Skip to main content


Transcript: Launch of ‘A review of the causes and impacts of unlawful file sharing’ April 11 2014

Posted on    by CREATe Team

Transcript: Launch of ‘A review of the causes and impacts of unlawful file sharing’ April 11 2014

By 23 April 2014May 14th, 2024No Comments

Date: 11 April 2014
Venue: Stationers’ Hall London
Time: 3 – 6 pm

Click for the event Resource Page (containing access to download the full report pdf, list of industry panel respondents, press coverage, social media reactions, event photos, and, a list of registered attendees.)

Watch the videos

MK: Martin Kretschmer
DZ: Daniel Zizzo
SW: Steven Watson
PF: Piers Fleming
AB: Alison Brimelow
DY: Dominic Young
RA: Robert Ashcroft
JK: Jim Killock
TB: Theo Bertram
JM: John McVay

Martin Kretschmer: Good afternoon. There are still people arriving and I thought I’d give them a chance to settle, but I think the academic 15 minutes have been reached, so let’s begin. Welcome to Stationers’ Hall. I am Martin Kretschmer, the Director of CREATe. Stationers’ Hall is an interesting venue for this event; it has a claim to be the birthplace of the modern copyright system, and I thought I’d start today with a few words about it to make your visit worthwhile, even if you don’t like what you hear later – which is very unlikely.

Anyway, in 1557 the Stationers’ Company was granted a Royal Charter from the Catholic Queen Mary, and essentially it was designed as a means to control the circulation of printed materials. It was very well established mechanism on the Continent: a guild-like structure. The only way you can publish is if you are a member of the guild. So the Charter established a monopoly to the Stationers’ Company and that offered good control of the circulation of printed materials. Through the internal processes of the Stationers’ Company, many dispute resolution and registration processes evolved. It was a turbulent time; there was the period of the Commonwealth, the Restoration, and the Licensing Act of 1662 which again was an instrument blatantly designed to regulate the press. It sets again the requirement for a book to be registered before it is to be allowed to see the light of the world. If you later enter the Court Room of this venue, on the wall on the left there a later version of the Royal Charter, issued in 1682 by Charles II to the Stationers’ Company. This is one of the most beautiful of the Royal Charters; and if you’re here it is worthwhile seeing it.

So how does all this give rise to the birth of the modern copyright system? Well, the Licensing Act lapsed in 1695. There was a period of time, almost by accident, where a free-for-all began; pamphlets turned up and could be printed and re-printed without permission, and there was heavy lobbying in the following decade, eventually leading to a statutory copyright system as we know it today. In 1710 the first copyright statue anywhere in the world was enacted, the Statute of Anne. It still contained a requirement that in order to obtain copyright you had to register your work in the registry book of Stationers’ Hall. That book is up there – you can still see it. And that requirement was in place until the 1911 Copyright Act. In terms of the Berne convention, it was a constitutive formality: you can’t have the copyright unless you register. And that obviously caused problems once the UK joined Berne (with its prohibition of formalities). The requirement was abandoned, and registrations petered out. But there are a lot of interesting data in that book which is available for researchers, so I would invite some of our economists to have a look at that at some point.

The hall you sit in today was built after the Great Fire of 1666. There was an earlier Stationers’ Hall on the place but not of this grandeur. And later on there were many additions, but this hall in its panelling and layout is pretty much the 17th century hall. So, if you deal with digital matters, a little weight of history is not necessarily a bad thing. We get very caught up in the topic of the day, and it is quite useful to step back occasionally and see the bigger picture.

For CREATe, this is the second big London event in a month. On the 12th of March 2014 we launched the Copyright User Portal, which is a digital resource aimed at media professionals, creators and users. This is an attempt to generate, from the bottom up, the questions which interest in practice. On the screens in the other room you can see some examples of the Portal we launched at the Creative Economy Showcase of the AHRC. In its first month the Portal had 2,000 unique visitors and 6,000 page views, and a quarter of visitors come back; and we’ve had very good reviews. Th approach is something which has not been attempted before. Normally, we think of copyright as something we need to tell people about; so we design resources as a tool to teach creators what they should know. We did it the other way round. We asked them what they didn’t know, what they wanted to know. We looked at the most frequently asked questions about copyright on the web and then systematised them. I see Dinusha Medis here and a whole family of other contributors – I don’t want to single them all out now – Bartolomeo Meletti, who was the lead producer of the portal, Kris Erickson, Hayley Bosher, Ronan Deazley, Ruth Towse. I warmly recommend to you to talk to these colleagues and to have a look at the portal.

In some ways, the design approach of the portal is not so different from the UEA file sharing study launched today. We try to understand what makes people tick before we make prescriptions – so behavioural economists may be quite comfortable with this approach.

CREATe is a team effort. It is a consortium of seven universities, and we have quite a few of the representatives here. My Deputy Directors Lilian Edwards and Philip Schlesinger here in the first row here, and we’ve got Management Committee members from all the consortium members but St Andrews, I think. Nottingham is here, Goldsmiths, we Edinburgh, and who have I forgotten? That’s pretty much the whole list – Strathclyde and Glasgow, obviously. Part of the purpose of today’s event is to probe what academic research is in the eyes of the user and also in the eyes of industry, and we are here to be challenged. We don’t want to play it safe, and if you have a challenging issue for us, come and challenge us.

Let’s now turn to today’s study: In my view, it is a breakthrough in the study of digital condition. And you may say, “How can that be? It’s only a review.” Well, it’s a systematic review, and it’s a systematic review in an area which has not been treated well by commissioners. So almost everybody who has commissioned a study had an agenda; they have a reason why they want a particular type of data examined for a particular type of result. And therefore, at times it is important to reset the state of the art. You can’t do it much better than in the study we have in front of us; as CREATe we are proud of that effort.

A word on the response panel. Every sector that is addressed in the study had the right to nominate a speaker to react to what the study finds. Remember, the study just reports what other studies have done. And so we systematically approached the umbrella bodies of each esctor – so, UK Music for music, the Publishers’ Association for publishing and so on – and we asked them to put up a speaker, and snowballed from there. The panel you will hear later has been self-constructed in that way, there was no deliberate intervention. It was an opportunity to give the industry a forum to respond to academic research.

Lastly, I would (very briefly) like to introduce Alison Brimelow, who probably doesn’t need an introduction – I think everybody in the room knows her as former President of the European Patent Office and as the former Chief Executive of what was then the UK Patent Office. We are extremely pleased that Alison has agreed to chair our Advisory Council, and Alison will also chair the discussion today. This will be probably the last you see of me today. It was a longer introduction, but I hope you’ll forgive me.

Daniel Zizzo: Thanks, everyone, for coming this afternoon. The work I’m going to present today is joint with Piers Fleming and Steven Watson, and has been conducted at the University of East Anglia. What we’re going to do today is to try to answer based on existing evidence, try to answer the questions what is the evidence that unlawful file sharing causes harm, and in this context, are the questions being answered. And much of the report is, in effect, a question, which is why do people choose to file share; or rather what are the classes of reasons by which people may choose to engage in unlawful file sharing. Now, my report is going to be very much on the evidence, and the way we approached the evidence has been to use a methodology that is used in the medical sciences and natural sciences, which is a scoping methodology, which is a formal procedure by which you can look at evidence by looking at keywords, the number of different times it occurs, and see what these are. Keywords are based on criteria such as, for example, things that the study measures, etc. And by this process, which Steven will be able to talk about later, we moved from over 54,000 different sources to a little over 200; and this represents the empirical, primary data about people’s unlawful file sharing over 10 years. Notice that the focus has been squarely on the evidence – this could be an important point. So it’s not been on, for example, economic models. That is a very relevant point for some of the things I’m going to say later.

One of the dimensions that we felt was very important was actually the kind of empirical data that we find in these studies. Not all data is, we would judge, of the same relevance or significance. There is obviously qualitative data that can provide some valuable information; but then really, one way of thinking about this is in terms of the extent to which a particular variable, a particular measure of unlawful file sharing actually can be used as a proxy for actual behaviour. As it turns out, you would see not many studies actually do that. So we’ve actually got quite a bit of a range of different measures that are actually used. So there is quite a bit of work which is on stated preferences and attitudes. This is simply about how good or bad an outcome is considered to be; right or wrong, preferable. We would judge this as being generally a fairly low quality type of data: the distance between that and actual behaviour is quite large. One step further, or one step closer but still very far away from actual behaviour is the number of studies that look at intentions; so they ask participants behaviour that they plan to engage in in the future – so intentions of future things. Hypothetical, counterfactual, however you want to call them, but not actual behaviour. There are studies that look at willingness to pay; that is, the amount of money that a participant states that they’re willing to pay in order to obtain a good. Now, usually this is done in a hypothetical fashion, which frankly reduces the value of these studies considerably because you can always state the numbers if you don’t actually have to have a purse to back it up. One step closer, there are studies of stated behaviour. Now, this is when you ask a participant behaviour that has been engaged in the past; so how much legal file sharing have you done in the past six months, for example. There are still problems with this type of data because of the type of responses that you get when you ask these types of questions, because people don’t remember, because people might not give you the right answer – but it is one step closer to actual behaviour. And finally there is observed behaviour; so that is behaviour that is either directly observed – such as in experiments – or else at a population level – say, sales data. So that is one important dimension.

