 I'm Stuart Schieber. I'm a faculty member in computer science in the School of Engineering Applied Sciences at Harvard University. Also the faculty director of the Office for Scarlet Communication at Harvard. And this session is the session, What Price, Open Access. I'm accompanied by two esteemed colleagues who will also introduce themselves, well, Shimronavi Anderson. And we're talking about open access. And like many of the project briefings at this meeting, we're not all talking about a single project. We're talking about individual projects at different locations. And my responsibility, I think, is to introduce this topic before my clever colleagues talk about their insights. So I want to start just by taking open access as a good thing from the point of view of the underlying goals of scholarship. I hope you'll all redistipulate to that. But if you don't agree, we can certainly talk about that in the question period. But there are many things that would be nice that just aren't feasible. And maybe open access is one of those things. So the question is, what is the reality of open access from an economic perspective? Let's take a simple example. Imagine a major research university, let's call it Harvard, whose faculty and research staff published something like 8,000 articles a year. That's, in fact, the number of ISI-listed articles for Harvard in 2003. Suppose these were all to be published in open access journals that charge article publication charges of, let's say, $3,000 each. Not an unreasonable number in practice. If you look at the UK Finch report, it puts the cost for open access journals averaging between 1.5 and 2,000 pounds, British pounds. That's about $2,800 to $3,200. And that's also about what the journal's Public Library of Science Biology and Science Advances charge and almost every hybrid journal on Earth, hybrid journals are subscription journals that will take an additional fee to make an article open access. They typically charge a vicinity of $3,000 for that privilege. So if you have 8,000 articles, a $3,000 article, that would cost the university $24 million a year. Well, where is the university going to come up with an extra $24 million? One place to look would be the subscription outlays that the university already has to pay. This is my hypothesis. The journals have all become open access, and so all these articles are open access. The university wouldn't have to pay subscription fees anymore. But in the same year, 2003, Harvard only paid about $5 million for scholarly journals. So it looks like open access is grossly unaffordable, cost increase moving from subscriptions to open access of something like a factor of five more expensive. So yes, it would be wonderful if we could afford it, but we just can't. Switching to open access would just be too expensive. So this particular example I've given is based on data from an important 2004 report of the Cornell University Library Task Force on open access publishing. They noted these numbers for Harvard and did similar analyses for lots of other universities and concluded this. They set a model in which institutions pay for their faculty to publish and referee to open access journals would not bring about cost savings. In fact, their library, the Cornell Library, would likely see its serial expenditures rise significantly if the library used its current subscription funds to pay for author fees instead. So the rest of the discussion today will take a look, a closer look at this kind of analysis to see how more detail and new data may change this picture. But I want to set the stage by examining some of the underlying assumptions in this type of analysis and principles on which a sustainable open access system ought to be built. Now, the first thing to note is that there are reasons to be suspicious of this five times the cost claim. If we look at the economics of the situation, not at the micro level, looking at a single university, but at the macro level, looking at the scholarly publishing industry as a whole, the picture looks quite a bit different. So here's some data from Outsell for 2011, similar numbers accrue in 2003 as well. The scholarly journal publishing industry revenue for 2011 was about $9.4 billion, according to Outsell, covering the publication of something like 1.8 million articles. So that's an average revenue per article, $5,000 an article approximately. Now, compare this average revenue per article for subscription journals, $5,000 an article, with the $3,000 revenue per article, we hypothesized just before for open access journals. Suddenly, open access looks like a tremendous savings overall. If you imagine that the outlays for a scholarly journal publishing went from $5,000 an article to $3,000, that would be a 40% savings overall. So how can we square this macro observation about substantial cost savings with this worry about open access being financially unsustainable? What's going wrong with this university-specific analysis? Well, it turns out it ignores lots of things. Suddenly, there's plenty of echo in the room. Oh, that's better. Thank you. Ignores lots of things. I'll mention four. You can see what the effect is. So first of all, looking at universities like Harvard and Cornell, these universities are not typical. Whether the outlays for payments on the writer's side would be more or less than outlays for payment on the reader's side will depend on the proportion