the platforms to deal with a variety of social problems, from child safety to fake news and intellectual property infringement. Partly as a result of that, those platforms have become immensely powerful in their decision-making and curation of themselves, and are having an effect on people’s enjoyment of their fundamental rights.
The Committee is absolutely right to identify the problem of balancing freedom of expression and regulation in that complex co‑regulatory framework, but we need to examine an institutional solution that would work better than the current fits and starts, where there are all sorts of problems of material not being taken down or being taken down too quickly, overblocking and a situation in which rights are not really respected.
Mark Bunting: I am a partner at Communications Chambers, an advisory firm specialising in media and telecoms policy. I worked at Ofcom between 2004 and 2008, so I have some hands-on exposure to the joys of content regulation and broadcasting policy. I then spent eight years at the BBC. Last year, as a visiting fellow at the Oxford Internet Institute, I worked on a project on the applicability and options for content regulation online, which is obviously one of the areas you are most focused on. By way of a brief disclosure, I should say that, as an adviser, my clients include technology companies, DCMS and broadcasters. I am currently working on a project funded by Sky looking at options for online content regulation.
To give you a fairly brief answer to the question, I hope, it is important to be specific about where we think the problems and gaps may be in internet regulation. The current issues generating so much attention fall broadly into three buckets: data and privacy, particularly with the Cambridge Analytica and Facebook controversies of recent weeks; online content and platforms’ role in regulating content; and competition, which I know you want to talk more about later.
To my mind, the most obvious gap is in content regulation, which is why I have been working in that area for the past 18 months. To be more specific about the gap, it is not so much in the rules about what kinds of content are legal or not; it is more in the creation of regulatory capacity to engage with platforms’ role in managing access to that content. As Damian said, we need to find ways of institutionalising a different kind of relationship with the intermediaries that govern access to content to ensure that the principles of good governance are met.
Q13 Baroness Kidron: That is very interesting. Can I ask you the traditional question? Are they platforms or are they publishers? In answering that, can you also let us know whether you think we are due a new definition, and whether part of the problem with asking the question is that we have not asked it in a comprehensive way? Perhaps you would deal with that first and then I will come to another point.
Dr Damian Tambini: Legally, they are platforms; they have a shield from liability until they are notified that they are hosting something potentially illegal. I think there is emerging consensus that we have reached a situation where the law needs to catch up in some way, and there needs to be an intermediate category between publishers and mere conduits. That is easier to say than it is to do. Legally, even in the UK, once you click on that narrow question and open it up, how the law treats you as a publisher depends on the different legal area you are speaking about, so intellectual property would be very different from defamation and so forth.
The idea that there should be a complete shield from liability goes back to the 1990s when there was a new thing called the internet and we wanted to foster innovation and economic growth by giving it a shield from liability. We have moved on to a situation where there is a small number of very powerful monopoly players, and oligopolies in some markets, and what has been called an indirect subsidy of the liability shield needs to be opened up and reviewed.
Mark Bunting: As a matter of law, things have to be capable of clear definition and we have to know which categories organisations sit in. As a matter of policy, I do not think the definitional game is a very helpful one to play, with no offence intended to the question.
Baroness Kidron: That is fine. That is what we are here to discover.
Mark Bunting: The starting point for policy is: what is the activity that is systematically likely to lead to consumer or citizen harm, or to missed opportunities for citizen and consumer benefits, which is the other side of the coin? The activity that is not well addressed by existing law or regulation is the role that intermediaries now play not just as a conduit for content but in actively curating, as the buzzword goes, that content, by which I mean that they select which content is presented to users; they rank that content; they recommend content; and they moderate content. You cannot do that in a purely neutral way; you have to do it by setting the values that you want to optimise in the searching of content.
In markets where exposure to content is so important to fundamental rights, as Damian says, and to all sorts of public goods, including the effective functioning of democracy, the values that intermediaries bring to bear on the task of curating content seem to me a legitimate issue of public policy. Interventions that enable us to get better at seeing how they do that, and what the effects are of that curatorial function, seem to be the things we ought to be trying to identify.
Baroness Kidron: A question that really interests the Committee is the design of services. You said that the problems lie in the three buckets of data privacy, content and competition, but there are worries about the actual design of services—for example, if you read last week’s testimony, design to addict. What role does that have in understanding what one is dealing with in terms of definition or non-definition—however you wish to answer it—and do we need to look at regulation, governance, or whatever version of that we come to in the end, in the design of services?