Now, if we focus now on the welfare side, it is important to think of welfare from both the producer side and the consumer side. And we have come to this study with no preconception; we have found studies that do both, or rather do one or the other. So there have been studies that have looked at the impact of unlawful file sharing on legal sales. The focus of these studies has been on the producers, which of course is an importance. There have been very few studies, I have to say, that have looked at purely the consumer side by asking for willingness to pay for files. So here they focus purely on the consumer side. Interestingly – and we’re going to see this pattern again – out of 67 observations in terms of studies that we saw, about two thirds of them are about music. Now, there are a number media, different media that you can think of: music, movies, software, games… But we only find a focus on music and, to a lesser extent, movies – very few other mediums. If we focus on the producers’ side, that is where most of the work has been, and what these literatures have looked at is the correlation between unlawful file sharing – that’s what UFS means in these slides – and legal file sharing. There are different studies which have reached different types of conclusions, sometimes sensitive to what you include, what you exclude, and the method of analysis. One problem is that in many cases there is reliance on stated, even hypothetical behaviour or counterfactual. That is a problem with these studies, and ignoring the consumer side is a problem. The producers’ side is very important; it’s extremely important, we do take that very seriously. But you can’t ignore the consumers when you look at welfare. There are only four studies that look at the consumers’ side that we could find, and their general claim is that unlawful file sharing is actually good for welfare: that’s the claim in these studies. However, we shouldn’t jump on the bandwagon too quickly because although the idea that you measure the consumer’s willingness to pay will appeal to many economists, it is actually unable to capture dynamic welfare effects; it is unable to capture the extent to which you may discourage the production of future creative goods – and that is not captured here, so it’s skewed as an analysis. It’s been mostly with students – just university students – and finally, all based on hypothetical data. And I strongly expect, as an experimental economist, that this will have biased upwards quite systematically in favour of unlawful file sharing. So, clearly limitations; clearly the need of evidence, evidence, evidence.

Now, why do people file share unlawfully? What you will find in the report is a little model which classifies different sources or different determinants of unlawful file sharing. We classify them in five groups. There are considerations of financial and legal utility and disutility, so financial and legal determinants: the differential price between legal and illegal, and the legal barriers, the legal framework. There are dimensions and determinants that relate to experiential types of factors; that is, the extent to which you have a preference, you perceive the good that you want to kind of engage in, you want to consume. There are technical factors: how easy it is to get access to legal or unlawful products. There are social factors, potentially there are social factors: you do what your peers do. And there may be moral factors or determinants; for example, to what extent do consumers think that there is harm for other parties by engaging in unlawful file sharing? There are moral types of considerations that may be/may not as well. So these may all matter, and they will lead to overall judgement that then will lead to a decision whether to do nothing, purchase a legal product or engage in piracy.

We have mapped the evidence we have found in a cubic model: this is the cube that you will find on the cover of the report. This cube has got three dimensions. One dimension is the type of the source of unlawful file sharing; so experiential, moral, social, financial and legal, technical utility, for example. In another dimension you have got the type of data – this is what I was talking about earlier, how close you get to observed behaviour. And then in another dimension – that’s this dimension over here – you get the industry: music, software, movies, videogames, books, TV or generic (we simply couldn’t tell from the study). Now, it should be obvious by looking at that cube how the evidence appears to be quite skewed. You see each of these little spheres represents the number of observations that we found in terms of studies for each combination of medium – that is, market – type of measurement of unlawful file sharing, and type of determinant of unlawful file sharing. And you will see that there are quite big spheres up there and they’re little down here.

Let’s look at this in more detail. So, stated preferences: this is not ideal, it’s not observed behaviour. But what there is, it’s primarily on music, followed by software – very little on anything else. This is less visible but it’s intentions, and basically it follows the same pattern; so, music, followed by software, followed by everything else. This is less visible but there is very little to see, actually, because there are very few studies on the music, basically, for willingness to pay. This we have considered before: stated behaviour. What about observed behaviour? This is the most interesting one: it’s the one that most accurately reflects actually what’s going on. And you find here that the data is skewed in two ways. First of all, again, there is a strong skew towards having studies… not many studies, so I’m not saying actually there has been a lot of studies with music because there hasn’t been; but comparatively there have been more with music, followed by movies, a little bit of software, and then almost nothing else – virtually no data in other markets. And what data there is tends to focus on financial and legal factors or around technical factors. There is observed behaviour – there is basically nothing down here. Moral/social factors simply haven’t been seen when it comes to actual behaviour.

So, some implications that were found: financial and legal utility. High prices do appear to reduce sales, lower the willingness to pay for content. Stronger laws do appear to reduce unlawful file sharing, but the effect may be temporary; we’ve tended to find evidence of temporary effect. And there is limited behavioural data confirming causal… a cause overall. Experiential factors: we found some evidence that deciding to sample new content/access niche content, build a collection may be relevant. Technical utility/technical factors: it tends to work as an initial barrier to unlawful file sharing, but again, there is the need of hard, observed behaviour in terms of evidence. Social utility: we do find some correlation between unlawful file sharing, mainly in terms of intentions and peer unlawful file sharing, but causalities are less in this area, and there is very little in terms of observed behaviour – and similarly in terms of moral factors.

So, let me conclude. Our key benchmark, our key message has to do with evidence, evidence, evidence: having good evidence, having more evidence, having evidence about behaviour, having causal evidence is absolutely key for a good understanding of unlawful file sharing, and therefore for affecting, for changing the incidence of unlawful file sharing. We have perceived it in a systematic way: that was the point of our survey. I’m aware, we’re aware that this is an area in which there are strong ideological views from both sides of the open access debate. That’s not what we’re interested in; that’s not what we want to do; it’s not what we really are there to do. We wanted to see what the evidence is, and the review methodology enables anyone who wants to questions some of our conclusions to actually… you know, they can do something similar. There is a description of the methodology, so if you think you can do better you actually can: the methodology is described in the report and you’re welcome to use it. As I said earlier, the “why” is essential for the “how”: if you don’t understand why, you won’t be able to understand how, or how you’re going to affect unlawful file sharing.

The cube is helpful in that context to understand where the gaps are in the evidence, and we have seen clear gaps in the evidence base, both in terms of the fact of concentration of studies in very specific… in specific industries but not in others, absence of studies in others, and along the lines of focus often on proxy, so unlawful file sharing that, from a behavioural experimental economics perspective – which is mine – frankly are likely to bear very little relationship with actual behaviour. These things matter. These things clearly matter as far as policy goes, as far as stakeholders go, because these are the kinds of things from which then, in relation to… depending on how they’re mapped you can use as a lever to affect unlawful file sharing: to reduce unlawful file sharing, to increase legal sales. So, in terms of financial/legal utility, of course you can think of ways of changing the price differential between legal and illegal – that’s where different business models may come in; you can think about changing legal consequences, but the worry is that the effects may be temporary, so that is a source of concern if you think about that – it may not actually be the solution that you think it is. Experiential utility: one of the things that emerged was that people who tend to engage in the greatest amount of unlawful file sharing tend also to be the people who tend to engage in the greatest amount of legal purchases – so the people with the greatest interest in the media, per se. So there is a point there of thinking about policies which target consumers with greater interest in the media specifically. Technical utility: that is, in principle, straightforward – make legal easier, make illegal harder. Sometimes, for example, there is a problem with niche content that may not be available legally, in which case unlawful file sharing is the only way to go if the consumer wants access to the content. Social utility: change the social norm, changes to the norm – that would be a recommendation. Moral utility: change the perception of moral value, for example the perception of harm. For some reason advertising campaigns have been along these lines, but the problem there – and of course that’s a more general problem –that applies the most to this type of potentially attractive form of interventions relates to evidence, evidence, evidence, or rather the insufficient amount of actual behavioural evidence.

Thank you very much.

AB: Right, I think I need to call the panel up to the front. Right, ladies and gentlemen. I note with some relief and satisfaction that we’re not hugely adrift from our timetable. My relief is that not only do I want to hear from the panellists who’ve come here to comment specifically, but I want to ensure that there’s plenty of time to hear from the floor and also plenty of time for you to go and get a drink at the end of the session. So, there’s a package to deliver here and it’s my job to try and deliver it.

As you’ll have gathered from the introduction, I’m Alison Brimelow and I have an interesting and varied background in IP but not a lot of it in copyright, but I am re-engaging at speed when I’m not doing other things that people in retirement do. I am delighted to welcome the panel, who have sort of slithered in. And could you just, I think from the end, say who you are for those who don’t know you? And then I suggest if you make your comments starting at the far end and moving down this way. I was doing my maths in all of this, and I think you’ve got between six and ten minutes each. And I’m pretty brutal about timekeeping – you have been warned! This is important because we want to hear from a lot of people. This is about hearing as well as saying today. Okay, from the far end.

DY: I’m Dominic Young. I’m the Chief Executive of the Copyright Hub.

RA: I’m Robert Ashcroft, Chief Executive of PRS for Music.

JK: I’m Jim Killock, the Executive Director of the Open Rights Group, which campaigns on digital rights.

TB: I’m Theo Bertram, the Head of Public Policy and Government Relations for Google UK.

AB: Right, well, I think off we go, then.

DY: Well, now; I wish I hadn’t sat here!

AB: It is entirely random!

DY: I know. I’ll remember not to sit at the end next time. I was thinking about this and I was fascinated to hear the presentation because it made a lot more sense to me listening to that. And I was wondering about the question, and it was a funny question really, in a way, because you can sort of split it up. One is is file sharing a good or a bad thing, and the other one is is the illegality aspect of it something which we should or shouldn’t be worried about, I suppose. And actually, you know, the Copyright Hub is largely attempting to do something which, in terms of this study, I suppose, would be called technical utility: it’s trying to make licensing easier, and therefore creates alternatives where the reason for illegal activity is because the legal alternative is difficult to impossible. But actually, it’s really easy to make an unlawful thing lawful in the area of file sharing and copyright, which is the rights owner just needs to say it’s all right. It’s quite simple. And usually, if somebody is in the business of creating work and putting it out there with the intention of people consuming it, then their default is generally that they want to say yes to people having access to it in some way.

Usually in the world of copyright and licensing yes comes with conditions: yes, if… Sometimes they’re commercial. It’s interesting that… I can’t remember which slide it was on, but one of them specifically mentioned sales as being the sort of influencing factor; but there’s lots of forms of value that get exchanged in return for licenses. We know, for example, that creative commons licenses have conditions. There is an “if” on a creative commons license, nearly all of them; it’s just not a financial one. It’s not “If you pay me”, it’s “If you do other things”: credit me, don’t change my work, or whatever it is someone chooses from the menu of available creative commons licenses; which in the commercial world, there is an “if” that is financial as well. But financial outcomes are not the only outcomes that people look for in return for permission.