of writing that gets done as opposed to reading. That is, universities like Harvard and Cornell, which generate lots of articles relative to the size of the subscription budget, are the least sustainable cases. Whereas universities that generate relatively few articles, therefore, hypothetically few article publication charges relative to their subscription fees, would be more sustainable. And if you look at this kind of figure, here's a proxy for that figure. This is Serial's Expenditures for Published Article for a random selection of range of universities. Sure enough, Harvard looks like a far outlier. Harvard is on the far right in this slide. Very, very low Serial's Expenditures per article published. That is to say, the researchers at Harvard are writing lots relative to their reading. So that first of all should give us a pause that Harvard will not be representative if it is unsustainable. It's going to be representative of the vast majority of universities. Second, the university analysis assumed that universities should pick up the entire tab for all the articles that university community researchers generate. To my mind, the appropriate principle to determine who should be covering the dissemination cost is this, the one on the slide. That dissemination is an intrinsic part of the research process. And therefore, those that fund the research should be responsible for funding its dissemination. Now, sometimes research is done in a university, by a faculty member, who's paid to do both teaching and research. And that research is not funded by any external funder. And therefore, a university implicitly is the funder of that research. And therefore, the university should be on the book to pay for the dissemination, whatever that turns out to be. But often, there are entities outside the university who participate. So the consequence of this principle will be that if research is funded by granting agencies or foundations, they should be on the hook to cover the open access article processing charges. And it turns out that that constitutes the vast majority of articles in fact. Similarly, if an article has external co-authors, the co-authors institutions ought to share in the expense. So that will mitigate charges considerably. The next assumption that may be questioned is this estimate of particle charges of $3,000. That estimate is based on the high end of charging. Lamer journals like the Public Library of Science journals, the open access siblings of science, nature, and cell. And also, most problematic with the hybrid charges of the major commercial publishers. But the high end is not surprisingly above average. That's why it's the high end. Here's a chart of publication fees, article processing charges for all of the ISI indexed open access journals in a scatter plot with article influence. And I've highlighted with this blue background those charging over $3,000, it's a very small percentage. And there's this clump along the left edge. Of course, those are the, as it turns out, majority of open access journals that charge $0 article publication charges. Those $0 journals often are not even averaged in and average publication charges are measured. There's research by Solomon and Newark who actually did an empirical analysis of the average article publication charge. This was a few years ago, not too many years ago. Turned out to be $900, plus or minus a few, whether calculated per journal or per article. Welcome Trust reports about 1.5,000, that is 1,500 pounds per article, but those are mostly hybrid fees. So using them as a basis for the finished report estimate, you're going to get a very high estimate. At Harvard, we run a fund to pay researchers open access journal article publication charges. And on average, we've paid $830 per funded article since we started the fund in 2009. On average of $3,000, it's not particularly appropriate. Finally, fourth, the analysis assumes that as the publishing world moves to open access, everything else will just stay the same. But the economics of the open access article publication charge market are quite different from the economics of the subscription market. So things are not likely to stay the same and prices in particular. In fact, we're seeing huge changes already. Recent entrants into this market are operating sustainably at dramatically reduced fees. I mentioned public library of science before. The bulk of their revenue comes from articles from the plus one journal, which charges just $1,350 per article. Ubiquity Press has spin off of University College London as a base article publication charge of 300 pounds. So that's about $500 per article. Most recently, Peer J publishes articles based on a lifetime membership, lifetime membership for author of starting at at least $99. Estimates of the average cost of publishing a first paper in Peer J are something like $450. So a far cry from the $2,000 estimate. So let's see how all these factors affect the analysis for Harvard and then I'll turn it over my microphone to my colleagues. Imagine we're publishing 8,000 articles per year these days but now assuming open access article publication charge of about $1,000 per article. And of those, perhaps 75% are externally funded. So those internally funded and therefore be covered by the university itself would be 25% of the articles. Of which internal authors encompass something like 60% of the authorship. So the total outlay