 Good afternoon everyone and welcome to another economic seminar on valuing the public domain. We are quite excited to have with us Chris Erickson from the University of Glasgow who was one of the co-investigators of a study that was commissioned by the UK Intellectual Property Office on precisely the topic of valuing the public domain. Just by way of introduction, as you know in the Intellectual Property System there is this inherent balance that policy makers try to strike between on the one hand giving incentives towards creative and inventive activity And on the other hand recognizing that creative works and inventions have public good characteristics and they should be disseminated as widely as possible. For that precise reason, Intellectual Property Rights are usually time bound. If one leaves aside trademarks and wants exclusive rights expire, in a sense they move into the public domain and they are free for everyone to use. And for economists that trade-off is quite obvious and I think it really goes back very much to the foundations of the Intellectual Property System. But it has always been difficult to quantify how important really is the public domain and how is it actually used by creators and by society at large. And for that reason we really welcome the study which originated in efforts I think by the UK Intellectual Property Office to improve the evidence based on which policy makers in the United Kingdom can make decisions. But it certainly has implication and is of interest going beyond the UK and in that sense we are glad that Chris has agreed to present the report that they have come up with here at a WIPO economic seminar. I probably should mention and you might elaborate on the precise origin of it. Chris is the one who is presenting this but this was a co-product with his fellow investigators at the universities of Glasgow and elsewhere and maybe you can say one or two words about that. Chris asked that we leave aside all the questions for the question and answer period. We will make sure that there will be sufficient time. Obviously if you have extremely pressing questions for clarification you can always raise your hand but let's leave all substantive questions to the end if that's possible. So with these introductory remarks let me give the microphone to Chris and I suggest maybe that you present for 40 to 50 minutes and then we still have time for questions and answers. That sounds great. Thank you Karsten. And if I do go over time which I will try to keep myself to time. I can indicate. Thank you. I'm going to stand if that's okay so I can see everybody in the room hopefully. Well thank you very much for having me here and thank you all for coming. Karsten informs me that this is a good turnout and indeed it looks like a great turnout so thanks very much and I'm really excited to share this research with you and get your input and ideas about the work as well. As Karsten mentioned this is a work of collaboration with myself and Professor Martin Kretschmer at the University of Glasgow as the lead investigators and co-investigators Paul Hild from the University of Illinois and Fabian Homburg and Denisha Mendis from the Business School at Bournemouth University in the UK so they're not here to present with me and I'll be presenting some of their findings and research on their behalf so please forgive me if I don't have all the answers to your questions. Some of the details were worked out in collaboration with them. Okay so we'll move ahead fairly swiftly to keep everything to time. I want to leave a good amount of time at the end for questions and comments. First of all can everybody hear me? Is it coming through okay? Okay thank you. So this research has been summarized in an 80 page report which was published by the Intellectual Property Office in the United Kingdom in March and you can download it, the PDF in its entirety from gov.uk so if you're interested in having a look at the report you can either contact me and I'll be happy to forward you along the study or you can download it yourself from their website. And to provide some background to the study and why we did the study this was commissioned essentially by the Intellectual Property Office to generate new evidence and new findings about how the public domain might contribute value to the economy and the society of the United Kingdom. And they gave us the following brief which was firstly to map the size of the public domain and the frequency of its use, a somewhat impossible task as a matter of fact. Other people have applied different methodologies for trying to get a handle on the size of the public domain and an academic named Rufus Pollak has done a lot of excellent work in that area mapping the size of the public domain. The second part of the brief was to analyze the role of the public domain in direct and indirect value creation for UK firms and the wider economy. And this is part of the political climate of the moment in the United Kingdom which is quite interested in this notion of the creative industries and the health of the creative industries and trying to promote innovation and growth in that sector. And this is in the context of an economic downturn and also in the context of a country which is a net exporter of cultural products so it's very much comes out of the UK political system if you like. The third part of the brief was to assist UK media companies in identifying and developing business models that draw benefit from the public domain so it has an industry focus. So how well we managed to achieve these objectives I leave that to you to decide. Our approach was to really select three very specific markets and domains of activity and try to get a handle on how the public domain contributed value in those particular areas. We didn't try to enumerate the value of the public domain for the entire economy which I think is a task really beyond our research team. But we tried our best to understand how the public domain works in very particular case studies and I'll talk about those in a moment. The first challenge was to define what we mean by the public domain and journalists have a particular definition of the public domain which is information which finds its way out of the private sphere and into the sort of the public sphere of knowledge. We're interested in the copyright public domain so these are works and ideas and bits of information which are free for uptake by all. That's our short definition but we built on literature. There's an extensive literature actually in terms of trying to define what we mean by the copyright public domain and trying to figure out where its boundaries might actually be. So in this table which I apologize is hard to see these are some of the proposed definitions of the public domain and the ones that we've chosen to adopt and the ones that we did not adopt for this study. So the main group of works which we consider are works which are out of copyright either because they never qualified for copyright protection because they were created in antiquity or because the copyright term has expired. And in the UK that's mainly works for which the author died more than 70 years ago is the best way of saying it. But we also include works which do not qualify for protection. So here