But if you’re doing something illegal because you can’t get permission, there are probably one or two reasons, maybe even three. The first one is somebody doesn’t want you to do it; they want to say no; you’re not willing to agree to their “ifs”, their conditions. Another one is it’s just not practical: you don’t know who to ask, the nature of the conversation you need to have with them is so sort of clunky and expensive to have compared to the use you want to make that it’s just a waste of time and you can’t be bothered and there’s no sort of sensible legal service set up. And the third one is that maybe the thing was never there to be shared in the first place. I think probably nearly… I would be… this is… I’m very un-academic, so I’ll just make up statistics and hope that they feel about right. But I would hazard that the majority of work created to which copyright applies is also private, and when people use the right to say no to use of their work, actually what they’re doing is exercising a privacy right as opposed to a commercial one, and so sometimes people just don’t want you to.

So if the question is… one of the questions implied by this is should the law change to make things that are currently illegal legal because you want to do them – and I’m not saying… I mean, obviously there’s no conclusion reached for that – and I’d say that’s probably the wrong question. I’d say the right question is how do we make it easier for those who seek permission to ask the question and to receive the permission: the yes, the no, or the “yes if”, which is the most likely response they’re going to get. And fundamentally, that is what the Copyright Hub has been set up to try to do. It has no particular sort of ideological view other than within the law copyright is a pretty successful thing. You know, it was very interesting, Martin’s introduction: it’s 300 years old, and fundamentally pretty similar now to what it was then. And it simply says if you create something and it’s yours, it’s up to you to decide what happens to it.

One of the reasons that I perceive there’s so much illegal activity – only one of the reasons – is that the process of exercising that right to say what happens to your work is actually rather difficult in the digital area. It’s very easy to copy things and pass them around; it’s very difficult to track down a person who owns them or controls the rights to them and then get permission, and it’s still a process that largely happens manually, actually. Very often you end up making phone calls and sending emails, and if all you’re doing is sending a file to your mate that seems like an awful lot of work and probably a lot more cost than the actual value of the transaction.

So I would just say I think improving the capability of the rights that copyright gives creators to be exercised by those creators in the way that they choose is the big challenge for the future of the way copyright works and the way the creative industries work; and that’s the thing the Copyright Hub is trying to correct. We are pro-copyright but we are not ideological about the ways people choose to use it, other than the rights that you are given as the creator of something, as the owner of a piece of property, is a right that you should be able to exercise in the way that you see fit. My observation is that that right has given us incredibly diverse media with thousands and thousands of organisations competing for the attention, time and money of consumers, and a massive choice for consumers. So it seems to work quite well in the interest of consumers, and making it work better in the digital sphere and era is the big challenge facing us, and that’s the one that we’re trying to address.

AB: Thank you very much: admirably succinct.

RA: Robert Ashcroft, again, PRS for Music. I’m very interested in the fact, first of all, that there were so few of the studies on the willingness to pay methodology, and certainly most work has been done on trying to evaluate the impact of unlawful file sharing on sales. But one of the things that struck me reading the report about the willingness to pay methodology is – and please correct me if I’m wrong – but there appeared to me to be a hidden assumption that the supply of music is a constant, and that bothers me somewhat because my concern is for the functioning of the market and the reward of the creators. And without a market, the functions to give them reward, then the supply of music runs the risk of drying up to some extent.

I guess I’m also surprised that we should conclude that further research is to be done on the sales side, because as far as I was aware – at least before this report was published – there was an awful lot of work that had been done, and I’m certainly aware of the study done by Stanley Liebowitz who had said – I can quote from him – “the file sharing literature has focused mainly on whether file sharing has decreased record sales with less attention paid to the size of any decline.” Although there’s still some contention, most studies have concluded that file sharing has decreased record sales. What has not been noted is that most estimates indicate that the file sharing has caused the entire enormous decline in record sales that has occurred over the last decade. This heretofore hidden result is due to the lack of a consistent metric that will allow easy comparability across studies. The task of this paper is to provide such a metric, translate the results reported in the literature into that metric, and then summarise the results from this exercise. The conclusion, therefore, was the kind of decline we see here in the UK – from 1.2 billion to just over 750 million – can largely be explained by unlawful file sharing. So, coming at this from a practical businessman’s standpoint rather than that of an academic, and coming here from the standpoint of trying to ensure the livelihood of our members, then it would appear to me that the answer from most studies is it’s pretty evident this is causing harm.
The last time that Theo and I shared a platform, although we weren’t physically present, was when we had jointly commissioned some research into the ecosystem of unlawful file sharing, and we had that work done by Detica, a division of BAE; and that essentially… the essential conclusion of that was that piracy was a business. We drew slightly different conclusions from it – Theo will speak for himself on his conclusions. Mine were that this is a business where people are making a lot of money out of not paying for the content that they ought to be paying for. We both agree that there is a “follow the money” strategy to strangle this supply: stopping advertisers, or at least giving advertisers the information that enables them to avoid advertising on illicit sites; stopping the credit card companies making means of payment available to unlicensed businesses. And I think that since the report was published in 2012 there’s been quite a lot of progress in this country on that, and we have seen the decline in the overall business sort of plateau since then. But it seems to me on many levels that there is evidence that says that unlawful file sharing does cause harm, and I’m very surprised therefore to see the conclusions that we’re unable to draw that conclusion from this comprehensive research. Thank you.

AB: Thank you very much indeed.

JK: I thought this was a rather good paper. I thought the best bit about it is probably the way it shows the skew of evidence to be to do with music, and also widening the notions of what kinds of harm and welfare we’re looking at, what kind of utility we’re looking at, and the way that suggests then strategies for dealing with the question. I think all of that is very, very helpful because policy makers, as well as business, want to see more lawful sales and greater licensed revenues. That is undoubtedly a good thing from an economic point of view, even if it isn’t the whole story about this problem. But I think that inasmuch as this study gives a good range of responses and suggests a range of techniques that might be used for both business and policy makers to think about the problem, that’s very, very helpful.

A few people have alluded today about an ideological question. I’m just going to say something on that because I think there’s a sort of unspoken thing going on there. I don’t think the debate, that ideological debate, is about pro- or anti-copyright, or inasmuch as it might be those people are not in this remit. They’re just simply not in this remit. And the debate rather has been between people who are worried about the future of copyright and copyright industries, legitimately worried about those concerns, and those people who, on the other hand, are worried about the impact of the enforcement and perhaps the extent of copyright in terms of length of copyright or the lack of exception, something like that. But that second group of people to which the Open Rights Group belongs – the people who are occasionally concerned about the impact of enforcement measures on things like human rights, due process, questions like that – those people often are portrayed in a rather… in a slightly different light, and I just want to reject that notion outright because I think it’s extremely helpful if we all understand that ultimately we’re on the same side. And perhaps then this evidence today can be used as a way for all of us to discuss this in a relatively rational and progressive and helpful way.

One other thing, a few other things about what this study doesn’t do – and I don’t mean that as criticism – but a few things that I think worth mentioning. One has been the use by policy makers of proprietary evidence in this space, by which I mean evidence where you can’t interrogate the methodology that underlies the results. We saw that used a few times in the Digital Economy Act debate. I think policy makers should have a simple rule that any evidence of any kind where you can’t disclose the methodology that comes up with a figure, that should be simply disallowed in terms of policy discourse. That would force… that would encourage, let’s say, people who wish to produce evidence from a commercial point of view to be absolutely fully clear about how they come to their conclusions in order that that evidence that they produce can legitimately be used in public debate in a way that is sensible rather than obfuscatory and scaremonger-y.

Secondly, I think probably because it doesn’t exist… I think it’s worth thinking about the impact of the enforcement measures and what kinds of study might be put on the other side in terms of the measures that have been put in place to deal with unlawful file sharing. So the obvious way that most unlawful file sharing is used and people try to reduce it is through greater licensed works: it’s through the sale and use of platforms. That’s where I think most of the positive effect has been. But equally, there are measures that have been put forward, including blocking websites, targeting users in various ways, and copyright trolling by groups like Golden Eye, and of course questions around what kind of legal redress is available or not to rights holders who want to see their rights enforced. I think all of those questions would be worth looking at, and I think they’re a key part of the unlawful file sharing debate.

Also, already mentioned, has been… and I think is a useful benchmark in this: what has the impact of legal file sharing, in fact? I think there are platforms like VODO where documentary makers, people who’ve got a social message to put forward are having… you know, they’re reaching new audiences, bigger audience, they are able to distribute their work cheaply and effectively; and then obviously that has social welfare benefits without having the adverse effect of unlawful file sharing. And just in terms of your study and understanding the impact of unlawful file sharing, I think it’s worth studying what the benefits of purely lawful file sharing are in order to understand what the impacts either way on file sharing might be.

So, I think that’s most of what I’ve got to say. I think, as I said, the object here to increase access to material production and creation. Commercial production is a very important part of that, but equally we also need to remember that social production needs to be encouraged, if only to help fuel the future of commercial production and platforms, because it’s often the grassroots stuff that feeds up into the big commercial successes of the future.

AB: Thank you very much. Theo?

TB: I’ll just start by saying a little bit about Google’s perspective, and then I’ll talk a bit about the report. I usually find if I don’t talk about Google’s perspective I have to answer the same questions again and again afterwards, so it’s almost like a kind of disclaimer at the beginning of any statement I make on copyright, which is that Google is committed to supporting the creative industries and absolutely committed to tackling piracy. We’ve supported the film and music industry to the tune of $1 billion through YouTube over recent years. And on copyright take-downs I think we now take down around 25 million URLs a month, and we do that at industry-leading pace and we work very closely with organisations like the BPI and FACT to make sure that we do that as effectively as possible.

So, that kind of health warning over, I’ll come onto the report. Martin, you kind of invited us at the beginning to be challenging and provocative, so I’m going to be a bit challenging and provocative. There are a few things that I… I mean, I think the report is great and there’s lots I could say that’s really good about it; but I’m going to take issue with the things that…

Now, the first thing I would say is – and maybe it’s because this hall kind of makes you sound like a vicar preaching from the front – I felt that the overall view of industry and Government’s approaches in these spaces was a little bit “holier than thou”. You know, it would be easy for me, in a way, to agree with the report and say, “Yes, the big problem with the copyright debate over the last few years has been the BPI’s data, it’s all wrong,” or to say, “Yes, it’s all the other side’s fault. Everyone commissions this data and they commission it specifically for the reason of trashing us, but our data is all perfect and everything we’ve commissioned is spot on.” But I don’t think that’s… if ever that was a kind of caricature of the copyright debate I think in recent years it has moved on, and I think the work that we did with PRS was challenging for all of us and I think it was data-driven. It was us and PRS, the BPI, the Premier League, the Publishers’ Association. We took data from Detica and it was entirely objectively done. And actually, we didn’t really have a conclusion because we didn’t want to each draw separate views from it. I mean, clearly industry and Government and the different parts of industry come at the issue from different perspectives; but I don’t think that means we can’t agree on the data. And I think one of the good areas of progress that we’re making at the moment between all sides of that is a kind of agreement on it being a data-driven approach. And if you look at something like the Ofcom survey – you know, the Kantar Media report that they do fairly regularly now – you might draw different conclusions from what the data means, but at least there is a kind of shared agreement on the value of that data.