would be then 8,000 times $1,000 or the 25% of the articles pro-rated to 60% for the internal authors. That's something like $1.2 million. So it would be an estimate for kind of more current or current estimate for the total outlay. And of course in that time, since 2003 when my previous estimate was predicated, subscription costs have gone up. Harvard is now spending something like $7.2 million for subscription journals. So if you open access model would involve a cost decrease now of something like a factor of six. So which is it? Is it a cost increase of a factor of five or a cost decrease of a factor of six? Well, we don't really know. This is all hypothetical. It all depends on what would happen if many things changed that led to a wholesale modification of the market from reader side fees, subscription fees to writer side fees, article publication charges. But we can start doing much more detailed concrete analysis now that we have some data. And this is just the kind of analysis that Ivy and Ralph have done for their respective institutions. Let me stop there and I'll turn it over to Ralph. Thank you very much. It's a great pleasure to be here. My name is Ralph Schimmel. I come from Munich, Germany, representing the Max Planck Digital Library. And I want to say only a few words as an introduction. At the Max Planck Digital Library, I'm heading the Department of Scientific Information Provision. So I'm responsible for all the publisher contacts regardless of whether it's along the subscription model or the open access model. We in the Max Planck Digital Library have a consolidated single budget for subscriptions and open access charges since 10 years, already in place since 2005. Is this a, no, this is not better. Is this better? Yeah, okay. The Max Planck Society is known hopefully also in this part of the world as a very big supporter of open access and we support the, and we believe heavily in the golden road to open access. We at the Max Planck Digital Library have quite a large processing record of APCs, of article processing charges. We tend to believe that we are the largest library in the world with having these processes and we monitor the data over the years for quite a long time. So I want to show you some graphs and analysis which is all based on our own data analysis. So I'm not citing other people's work. You will see our own data analysis in this presentation. And my part in our setup here is the yes we can part. We believe that we can show that there is enough money in the system that we can make and finance a transition from the current subscription system on a large scale basis to an open access business model. So this is what we think we can show. There is enough money in the system and that the transformation can be organized on the basis of the current subscription spending without extra cost. And quite to the contrary, substantial savings in the market are to be expected. I want to show you this on the global, on the country and on the institutional level in the subsequent slides. So these are figures that we could analyze ourselves, but these are also frequently quoted figures in many, many studies in the last year. So this is what we know about the annual spending figures currently in the market. The market of academic research papers as they are published in scientific journals is in the order of magnitude of 10 billion US dollars at the moment. We know that the output volume is in the range of two million articles. 1.5 at the moment are indexed in the web of science and the rest is calculated by many others also based on some calculations that come up to in the range of two million total. So you can make this calculation from these given and known figures that there is at least $5,000 eight per article if we count the two million. And if we do limit the calculations only to the web of science share then we have even 6,700 US dollars per article that is currently spent in the subscription system. When we look now and go down to what do we know about the current APC charges? And Stuart has mentioned several and they are predominantly high in quoted and reported in the UK context in the FinF report and subsequently we believe that these figures are not very sound because all the published evidence in other instances where Solomon have been cited but also in other contexts indicate a much lower figure. Our own figures in the small funds in the bottom part is in the range of 12 hundred videos at the moment. And this is based on evidence across several years of record and observation. So we think for calculation purpose which includes already a substantial margin of error. We set it here for calculation purpose to 2,500 US dollars per article. Let's just see there's the use on that's meant to be dollars. I will just have some calculations. A few slides later that's based on yours. And so you can make this calculation with two million articles we will believe that with half the size of the current market open access could be financed. The output could be financed in an open access model. So what we believe the exercise, the transformation that needs to be organized on the global scale is that we now have the big spending on the left side. That's the current annual subscription big pocket. And on the right side, we know the current share of gold open access is in the range of 13% annually. Of course there's also some spending that goes in there but the