Mark Bunting: The interesting question is about the process that intermediaries have gone through in designing new services or features. I would be very cautious in specifying rules about what design features platforms must have, or the particular tasks that they have to incorporate in design processes, but there is the concept of responsible design, which I think is partly what you are alluding to. For me, the interesting part of that is whether intermediaries, in designing their platforms, have taken reasonable steps to think about what the unintended consequences of those design choices might be. Have they been open and accountable in deciding what to do about those unintended consequences?
To give a practical example, the chief executive of Instagram has been very public about his attempt to shape the Instagram environment in a way that makes it harder for trolls to abuse people on that platform. That was partly for commercial reasons, because a clean environment is one that people want to spend time in, but there was also an ethical dimension in the way he talked about it. There were choices that he could have made or not made. We want to try to find ways of encouraging platforms to have design processes that enable them to think about those sorts of consequences, and make choices about them ahead of services being launched, rather than waiting for problems to emerge later.
Viscount Colville of Culross: How would you suggest doing that?
Mark Bunting: It is probably hard to avoid coming back to codes of practice and statements of principle. For example, there could be a code of practice that specifies the sort of external third-party engagement platforms might be expected to have in their development. There is expertise in the effects of technology on products. Have platforms taken reasonable steps to understand that research and evidence? Can they demonstrate how they have taken it into account in their design choices? You can set general expectations of engagement and anticipation of potential problems without necessarily specifying in great detail actions platforms should undertake.
Dr Damian Tambini: There is a slightly more direct and less hedged way of answering the question. How would you get them to do what you want them to do, or what the public want them to do? In general, we need to tread very carefully, because there are issues of media independence and the autonomy of those organisations from the state wrapped up in any of the moves that are made. There is a wider bundle of issues in platforms developing their ethics. This is an example of them developing their ethics and sense of social responsibility in a dialogue with society through institutions such as Parliament.
It is not the case that there are no levers in the hands of Parliaments. We could consider this a historical moment. There are historical moments when companies become too big and powerful, and they are regulated as monopolies. If that does not work, they are broken up. In a sense, Mark’s buckets are, unfortunately, overflowing into one another, because that is when the question of competition runs into all these other questions. If in five years’ time this and other Committees are still aware that there is a significant consumer and citizen detriment in relation to the platforms, other things—taxation, competition law and changes to the regulatory framework—might need to be brought into play, so there is a big stick in the background.
Baroness Kidron: Dr Tambini, as you are being very direct, there are not only unintended consequences of design but intended consequences of design, which I think was what the previous witness was getting at. We have design based on addictive loops. In your direct answer, would you say that is something one should look at in a societal way, whatever the framework?
Dr Damian Tambini: As a personal view, I think that is an issue, and it is coming out of the research, whether it is research on children or on social media use more generally. As a policy expert, I am quite intrigued by the question of how we have that conversation. Institutionally, something is not quite working when a moral panic blows up, and Select Committees are asking for the removal of certain kinds of content. We have small moral panics and we do not have an effective regulatory process whereby there is clear articulation by public authorities of what the ask is, with the clear involvement of civil society and the public.
At this point, I have to raise a question for the Committee about something I have attempted to look at as a private citizen. What is happening with the digital charter? Is that such a process? There does not seem to be a huge amount of transparency to the public. I tried to find out what was happening. How legitimate can that process be if even experts do not know what is going on with the digital charter? Historically, those kinds of processes, such as the royal commissions on the press going back to 1947, have tended to be cross-party and appointed by Parliament, but they involve other parts of civil society and are utterly transparent.
Baroness Bertin: I should declare an interest: I work part-time for BT. You talked about breaking up monopolies. I would love to know how you think the break-up of Facebook, for example, could realistically work.
Dr Damian Tambini: Obviously, this is different from how BT is regulated, or from the historical break-ups of telecoms monopolies that were in one country. We are in uncharted territory, but services provided through specific markets such as advertising, in which these companies are operating, could be separated through obligations that they are, for accounting purposes, separate from other divisions within the company. This has only relatively recently been on the agenda, so I posit it not as a fully developed policy design but as an idea.
It would be difficult, with the possibility of companies such as Facebook, Google or others withdrawing their services from a particular market. They have done that in certain instances where regulatory burdens were inappropriate, whether that is China or Spain. Certain services were withdrawn. I absolutely take that point. Effectively, a regulator working within one market cannot break up Google or Facebook globally, but it can enforce certain kinds of accounting separation within their operations in this market.
Baroness Bertin: I am not dismissing the idea; it sounds really interesting.