we're interested in the idea expression dichotomy. There's lots of aspects of a work which are not actually protectable by copyright and which are actually able to be taken up and used as inspiration for other works. But the line at which something goes from an idea to an expression and becomes an infringing use is determined in the courts. It's been proposed that we consider privileged uses of work. So these are uses which are permitted under copyright exceptions like the exception for criticism and review. And in a sense that those types of uses are available to certain members of the public in certain circumstances. But we did not include privileged uses in this study because they're not available to all users at all times. And we're interested in commercial use as well as noncommercial use. Then similarly there are permitted uses of works permitted by private ordering schemes like the Creative Commons and OpenGPL. And we do include some of those types of works but only the works that don't specify who the downstream user needs to be. So we don't actually include noncommercial licenses and we don't include aggressively viral licenses because those actually restrict what can be done with the work in downstream. We only include what we call free and open types of licenses. Finally we might think about works, tolerated uses of works or uses of copyright works which are actually unknown to the rights holder as a kind of pseudo or quasi public domain. So fan fiction which exists outside of the purview sometimes of the rights holders. Does this constitute a public domain? Well no, from our perspective because those tolerated uses can become untolerated if the rights holder emerges and discovers the use. So we did not include that type of infringing use in our study. So the three empirical cases which I want to talk about today really briefly are first a study of small and medium sized creative businesses in the UK who've used public domain works successfully in a commercial setting. We performed interviews with some 23 of those firms and examined their managerial decision making around uptake and use of the public domain works. Second crowdfunding platform we looked at Kickstarter pitches which incorporated public domain inputs in their pitch and we have some quantitative data about that which I'll share with you. And thirdly we looked at the inclusion of public domain imagery on Wikipedia sub pages to try to estimate the total value to Wikipedia and society that the public domain availability of those images add. So quickly moving on to study one which is on the uptake of public domain materials by UK firms. So here we're interested in the factors that prompt decisions by creative managers to take up and use public domain materials in the first place. And this is responding again to the brief from IPO to get a grip on the type of business models that small firms might be employing when they use non-excludable, non-rival public domain inputs. And we look at their strategic decision making in that context. So we generated a sample of firms by searching backwards on known public domain works like Sherlock Holmes, like Darwin's publications. We have a long list of public domain works which we searched on. We found commercial exploitation and then we sought out those firms and sought to interview them. So in terms of the academic context of this research this fits I think in two broad literatures. The first literature is a sociology of media work and management studies which looks at how creative firms manage themselves and how they view their use of intellectual property. And there's an interesting kind of prevailing debate in this literature about the make or buy decision in small firms with Hotho and Champion pointing out the dilemma that small studios face when they either decide to do work for hire which essentially is building on someone else's in licensed IP essentially as having no further intellectual property right in the finished product as you would do if you created the video game for Toy Story 3 for example. You wouldn't necessarily have any secondary rights in that work. Or to create original works from scratch which has been highlighted as one of the aims of creative studios because these people want to express themselves creatively and do creative work. We propose a third option which is that these creative firms actually might use and build upon work which is in the public domain. So they don't have an excludable right to the source material but they might build upon it anyway for a variety of reasons. And the other literature which we think contributes to this understanding is literature on user led innovation on private collective innovation and on the management of commonly held resources in other types of innovation industries. So we're essentially proposing to think about this decision within the firm as similar to other types of open innovation decisions in other firms in other industries and we're looking for a fit or not with our propositions there. Okay so why do the SMEs actually take up and use work which is in the public domain? There are a variety of incentives that we identified. One of the important incentives is that the public domain work already has a pre-existing audience. So in a sense rather than coming to the market with an untested idea the small studio can come to the market with something that already has wide appeal. And that was expressed by this company that did a digital interactive theater adaptation of Dracula saying it's so well known, it's lodged in our imagination, it connects with our subconscious and it was really a love of the book of the original material. The production seemed to connect with people and it's good to capture the magic of something. So you see a lot of affective language in here and you see a lot of references to the power of a pre-existing audience and that comes up quite a bit in our qualitative research. Here's another company which uploads fairy tale songs or sorry what do you call them, nursery rhyme songs to YouTube of which the nursery rhymes are in the public domain. And this person worked in the music industry in rights clearance for about a decade before she struck out on her own and she comes to the table with pre-existing knowledge of rights clearance. And she voices this other incentive which is that working with public domain materials doesn't carry the transaction costs of tracking down the rights holder and asking for permission. So here she was able to build upon her knowledge of what was in and what was out of copyright in order to essentially lower her operating costs. But the downside was because it was non-excludable she quickly found a raft of imitators which joined her on YouTube also uploading nursery rhymes based on public domain songs. So these are largely the two incentives on the incentive side. In terms of the protectability, so the problem of the protectability of the underlying inputs, here's a medium-sized video game studio discussing the problem of protectability and saying, well, you know, I think that within the video game sector you're in a different position compared to film and TV in that you can't really copyright gameplay ideas anyway. So really the only IP that we're going to create is the product that we