two sides do not talk to each other. The open access is taking aware shares from the publishing system but it's not taking aware shares from the subscription spend. And we believe the exercise to be organized must look like this. We have to shift from the left to the right and by that shift we see the potential of substantial savings which would be at the same time or potential for new services in what area whatsoever. So in order to further substantiate this we now want to look at the country level. And the key message is here. You can never count the annual output in a given country all by itself because papers are co-authored also across national boundaries. And this graph shows you for the three countries Germany, UK and France in the last 10 years their output and the duplication and the limitation only to the corresponding author papers. And you can see if you look at the graph you can follow it from the left to the right and you see the pattern is very persistent over 10 years in all three countries. It's always in the range of around two third between two third and close to 70% that the papers are from corresponding author in this country. That means, and this is already a very fundamental limitation in the Finch report and in all the monetary calculations that have come out of the UK so far that they have failed to deduplicate in their cost analysis. So we would propose that the UK should reconsider their monetary calculations. All the figures and the costs that they report are tentatively significantly too high. In the next slide, yeah please look only at the, I take now the 2013 column and we do a close up on the 2013 column for the three countries starting from left to right. So you see in the left column you see the total figures of output. For Germany it's 103,000 publication index in the web of science for 2013. Counting only the corresponding author's paper reduces this figure to 70,000 papers. And if you multiply that this is now Euroma dollar anymore so then we assume the 2,000 euro level here we would come up with a total spend for Germany of 140 million euros excuse me. And the same calculation for the other two countries. When you look on the right side, what do we know about this country in terms of current subscription spending? The UK is well reported. There are so many reports out there. This is already a figure that the research council and GIST collections and various other reports have repeatedly stated in the last year. So it's from Great Britain Pounds converted to Europe, it's 260 million compared to what they would have to pay 144 in an open access arena. And the same for Germany, 140 compared to minimum, this is a very conservative estimate. We don't in Germany, we don't have the many reports as in the UK but there are several working groups of which I'm a member where we have done some calculation and this is a very, very conservative estimate. We are just working on a data analysis project and by the end of the year we hope that we can collect hard evidence on our spend across the universities and research institutes in the country. So this is a very, very significant finding and this has not been reported and published so far in any report that we know and we want to bring this evidence and enlarge this in the coming months and years. So now I want to do another close up to the German figures for 2013. You see the same 100, 3000 publications for 2013 as index in the web of science. So this is the right a pie chart. The middle pie chart would do the deduplication. So the pinkish element, this is the element of articles that would not appear as APC invoices in Germany at all and the big publisher normalization. So the Elsevier share would go down significantly 21,000 total only 14,000 with corresponding author papers the same with Wiley and Springer. And so this is the share they would in terms of APC invoices Germany could expect like one third of the total output to go to those three big publishers. On the right, this is like an ideal of prototypical library. These are very much our own max-plank digital library figures that would represent a pie chart of the subscription costs as they are currently paid. And you can compare these to your own figures at your local library, your own institution. And I'm sure that your Elsevier, Wiley and Springer share is relatively similar to this here. And we know from some libraries that they will even exceed 70 or 75% even higher than this. So when you see this, you can already see where the savings will be secured. Now we want to break this down to our own level at the max-plank society. Similar to Harvard, we know our output. We have an annual output of 10,000 research articles per year and we have been able to monitor and observe the share of our corresponding author papers. It is in the range between 40 and 60%. With the open access publishers, it's closer to 60%. We are just in the process of doing a data analysis with Springer and there we have found that with Springer we have only 40% corresponding author papers. So this is a very significant finding. So in this pattern, we have been able to observe persistently over the last year and over the various publishers. It's recurring wherever we look to. So the average APC level that we monitor and you see a link here on the slide, we document our APC spending. So you can visit this website and you will find all our APC invoices documented and we continue