Baroness McIntosh of Hudnall: I want to go back to the issue Baroness Kidron raised about design to try to tie it to the point Mr Bunting made about a values base. Are there any models outside media, not necessarily specific regulatory models, to which we could look for ways of thinking about harm and that we could apply to the question of design? For example, nobody thinks it is a bad thing that there is no lead in the paint on children’s toys. Once upon a time there was, and it had to be regulated away; people were not allowed to sell toys if they had lead paint on them. There may be something in that area. It is not to do with the content. A toy could be in many different shapes and sizes and intended to do many different things, but if it had lead paint it could not be sold. Is there something about design in other kinds of regulation that can be viewed in that way and drawn into this discussion?
Mark Bunting: I might give you a yes or no answer. Last week, you had before you Lorna Woods, who is working with Will Perrin. They are exploring whether there are analogies with workplace safety legislation, which is quite an attractive analogy. Workplace safety legislation is not so much about specifying that there have to be so many fire exits and this kind of smoke alarm; it is more a general principle that workplaces should be safe places in which to operate, and then it is up to companies to work out how to do that. That is potentially quite attractive.
The limiting thing, which is where the “no” part comes in, is that I am not sure we yet have a very good grip on the range of potential problems we might be dealing with. In workplaces, by the time the legislation was passed, we knew what sort of industrial accidents happened, and companies could be reasonably well placed to develop policies to address them. In this climate, we do not necessarily have a very good picture of exactly what the problems are, or how to identify them when they arise. That is not to say that the analogy does not work, but that we almost need to take a step back and find ways of engaging with companies about the actual risks associated with their products, and somehow incentivise them to have an open conversation about that, including, as Damian says, civil society and other stakeholders. That is a very difficult task, but it seems to me that in a way it is the heart of the regulatory challenge.
Baroness McIntosh of Hudnall: It may be difficult, but it is not unprecedented. That is all I am trying to inquire about. Is that right?
Mark Bunting: In broad terms, I agree.
Dr Damian Tambini: To extend the analogy, if the toy is an online game, one of the problems is that “Grand Theft Auto” may be bad for children, but a number of adults will think it is their right to use it, so you cannot simply remove it from the market.
Closer to home, there are some interesting analogies in how these ethics emerge. Even in newspapers, the separation of editorial from advertising and the separation of comment from factual reporting emerged, and they help consumers to know where they are and they protect democracy. The question is how to have conversations about how that ethic works and how it is communicated to a very confused public.
The Lord Bishop of Chelmsford: I want to take you back briefly to your very first answer to Baroness Kidron’s question about whether it is a publisher or platform. I am still not convinced by your answer. I understand that it is not a publisher as we understand that word, but neither is it a platform in the way I think most of us understand that word. It likes to present itself as a library. When I go to the library, it is a democratic space where every book is equal, and to navigate my way round the library I have to work out the index system.
It is not like that; it is like going to WH Smith, which I thought was a stationer or bookshop, but now, when I go to the till, I am bombarded with all the other stuff they are trying to sell me. Do not tell me it is just a platform. I wonder whether we need some new language, rather than just sitting behind the old language. Maybe one of the things we could usefully do is to ask whether there is a new way of defining what that space is. That might be the key to unlocking how we might do the things I think we all want to do.
The Chairman: Do you think that is something the Committee could usefully do?
Mark Bunting: Yes, but I urge you to focus on the activities that are causing harm rather than getting too caught up in the definition. I agree that, for example, Facebook is neither a straight publisher nor a straight platform, but there are many different companies in this space with very different business models that operate in different ways.
One of the regulatory tasks is to find a way of capturing potentially everybody from Google Search at one end, through Facebook—they are two very different services but they tend to get bundled together—all the way down to much smaller sites that may have a particular task; for example, sites that enable teens to upload videos about their homework may have particular obligations. They may not look the same on the outside, but they manage, or curate, our access to content, which is of vital importance to fundamental rights. To my mind, that is the activity we need to find ways of engaging with.
Dr Damian Tambini: I agree. You put your finger very neatly on the challenge, which is to come up with definitions that people generally understand and find intuitive, but which also represent in some way how the regulation and the institutions are working.
To come back to my original point, legally, platforms are not liable for content that they do not know about. That is the key thing. The way that people use platforms such as Facebook and others has more to do with the way they understand television and media historically. There has always been a slight catch-up in media literacy and people’s understanding of the risks associated with the environments in which they find themselves.
Q14 Viscount Colville of Culross: We have talked a little bit about regulating content. I would like to ask about the way platforms moderate content and whether it is fair and transparent. Mr Bunting, you have written about accountable design for algorithms. Is that something we should be looking at? We keep hearing about algorithms and the way they drive users in certain ways. Is that an area we should look at making more transparent or, if that is not possible, persuading platforms to make it more transparent? Should we be getting involved in that area?