make. There's nothing stopping anybody saying, oh, they did that really well. We'd really like to do that and make our own Jack the Ripper documentary and, you know, reusing the materials and heavily taking the gameplay style and things like that. So for this interview subject he already deals with the problem of protectability in a copyright industry that in the video game sector he and other studios already struggle with protecting the core ideas of their gameplay. They can protect the source code as a literary work but as we were talking about earlier, protecting the essence of the game, the look and feel and the interface and what not is already a challenge. So they already face a challenging IP environment. An important theme as well emerged in terms of searchability and quality. So the firms that did successfully commercialize public domain inputs really relied heavily on the availability of the public domain work either with a memory institution or with some other archive or museum or place where they could locate and actually use the public domain input. And this person was an author from Bristol who is quite keen on the annals of Bristol in the 17th, 18th and 19th century by John Latimer and he says they're very expensive to find them in print but all of a sudden they're available online through archive.org, an internet repository. And now anyone can go and use this material and draw inspiration from it. And he talks about an example of that where a bunch of steampunk authors showed up at his house one day and asked him about the history of Bristol and he was able to point them to this resource which they used as inspiration for a product which is this volume of short stories which was published last year by Wizard Tower Press on Bristol, a sort of science fiction novel if you like about Bristol in the 19th century. But it draws on this public domain material which is the annals of Bristol in the 19th century by John Latimer and here you can see it sitting on internet archive and you can flip through and actually access the digitized material. So again, you know, searchability, quality, metadata, you know, availability of public domain works is really something which was stressed by these firms. So we analyzed this using a business model and a strategic management orientation and we've built on the concept of the value chain, the generalized concept of the value chain to examine where precisely in the activity of value generation public domain inputs actually fit in the cases of the companies that we use. And we found four particular kind of business model types, if you like, following from Badden Fuller and Morgan in their recent call for business model abstraction. And so the first type of business that we examined what we call portfolio builders, these are essentially firms which are still working in a work for hire mode. They're essentially partnering with a public institution like a university or an arts council and they're being commissioned to create a project of some public significance. So this might be the 150th anniversary of the publication of Darwin's Origin of Species or something similar to that. And they're being commissioned on that basis. Their competitive advantage, if you like, comes from their creative capabilities and their investment in the quality of the creative goods which they make and which they hope to win future commissions. But it's not a particularly sustainable approach unless they're able to connect the dots and line up for their commissions. Interestingly, the user is quite important in this model because it's the user who has the interest in the public events, if you like. It's the local community where the firm is located where they're essentially serving a latent demand in the community. And the value proposition is meeting the public demand at the appropriate place and time and context. Technology platform developers essentially have proprietary IP in the underlying technology platform. So there's one company we looked at called Onilo and they sell whiteboard animated software system for primary schools that let teachers show animated school books to children. And they offer a subscription based model and they have a catalog of copyright but also public domain works which you can buy on this platform. The public domain works were some of the first works that they offered because they were free and they were investing mainly in the technology platform, not in the content. So in that sense the public domain material was a placeholder and later they invested in licensing copyright books and materials. So they are a hybrid of both copyright and public domain work. And here they're integrated across the value chain, right through marketing and distribution, right to the user, more so than some of the other firms. Fan community engagers, these are people who are, for example, fans of a literary work like Sherlock Holmes or fans of Shakespeare who essentially build a business on the back of their involvement in the community. And here there's tremendous importance in terms of their relationship with the users and the other fans who they serve. So their competitive advantage actually emerges from the relationship that they're able to build with the fan communities that purchase their content. Because essentially, again, it's non-excludable, anybody could come in and offer similar content but their advantage comes from having a pre-existing relationship with a fan base. Finally, these entrepreneurs, these are traditional media companies, they invest in both copyright and public domain and original works. And they're very good at spotting market opportunities, spotting demand and offering products to meet that demand. They're very sophisticated in their knowledge of IP, they will license when they need to, if they don't need to, if something's in the public domain and they see an opportunity they will exploit that as well. And they, when it comes to protecting their investment in public domain works, their strategy tends to be in terms of investments in marketing being bigger than others and also being first to market with quality products. So when we examine this qualitative data in respect of this theory from Von Hippel and Von Crow on private collective creativity, I think we find some interesting connections. For Von Hippel and Von Crow and the subsequent authors, the incentive for engaging in this kind of private collective innovation, innovation where you don't own the underlying inputs, might be, for example, to diffuse an innovation which increases the innovator's profits through network effects. And we see this somewhat with the technology platform's strategies. We might also see a strategy of incorporating free software to help fulfill the credible promise of a prototype. So putting in placeholder public domain content to show investors what the product will look like is a strategy that we see the firms engaging in. We see Von Crow and later authors, Sturmer and all, talking about lower costs of IP protection. There's nothing to protect other than the underlying technology IP, but that's a different system in the video games sector. And an increased speed to market, which we do see occurring across the board of firms that we've looked at. One thing that we need