to upload this as to some other German universities at this time. And this is how we come to a current spending level of 1,258 euros. And this figure even includes almost 20% taxes. So this is after taxes even, the figure. So our expected total expenditure is that we would think that we would have to pay a maximum of 6,000 paper in a year and if even multiplied with an average APC and we built in some safety margin, that's why we calculate with the 2,000 euros instead of only 1,200 that we observe at the moment. And so we think that the maximum cost to be anticipated for the Max Planck Society would be 12 million euros. And even that figure is already comfortably within the range of our current spending limit. But in reality, we would of course, according to what I said, expect even lower costs. So this is already with enough safety margins. So for us, the conclusion is very clear. We as a heavily output oriented organization think it is possible to make this transition and we are also committed to make this transition. This has already been a very high level decisions in the Max Planck Society, our president and all our boards with the research directors are pretty much united behind that goal. So we see a lot of discussion and reporting in the arena of open access and publishing. We would propose that we just leave all these functions and elements with regards to publishing alone. These are all important and shall be and must be discussed. But we would suggest that we just concentrate on the underlying business model and that we see that we shift steadily from the current tall access system and the cash flow for subscriptions to open access system as full as possible open access system. For us, we believe it's time to disrupt the subscription system. Ultimately, all subscription spending must be stopped. Otherwise, we will never get to a real open access future. The money currently locked in the journal subscription system must be withdrawn and repurposed for open access services. This is our dedication and this is what we are going for and we would hope that other institutions would get to similar analysis of their own and also then eventually join into such an initiative so that it can finally be accomplished. Thank you very much. I'm Ivy Anderson. I'm the director of collections for the California Digital Library. And in that, which is part of the University of California system, as most of you probably know. In that role, I'm responsible for system-wide licensing activities of the CDL and of many of the University of California campuses. And I'm also involved in business modeling around open access and this conversion. So that's how I come to this conversation today. So I'm gonna be talking about a project that's intended to be an exploration of the ideas that Stuart described and in many respects a response to a response to what Ralph had discussed here today as well. So this is a project called Pay It Forward. The Pay It Forward project is a grant from the Andrew W. Mellon Foundation to investigate the impact on North American research-intensive universities of a large-scale conversion to Gold Open Access funded via article publication charges. And I should just say at the outset, we know that Gold Open Access doesn't necessarily always equate to article processing charge funded publication, but in the context of this project and this talk, I will use Gold Open Access in that way. So this is an 18-month project, January 2015 through June 2016. We were fortunate to have a planning grant from Mellon in 2014 to plan this project and now we're in the initial phases of making this project happen. So in this project, we're attempting to answer one fundamental question and that is, as we've just been discussing here, can a large-scale conversion to Open Access be funded via APCs be viable and financially sustainable for large North American research-intensive institutions? And we chose the words viable and sustainable purposely to signal a couple of different aspects that we wanted to approach in this grant. By viable, we're really talking about author willingness to publish under the right conditions and also a scholarly publishing system and infrastructure that can support high-quality research and allow scholarship to flourish. And by sustainable, we're talking about the financial side of things. Can we in fact afford this kind of publication? So why this project and why now? Well, I think we just heard a big part of the why now. There's a tremendous disconnect right now between North American and European and UK approaches to Open Access. In North America, the policy preference has been and continues to be towards green Open Access with things like the NIH Open Access Policy, USTP directive, Canadian Tri-Agency policy and so forth. But in Europe and the UK as a result of things like the Finch Report, the EU Horizon 2020 Project, the direction is very much a gold-funded Open Access. And so there's an enormous disconnect that's developing between our two regions which together are about two thirds of the world publishing system. So I think in the US is about 31% of worldwide publishing. The EU is actually a little bit more than that. So we're talking about a very large percentage of the scholarly publishing system that is moving in different directions. And so this is creating a very confusing economic situation. Institutional license fees and APC revenue are both