And also, for example, one of the things you were drawing out was the idea that music shouldn’t be treated in the same way as different types. And I think that is kind of self-evident to us already because the conversations that we have, the way in which we deal with content, the way in which we see this happening we know is different. So if you want to… if someone is trying to get hold of the next Premier League match, it’s a very different experience than if they’re trying to get the latest album from someone, because the Premier League match only appears at a certain time and is much more likely to be streaming, and the whole way in which that is promoted is entirely different. So we see that and we know that from our own experience as well.

On the methodology, it’s interesting. I mean, I think there are lots of good surveys that were done using this methodology from medical science, where you look over the years at the different reports that have been done. I can remember looking at… There was one that was around how repeatedly, over a number of years, journals had shown up that protein kinase, reports on protein kinase had a correlatable incidence with reports on problems in pregnancy and infant mortality, and this was the way that the researchers were able to conclude that maybe there was a connection between these two things. So it wasn’t something that was discovered in the lab, but through research of the literature of the time. And so I can see why kind of looking back over literature over time and all these different surveys, what does it tell us about what we’re studying. But the reason I don’t think it works quite so well with the topic that we’re looking at is that the human body doesn’t change or doesn’t evolve very fast; certainly not at the pace at which the industry that we’re in evolves. And the studies that you would have had from five to six years ago would virtually be out of date now in terms of the way in which content is shared and delivered, and also how it’s pirated. So it would be interesting… Obviously I’m at Google, I would like to see this animated in a nice Gif as well if possible. But what I’d be interested to see is how do those bubbles change over time; and if you cut it year by year, is one of the reasons why we see music being bigger because music has just been pirated for longer or has had more of an impact for longer? I think movies and TV, you can correlate kind of piracy, I think with the speed of download and people’s access to it. It’s always been a lot easier to pirate music because it’s a smaller file size and easier to share than a big film. So it would be interesting to see kind of how did that change over time. And even the term “file sharing” already feels fairly out of date. The conversations that Eddie and I from FACT were having this morning were all about streaming, and there’s a kind of sense in which downloading already feels like something that was some era ago and we’ve already moved on. And so I think part of that is what creates the big challenge that we have in trying to make sense.

And just to come to your point, finally, on evidence, evidence, evidence: I agree that we do need more… we need evidence that we can all agree on, and we need evidence that is neutral, and we may not agree on kind of what does that mean. But I would also say that we’re not with a dearth of evidence; there is a fog of evidence. The problem that we have is you look at all of… as a business, if you’re PRS, if you’re Google you have so much data, and it’s trying to work your way through that and understand what is happening, particularly with the pace of change in this industry and the volume of consumers that we’re dealing with, the amount of content.

And finally, I guess my challenge to you and to CREATe is this is a great academic study, but within that fog of evidence we know there’s a fog of evidence and we know that it’s sometimes very hard to tell what is the definitive outcome, what’s the cause and effect here; but that doesn’t stop the fact that we have to act. You know, we have to do something, and if you’re Government you need to do something about it, and if you’re PRS you need to do something about it, and if you’re Google you need to do something about it. So how do we move from… I agree we need longitudinal studies, but we also need to act now. The problem is kind of immediate. So, those are the challenges I put your way.
AB: Theo, and indeed all of you, thank you very much. If I can comment in my capacity as the PAC Chair, I have a lot of sympathy with what you’ve just said and I think that the difficulty – and I’m looking at Tony Clayton at this point – the difficulty with policy makers in a sector that moves so fast, and where indeed there is so much evidence but also contest, is to work out exactly what makes sense and in what order you respond. And legislation is notoriously slow, even in sexy areas of policy – I don’t know whether copyright is sexy or not, but… Yes, I think so, Tony is saying! Unlike other bits of IP which aren’t.
Thank you very much to the panel, and I particularly thank you because you’ve given us lots of space for people to contribute. Before I move on to comments from out there, would the panel like to comment on what you’ve just heard… Sorry, would the authors like to comment on what they’ve heard from the panel?

DZ: Yes, absolutely. I feel we received some very constructive feedback, and some of this feedback is particularly helpful because I’m sure it’s going to help us write our next grant application! So it will be especially helpful in identifying the next steps in terms of research. So, I will just answer a few points, and then I’ll give… I think I’ll let Steven answer some more, but very briefly.

Most definitely there is benefit… there is certainly an issue about accessibility of the legal content, and therefore absolutely the extent to which we can make that easier – that is definitely going to help and it relates to the technical utility.

In relation to the comments about the difficulty of – Theo, this is in relation to your comment about the difficulty of dealing with data in an ever-changing situation. It kind of reminds me of something that Albert Einstein once famously said; and Albert Einstein said that the reason he became a physicist was because social sciences were too hard. And it kind of reminded me of that! So yes, absolutely, we live in an ever-changing world and the pace of technical change makes it harder. You’re absolutely spot-on in terms of interesting lines for future research – I think that’s absolutely spot-on. For example, how do the studies change with time; or, for example, streaming, the ever-increasing importance of streaming, and how do we deal with that in a context in which we actually need to make sense of it for policy purposes? And true, as well, that we’re a bit like in the conditions of doctors, and doctors need to treat the patient even if they don’t quite know everything about what is happening to the patient; so I think that’s also a point absolutely well taken.
My sense is that there is useful research that can help with that understanding that’s not been done. Part of the problem has to do with all the gaps in that cube, really, and identifying where the key differences lie – where we suspect that they do lie – and also exploiting some of the methodologies that are now available in terms, for example, of combining behavioural experiments that can help with causality, and with the actual data there are ways of doing that, and I think there is a lot of potential there to try to identify causality in a better way than, say, in an Ofcom study, a type of study, a survey type of study in which you simply ask about stated behaviour in the past. And yes, I mean, there are lots of limitations there, for example.

Now, about the comments about the fact that there are a lot of studies… I mean, there is a lot of common understanding now that a lot of the work… you know, this does not have a secret agenda – yes, absolutely. I think that’s absolutely agreed and understood, and I think that’s important that we all work together to get a better evidence base. That’s absolutely true.
Steven, do you want to answer some of the points?

SW: Thank you. I think there are three points I want to pick up on. The first one is why is it we say we can’t definitely say there that there is an effect when other studies have. The second is there we were talking about the blocking of websites, so I might make a brief comment on that; and finally, in terms of the value of the methods that we adopted.
So to assess why is it that we say that we can’t necessarily say that there’s a definite negative impact upon sales with file sharing when the review you point out – I believe it was Stan Liebowitz – and what he… There’s two issues or big differences. The first, of course, is that he looked exclusively at music, and I think we do say in the report that the predominant effect in terms of music is that there may be negative file sharing. But partly we also look at other industries where the evidence is far more ambiguous; but there’s also the effect that in that study by Liebowitz there were two additional problems. The first is that what he doesn’t necessarily discuss that I think we do a little more is the limitations in the type of data and the different types of model which are used. So yes, the studies that have been used have identified that there are negative correlations, but they’re almost exclusively just physical sales; whereas we’re saying that actually the dynamic effects are much more interesting. I think you said yourself we need to know what’s going on in terms of music production as well, and that literature is very much in its infancy. So what we really need to start trying to do is to develop empirical studies which do look at content creation as well as necessarily just sales, but we also need to look at sales in terms of not just physical sales; we need to look more at online sales, we also need to look at the new and different business models that are coming out in terms of, say, Kickstarter, PledgeMusic, things like this. I think there’s a study in the US where they’re looking at all of the different ways you can make revenue from music, and almost none of the studies that we’ve examined have begun to scratch the surface of the different ways you can make revenue through music. So this, I think, is one of the reasons why we’re saying the evidence is ambiguous. It’s partly because they just looked at music; it’s partly because we’re also discussing the other ways in which the evidence is limited.
In terms of the blocking of websites, I just thought it was worth discussing very briefly that there are a few studies on this and they tend to find similar effects as the introduction of new laws, which is there is an initial shift in behaviour, but generally we tend to find unlawful file sharing tends to be fairly habitual. If you do unlawful file sharing, often your reason for doing so is because… you know, over time. There’s an initial reason you do it, and then you do it because it’s what you’re used to. So if you close down your favourite website, often there’s an initial period in which you shift to legal sales and then a new different website. But perhaps some of those people who shift to legal sales in that interim, because it’s habitual behaviour they’ll stay on legal sales. And so there’s certainly more evidence we could try and explore of that, because again, a lot of the literature on that is somewhat limited. But I thought that was worth commenting on anyway.
And finally, in terms of the value of the methods, I think there’s a couple of things I want to say here, the first in terms of how do we keep up to date. Obviously one of the values of these more systematic type approaches is that they are phenomenally easy to update compared to almost any other review. So again, if you think of the Liebowitz report, he excluded some studies on… it wasn’t entirely clear why some were included and some were not; whereas with our study you know exactly why anything else was excluded and why it’s there, and if you want to you’ve got an Excel file when you get to the full text stage, so all those relevant studies you can see exactly why we did or did not put it in the review. And you now, if you want to replicate what I’ve done or expand upon it, you don’t need to begin where I did. You’ve got my search string, you’ve got my sources. You can easily update that. So if the literature moves on, you can use my work as a springboard: that is the hub. So you can update this and keep it relevant. And one of the things we were talking about, compared to, say, a systematic review in the literature, so we’re saying things change and it’s much more nebulous in social science compared to a human body which doesn’t change; well, that was why we adopted the scoping review rather than, say, a full systematic review. So systematic reviews in medicine tend to adopt a very strict hypothesis and they test that hypothesis: does this drug help blood pressure better than this other drug. We obviously have a much more diverse set of questions that we’re trying to answer; and therefore the scoping with the methodology, it’s not so much about “answer this one specific question”, it’s “what is the state of evidence that has been used to adopt this question?” And the real aim there is theory generation. So the idea is if we explore file sharing this way we’ve got a framework. You could then make some predictions about what will happen with streaming, how might this change; and you could then test those predictions by developing your own studies or assessing that literature. So hopefully this type of approach will help to address those questions, although I agree that obviously we can’t do all of them at once! I think that’s all I wanted to add.