Mark Bunting: I think you should. To my mind, the concept of accountability in algorithms is more important than transparency. Another way of putting it is that it depends on what you want them to be transparent about. I do not think that transparency of an algorithm in itself means very much, because they could publish the billions of lines of code that make up the algorithm and none of us would be any the wiser.
The aspiration would be that we know a bit more about what the algorithms are trying to solve and what are the data on which they have trained those algorithms. For example, if there is an algorithm to detect extremist content on YouTube, my question would be, “YouTube, what have you done to assess whether that algorithm is working effectively both in capturing content that genuinely is extremist when qualified people look at it, and in not capturing all sorts of material that is legal content and has just inadvertently fallen foul of an algorithm?” It is not the algorithms themselves that policymakers should be exercised about; it is the steps platforms have taken to ensure that the algorithms are working as intended, and how they are measuring success and reporting it against those objectives.
That is a legitimate question, partly because the platforms have said that they now do a lot using automated tools to detect all kinds of illegal material, but they do not say very much about how they have evaluated the effectiveness of those automated tools. Finding ways of incentivising platforms to adopt principles of good governance, accountability, openness and impact assessment is the most important task.
The Chairman: Do you agree, Dr Tambini?
Dr Damian Tambini: It is important to open it up a bit. There is the general process of curation, which includes the positive promotion of certain kinds of content, and may be done by the special source relevance algorithm on any platform. There is also a narrower category, which is the removal of content that breaches community guidelines or the law.
In relation to the second category, people very often categorise the platforms as censors, and campaigners and policymakers are keen to point that out. They say that taking down material impacts fundamental rights, so it should be subject to due process and appeal, and opportunities to put back content. Then you get into questions about scale and practicalities. The platforms say it is very difficult to scale, because they do a lot of it automatically and with low-paid moderators. There is thus a practical problem that the big platforms in particular have something like a censorship role, and policymakers and others do not quite know whether they are editors or censors. That is a really important principle to take into account.
When we come back to the question of competition being linked to content regulation, the principle is important. A platform that does not have very many users and has a small market is much more like a publisher, a journalist or an editor. On the other hand, a platform with a large market share is operating something much more akin to a censorship role, whether that is taking down content or the right to be forgotten or impacting on fundamental rights in another way. What we do not have in the regulatory framework is the possibility of linking regulatory obligations to size. That goes across many different areas. One of the institutional issues we pointed out at the beginning is how to have some kind of regulatory institution that is able to link those competition and other public policy issues.
Baroness Kidron: I want to go back to the issue of responsibility. Mr Bunting, you have referred a couple of times to the intended consequences, or the idea that there is only deliberate good happening, and you, Dr Tambini, have just referred to low-paid workers. That is a choice, too. Everybody wants this technology to be wonderful, accessible and available. That we are all agreed on, but we are trying to imagine another world where the status quo is not automatically assumed. They could have high-paid workers who might have more skills. I am interested in responsibility. Regulation is not the only tool in the shed.
Mark Bunting: On issues such as labour rights and employment conditions, those firms should be held to the same standards as other firms. To the extent that there is regulation of international labour rights, of course they should apply equally, and there are processes in place to enable investigation of those things.
I hesitate slightly, because I feel that sometimes we are at risk of holding these companies to a higher standard than we would have held companies in the past. The companies have a very strong incentive to maximise their value to users, and I take the point that that can tip over into addictive behaviour. Where there is robust evidence that products are addictive, there is a role for regulation, in the same way as there is for gambling and alcohol. I do not know enough about the area to know how conclusive the evidence is at this stage, but we should be careful not to detract from the value of platforms for the majority of their users because of concerns about a minority of users who may be using them inappropriately. There is a balance of responsibility between the users themselves and the platforms.
Dr Damian Tambini: I agree with my colleague about labour rights; other aspects of what the platforms do are effectively regulated in other ways. I am arguing that perhaps we need some sort of new institution, even a regulator, but we should be wary, because one of the arguments against that is that it will be a Christmas tree. Everyone will hang on their pet issue, so anything that is done needs to be very closely circumscribed and not to overlap with other issues.
Q15 Lord Gordon of Strathblane: What role can users play in establishing and maintaining online community standards other than simply boycotting the service and hitting the share price of its provider?