to emphasize though is that this all comes with cost. It's not like public domain materials are just free. There are search and integration costs, as Haifleger and colleagues point out. There's difficulty in differentiating from the competition. When you're engaging with user communities in a close relationship, there's the difficulty of holding back certain proprietary business information and other strategic information. The cost of managing communities and generally organizational inertia with respect to working in this mode, which is quite different from working, for example, with a fully licensed IP. How are we doing for time? We have at least another 25 minutes. Perfect. Okay. Great. I'm sorry if I've been rushing through. There's a lot of material in the report. I'll try to slow down a little bit. This next study is a quantitative study of pitches on Kickstarter. A quick show of hands. Are you all familiar with Kickstarter? Yeah. Yeah. Okay. Kickstarter is a crowdfunding platform. It offers the ability for strangers to fund products in the pre-production phase. And the two types of products, if you could broadly break them down into physical design products, which might be protected by design right or patent, and then copyright products, like essentially media products. We're focusing on media products on Kickstarter in the first quarter of 2014. We developed a computer-assisted extraction tool, which gave us all of the products in four categories for those three months in the first quarter of 2014. Those were comic books, theater, publishing, which is essentially print publishing, and video games. So a mix of quite different sectors represented. And we looked at those works using the following perspective. We imagined that a pitch creator, that somebody who wants money from the crowd, when they go to Kickstarter to try to raise money, they have essentially four different types of project they could pitch. They could develop and publish original content. They could obtain a license and republish an existing copyright work in a new setting or new format. They could reuse work from the public domain. You see our interest in that. Or they could do some combination of the three. They could remix copyright work with some new public domain inputs. And in fact, we see some evidence of that happening as well. So we think we've covered, at least in the media products, we've covered the different approaches that pitch creators could take. And we want to know to what extent the IP status of those pitches influences their chances of success in the funding that they receive. Does it help if you have a copyright license when you go to Kickstarter? Essentially. Most of the existing literature, and there's a growing literature on crowdfunding, focuses on the problem of information asymmetry between pitch creators and pitch funders or backers. Backers don't have very good information about the likelihood of the seller to deliver a product. And that's sort of the danger of crowdfunding. When you invest, you don't know if the funder is just going to take all the money and run off to Thailand or Cuba or whatever. Or whether they're going to deliver the product. Or even if they do deliver the product, whether it will live up to expectations. And here's a bunch of recent literature on that. Essentially, what we get from that literature is that the importance of reputation signaling. So these signals that pitch creators could send to backers to signal their likelihood of succeeding. These are things like your social network profile. How many friends do you have on Facebook? Do you trust people that have a lot of friends on Facebook? Well, some of the empirical evidence suggests that is important. But we think that the IP status might work in a sense as a kind of reputation signal in the transaction over the crowdfunded good. Another problem is that these are non-repeated transactions. It tends to be a hit and run where you just get funded once. You don't keep coming back to Kickstarter the way people come back to eBay over repeated transactions which would allow you to rate sellers, for example. You can't really do that on Kickstarter. There are low incentives for individuals to do due diligence because the amount that you back is very small. You can only give ten dollars or ten pounds or what have you and you have low incentives to do diligence on that amount of money. In general, and we spoke about this earlier, the ethos of crowdfunding is to fund people who are not mainstream. The point of Kickstarter is ostensibly to fund new, original, untested, independent projects. These creators, just by definition, are likely to be somewhat riskier than corporate providers. One of our hypotheses is that IP status of projects serves as a quality indicator to potential backers, increasing their confidence in the quality of the good. Here we're looking at funding level, which is one of the variables that we were able to extract from the pitches using our tool. Our second set of hypotheses relates to just the transaction. So essentially when you go on Kickstarter, you ask for a set amount of money and if you raise the money within the threshold period, then it's a success and you have to deliver the good. But you've set the price as the seller. So in a sense this success is independent from the total amount of funds raised because I could go and ask for $500 and have a better chance perhaps of raising my $500 than somebody who goes and asks for $25,000. So here we're interested in simply whether or not the IP might work as a lubricant to help these transactions essentially meet with success. So our data consists of nearly 2,000 projects which were pitched, which were just the natural flow of sample of projects during those three months which were successful, unsuccessful or suspended for various reasons. Again it includes those categories. And of course we excluded pitches where there was no copyright good. Like for example theaters that were asking for new lights would not have been included in our sample. So the variables that we collected on were first of all success, whether the transaction succeeded and the funder got funded within the amount of time. The amount of funds raised, the number of backers attracted to the pitch, the type of media, the source of the intellectual property. So we used human coders to determine the presence of either licensed or public domain intellectual property in the pitch. And this was actually more straightforward than we thought. Pitches were very explicit most of the time about what they contained, what the project would contain. And then also the public domain rationale. So when public domain works were used, how were they used? Was it because they had expired? Because they were never protected? Because they didn't meet the threshold of originality? Or because they pertained to a Creative Commons license? That is a mistake, should not say exceptions, it's Creative Commons. And when it was licensed, when it was used a third party copyright work in a licensed way, whether or not the license had already been sought from the license holder. Or whether they said they would pay a license after they raised the Kickstarter money, that's obviously important to know. We also wanted to, as much as possible, control for the expertise and reputation of the pitch creator. So we recorded information about their status in the wider community and their previous experience, the size of the team involved, the gender of the creator or creators, and whether or not the pitch was accompanied by a video which has been shown to be important in previous studies. So what we find just descriptively is that 83% of all projects were original, primarily original. 