PF: Okay, I think I’ve just got one more thing to add – I think my colleagues have covered most of the points that I was going to go over. But again, it’s about the changing industry issue, which I think is a very important one and a point well made. But I think that the fact that the technology changes doesn’t invalidate our research or the research that’s been done already, because people don’t change that much. So it’s still going to be the case that if your friends file share you’re more likely to file share; and similarly, if you don’t see any harm in file sharing you’re still more likely to file share. So those determinants that we identified are going to remain important, and I think actually that’s definitely worth taking note of and thinking about what their impact will be on the industry.

DZ: Thank you very much again for all the valuable feedback.

AB: Before I open the floor, any comeback from the panel before we move on? Are you happy with that?

I just was intrigued by that last comment. I come under the heading of that social cohort, i.e. ancient, who are apparently law-abiding. I live in a town that is full of incredibly intelligent retired scientists – people who invented radar and things – and there is a group there called Growing Old Disgracefully. So don’t make too many assumptions about where compliance can be assumed in society! The other point I would make is what may be a break on the elderly, is they don’t know how some of these sophisticated games, and if they did they probably would. So just be careful about your assumptions.
Could I then move on to the floor. There are a couple of mikes. Take the first question there. I can’t see your face but I can see a blue arm. Right, could you please say who you are and what your… I was going to say where you come from, but I don’t mean Surbiton! Where you work and what your affiliation is.
EL: I’m the guy who was having the coffee with Theo earlier. Eddie Leverton from FACT. It is a bit like being in a church, isn’t it! There are quite a lot of points I would make. One is just thinking about my children this morning have gone to the Natural History Museum to see the dinosaurs, and in that kind of reference some of the research here – you know, 2009, 2010 – is pretty much ancient. So what Theo was saying, actually, things have moved on so much in terms of people’s pirate behaviour as well as their legitimate behaviour; and whilst I appreciate the study is about file sharing and file sharing is still a very predominant way that people are accessing pirated content, there is a whole raft of other means that aren’t taken into consideration in this, and also a range of legal services that have come on board. Even since you started thinking about this report, there are probably a dozen ways that you can access content in different ways that didn’t exist before, and for film, TV, for books, whatever that might be as well.

The other thing is I’m very pleased to hear that Jim isn’t anti-copyright, but I do take on board everything that he said. And one of the things that we have done in the film and TV industry is done some robust research that is a tracking study that takes place every year, and that doesn’t… you know, it’s robust, it doesn’t assume that every downloaded copy or streamed copy means a lost sale. I’m not sure how much that was taken into account in your research as well.
And also, I guess it’s not just a number game as well, the number of studies. Again, Theo’s point is exactly right that music has had a long tale of suffering from piracy; film and TV is more recent and books is even more recent as well, and there’ll be other sectors that will suffer. 3D printing will have an impact on manufacturing and designers, for example. So those will all need to be taken into consideration.
I encourage the research, but I think it’s just the start, probably. That wasn’t really a question, actually, was it.
AB: Observations are welcome as well. Thank you very much. One behind you and then you.
GB: My name is George Barker. I’m from the Australian National University, and I’m also a Research Associate at UCL. My doctorate is in Economics, and I’ve got a Law degree and have been studying copyright and the evidence, and I’ve published some of the works that you’ve cited. I wanted to respond perhaps a little harshly, because I think you’ve been a little bit harsh about other people’s work. And take it as it stands, because I’m just speaking my own mind as an academic about how I see what you’re saying.

I feel that no one should walk away with the conclusion that you’re promoting, which you say the question that you seek to ask is whether UFS – the file sharing – is bad for welfare, and you’re saying it’s unclear at this stage. Be wary of ideology. Evidence, evidence, evidence. Now, that’s not new. We’re all saying evidence, evidence, evidence, actually.
So, the first point I’d make about your study is that it’s not research, actually, but you present it as research. It’s a literature review. It’s a review of other primary research. So let’s get that into perspective. There are other literature reviews that say the exact opposite, probably, of what you’re saying; so there are competing literature reviews now. The first point, though, is also when you say it’s evidence, evidence, evidence, I’d say actually first it’s about theory because the Hadron Collider wasn’t built and they didn’t go looking for a particle for no reason; they did it because of theory. So theory should be the starting point. And I think the core question here in terms of “Where as we at?” is whether sales and revenue from a point of view of economic theory is likely to have a dynamic effect on investment. And does economic theory predict that? Well, yes it does. The rate of return on investment drives investment to the extent you can hold costs of investment constant, but it’s the return on investment that drives investment. So therefore we know that dynamically a fall in sales, by economics, is a critical thing to focus on for an empirical research strategy if you want to test the hypothesis whether welfare has declined dynamically… i.e. dynamic welfare; not just welfare today, people who are buying things today, but welfare into the future. Because if we contest whether sales have fallen today and revenues have fallen today, then we haven’t refuted the prediction of economics that piracy, i.e. theft, reduces sales; and then the second prediction of economics that reductions in sales reduce investment; and then the third prediction of economics, which is that reduction in investment will reduce output and welfare in the future. So obviously the best place to start with the study – and so my question is – surely from a theoretical point of view, with the impact of sales.

And then the question is… my second question for you… There are eight peer-reviewed economics journals. That’s a screen for me about which ones to look at; not 5400, but with the articles peer-reviewed. Of those eight, there was one that never made their data available, so there are actually only seven. Of those seven peer-reviewed economics articles, you are aware – as Liebowitz has pointed out – that of the top journal articles – seven – six of the seven show that sales fell and that the entire fall in those sales, as Rob has pointed out, was due to piracy. The seventh… six of the seven said there was a reduction in sales, and therefore all the follow-on occurs about welfare falling likely; i.e. we haven’t refuted the hypothesis because it’s a scientific methodology. The hypothesis is piracy has a negative effect on sales, a negative effect on investment; so we’re looking for evidence that refutes it. At the moment, six of the seven studies are saying there’s no evidence to refute that hypothesis. Actually, the evidence is in favour of it.
The seventh study is the one that I’ve done a lot of work on, where the study only said that there was no effect – and that was Anderson and Frenz (2010) – that they could find. But when I went and got the data and I analysed it, I found they’d excluded from the analysis everyone who was not buying music, and that biased the result. They did it on the assumption that those people that weren’t buying music weren’t participating in sales and shouldn’t have mattered. Intuitively, yes. But the survey itself, when you looked into it further, asked people, “Did you buy music last year?” and people said yes, about half of them. So when you put that group back in you get a statistically significant result, which is that a 10% fall in sales causes a 0.4% decline in music sales. Sorry, a 10% rise in piracy causes 0.4% fall in music sales. So you’ve got 5,400 studies out there. You’ve only got six or seven that are really worth looking at on the sales issue. The sales issue is probably the only thing we can focus on.

So why isn’t music sales the one to focus on, given Google’s point that really, because of broadband, movie sales hasn’t been greatly affected? It’s music sales that are first off the rank in this piracy effect, so that’s the one to focus on. And if we can show what’s happened to music sales, why can’t we predict that that may be what will happen in the other sectors? But I just can’t understand why at least your paper doesn’t show that the evidence for music – first off the rank industry – clearly shows in the published economic journals a negative effect from piracy on sales; and the economic theory predicts that that will reduce welfare. Why don’t you say that?

AB: Okay, thank you. Would you like to pick that up, Daniel, at this point?

DZ: I’m happy to answer this question. I may ask Steven on a couple of points, but that’s fine. That’s all valuable feedback and I will be happy to answer it or for the team to answer it.

So, let me perhaps begin from the comments that were made initially. There was a comment about unseen behaviour – behaviour in the past – and of course the extent to which that may affect the size of some of the spheres. The answer is yes, of course it may affect the size of the spheres, and so clearly in a dynamic context, as this review gets updated we will be able to see where actually there are studies that try to pick up on the things that are now evolving. Nevertheless, even in the context of eBooks – we can probably say that eBooks have been around for three years – our studies are up to 2013 and include grey literature, so the fact that very often there is actually no study, no observation, seems to be quite telling. And it’s not just a matter of very few studies, but it’s a matter of very… of, indeed, close to zero studies. Nevertheless, and I agree, clearly it has to be updated; it’s a process that has to be regularly updated.

In terms of the point about the relationship between unlawful file sharing and lost sales, yes, of course; this is an area that we do spend quite some time in the scoping review to discuss, and it is a crucial point. Indeed, the study by George that he was referring to, and some of the study by George, some of the studies by other researchers have looked at these issues. So yes, there is a literature that provides that and it’s one that we discuss as a key part of the sales approach.

Moving onto the point by George on the relationship between theory and evidence. Yes, we made in the scoping review a deliberate choice to focus on evidence. Of course, in the context of the evidence we would… our scoping review would become a study that does relate to any dynamic effect, and we haven’t really found one except potentially indirectly through sales approach types of studies, which we do discuss. But we do mention dynamic effects specifically as a mutation of certainly the willingness to pay approach, and there was a slide that mentioned precisely that. Having said that – and I’m talking as a behavioural economist – the fact that there is an economic theory doesn’t mean that that actually translates into reality, by any means or imagination, can make loads of counter-examples of this kind. Indeed, as shown by the study by Waldfogel in 2011, it is even the case that sales and consumptions tend to follow necessarily the same trend; so you can have that sales go down goes while consumption remains the same. And indeed, the way that the market works dynamically in terms of evolution goes beyond the scope of those economic models.