Mark Bunting: There are a few ways. The most powerful way is as a user of the service. One of the things that platforms do very well is fine-tuning the way they operate to try to ensure that they deliver a service of value to users. There are some suggestions that Facebook’s growth may be slowing; usage is declining in some markets. It is impossible to know the extent to which that is to do with some of the public issues that have arisen over the past few months, but if users feel that the platform is not operating effectively for them they will go elsewhere. That was how Facebook came to replace Myspace in the first place, so that is important
Beyond that, there have been efforts, particularly by Facebook, to consult users on changes of policy and terms of use, although they have been somewhat variable over time. This morning, I happened to look at the Facebook governance page, as I think it is called, which has had a recent update of its data policies. The previous update was in 2015, so perhaps the engagement is not quite as frequent as we might like.
The third area, which in a way comes back to the point about design, is that users play a very important role in establishing the norms of platforms.
Lord Gordon of Strathblane: But do they? Surely, it is just signing up to the terms and conditions, and those are devised by the platform provider.
Mark Bunting: They are. Clearly, they are very important rules and documents, and the power of algorithms in shaping the ways we interact is greater than the ability of users to force change. Part of what we are seeing is a gradual evolution in people’s understanding of the right and wrong ways of using services. It is a slow process, and it is not as rapid as the platforms changing themselves, but users have a role to play in helping to set norms, and platforms have a role to play in trying to enable the formation of norms that are more responsible rather than more damaging.
Lord Gordon of Strathblane: I do not think anyone would accuse platforms of encouraging violence, but they certainly seem to enable it. People who one hopes would behave perfectly properly if they met somebody in the street behave like absolute morons online. Is there some work being done on that?
Dr Damian Tambini: It has been widely recognised that there is what Michael Ignatieff calls digital disinhibition. Other than that, there seems to be a pattern that people are much less inhibited, and the usual social norms do not restrict behaviour.
Going back to the question about what users should do, I agree; they should switch. They should be able to switch between platforms. One of the problems, which is linked to some of the competition issues, is that there is quite effective consumer lock-in for reasons of data portability, barriers to entry and the costs of switching. Those appear to be very high, which is a competition policy issue in itself.
Many users are children. There are specific issues around how the process of learning works between children, parents, schools and platforms. It is very much fits and starts. One of the reasons is that different platforms have very different approaches to children, whether that is consent age to join the platform and how they are treated for data protection purposes, or content moderation standards. They all have very different policies. That creates difficulties for parents and schools in working out appropriate rules for children and effectively communicating them, which is an area that I think the Government are looking into at the moment.
Q16 Baroness Bertin: My question is about balance in online protection, child protection and hate crime. How do you get the right balance? It is such a grey area—not child protection—but I would like to have your views as technical experts about how realistic you think it is to put the genie back in the bottle.
Dr Damian Tambini: It is very difficult to get the balance right. It is a question of procedures. How do you create a pyramid whereby complaints and disputes that are clear and easy to deal with are dealt with by the platforms at a low level, with the potential to escalate to a co‑regulatory body and ultimately, for a very small number, to the courts in setting standards, if that involves illegal content? The challenge is to come up with effective procedures.
The German hate speech law, for example, which has been much commented on, is controversial because it is seen as an infringement of free speech rights. It establishes a procedure. In Germany, as here, authorities were getting impatient with platforms being very slow to take down hate speech, harassment and violent content that was illegal and breached German law. They set in law clear guidelines for different categories of content. It needs to be taken down within 24 hours if it is clearly infringing the law, and within a longer period if it is more difficult to categorise. A small number can be referred to a publicly appointed board if they are too difficult to deal with and the platforms cannot reach a clear view on them.
There are possible procedural solutions, but it is a question of getting the balance right in how much you push back to the platforms to adjudicate those rights and where the rules come from. Should the rules be set by Parliament or by the platforms? Should consumers have the ability to switch on the basis of different policies, or is there some other model? At the moment, there is a bit of confusion.
Baroness Bertin: We know there are millions of child images at the moment. Do you think companies could do more to stop those images going on to platforms in the first place? I think they are hiding behind, “We have referred thousands of these images to the authorities”, but that clogs up the whole system, and one could argue that in some ways it is not helpful. They need to stop the images going on there in the first place. Do you think they could move more towards that?
Mark Bunting: There have been moves towards that. Collaboration in that area goes back a long way, as you may know through your BT relationship. The Internet Watch Foundation was established as an industry body in the 1990s, under very significant political pressure and the risk of regulation, to try to find ways of dealing with those sorts of problems. Initially, it very much involved that sort of notification and action, but BT was very active in developing technological solutions to enable ISPs to identify flagged content. The question of whether that is enough and whether more can be done is virtually impossible to answer without being much closer to the detail.
