Katie and Steve speak with WIPO economists Alexander Cuntz and Matthias Sahli about their recent article, Intermediary liability and trade in follow-on innovation, published in the Journal of Cultural Economics in February 2023. Their research looks at how intermediaries in the art market altered their behavior after the Second Circuit’s decision in Cariou v. Prince, which was seen as greatly expanding permissible fair use in appropriation art. THIS PODCAST WAS RECORDED BEFORE THE SUPREME COURT’S DECISION IN THE WARHOL CASE, WHICH PULLED BACK ON THE TRANSFORMATIVE USE ANALYSIS FOR COPYRIGHT FAIR USE.
Steve Schindler: Hi, I’m Steve Schindler.
Katie Wilson-Milne: I’m Katie Wilson-Milne.
Steve Schindler: Welcome to the Art Law Podcast, a monthly podcast exploring the places where art intersects with and interferes with the law.
Katie Wilson-Milne: The Art Law Podcast is sponsored by the law firm of Schindler Cohen & Hochman LLP, a premier litigation and art law boutique in New York City. Hi, Steve.
Steve Schindler: Hi, Katie.
Katie Wilson-Milne: How are you?
Steve Schindler: I’m great. How are you?
Katie Wilson-Milne: I’m just dying to talk about something other than NFTs. So we thought we’d revisit our old favorite topic, equally confusing, of fair use.
Steve Schindler: Right, but this is going to be a slightly different twist on it.
Katie Wilson-Milne: Yeah.
Steve Schindler: So what are we doing today?
Katie Wilson-Milne: So today we’re going to talk about an analysis of the economic impact of certain copyright law developments in the US, particularly with respect to the uncertainty around appropriation art’s legality, and the parameters of fair use, which we’ve talked about in terms of the Warhol case and Cariou v. Prince before on the podcast.
This is a hotly contested issue. As we know, the line between the derivative work right protected by copyright and the copyright infringement exception of fair use that has allowed copying and appropriation in fine art under certain shifting circumstances is still in flux. So not only has the standard for when appropriation of existing expression is permitted in the fine art context changed over time, it’s inconsistently applied within the same jurisdiction. So, for example, in the Second Circuit, there’s been some inconsistent law on this, and certainly between circuits there’s a disagreement and inconsistency all leading to understandable confusion on the part of artists, the art market, and their lawyers who advise them on sort of where the rules lie here.
But one thing we have not talked about on the podcast, we realize now after meeting our guests, is the clear concrete changes, both in terms of economics and market behavior that such legal uncertainty might cause, which is a, you know, very law-and-economics-type analysis. So today we have two guests, economists Alexander Cuntz and Matthias Sahli, who will describe their analysis of the impact of one particular blockbuster case on copyright fair use, Cariou v. Prince in 2013, and its impact on the art market, specifically in terms of what they call intermediary risk or indirect liability, and we’ll hear from them more about the research.
Steve Schindler: Sure. So we’re here today talking with Alexander Cuntz and Matthias Sahli about their recently published article in the Journal of Cultural Economics called Intermediary Liability and Trade in Follow-on Innovation, which appeared online on February 12th, 2023. Alexander Cuntz, who heads the Creative Economy section of the World Intellectual Property Organization, sometimes known as WIPO, located in Geneva.
Before joining WIPO, he worked as a senior consultant at the Federal Expert Commission for Research and Innovation in Berlin, and as a research fellow at the European Commission in Seville. His research focuses on creativity, the role of intellectual property, and the impact of digitization on creative industries.
Also joining us is Matthias Sahli, who is a full-time research fellow also at WIPO, and a PhD candidate in economics at the University of Neuchâtel in Switzerland. He earned his Master’s degree in economics from the University of Bern, and previously worked as an economist at the Swiss Federal Institute of Intellectual Property. His research interests lie in the intersection of intellectual property and economics, and his current work focuses on the empirical analysis and impact of intellectual property rights in the creative economies. So, welcome to the podcast, Alexander and Matthias.
Katie Wilson-Milne: Yeah, welcome.
Alexander Cuntz: It’s a great pleasure to be here. Thank you, Steve and Katie, for the kind introduction.
Katie Wilson-Milne: So let’s start with having you give us a summary of your research on this topic, and then we can dive into some of the details.
Alexander Cuntz: Yeah, let me maybe start off with a disclaimer. As you mentioned, we work for the WIPO here in Geneva. So this is of course also political work, a political environment. So let me put a disclaimer that we are independently doing this research, and this is not an official position of our organization, nor of any of our member states of course. Still, I think it may be interesting, and let me kick off with the research itself. So we started off with a largely under-researched area, the visual arts market. And basically our starting point was can we find out how appropriation artists would respond to the current legal framework? Would they care about copyright or not? And would that kind of affect the ways they set up their artistic practice? We found out in the course of the research that this would be difficult to actually get to, and we saw from at least anecdotal evidence that even more than scientists, artists wouldn’t care that much naturally and would kind of— if it’s their destination to appropriate art and make that their practice, they would continue probably to do so.
What we then thought to start thinking about was actually maybe it doesn’t matter to the artists themselves that much, but maybe it does matter to a market that invests in them, and that actually puts a lot of money on the table to commission the work, to trade it, to curate it. And so we started to think about liability, which is again, another topic that is largely under researched and that we would want to find out basically how liability would impact the behavior of trade intermediaries in the visual art market.
Steve Schindler: And who are the intermediaries that you’re talking about?
Katie Wilson-Milne: Yeah, and maybe we should also define how you define appropriation art.
Alexander Cuntz: Yes, of course. Maybe I start with intermediaries. So what we looked at was actually secondary intermediaries in the secondary market. So for example, galleries, auction houses, but we also had an eye on museums that would curate artworks created by appropriation artists, so that was the other angle. Maybe over to you, Matthias, if you want to explain a bit kind of the art historic background of this, if I may ask you.