There are various comments you have made about the value of a scoping review. Well, I would say that national scientists would be very surprised by what you stated because they would actually think that the Cochrane Review, that the methodology, systematically looking at the effectiveness or not of, for example, a given medicine is absolutely crucial in order to evaluate the effectiveness of the medicine. So actually, we need more scoping review, more systematic review work in social sciences, not less – and that’s absolutely important. We need less models and more review work, in my view. And we have looked at every single of those papers, and the basis and reason to which we assess their quality are identified in the spreadsheet. So that is all available, and if one disagrees with that conclusion, well, one can go back to the spreadsheets and one can see the extent to which we have made a mistake – which we may have in specific cases – and therefore review our conclusions. Reputability, as pointed out earlier by Steven, is actually something that is allowed very easily by our methodology, and I think it’s a strong plus.

Clearly a pure sales methodology does ignore any consideration of the consumer side, and that seems to be absolutely crucial as well to consider. And indeed, when you stated, when you mentioned your own study, about 10% rise in effectively consumption connected to 0.4% reduction in legal sales, but actually that suggests that the amount of consumption went up considerably. So actually the consumer surplus is likely to have gone considerably higher, and so the consumers need to be taken into account as well as the producers.
Steven, anything you want to add?

SW: Just a couple of things. Two small points still left. Why don’t we focus on music because that’s the one where the most evidence is? Because I think it would be wrong to make the assumption that what is true for music is true for all other media. We can’t rule that out, but we obviously can’t assume it so we need to check what the evidence is on that.
Secondly, in terms of why not focus on just peer-reviewed articles, I would suggest that… you know one of the first things when you start looking at how to do systematic review methodology, you would rule out that assumption first of all because you cannot assume that the quality of the journal that something is published in is equal to the quality of the research within it. You have to judge each body of work on its own merit. I mean, for example, if I’d only looked at the peer-reviewed literature I would have found the Anderson and Frenz (2010) literature but I wouldn’t have been able to find Dr Barker’s work that corrected some of the work in that article. So I think you have to take as wide a scope as possible, you have to find as much of the evidence as you can find, and then you judge that evidence based on what it is.

And the real other advantage of the systematic review methodology as opposed to a literature review is we examine it. Well, the problem is that even if you don’t believe you have a particular bias, you have no desire to seek a particular answer when you’re conducting your review, humans are humans. We all buy newspapers that tell us what we already believe. You don’t buy The Guardian if you don’t like hummus. Do you know what I mean?! And in much the same way, if you are a researcher and you have a particular theory or you have a particular conclusion that you tend to believe in, you are much more likely to first of all read the literature that finds that answer, you’re much more likely to remember literature that finds that answer, and you’re much more likely to have a body of literature that you collect which supports that view. So the power of our methodology, I believe, is that it doesn’t matter what I believe; I had to confront every piece of evidence that came my way and I had to account for it and had to present it to everybody who reads the work. So therefore, if you don’t believe what I say, well, you’ve got all of the evidence and you can have a go at me, frankly, and that’s the useful part. That’s the use of this methodology. That’s all I wanted to add to that.

AB: Thank you very much. If you want to come back can you come back briefly, because I’ve got several people now waiting. You’re two and you’re three.

GB: Very briefly, I think literature reviews are a good idea. I agree with going through the 5,400 studies. My debate here is only with the conclusion and how it’s been read. I noticed on your website it said, you know, “Myth-busting.” So you’re sending a very strong message here, I think, and if you didn’t intend to you need to revisit how it’s being interpreted. You’ve read all this literature, it’s all of equal merit, or you’ve only focused on that which is any good, and you’ve concluded that there’s a bad outcome for the assumptions of industry and governments alike. So I’d just be careful of that.
The second one is I, of course, agree with lots more research and lots more data. But I think people have to make policy decisions in real time about laws, and they would look for researchers to look in that area where you can really test the critical assumptions and hypotheses first, and report on that and emphasise that if there is a strong result. And I just wonder whether you might want to emphasise that music sales result a bit more when you say this has got an unclear outcome.

AB: Thank you very much indeed. One of my pet hobbies in retirement is pursuing policy by mantra, and I think that lies underneath some of your comments as well: beware of the mantra. Thank you very much. You were next – very patient.

JM: Thank you very much. I’m John McVay, I’m Chief Executive at PACT, the UK Producers trade association. We represent film, TV, animation, children’s producers. And it’s interesting in your research that you actually draw attention to this, that there’s very little research for television – the use of unlawful file sharing for television – and I think Theo or Jim said earlier on, “Well, that’s because of bandwidth.” But there is a corollary to that which goes we’ve just seen the decline of one of the most important sources of financing for British television, which is DVD sales, which were fantastic, wonderful margins. The reason why they were so good is because it’s quite hard to copy them, other than the Vietnamese or Chinese copy farms which got closed down in the Midlands. But as soon as bandwidth came on where it was easy to file share, it’s completely killed the market – so there’s a smoking gun there. The reason why you’re not getting the same research generally from all British television is because, unlike music and film, we give you it free every night. You get thousands of hours of free content every night you can then get in catch-up through your VoD service, or you could get it on your PDR. Our monetisation is through advertising, international sales or selling onto third party distributors, and DVD was a way that we funded making great British content. We are the second largest exporter of content in the world, which for a small country is, I think, a good story. So I’m concerned about… I agree with that gentleman’s point about sales. There is a smoking gun in television: it’s called “look at DVD sales and what happened, and correlate that with bandwidth” – and that’s why that industry has declined and that’s why we’ve seen all those margins.

The other point you mentioned but you sort of skip over in the report on this, and you skipped over in your piece before, which is about stated behaviour and actual behaviour; because that gets me to the point which I think is what all of us – Jim, everyone on the panel – and what I deal with every day, which is the morals of this. I’ve got three teenage boys. If I ever caught them file sharing they’re dead – simple – because that pays my wages, my members make money there. And if you want to engage with politicians on this and a little bit more with industry, we have to live in the real world. I think the point made about the human body doesn’t evolve as fast but human culture does and human values do, and that is what we deal with in order to market, monetise and make things available to society, and to invest in that. And given you’re economists and I’m not, I’m not an academic… Hands up any other non-academics in the room. Ah well, okay. But I am from industry, and I’ve been in this business nearly 30 years, and I’m a bit upset that yet again here’s academia not actually presenting something which industry can properly engage with. But I would encourage you to read a very good report that came out yesterday – and I was over in Brussels at the launch of it – on a review about IPR in Europe done by a Professor of Economics at MINES Paris Tech and a senior researcher at CERNA called Olivier Bomsel. I would check that out because it’s a really good report about how creativity – he’s French so he comes up with fancy words – the media-isation of creativity; basically, how do you create the single creative act, and what does industry do to make that a social welfare, make it a welfare good. And I would encourage you to read that because it might help your next report.

AB: Thank you very much indeed.

DZ: Answer now or take another question?

AB: We’ll collect two or three. You’ve been waiting very patiently.

JH: I’m not sure about my patience, but I’ve waited a bit. Jeanette Hofmann, Social Science Research Centre, Berlin. I have a very simple question based on what Steven just said. What were your expectations and were you surprised about the outcome of your work?

AB: Well, I think that does deserve an answer, so if we take those two. Comments on the smoking gun, which I find absolutely fascinating – I think your teenage sons will take it as a challenge to find a way to do what you’ve told them not to do, but we’ll have that discussion separately – and also were you surprised.

DZ: Okay, so I’ll let Steven answer the question about expectations, and I will begin answering the other questions. Of course, the message in the report is nuanced, so George I mean, obviously we do try to nuance the message in the report by mentioning aspects from different angles – so absolutely. Having a nuanced message is very important.

The comment about stated behaviour and actual behaviour: yes, it may be where I come from as a behavioural economist, and the fact that behaviour is the thing that matters most. It’s extremely difficult… I appreciate it can be very difficult to measure behaviour, but I think now there are techniques to try to do this more than has already been done. Sales data can be useful, no doubt about that – it’s part of the story. Part of the problem has to do sometimes with causality, and identifying causality.
You provided a reference which I will certainly want to check out. And similarly in terms of smoking guns about what may be the reasons why we have more or less of probably different media: absolutely, you’re entirely right. There may be, for example, different technical reasons that may explain why that’s the case, and you’re totally right: it’s something that needs to be taken into account. And finally, expectations, and Steven, if there are any other comments that you may have?

SW: I’ll deal with the more controversial one at the end, shall I, so we can talk about other things first. In terms of the DVD sales and so on, as bandwidths go up DVD sales have gone down. Has… I mean, I don’t know because this is relatively new stuff and there were very few studies we identified on TV. But have DVD sales gone down because Netflix is a thing? Free Netflix. I mean, this is the sort of useful thing, so if we can get more studies on TV then we can start explaining this stuff, you know. So I’m not saying there is or is not an effect because I don’t know – which gets us to the controversial thing.

I wasn’t particularly surprised because I didn’t have a bloody clue, is the simple solution to it. I mean, I wasn’t working in this area originally. I came from doing a PhD looking at why people took their medicines and so on. So it was the method I brought to this rather than prior knowledge, so I didn’t know anything much. So can I be surprised at the end of this when what we can say is well actually, the evidence is insufficient for me to tell you what was going on? No, not really! So, no; I mean, I was maybe surprised… There were things that did surprise me when I was going through this. I was surprised that coming from medical literature everyone always had to declare where their funding’s come from, if they’ve got conflicting interests. That wasn’t standard. I was surprised that there were so many working papers and so many that weren’t published and peer-reviewed articles and so on. This sort of thing surprised me. But having considered all of that and coming to the conclusion that we didn’t necessarily have a definitive answer? Not especially surprised, no!

Finally, to touch just on the morals, because I think this is one of the things I’m most interested in personally. So what we’ve tended to discover in the literature is that there are different moral frameworks when discussing this problem, both within and between cultures. So there are those of us who have absolute red lines on this: unlawful file sharing is theft. There are those of us who have a completely different appreciation, which is that unlawful file sharing is sharing. And it’s very difficult, so what we have is that people who have these different moral frameworks are actually having different discussions. You know, the guys who say it’s stealing are saying, “Well, this is the red line. Why can’t you say you’re not crossing this red line?” whereas the guys who have this more sharing view, they have the exact same: “There’s all this culture that people aren’t getting access to – why can’t you see that?” And what we’re having is they’re just two people with different views talking at each other rather than a genuine dialogue. And like I say, you’ve got to expect these views will change both over time and between cultures. So for example, we find a study in… the same authors used the same method in both China and the US, and what they found was that in the US there was much less file sharing of movies but they found a much greater displacement rate; so people who did unlawfully file share movies bought less movies, possibly as a consequence of that. In China they found that almost everyone was using unlawful file sharing but it didn’t actually have much impact on the number of movies that they bought. So, the impacts might be different, the causes might be different between cultures, and that may change over time.