Baroness Bertin: You mean the technological detail.
Mark Bunting: Yes, but also evaluations of the effectiveness of what currently takes place. To touch on a point we made earlier, there is not a huge amount of external accountability in these areas, in the sense that those sorts of evaluations and assessments of what more could be done are generally not publicly available. The answer to your question is, “Possibly”, but it reveals a broader issue, which is that there is expertise working very hard on those questions, but we find it difficult to hold the activity to account and understand how effective it really is.
Q17 The Lord Bishop of Chelmsford: I want to focus on what is often referred to as the TV-like content that now appears on the internet in video on demand. A couple of years ago, the authority for television on demand shut down and its functions are now with Ofcom. I have two questions. How should we regulate video-on-demand services, and who should the regulator be? Should it be Ofcom, or do we need a new regulatory body specifically designed to regulate this sort of content?
Dr Damian Tambini: We are in a period of transition. In the past, obviously anything that was TV-like required spectrum for broadcasting and was licensed. In the future, just about everybody will be providing video and it will be much less regulated. We are somewhere in between, where there are categories of TV-like content. That is why in the audio-visual media services directive a specific package of very basic rules applies to that category.
Consumers are catching up. They expect on-demand platforms to be less regulated than broadcast platforms. There seems to be a space emerging for a co‑regulatory body with multiple roles, including auditing various forms of editorial content online. It may be an on‑demand regulator, in the sense that platforms ask to be regulated by it. There may be a voluntary aspect, but I would not base it on any type of content or medium of delivery. That would be a mistake, given the rate of change. We need a system based on opt‑in or services such as news versus other kinds of service, rather than types of service as in video versus radio or text.
The Lord Bishop of Chelmsford: You said that the viewer has different expectations. I am not suggesting that you are wrong, but I would be interested in the evidence for that. People increasingly receive all these things through their smart TVs. Do they have different expectations, or do they just think, “I’m watching TV”?
Dr Damian Tambini: You have caught me there, because it is some years since I looked at the evidence. Ofcom surveys these things annually in terms of consumers’ expectations. This is going back a long way to the time I was involved in setting out policy before the Communications Act 2003. That was done very much on the basis that consumer expectation at that point remained that video platforms, TV and TV-like content would be regulated. That was where consumers were. I am offering my judgment and estimate, and the Committee should probably look at the evidence, but my sense is that consumers, particularly younger people, have moved on considerably.
The Lord Bishop of Chelmsford: I am sure we got it from Ofcom, but it would be worth looking at it again. Thank you.
Mark Bunting: For me, the big distinction is not between broadcast TV-like services and online TV-like services but between commissioned and editorially controlled services, in which I would count Netflix and Amazon as much as I do the BBC, and open video platforms of the likes of YouTube. Historically, YouTube has not been in scope for any of this regulation because it is not a TV-like service, although there is some debate in the AVMS directive review about what obligations platforms such as YouTube should now be subject to.
In the former category, the Netflixes and Amazons, personally I do not see a strong case for changing the current regulatory regime. The rules for those services are broadly similar to those for broadcast services, and Ofcom has responsibility across the piece. If we move beyond the current regulatory requirements and think about how to regulate YouTube, it will be a very different kettle of fish, and all the issues we have been talking about today will come home to roost. There is a separate question about what role Ofcom should play, and whether a new institution is needed for that.
Lord Gordon of Strathblane: Concentrating on the services that are like television, is there a case for making the regulation identical? I thought that might be the implication of what you were saying. In which case, do we end up regulating online as we do offline at the moment, or do we deregulate offline to the equivalent of what online has become?
Mark Bunting: Personally, I do not think that the principle that the same rules should apply to everything is necessarily a useful starting point because of Damian’s point about how expectations may or may not vary. I do not know the answer to that question. To the extent that audiences have different expectations of Netflix or Amazon, there would be a reason for having different standards from those that apply to, say, the BBC, but that is an empirical question. I do not think it is one where you would start from the principle of saying that the same should apply across the piece.
Dr Damian Tambini: To underline that, Mark has advanced one idea of what should guide regulatory design in editorial control, but I want to return to the point I made previously about size. These are not rules that should apply to every publisher, including the Facebook accounts of everybody in this Room. The rules should apply potentially to large and powerful companies that have a huge impact on our national life.
Q18 The Lord Bishop of Chelmsford: What do you see as the future role for public service broadcasters? Netflix will be spending $8 billion on content this year, and the BBC will be spending £1.6 billion. The big are getting very, very big, and one wonders what the future is for PSBs. Do you have any thoughts on that? What might regulation do to help that economy of broadcasting?