11% contained third party copyright work, 6% contained public domain, giving us our total of $1,993. Now the concentration of public domain works differs across with a higher proportion, interestingly in graphic novels and comics, as well as theater, and a lower proportion in video games and publishing. We find a high degree as well, so this previous simply records the primary status of the pitch. This shows us in the light purple the amount of our sample which constituted work which actually contained multiple inputs. And this could be original work by the author, third party copyright work, or in 12% of those cases, original work, public domain work, and third party copyright work. So we see a quite high degree of transformative recombination. Here's in response to a hypothesis in Group 1 related to the overall funding. We do find support for the hypothesis that presence of public domain material raises the total funds attracted to projects. So this shows the red line is the mean across all. The first group of a bar chart is all projects. The bottom is those which only include public domain works and you can see a slight increase in total funds raised in certain categories such as in video games. But that does not equate to success as we've already covered. So when we look at a binary logistic regression on price, this is funds raised essentially. We see significance for projects where there was permission already sought from the rights holder and use of a third party work. This is significantly associated with a higher funding level achieved. When we look at public domain works compared to the reference category of original works, we see also significance for works where the primary input was public domain for a higher funding level achieved. Here's the binary regression for success and we see significant influence of both public domain inputs and third party copyright licensed work in predicting success of Kickstarter projects in our sample. So seeking permission is significantly important here for those projects which do use third party copyright work. So it does appear as though that is acting as a signal. We see that the most significance is in the theater category for the success rate of public domain projects. And I think this essentially reveals the particular effect of Shakespearean theater in this category that there's a tremendous affection for Shakespeare in theater production. But here we show generally the odds plots across the four categories and what we're interested in here is the first, so I don't want to shine a laser in anybody's eyes, but if you look at the top plots in all four quadrants, that relates to the public domain success rate and the red line being just one. So we see a higher success rate across comics, theater and games, but not significantly in publishing for a success rate of works which incorporate public domain inputs. So I think that's quite interesting that both public domain and copyright works were significant in likelihood of success on Kickstarter. And I think that's somewhat counterintuitive because Kickstarter's ethos again is about untested and original work that actually known public domain works and known copyright works should perform so well there. And we see difference across the categories which is interesting and I think this warrants further study. We suggest that one of the reasons why the publishing category does not seem to be significantly influenced by the public domain inputs is that most of the public domain works that we're referring to are already books. The vast majority of public domain work is literary work because literary works have had time to expire. So it could be that the funders on Kickstarter are rewarding adaptation and recombination by seeing those literary works adapted to different genres like video games and comics. That might be what we're observing. But we did also find a significance of importance of the status of the pitch creator and their reputation as well. Moving to the third study which examined the value of public domain photographs on Wikipedia. This was led by Professor Paul Hild who's a professor of law and economics at the University of Illinois. So the methodology here was to essentially build a data set of Wikipedia pages relating to people whose lives took place either before or after the line separating works which would be in the public domain and works which would be in copyright. In the United States that's more or less equivalent to 1923. And remember here that although we're using authors, composers and lyricists as the sample of pages, the copyright work in question is not relating to their work as authors but relating to any photographs of them while they were alive. So essentially by knowing the birth and death dates of a large sample of creators who are likely to have Wikipedia pages, we can figure out how important the public domain status of the imagery about them is to populating their Wikipedia pages. And if we imagine that Wikipedia pages are enriched by the presence of illustrations and graphics and photographs, then we essentially can get some sense of the value that's contributed to Wikipedia by the availability of those public domain images. So essentially the first thing to see is simply the distribution of 362 pages on Wikipedia relating to famous authors that the availability of images about them declines over time as we get closer to the modern period. But this is counterintuitive because there are more cameras in the world, there are more photographers in the world since 1890. So why are there less photos available on Wikipedia of these famous authors? Well we think it has to do with the copyright status of images that might have been taken while they were alive. And the reason why it drops off after 1900 is because those authors would have been in their 20s in 1923 and so photographs of them when they were working as authors are likely to still be in copyright and held by the original photographer of those creators. So that's why we think we see an actual decline in the amount of availability. And when we look at the legal status of the images of authors, one of the great things about working with data from Wikipedia is that page builders, editors on Wikipedia, when they upload images they have to be very judicious about explaining the licensing of the image that they use. Wikipedia doesn't like to use copyright images, they can rely on fair use but they have very strict, they place quite strict restrictions on editors about when they can make fair use of copyright images. So some 80% of the images that were available of all the authors we looked at were actually public domain images. In terms of the justification that editors gave when uploading images to our sample of pages, half of them were uploaded because it was an image which was in the public