increasing as we just heard. There's about 13% of Open Access gold, Open Access publishing, right? Actually, I'm not sure if the 13% is all gold or just it's all gold. Open Access publishing and that's forecast to accelerate at quite a rapid rate with the policy developments elsewhere in the world. There's a lack of transparency. There's a tremendous potential for publisher double dipping. And there's confusion about what we are actually paying for when we're licensing content. So that's a huge driver for this project. Another driver for this project, and this is maybe a local story, was a UC faculty Open Access policy that was passed in 2013. It's a green Open Access policy, but when that policy was enacted, faculty began to ask the libraries, well, does this mean I have to pay to publish? Well, it doesn't really mean that they have to pay to publish, but there's a lot of confusion about this. And at the same time, we had launched a series of Open Access pilots throughout all of the 10 UC campuses. And so we were also shelling out money for Gold Open Access, and the university librarians at the campuses wanted to know what the economic impact of this was going to be. So it really became important to us to understand what this cost might be. So in 2013, at the request of our university librarians, CDL did some preliminary modeling of APC costs and scenarios. And we took a range of publishers, about nine different publishers, and we looked at our publishing output and our licensing costs for those licenses. These are all system-wide licenses that you see. So I'm not giving the publisher names here because I wanna be careful about implications that might escape into other flora, but we looked at nine different publishers, most of the large commercial publishers, as you would imagine, and a handful of others. We also looked at a university press publisher, some smaller biomedical publishers. So we tried to look at a range of publishers, and we looked at our article, UC article publishing output with those publishers. It turned out that those, or it turns out that those nine publishers consume about half of, or capture about half of the publishing output at the University of California, about 18,500 articles or so out of a total of 36 or 37,000 articles. That was in 2010. So we did calculations that were very much of the type that you've just heard here today, looking at what we're spending with these publishers, what a variety of APCs and combination of APCs and grant funding might look like. And then what we wound up with was kind of a heat map that showed us which scenarios of APC and grant funding would produce either a more expensive or a less expensive cost per publisher. So you can see that we have scenarios of a $3,000 APC with no grant funding, $3,000 APC with 50% grant funding, $2,000 APC with 50% grant funding. These numbers were fairly arbitrary, but they're also within the range of what we know about the ranges of APCs. And we just scaled that down. I've only shown a few here, but we kind of did a sort of a normative series of different APC and grant funded combinations. And one thing you can see in this chart, the horizontal lines represent particular publishers. And you can see where the horizontal line is read all the way through that no amount, even as low as we might be able to drive the APC down and as much grant funding as we might be able to bring to bear, we're never going to make that particular publisher's cost less expensive than it is right now. And conversely, where the bar is almost entirely green, you can see in those cases that publisher, the open access strategy would be actually much more financially sustainable. And the sort of pink and the light green represent variances here. So if it was 15% positive or 15% negative, we sort of looked at that as a number, if you will, of an acceptable range of variation. So that's just looking at some very straight collaboration analysis by publisher. But what about if we look by discipline? The APCs and percentage of grant funding and so forth, many of the characteristics of publishing vary by discipline, not by publisher. So we wanted to take a look by discipline as well. We did this in a variety of ways that were a little bit rough. So we do have a categorization that we've done of all of the journals that we licensed system wide by LC class and we use that LC classification as our subject mapping. And then we took the publication output by discipline from an accountability report at the University of California prepares every year. So we took sort of the best data that we had, but it's somewhat rough data. And we looked at disciplinary output in three buckets, life and health sciences, physical sciences and engineering and social sciences and humanities. And then we did the same kind of calculation. And this produced a different kind of result that was interesting, which suggested that in the life and health sciences, even with very high levels of grant funding, and you can see that one of those, I have two models there, one of which is probably more realistic than the other, but a 1500 APC with 50% grant funding or $2,500 APC with 75% grant funding. Again, these are somewhat arbitrary figures, but they're also grounded in what we know about average APCs and the range of grant funding in those disciplines. In the life and health sciences disciplines, the gold open access looks like it's gonna be more expensive. But