Mark Bunting: I can give you a brief and high-level view. Public service broadcasting continues to be very important, not only for the public policy and social considerations that weigh on policy-makers but because it still accounts for the majority of viewing in the UK. Despite the vast sums of money that Netflix and Amazon spend, they still account for a relatively small share of video consumption.
To go back to the previous question, it is right that we expect a different level of commitment or obligation from public service broadcasters than we would from purely online services. The key question is funding and the sustainability of the obligations that they face. I am not in a position to give you a view about how sustainable those obligations are, but you can see risks coming down the track. It is not just about content competition; it is about competition for advertising and talent. In those areas, all the PSBs find life harder than they did in the past. Parliament, the Government and Ofcom all have important roles to play in monitoring the health of that ecosystem and trying to find ways of propping up what continues to matter.
Dr Damian Tambini: The Committee should not miss the opportunity to mention the importance of the security of BBC funding and of independence protection in the process of review of BBC funding, which was a serious problem in the last funding rounds. We need to avoid that happening again.
As the focus is on internet platforms, the issue of prominence and findability is hugely important. It would be interesting to think about that if one of the things the Committee is doing is trying to articulate the beginning of a societal ask of the platforms. They develop algorithms to make certain kinds of content prominent; for example, Facebook is discussing how to develop in the United States a way of recognising quality news. It is also looking at crowdsourcing through user recommendations and surveys, and a way of bumping certain quality services up the news feed.
That is hugely important after the recent emergencies we have had about fake news and disinformation, and what the BBC is doing is more important than it has ever been. That activity on the part of the platforms should, surely, incorporate what Parliament and society have agreed is to be viewed as socially important quality content, which is funded and legally required to be universally available. There is an area around prominence and findability that needs to be worked into the framework.
Q19 Lord Goodlad: You mentioned transparency a short time ago. First, what information do you think online platforms should provide to users about the use of their personal data and how it should be presented? Secondly, do you think that the general data protection regulation provides sufficient protection for people on transparency in the collection and use of personal data, or do we need further regulation?
Dr Damian Tambini: The GDPR is definitely a step in the right direction. A lot will depend on how it is implemented. There is obviously discretion, with new legislation being passed here. There will also be some discretion in implementing it by the Information Commissioner’s Office.
A key area that I would like to highlight is data portability. One of the objectives of introducing data portability was to give consumers the ability to download their data to bring down switching costs so that they can move to other social media platforms. It is hugely important that that is effectively implemented, and that we develop common standards for the formats that will, effectively, feed into competition between social networks so that it is practically possible. It will be interesting to see how that develops over the next weeks and months.
There are difficulties with transparency when we are talking about privacy. In some ways, the principles have always been there, and some of the rules will not make a huge amount of difference. It may be possible that the Committee can help by making consumers and the public more aware of the rights they have. In a sense, it is too soon to tell.
Mark Bunting: I agree. I was just looking at some research that Doteveryone published recently that shows the limits to users’ understanding of how data is used now. It found that 45% of members of the public were unaware that information they enter on websites and social media can be used to target ads; 32% do not realise that their search data is collected; and 30% do not realise that their purchase data is collected, so there is a significant comprehension gap. Mark Zuckerberg himself said at one of his congressional committee appearances that no one ever reads the terms of use and end-user licence agreements that they sign. At that level, there is common recognition of an issue.
What is unclear is how much users really care about privacy, whether they will change their behaviour as a result of more information becoming available, and, if so, how quickly those effects will work through. That remains to be seen, but it is at least possible that making more information available to users may not in itself do very much to ensure responsible use of data, which is where GDPR comes in.
Q20 Baroness McIntosh of Hudnall: This is a rather big question late in the day. You have talked quite a lot about competition law as we have gone along, and my question is about whether our current competition law is effective. Dr Tambini, I know from your evidence that you do not think it is, and that it is drawn too narrowly; you made some interesting observations about Amazon, for example.
Do you think that the law can be effective, or does there need to be new law to deal with the growth of platforms and their reach? Secondly, given that a lot of the current regulation is on a Europe-wide basis, what risks will we be exposed to in that area once we leave the European Union?
Dr Damian Tambini: The second question is easier. There are risks, if you crash out without a deal, that the rules are simply not clear, and there is a combination of directly effective EU legislation and domestically-passed law, with an unclear relationship between the two. There are also risks post Brexit of fragmentation, given that the Commission has in some cases been big enough to stand up to the legal power and expensive lawyers of the global giants in ways that may be more difficult for one country alone.