domain due to copyright expiry. There are other ways that the work could have found its way into the public domain however, such as being dedicated by the creator through, for example, Creative Commons type attribution or through reliance on fair use or through some formalities in the US copyright law about which I'm not entirely familiar in which Paul Hild understands better than I do. With respect to registration of copyrights and there was a change over a period of time. So we clearly see that the public domain status of images over time does seem to increase the likelihood of an image appearing on Wikipedia. How much value does this add to Wikipedia or by extension to society in terms of other perhaps commercial information service providers? That might be like if you can imagine a commercial version of Wikipedia that could make use of the same imagery in their own business. One thing we've used here to try to estimate value is an interesting phenomenon that we discovered which is that many of these images, this is Rudyard Kipling, are available for licensing from Getty and Corbis. They offer essentially the identical image as commercially licensed work. But the work is also available free on Creative on Wikipedia Commons and tagged as out of copyright due to expiration of the copyright term. So we found this puzzling but it gave us an idea about how to perhaps place some kind of commercial value on the used imagery on Wikipedia. Which is to use this yearly digital license fee as a benchmark for trying to establish some form of cost savings to Wikipedia by using equivalent images which are in the public domain. So one important thing here to determine is whether or not the copyright public domain, the term expiry public domain is the only source of a photo compared to for example Creative Commons. Because if we're talking about cost savings, we have to imagine other ways that Wikipedia could obtain the image, right? So if the author is still alive, we could pay a license fee to Corbis for a digital image of the author. Or maybe we could get on a plane, fly to Canada, track the author down, take a picture of the author with their permission and then bequeath that image to the public domain. That's another way that an image of that author could potentially enter the public domain. And the costs for generating those sort of Creative Commons licensed images depend a lot on the nature of the subject matter that we're trying to photograph. So this is very much just a rough estimate because Wikipedia covers such a vast amount of different topics. Looking at a random sample of 300 Wikipedia pages in all shows that about half of them contain images and 87 of those containing images cite the public domain, the copyright public domain as the source of the image. So Paul Hild taking the license fees which are available for images from Corbis and Getty which are identical images to the ones which Wikipedia is already using. And multiplying those across all English language Wikipedia pages where public domain images could be used finds a total cost savings equivalent license of 200 million US dollars. Now of course this is just essentially a thought exercise but that is one way that we could calculate the value of those public domain pictures. Another way to think about how to calculate the value is to think about the value that the use of those public domain images actually adds to Wikipedia as a resource. And one proxy for this value is visitorship. So if we believe that public domain or images of any kind add usability to a page and attract more visitors then perhaps we can calculate the total amount of visitors which come to Wikipedia as a result of the added images which they've used from the public domain. And that's what we tried to do by using a matched pairs methodology which controls for the different levels of popularity and interest in the authors, lyricists and composers that we used in our sample. So we tried to pair unsuccessful authors with one another and popular authors with one another. And we used two ways of doing that. One was to take their number of reviews on Amazon.com as a measure of their overall popularity. The other way was to look at their lowest traffic ever received on the author pages and we call that their sort of ambient popularity. That's their popularity if no university course is happening that week. They haven't suddenly been featured in a film because Wikipedia's traffic is quite volatile. So by taking the very, very lowest level of traffic achieved we think that's one way of pairing them off. And what we find here, so apologies this graph is sort of intuitively reversed, but we find that for author pages where there are images, which are in the public domain, we find depending on whether they're an author, a lyricist or a composer, we find a 17 to 19% increase across the sample of authors who received an image on their page since 2009. Paired with authors who never received any image on their page. So authors' pages that received images which were in the public domain since 2009 show a 17% increase in traffic received. And by calculating this all out according to the percentages of presence and the 300 random pages of Wikipedia that we sampled, we come to a total yearly contribution of $33,900,000 in terms of the additional traffic which Wikipedia enjoys as a result of the presence of those images. And that dollar value if you like is essentially the equivalent that a commercial web operator might hope to earn from advertising revenue on a per-click basis from the added traffic that their website would receive. And to use that we calculated, we used the figure of 0.053 cents per visitor, which is a widely estimated value per visitor on commercial websites, although that of course fluctuates as well, so this is also a rough calculation. But the purpose of this was to give you a general sense of two perspectives that we could have on value in Wikipedia. And one further thing which we're quite interested in understanding more about is what happens to all of this Wikipedia content downstream of Wikipedia. Because the website has essentially cleared the rights and some of this material by licensing it on a free and open source basis, and because it incorporates free and open licensed images or public domain expired images, then it's free to be taken up and used by journalists, software developers, other innovative companies. We don't have any sense of the size of the downstream use of the Wikipedia resource, but we think if we could extend this finding to looking at what other users are building upon Wikipedia, we could perhaps come up with a third way of calculating the value that this contributes. Okay, so to quickly close, I hope I've convinced you that the public domain and its presence is important to innovation and creativity, that there is uptake of it taking place and that there's lots of potential for innovation. I think the finding suggests that there's a lack of knowledge among some users and practitioners and that there could be policy implications in terms of improving information systems, improving archives, searchability, metadata, and other forms of collecting and sorting and making available work which is in the public domain. I've argued elsewhere that I think what this