conceivably, there might be other disciplinary areas and social sciences and humanities and possibly in the physical sciences where the balance might be positive. And so in this particular hypothetical scenario, the bottom line number, which is that pink shaded $26 million figure is very close to what we're spending today. Well, none of this takes corresponding authorship into account. So what if we do factor in corresponding authorship, which is both Ralph and Stuart have talked about is a very relevant factor. Our proxy for this was one data point. We did some hand tabulation around that time of our corresponding authorship rates in nature journals for the University of California authors. And we found when we did that hand tabulation that in fact, about 43% of those articles were fourth, 43% of those articles, the corresponding author was from a non-UC institution. So that kind of mirrors that 60% figure that seems to be fairly, fairly common. And so we used that figure just from that one data point we extrapolated against, again, this was sort of an indicative model. A 60% corresponding authorship in the life and health sciences and also the physical sciences, but no 100% corresponding authorship in the social sciences and humanities where those factors might not apply. And in this model, the picture looks much more sustainable. So in this case, we looked at, again, a $2,500 APC with 75% grant funding in the life and health sciences, a $2,000 APC with 25% grant funding in the physical sciences and admittedly a reasonably low APC in the social sciences and humanities, but that's not unrealistic given the kinds of fees that some of those journals are charging. And this looks like a much more sustainable path. In fact, it's about 20% lower than our current subscription costs. So again, this was just some early, fairly, I won't say it's entirely simplistic. It used, tried to use some nuanced factors, but it used them in a very gross way because we didn't really have a very rich set of data points. So coming back to pay it forward, in this project, we want to build out that sort of model but do it in a much richer and deeper and much more nuanced way. So this project is now underway. We're in the first quarter or we're heading into the second quarter of the project. It has two main tracks. One is a qualitative track where we're investigating the perceptions and aspirations and behavior of academic authors, both faculty, graduate students and postdocs. And we're doing that through focus groups and also that we're planning a survey that will follow on survey to the focus groups that are now underway. And then we'll also be doing a publisher survey where we'll be surveying publisher attitudes, plans, perspectives on their current and future plans for open access. We're also doing quite a bit of quantitative modeling that will take the kind of work that Ralph has been, the German University has been doing and that I've just talked about but make that a much richer picture. So we're first going to be looking at the publishing costs and looking at the cost of publishing both by traditional and also non-traditional publishers to build more of a ground up model of what should publishing cost. And then we're going to be looking at both the publishing output of the partner universities and I'll get to the partners in a minute and also the subscription and licensing fees of those universities and do a much richer quantitative model. So the project is co-led by UC Davis and CDL for project team, you see here McKenzie Smith and Lane Farley are the co-PIs. Greg Tannenbaum is our overall project manager and I'm working with and coordinating a lot of the quantitative analysis components. We've assembled a tremendous group of partners. We didn't want to make this a University of California project. We felt it was important to embrace a broader set of institutions but again we're talking about institutions that contribute a very significant portion of the publishing output where that economic impact will be the most felt. And so Harvard, Ohio State, Ohio State University, University of British Columbia and then the 10 University of California campuses. We have three industry partners working with us in this project. So both Web of Science and Scopus are contributing bibliometric data. The ALPS, the Association of Learned Society and Professional Publishers is our publishing industry partner and will be working with us on a publisher survey. We've assembled what we think is an impressive and outstanding team of consultants on the project. I mentioned Greg Tannenbaum, also Carol Tenebier at University of Tennessee is running our user surveys. David Solomon and Brokester Bjork who both Stuart and Ralph mentioned and cited have done a lot of work on open access modeling and they are gonna be our modeling consultants. And then Mark McCabe at Boston University is also working in the scenario modeling but he's bringing an economic, sort of pure economic and market analysis approach to the project. Project Timeline, January through March is sort of the phase that we're coming out of where we're finalizing data specifications for the project. Carol Tenebier's group has been conducting focus groups and Greg Tannenbaum has been developing the publisher survey that will go out in the next quarter of the project and we'll be doing our data collection and refining and building models as we go. The