Now, there’s an interesting debate as well in terms of whether or not the internet itself is a collectivist culture. I mean, that’s getting into a really sort of nebulous, difficult argument, so I’m going to leave that as an open question for you to fight amongst yourselves with rather than try and give you a definitive answer. But I find this interesting, anyway.
AB: The Chair is facing a taxing task. I think I’ve got two people who are in hot pursuit on a specific point. You two, yes. Were you in hot pursuit? I’ve got Roger waiting, but you’re well down the queue, Roger. You’re throwing light on something that’s just been said, yes?

RB: Well, on the presentation, on the report.

AB: Well, no, because I’ve got a queue. I thought you were coming in on that specific point. I’m sorry, but I think it’s important to keep a bit of order. You were next.

CH: Thank you. Christian Handke, Erasmus University. First, let me congratulate you on the systematic and transparent literature review that is clearly a step forward. I know a couple of other literature reviews and I can say, so I think it will inspire others to be expanded and you can always do more – that’s clear. I think everybody understands that and so do you.

One expansion I’d like to suggest, of course, is that sooner or later hopefully we will be talking about the effects on the supply of creative works rather than rights holder revenues, because I suppose we all agree as we work towards a welfare analysis and consider the changing costs of production that is a decisive issue. Now, I grant you immediately that that literature is so miniscule that the methodology you present here might not be applicable, but hopefully it will be in the foreseeable future.

What also happens when you produce something that is systematic and transparent is that you enable others to provide you with constructive criticisms, so let me give you one item. You make the methodologically clever point that you check your results, your search results, against the results of previous literatures reviews or the contents of previous literature reviews. And let me point out that you use three previous literature reviews there, none of which is particularly recent – the last one was from 2010. Secondly, two of them have practically the same authors, and the third is specialised on consumers. And so I would suggest you expand slightly and use more recent literature reviews, perhaps, in that area, just to double check. I don’t think it’s going to change much, but I think that would be a worthwhile exercise. Thanks.

AB: Thank you very much. Can you hold that for the moment? Tony, I think you were next, and then it comes to you.

TC: Thank you very much, and thanks for a very interesting presentation. I have to say, I’m immensely heartened by the discussion here this afternoon.

AB: So far.

TC: Well, because it seems to me that we’re all in the business of trying to improve and understand the evidence that is available. You may have seen, some of you, on the… Sorry, I’m Tony Clayton, I’m the Chief Economist at the IPO. You may have seen on our website we have just updated standards of evidence, which essentially mirror precisely what’s been said in the discussions here. We will take account of evidence if we understand where the data has come from, if we understand what you’ve done to it, and if it’s available for peer review. If it’s not, we won’t. And I think that to try and set those standards is really very valuable. I mean, this is a pretty tough thing to do, this type of review, partly because – as has been said – the area that you’re trying to look at is changing very fast. You know, 15 years ago it didn’t exist. So as has been said, the evidence from even five years ago is now looking at bit dusty. There are lots of causes for the changes that we see. It’s not just changes in what’s technologically available; there are changes in cost structures, there are changes in attitudes, and there are changes in incentives – and it’s the incentives I think that we need to think about. And I’d also suggest that we need, in looking at this whole issue, we need to understand both demand – which I think is where most of the work here has focused – and supply. You talked about Waldfogel’s work, and I’ve had a chance to see a private presentation of his next paper which will do, I think, precisely what you suggested it did, which is look at the production and the welfare-based consumption of music – music again, I’m afraid – and look at just what the effect has been on creativity. And it takes account of the fact that the cost of production, the cost of reproduction and the cost of distribution has fallen like a stone in the course of the last ten years. And if you look at these things over a long period, then it’s not just sales; you need to look at value added and you need to look at investment, creativity and investment – and that’s what tells you what the overall long term impact is. The sales number itself doesn’t necessarily do that.

AB: Thank you.

TC: Thank you very much. And I’m keen that people do focus on the production of new data. That’s why we spent half a million pounds’ worth of patent holders’ money on creating copyright evidence in the OCI Ofcom study last year. I don’t think we can do it again; I think the patent holders might complain. We’d like somebody else to step in and fill the gap, please, because we’d like that type of study to continue.

AB: I think IPR should be seamless. There’s plenty of overlap. A clutch of people waiting there – your turn now.

MS: Thank you. Maurice Samuel, PRS. In the report you say, “We conclude that caution drawing policy implications is warranted,” in terms of the cube. I’m just wondering, visually using it as a filter to identify trigger points for policy change, what should we expect the pattern of work to be? Should all of these spheres be distributed along the ceiling of that sphere?

AB: I think that question does deserve a quick answer and then we’ll move on.

DZ: Obviously, given all that has been said before, we may expect that the spheres may not be all the same throughout. And also, what a sphere does not tell you sometimes is the intrinsic quality of the study in different dimensions; it only captures specific dimensions. But certainly I would like to see studies that span all the way across, more of them; more with observed behaviour, more looking at causality – correlation is not causality. And it could certainly be useful to have more studies through time – that also relates to some of the other comments – because through time also can help unpick causality. That is what I would like to see.

AB: Thank you. You’re next and then Roger behind you.

RT: Hello, I’m Ruth Towse, I’m Professor of Economics of Creative Industries at Bournemouth University. Well, like other people in the room I’m terribly encouraged by the progress that has been taking place over the years. I’ve worked on the economics of copyright for well over 20 years and it was impossible to get any evidence or data or do these kinds of studies, and one reason for that was because industries were not willing to offer their data or let you inside. And that’s now happening and I think that’s very, very encouraging, and I think CREATe will help with that sort of development. Now, I’ve got a couple of perhaps slightly trivial points to make about this, and then a sort of big question, if you like.

The studies that you sample are very variable. I’m still worried about the fact that a lot of them are on students. They have very varied sample sizes. Well, why were they on students? They were on students because you could get data on students when you couldn’t mount a big survey on other things; so that explains a lot of this from the past. But I wonder if you’ve taken into account in your review the actual size of the samples in the studies themselves, because, in a way, relating sort of to the point George is making, I mean, some of these studies are bigger and heavier than others. Some of them take large samples and some take small samples, and so on and so forth. So I think that’s a relevant thing to just… That’s a technical question.

But my other question is much bigger. I mean, as a behavioural economist, do you actually have any sort of take on the, as it were, rate of illegal behaviour in society as a whole? I mean, what about speeding or paying television licenses? I mean, are people more illegal in their behaviour in relation to downloading than they are in relation to other things?

AB: Roger, do you want to throw your question into the field, and then we can get Daniel to wind up.

RB: Thank you. Roger Burt, retired from IBM, Intellectual Property Law. I’m very interested that this is based on drug study methodologies. Drugs studies are always put in a context, and what worries me here is that you’re treating file sharing as something in its own context. Is unlawful file sharing a substitute for previous low cost/zero cost lawful activity? If you were doing a drug trial you’d compare against previously known drugs and what the effect of them was. And previously… I’m an old person, I used to listen to Radio Luxembourg, Radio Caroline, Radio London which were unlawful…

AB: You’re showing your age, Roger!

RB: You’ve limited your decisions to conceptual framework, consumer engaged in unlawful downloading, purchasing illegal copy or doing nothing. But actually, what did they do before unlawful downloading? That’s what you should be comparing against, I think, and that would give you the context. Because actually, we used to accept that it was fed to us; we didn’t go and get it, it was fed to us. And does that change the way people behave? Thank you.

AB: A close relation there with Ruth’s question. I know you’re waiting and I’m going into your drinks time, but could we have a quick reply from Danny and we’ll come to you. It’s related to Ruth’s point, which is are people changing in their willingness to comply; you know, is it age-related and so on. I think it’s an extremely important point in relation to what happens to DVDs as well.

DZ: Our review was quite large as a class. So, one aspect that would certainly be interesting to do in our further work is actually to do a scoping review that compares this with other types of criminal activity. We haven’t done it, but if anyone wants to give us the funds to do it we will gladly do it. My expectation – but this is just an expectation, I may be proven wrong when the scoping review is done – but my expectation is that because of the social and moral acceptability of unlawful file sharing in certain segments of society and certain cultures, you would expect a higher incidence of such unlawful behaviour than, say, if you were talking about theft. So, Steven was talking earlier about China where unlawful file sharing is quite endemic in a way in which I would expect theft is not endemic. So that’s a brief answer to your question.

AB: I think… that says to me there’s more work to be done. I’m immensely interested as a former policy maker because people’s willingness to comply in the end is how you make law work. And in areas like drink driving I think there is a different social mindset compared with… you know, across society, compared with under-25s on UFS, and I think that has quite significant implications for making the law work.

DZ: But of course there is…

RT: The TV licence story might show us some way…

AB: …The TV licence story I rather fear is all about kicking the BBC and not modifying social behaviour, but I may be wrong.

DZ: But of course, I mean, and that connects also to another point that was made. There is heterogeneity in society, and different parts of society will have different social and moral norms as well as, in some cases, as your previous example, maybe technical awareness and ability and willingness to engage with particular media. So in that sense, talking about a prototypical consumer is always a bit deceiving; and I think quite a lot of research that would be worth doing, looking at different segments of the population in a more fine-grained way in terms of observed behaviour. Now, that’s not easy at all. It does require a lot of resources and that, of course, relates to the point you made about students. I’m an experimental economist, I do a lot of my own research with university students because that’s simple resource-wise. So I appreciate, of course, that that is the case; but we also have to appreciate that when it comes to issues like moral and social norms they may vary depending on the segment of population.

AB: Richard, do you want to queue-jump to come in there? There’s one waiting, but given your distinctive position I’ll let you queue-jump.