shows is that the public domain is constituted by use. So the use constituted public domain means that you can change the law to allow, in theory, a whole group of new works to enter the public domain, but unless the businesses and actual users can locate and make use of that work, then it's not really in the public domain. I mean it's theoretically in the public domain, but not unless it can be used in new innovation. It's interesting that across at least the first and second studies, we do find that use of public domain material seems to go hand in glove with licensing in of copyright material. Successful firms use some public domain stuff, but they also pay a license fee and they tend to be improving their knowledge over time about the opportunities which exist in terms of licensing in other intellectual properties. So this isn't simply a case of taking money away from rights holders and giving companies stuff for free in the public domain. It's actually about companies that are quite strategic in terms of their use of different inputs when they can. And finally, the overall point that I think ought to be taken from this study is that previous accounting for the economic size of, for example, the copyright industries is missing the important contribution made by public domain and other alternative inputs to creative production. And that if we're going to engage in discussions about the overall value inputs and outputs of the creative industries for the economy, we do need to consider not only the work which is copyright and excludable, but also the non-excludable inputs and outputs. Thank you very much. Thank you, Christopher. One thing that I completely forgot to do at the outset was to properly introduce you. I don't know why I did this, especially because you have, I think for any economist, the enviable title of Lord Calvin Adam Smith Fellow in Social Sciences. So apologies for that. Chris Erickson holds a PhD in political geography from the University of Washington in Seattle and his research interest centers on intellectual property in the creative economy. And he's affiliated with CREATE, which, understand, is a consortium of researchers which is led out of the University of Glasgow that studies the creative economy. So apologies, I should have mentioned these biographical details at the outset. But let me open the floor. Let me first of all thank Christopher for not only presenting one studies, but actually three studies in one seminar. I think they all gave really interesting and different perspectives on the value of the public domain. But let me open it to questions and answers, comments. Please, if you want to raise a question or make a comment, please use the microphone and if you could identify yourself, that would be appreciated. Somebody wants to break the ice? Marcella? I think we have to cut. Can you? Oh, okay. There we go. Well, first of all, I'm sorry, I was a little bit late and thank you very much. Waipo and Arthur for this study. My name is Marcella Baiba and I'm from the permanent mission of Chile to WTO and WIPO. Well, I would like to thank you for your study. We have a very deep interest in a principal basis to public domain, but I believe that this study clearly gives us more sound information regarding the economic value as well that it might have for us as countries. So this is very important. So a lot of findings are very interesting. The combined models of exploitation that you could identify in one of the studies. Also the low knowledge of corporate principles from users. I think that's something that's across the board, not only in public domain, but in general, and maybe the role of libraries and archives that I also think it's very important, that was highlighted, might be an answer for us as government, as a door to how probably engage with communities and promote the sound information. I agree with the idea that the use is one of the main concepts that we should aim for, that the identification, the status of the material is definitely a challenge. This would require local efforts obviously because in each country is a different situation and I know personally that we've been struggling in Chile how to solve this situation to give clarity and this is something that we're still working on. But I thought that there was also an interesting suggestion in terms of improving access regarding the more sound databases. And I would like to ask you if there are any examples. I think that in your conclusions you cite a British library mechanical creator initiative, but I don't know if there are other examples that you could cite or probably develop a little bit on that so that we can share that with our capitals. Thank you very much. Thank you, Marcella, for your questions and for your comments. Yeah, so with respect to the existence of initiatives in the United Kingdom relating to databases of public domain works, this is rather new. I mean in the UK we have an ongoing challenge with respect to digitizing holdings of memory institutions because these holdings often contain copyright material and the rights clearance costs represent insurmountable costs when we're talking about hundreds of thousands of individual works. There's recently been initiatives in that direction with respect to orphan works legislation in Europe and in the United Kingdom we have a licensing scheme through the Intellectual Property Office to try to offer a license for institutions that want to use a work after conducting diligent search which still resists diligent search and which is essentially an orphan work where the copyright holder cannot be identified. So that's an innovation that we're studying actually at the Create Center looking at how it's going to actually play out and whether or not it's suitable for all types of institutional users because it has a small but still significant fee. On the positive side it does appear to allow commercial exploitation of the work during a seven-year period after the license has been paid. But it's an interesting innovation. We're interested in learning more about that. I think the Mechanical Curator Project is an effort by British Library to make available images which are inside of books, illustrations and they automatically digitized a huge amount of books in the British Library but all of those books are out of copyright. They're public domain books and so the Mechanical Curator team are fairly confident that the illustrations contained in those books are also out of copyright but they've made them available on Flickr in a searchable way and then they've used the crowd to add metadata to the digitized illustrations in the books. And I did something fun with it a month ago. I made a font using, you know how at the beginning of an old book sometimes the first letter is quite baroque and ornate. You can actually string those together and make words out of the old typefaces of these old books. So there's lots of interesting, you know potentially interesting downstream uses that would come out of that project and we are certainly interested in those types of innovations. And I'm happy to speak with you more offline about Chile because that sounds like a really interesting case study. Oh I'm sorry I got to turn mine off. There. Perfect. Hello. I'm also from