On the question of whether new legislation is necessary, there is an article by Lina Khan, a US academic, in the Yale Law Journal that I recommend to the Committee. She tells the story of how the enforcement of competition law and the tests applied by competition regulators in deciding whether there has been consumer detriment have changed in the last 20 to 30 years. She is speaking about it in the US context, but it also applies in Europe.
It would require some kind of legislative change to deal with the issue. Something could be done using the discretion of competition regulators post Brexit. Last week, Lorna Woods, one of your witnesses, referred to the Enterprise Act provisions on public interest in media mergers. In a merger situation, some kind of additional public interest could be taken into account by a Minister. That would not require new legislation; it would require some kind of clear signalling and clarification of the policy on the part of the Minister.
At the moment, in a merger situation, it is possible to have a reference to the public interest for almost anything that the Minister decides, but that requires merger rather than the organic growth of a company, which is very difficult for competition law to deal with. There are real problems, and legislation to change the Enterprise Act would be part of the solution. How the competition authorities advise Ministers on competition decisions and how they make their own decisions could also help.
Baroness McIntosh of Hudnall: You talked about the difference between merger and what you referred to as organic growth. The growth of these platforms has been to a large extent through the absorption of smaller entities, has it not? Start-ups have been sucked into the big platforms, and that is where a lot of their growth has come from. Do you think that in the application of competition law there is any particular advantage in the fact that their growth is not, as you make the distinction, organic in the usual way?
Dr Damian Tambini: Most of those mergers happen elsewhere in the world. In the case of the small number that occur in this country—I cannot think of a specific example right now—it is unlikely that they would meet the threshold required under the Enterprise Act for them to be referred to a Minister, so on reflection I do not think the legislation would be particularly useful in those cases.
One very interesting area in media pluralism, which has not come up yet but could, would be if a platform wanted to buy a broadcaster, for example. If a newspaper buys a broadcaster, there are special public interest requirements. If Sky wants to merge with another company, or have a change of control, there are broadcast licensing concerns, but those would not apply if a broadcaster was being purchased by a platform, rather than by another broadcaster or newspaper.
Mark Bunting: I have a slightly different view from Damian on the relationship between competition law and other issues. I certainly agree that competition law does not effectively address many of the concerns we have talked about. It does not have the tools to manage harmful or illegal content, addiction or any of those sorts of things. I agree with Damian that scale is very important. Having a regulatory regime that allows for differentiated responses to companies of different sizes is very important.
Where I would differ, in the interests of plurality in this session, is on the desirability of using competition law as a way of fixing issues of social welfare, in the sense of the externalities that we are dealing with. There are three brief reasons for that. The first is competence. Competition regulators find it hard to balance issues of social concern against competition. That is not surprising because it is very hard, but it is not clear that competition regulators are best placed to do that job. The second reason is to do with pace and reactivity. Competition law is essentially, not entirely but very often, an after-the-fact mechanism, and one of the things we want to try to achieve is a more forward-looking approach to some of the issues we have discussed.
The third reason is the most fundamental. It is not clear to me that the remedies of competition law really address some of the problems we are talking about, particularly in the area of harmful content, which is where I would be most concerned. They might make things worse by fragmenting the problem rather than consolidating it. Last week, I was at an event with Tony Curzon Price, who is now an economic adviser to the Business Secretary. He made the good point that content regulation has always relied on there being a good monopolist who can set standards across the whole of a sector. To the extent that competition law tends to be opposed to monopolists, those things cut against each other. My personal view is that you need different frameworks for different purposes.
Baroness McIntosh of Hudnall: You do not think that the addition of a public interest or public benefit element in the way decisions might be taken would go any way towards meeting your point.
Mark Bunting: I am open to being persuaded, but on the face of it, it seems hard. We currently have quite narrowly defined public interest grounds for intervention, and those have not historically been a recipe for rapid and clear decision-making processes. If we were to broaden them substantially and make them apply to a whole range of different conditions, we could find ourselves getting tied up in endless CMA-led processes and trying to reconcile very difficult issues about the balance between competition and the protection of children, or whatever it might be. That sounds as if it could be painful.
The Chairman: I thank our witnesses. We have a very broad inquiry, and you have brought us broad knowledge and expertise very early in our inquiry, which has helped us a great deal. Dr Tambini, you have sent us some very useful written evidence. We would welcome correspondence from you if you follow our work as the inquiry continues, particularly in the area of international developments. If you see developments globally that you think may be of interest and relevance to the Committee, we would very much like to hear from you. Thank you again for taking the trouble to come here today, and we hope to hear further from you during our inquiry.