RA: Thank you so much. Richard Arnold, High Court Justice. I want to go back to what we were talking about earlier, which is the impact on lawful consumption; because so far, all the debate seems to have been on the volume of lawful consumption and whether that has gone down as a result of unlawful file sharing. My question to you is this: have you studied, have you found studies that examine not so much the impact on the volume of lawful consumption but the pattern of lawful consumption, the distribution of lawful consumption? Let me explain what I mean. Take music as an example, and let’s just assume, just for the sake of analysis, that lawful consumption remains constant; so the number of sales remains constant. It might be hypothesised that despite that fact the distribution of sales changed as a result of unlawful file sharing, so that whereas if you went back in time 15 years you would find that there were best sellers, not so big sellers, medium sellers, small sellers and tiny sales; now it might be hypothesised you’ve got a lot of mega sellers, no medium sellers and then the negligible sellers. So you’ve got a very narrow, sharp distribution as opposed to a broad, low distribution. Now, the reason I pose this question is it impacts on the question that’s been raised as to whether there is an effect on production, because if there is an effect on production a change in the distribution of consumption may be as equally powerful – or may not – as a change in the volume of consumption.

AB: Richard, what an interesting though, thank you. Hold it there, Daniel. There’s a very patient woman sitting there. I’m sorry, you’ve been most patient. Your turn.

MH: Thank you. I’m Monica Horten, I’m a Visiting Fellow at London School of Economics and I’m also the author of two books on copyright. I’d like to put my remarks actually in the context that we sit here today in Stationers’ Hall, and if these walls could talk they would have ghosts of disputes between rights holders and pirates. And just to give you one thought: in the 18th century the pirates were the Scottish booksellers. So, in that context I would like to make my remarks.

You’ve positioned the study in terms of unlawful file sharing, and I’m just wondering if this gives it a normative, even places a normative judgement on what you are doing. Immediately, when you use the word “unlawful” you place a normative view on the activity of the person doing the file sharing. And I just would like to twist the issue slightly here, and I wonder if you talked about the notion of “free” and the notion of getting stuff for free, if that would actually change the way you would look at the question. And I say that not in the context of people just wanting it for free because that’s what they want, but in the context of some reasonably established business models – like the advertising-funded business model for public television, for example – which is free at the point of consumption. So, free doesn’t necessarily have to mean bad or stealing; it means there is a business model, there is a big industry business model behind the “free at the point of consumption”. And I just wonder – and I pose the question only – that if you twisted the discussion and looked at it in that context, if you would come up with a different set of answers, and if that would enable you, for example, to explore the broken channel that I think was referred to by one of the previous speakers, because I do wonder if that’s really what the discussion is about. It’s about a broken distribution channel, a broken channel to market, and if that’s what really needs to be fixed so that people can get money, including the authors.

AB: Thank you very much. With your consent – I’m looking for a nod from Martin – I think if we have Daniel’s response at this stage and then we can move on to a more informal discussion. Are you happy with that, Martin?

MK: I am very happy for you to decide.

AB: How many more questions are waiting? One more there. That’s all right.

AG: Thanks. Andres Guadamuz, University of Sussex. It’s more a comment. Since 2002 I have been conducting a very unscientific study of my own on students, as it happens. I’ve been teaching a class on file sharing since that year, and every year I ask… I begin the class asking the questions, “How are you getting you getting use of it? What technologies are you using? Are you file sharing?” And the shift has been very interesting. You can sort of tell what’s been happening. You may think what you want about Law students, but the generally, most of the time, they want to do the right thing. And what’s been happening in the last four years is quite amazing: the shift in consumer patterns – their consuming patterns – is astounding. They don’t even consider owning files, let alone a piece of plastic, even remotely viable. Everything is streaming. I know this has been mentioned several times, but this year particular about 90% were saying, “We don’t own our music, we don’t own our TV; everything is streaming.” And this is something that I wanted to make sure that people are aware of, that this monumental shift has happened. And the business models have to reflect this. Thanks.

AB: Sorry, I was just having a quiet word with Theo about whether streaming offered better ways of attracting revenue, to which I got an, “Mmm” reply! Do you want to come in?

TB: You might be asking the wrong person because the person who knows about that would be the Performing Rights Society. I agree that there is a big move towards streaming. Streaming… it’s very interesting. Ad-supported streaming typically is what is made available; it’s just ad-supported streaming services on the PC. There’s a big desire to have that on a mobile phone, and very often that is subscription-only. So people are beginning to pay for a different kind of access, and what I think is observable – and there are some unfortunate elements to it – is that you’re paying for a music service, you’re paying for a user interface, you’re paying for the type of recommendation engine or interface with which you’re comfortable; the underlying content, it’s assumed that that’s free. Where people still are file sharing, they’re just grabbing content, often in indiscriminate quantities. The capacity of mobile storage, MP3 players, Smartphones is just enormous, so there’s no need to discriminate, there’s no need to choose; you just suck it up and then maybe decide what you want to listen to. So I think it is really important to begin to look at what are people listening to, on what platform are they listening to it, how are they getting access to it. I know that my son – he’s at UCL – uses a streaming service, and he wouldn’t have had that on his Blackberry had it not been for the fact that there was a discounted price for students. So there’s all kinds of stuff going on here, and I think you’ve asked a very, very interesting question that does merit further study.

AB: Which I think we can take due note of. Anybody else want to contribute or ask, or can I ask Daniel to bring us into drinks time? Daniel?

DZ: Okay. As I begin, let me thank everyone for all the useful feedback. I did feel that the discussion today has been very constructive and there has been a lot of very useful feedback. I don’t know, maybe I’m being optimistic, but I feel that broken dialogue… I haven’t seen much of a broken dialogue today, and I think that’s something positive.
So, trying to answer some of those questions and then letting my co-authors answer other of the points, in relation to the point about the importance of production, the supply side: that must be absolutely right. Again, the scope of our study had, by necessity, to be limited by an issue of resource. But undoubtedly there is scope there for using the same type of methodology to look at the supply side, and I think that would be very interesting. So, absolutely: that’s certainly work that is worth doing.

In relation to the point about the relationship… you know, how creativity is affected, the relationship between sales and creativity, and the relationship between these two and the business models, I think that’s also a fascinating issue. Yes, absolutely, we are in evolving business models. And yes, absolutely there is a question there about how the incentives for creative work change depending on the nature of their economic environment. So I was referring before to that work by Waldfogel showing a discrepancy between trend in sales and trends in consumption; and that, of course, points to – that was Waltfogel’s interpretation – to independent creators starting to come in and using different business models. So again, that is something definitely worth further work.

In relation to the point about heterogeneity, we’ve already discussed that so I’m not going to focus on that more. The general point about the language: we meant unlawful file sharing, if you want, in a legal way. We could have spoken about free content, but that would have made our analysis rather more general, and so there would have been an issue of focus of our analysis. Now, already we have to go through over 50,000 studies. I shiver at the idea of how many studies we would have had to look at had we considered all possible business models in thinking about the consumption of creative goods. Undoubtedly there is scope there for some scoping review that looks at different business models in the consumption of creative goods, and that may mean, of course, in some business models that those are consumed for free. I did refer in my presentation to other business models that change the gap between proprietary consumption for a price and the unlawful file sharing dimension. So, absolutely, that is something that clearly is there.

There were, again, comments about the technological pace of change; and yes, those are extremely important. Technology changes all the time. What we are thinking about in our framework is the way in which it may affect, for example, technical utility and the extent to which it may affect the ease of access. Now, it’s extremely easy, maybe, to download large volumes of files in a way that ten years ago, even five years ago, it was not in many cases. So, as a direction of future research, that is clearly the case as being useful. Nevertheless, as pointed out earlier by Piers, we also have to recognise that while we do have different economic and technological environments – and these are changing all the time – at the same time we, as human beings, even in terms of cultural change, it’s lower relative to some of those technological innovations. And so there is certainly feedback there to be considered, but it doesn’t mean that we can’t learn useful stuff by looking at work, say, of four years ago. I think that’s just an assumption too far. And equally, we do have the need of those inter-temporal studies very much, and those technological innovations can help with causality, in identifying causality in those inter-temporal studies; so they can actually be used actively to try to better understand the determinants of unlawful file sharing.
Now, I’m going to pass the microphone to Steven.

SW: Thanks. Just a couple of points. I suppose first of all, of all the criticism I never thought someone would say, “I wish you’d read more studies” – thanks for that! But to try and answer the very interesting question posed in terms of distribution, I’m going to frame this question with the huge caveat that there is not very much evidence on this – there’s a huge disclaimer upfront. But what there is is interesting. So what we find – again all on music – what we tend to find is the number of live performances and the price of live performances have both increased, so some of the money that was perhaps being spent on CDs is now being spent on going to see music live. Even more tentatively, I believe a couple of studies in Taiwan have found that the purchases of CDs have gone down but the purchases of clothing and wallets and other superfluous goods that go along with your identity as being a fan of that artist, they’ve also tended to go up. So there does seem to be some tentative evidence that yes, the distribution of what people are spending their money on has changed. And that may be because getting hold of the music… you know, the music has just sort of become kind of a loss leader to get you to start looking at all of the other things you can do to sort of build your identity around that artist. But again, I will end that discussion with the same caveat I started it with: very little evidence, very tentative, but it’s an interesting question and hopefully more people will look at it in future.

PF: Okay, I’m aware that I’m keeping people from their drinks, so I’ll be brief. Firstly, Ruth, I’ve got a little bit of preliminary data looking at subsets, looking at the effects of age, so I’ll catch you in the drinks to tell you about that fascinating research. And I would also like to reiterate the point that actually Alison raised. So I’ve been saying that I think the determinants are very important, but we need to understand why people file share; and I think your point that drink driving didn’t stop just simply because it became illegal, it stopped when the social and moral norms changed. And so we really do need to consider those factors as well and they’re highlighted in the review. Thank you.

DZ: So, to conclude: evidence, evidence, evidence. And thanks a lot for all the very useful discussion tonight. It was a pleasure to have it. Thank you.

MK: I won’t add anymore words. Just to thank the speakers, really, and the panellists. And I think it has been an encouraging, quite intense afternoon. Let’s continue over there. Thank you.