University of Washington in Seattle. My question is how you came up with the value chain model for creative business. I was wondering if you could elaborate more on how you came up with these six categories from procurement to users considering which you know like factors that you are considered if there's any unique to creative business. Thank you very much for your question and hello. Yeah so the value chain model I think was first developed by Porter in the 1980s around about 1985. And it has been fairly widely used in the media management literature which focuses on the management of media businesses. And this one is adapted from a variety of different sources. I recommend having a look at Eris and Bergen 2009 if you're interested in this concept. They've got a quite sophisticated approach to value chain analysis of media vertically and horizontally integrated media businesses. Here and we're really just drawing from other literature here in terms of defining what happens at the different stages. One innovation perhaps is that we've added the user in position five that's a relatively new concept. And we see that discussed in more recent literature on the sharing economy that a step in the value generation process and value capture process might be actually in the hands of users and outside of the boundaries of the formal business. And we think we are observing that in some of the business models in our public domain work and we do find that quite interesting. My name is Marie Silke. I'm a consultant for WIPOX at WIPO. And my question is more practical one because much of your data relies on public domain relies on Wikipedia. I was wondering how reliable that data are because, for example, I'm a content producer and I wanted to use a certain image and I don't go to Wikipedia. Can I actually trust them? Giving especially your example with the picture of your keyboard. Thank you very much. Thank you. That's a really great question and a tough question. There are two issues with reliability of Wikipedia on our side. One is the reliability of the data that we were able to access because as you saw a lot of our findings depend on the reported amount of traffic that was received before and after the magic date of 2009. That's how we derive the value generated by our public domain sample. So we used Wikipedia's publicly available tool which allows you to search on any Wikipedia sub-page and it reports traffic on a rolling historical basis. And it spits out a figure of 12,000 visitors in April of 2012 and that's what we've gone with. Now I suppose in the reported documentation for that tool it says it's pulled directly from the Wikipedia logs but we don't know precisely how many of those visitors are in fact robotic crawlers or actual page views or we just have to use the reported numbers that we've got from their tool. As a result of this study we have made contact with Wikipedia, Wikimedia Foundation. They have validated our findings internally and they've all offered us an opportunity to work with them more on a similar type of study of mutual benefit to both parties and then we would have access to their internal traffic reporting so hopefully then we'll have even more precise and reliable traffic data but we don't have that yet. On the case of liability for copyright infringing uses of work which is reported to be in the public domain on Wikipedia but may not in fact be in the public domain that I think that's a very interesting question and I don't know the answer. All that I can say is that and we talked about this over lunch the whole system of Wikipedia is based on crowd wisdom and so the uploader of the image is the one making the claim about the copyright status of the work and there's no one single rights clearance specialist at Wikipedia whose job it is to take responsibility for the status of the images and so it comes from the crowd but on the other hand when we are talking about huge collections of work crowdsourced diligent search and other crowd forms of aggregating content may be a useful solution but finding a regulatory solution to the problem of liability when we're talking about crowdsourced content is a really interesting paper and I encourage you to ask more questions about that. Thank you. Let me maybe ask one question which is really not that much related to your study more to what you said at the outset what your overall objectives were going into that and I think the work that you've done and one has to start somewhere and I think you've really broken your ground here really has a focus on creative works if you think of the public domain that of course also includes patented inventions of which the patents have expired for which there's a great interest especially in a place like WIPO where especially in developing countries many of the patents are not many of the patents that exist in the developed world don't even exist in the developing world are you aware of evidence of any attempt to quantify first of all the importance the size of the public domain but also potentially how it is used? Thank you for that tough question as well. Of course you're more familiar than I am about the research on patents and I'm sure that there is research on what happens to patents after they expire and actually it's a good nudge to encourage us to go and look at some of that literature because I'm sure that there's interesting insights there that might be able to be translated to this copyright related work so thank you for that suggestion and we will follow up on that I think that's a good one in terms of other research that I'm aware of that tries to get a grip on the size of public domain I said at the beginning of my talk Rufus Pollock's work he looked at I believe he looked at all of the holdings of the Cambridge University Libraries and look and essentially put an exact number on the number of books held by the library which are in the public domain and it's vast and it outnumbers the works which were in copyright and we've also been encouraged to think about other kinds of public domain information and in our definition at the beginning it's clear that we're focused here on traditional creative outputs like books and movies and films and video games but you know an increasing amount of material that we all rely on in our daily lives is open source software and some of the literature in this user led innovation by Von Hippel and others I mean does talk about software innovation where there's proprietary code and there's open source code but that negotiation of that tension for organizations is actually quite interesting and I think does shed light on our research and as we go forward with this study we're going to be looking at some of that literature in particular the decision of firms to drop their proprietary software and instead pick up and incorporate open source platforms and the costs and benefits associated with that. I hope that answered your question. Wow, it was a challenging question. Other questions? I think you impressed them all are you left no doubts in their mind about the value of the public domain so let me thank you once more for your presentation and we look forward to the future research work that will come not only from you but from create more generally so thank you Christopher.