expectation is that in 2016, the first half of 2016, we will really have our model to share and distribute and write up our findings. So what is a real deliverable here? Well, we think the deliverable is kind of a rich, detailed, flexible and publicly accessible financial model that can build a very rich picture of what the emerging APC model would look like in a North American environment, particularly for the larger research institutions in North America. And we plan to do this by rigorously modeling all of these different components and trying to really understand not only a static picture of what does the picture look like today, but also how is publishing growing over time and how do you factor growth over time into this kind of a model? So we hope that we will build a very rich model and a replicable methodology that then other institutions can use and apply in a local context. Just a little bit about work to date on focus groups. That's the work that actually is beginning to bear a little bit of fruit. So at this point, eight focus groups have been conducted with 60 participants, faculty, graduate students, postdocs at four of the project partner locations. So University of British Columbia, Harvard, UC Irvine and UC Davis. And OSU will also be doing a focus group fairly soon. As you can imagine, as we hope, there are a diversity of disciplines, ages and perspectives. And the group that's doing this work has begun transcribing the results and will be drafting the questionnaire soon. Just some early impressions from the focus groups. Not surprising a range of perspectives. So there's some true believers, there's some skeptics. There are a lot of faculty and students who don't really know a lot about this whole area and are just trying to understand it as many of us are. A lot of senior faculty said that they already post their articles and various repositories are on their websites. Many of the folks in the focus groups felt that they had very good access to journal content. They weren't deprived as readers, but they supported open access for others as readers and as a moral good, even if they didn't have a direct need themselves. Lots of concerns, of course, surfaced and these would be concerns in any transition. So where will the funding come from? Will they have to pay? Obviously a big concern for graduate students and postdocs. Concerns that richer nations might dominate publication, that there was a potential for APC prices to rise, just as subscription prices have risen over time. Concerns about lack of transparency, feeling that publishing is broken, but not necessarily a lot of clarity about how it should be fixed. And then concern that APCs are too high, a lot of folks seemingly not understanding or crediting why APCs should be as high as they are and feeling that publishers probably are just charging with the market will bear. There was some support expressed for library's role in coordinating this new area of activity and in negotiating around open access. So I just wanted to close with some thoughts about how this trend and the directions that we might be going in in the US and Europe and how does this look from a global perspective? And one of the things that Stuart and I both have talked about is the role of grant funding in potentially supporting open access charges. And in fact, if you look at the shifts in global R&D, you can see that in the last decade or so, the US share of global R&D is really declining as a share of the world's research. And that's largely because of the rise of the Asian countries and the developing nations in general. And so there's really a shift in investment that's happening globally. And that shift in investment is also manifest in global article share. So the chart on the right is from NSF's analysis of science and engineering global article share. That's, I'm guessing that's a little bit hard to read, but the EU global article share is the highest. US is below that. And there's quite a rising line for the developing countries and in particular China. And that's something that publishers will tell you and that our scholars have observed as well. And so really there's a shift in investment and a shift in activity that's going on around the world. And we've seen this in scope three also, which is a project that I'm involved with, where we've been looking at article share around the globe and high energy physics. And even in the last five years or so, there are palpable shifts in where the balance of publishing is happening. So whereas if you look at subscription distribution, the US is still by far the highest share of revenue in the STM global STM revenue. So it's interesting to contemplate and I think as we think about where we're all going in this environment to think about how the shifts in global output might drive and be a shift in the balance of investment around the globe. And part of the unsustainability that I know many of us feel is about the growth of publishing output in general. And there are some real questions about how much we in the US should be bearing the brunt of that growth when the growth is coming elsewhere, coming from elsewhere. So of course it is a global system, but it's interesting to contemplate these shifts in R&D versus where the balance of subscription revenue is coming from now. So that's just something to think about for the future. And with that, I think we can stop and open this up for questions.