Matthias Sahli: Yeah. Maybe— I think that we are in the middle of this study already. If we would talk about what we have considered as appropriation art, that’s part of our data selection process as researchers. And sometimes you have to do assumptions and you have to start from a certain point where you face limitations, but you go on with that. And we were relying for this study on Artsy.net to obtain so-called genomes of artists. So for us it was not really interesting, for example, to have a list of artists considered as appropriation artists or not with maybe 10 or 20 or so. We want to have a bit of a larger data set, and Artsy provided us, or we were querying their programing interface with their genetic information that they collect on an artist. And then they select over thousands of characteristics of artworks and artists as well. And we started this research with querying appropriation genetic information that art historians and also the AI and so gave to the artists and artworks. And that was our starting point to have a group of actually, of artists that we would say could be affected of this liability issue that we are discussing in the study.
Katie Wilson-Milne: Yeah, and maybe to back up— I mean, presumably before you were creating your data sets, which we’re definitely interested in understanding, you identified one event in terms of copyright liability that you wanted to measure, the Cariou v. Prince case in the Second Circuit, that case that dealt with fair use. You could tell us a little bit about why that case was sort of the moment you chose to research around. And also, you know, when we’re talking about appropriation art, I think, correct me if I’m wrong, you’re talking about artists like Richard Prince using the creative output of other artists in their own artwork in a very obvious way.
Matthias Sahli: I think this was laying on three pillars basically. So we were screening the law and economic literature as far as we can understand and screened for importance where we would see where scholars are discussing these appropriation art cases the most. And we were specifically interested in this, what we call contributory liability. So we wanted to have a case where intermediary was involved as well. And the Second— higher court decision actually seemed to hold liable as well the Gagosian Gallery in that case, so that’s why we chose it. And as well, we were comparing, for example, Google trends searches as well with other cases. And we wanted to understand whether the audience is reacting to such a case and not only like researchers as we or like legal scholars also to have a bit of a bigger picture. And we were talking to market intermediaries, actually to people from galleries.
So we were involved in a museum in Geneva here that held a Richard Prince exhibition, for example. And we wanted to find out like how important this decision actually was. And we thought that this would have like probably the highest importance.
Katie Wilson-Milne: I think you’re raising an interesting preliminary question that Steve and I think about a lot, which is you’re right, the Cariou v. Prince case, which we’ll give a little summary of in a minute, gained a lot of attention from legal scholars and lawyers and people who deal with licensing and copyright. And the extent to which market participants care about that has never been that clear to us. Especially in the fine art world, I think it’s our observation in practice that often market participants in the fine art world, maybe as opposed to the performing arts, maybe don’t react that much to legal changes, because it’s a pretty under-contracted, under-lawyered space. So I don’t know Steve, what you think about that?
Steve Schindler: It’s interesting, because when the first Cariou v. Prince decision came out, I think in 2011—
Katie Wilson-Milne: At the trial court level.
Steve Schindler: —At the trial court level, and certainly among art lawyers, this was a very important case and one that received a lot of attention. But if you recall on the district court level, Judge Batts ruled in favor of Patrick Cariou and against both Richard Prince and the Gagosian Gallery, and ordered that all of the outstanding appropriated works be delivered to Patrick Cariou, who is then empowered under this decision to destroy them. And that got a lot of attention.
Katie Wilson-Milne: Yes.
Steve Schindler: Even in the art world I think as participants, because all of a sudden you have this judge saying you’ve got this appropriation— a well-known appropriation artist— and you can destroy the works, because they violated your copyright, which is of course a remedy under the Copyright Act. And then of course, the case was appealed to the Second Circuit who came down mostly in favor of Richard Prince, as we know, in 2013.
Katie Wilson-Milne: And so opened up a very permissive space for appropriation art. And we’ll backtrack a little just to give some more background. So Patrick Cariou is a photographer. He spent years, I believe, in Jamaica living among and photographing Rastafarians, and published a beautiful book from his work there.
And Richard Prince came upon this book at some point and blew up certain photographs in the book. Actually, I believe 30 or more photographs, so a significant number from the book, and made some changes to them of various types. The most visible in this case, if people have seen the photographs, was that he painted some blue blobs and lines.
Steve Schindler: Lozenges, I think, is the way the court decision describes them.
Katie Wilson-Milne: I’m not being very respectful, but yeah, there’s some blue paint on—
Steve Schindler: There’s a guitar, too.
Katie Wilson-Milne: Yeah. A guitar, right. He painted a guitar in one of them. So some of them were just a line of blue paint, some of it was figurative like a guitar, but the photographs were certainly not obscured. You could absolutely see almost the entire photographs. And so Patrick Cariou sued in the Southern District of New York, sued Richard Prince and the Gagosian Gallery, and I think Larry Gagosian himself, who had displayed these works as part of a Richard Prince exhibition claiming copyright infringement of his photographs.
I don’t think there’s any question that the photographs were the works of fine art, that they were creative, that they were protected under the Copyright Act. As Steve said, the trial court on summary judgment found that there was copyright infringement and that there was no fair use because Richard Prince’s use of these photographs didn’t comment on the work, wasn’t parody.
It didn’t reference the work in any way, it just used the work for something totally different, his own artistic practice. That got appealed to the Second Circuit and the Second Circuit said, oh no, no, that’s not the right test for fair use. We’ll just tell you what we think part of the test should be, which is that the new works don’t have to comment on the original work. They just have to have some kind of new expression, meaning, or message. As lawyers, we find this kind of decision very frustrating, because what does that mean? Like what’s the distinction? How do we follow that, and how do we advise clients based on a standard like that? And as kind of further confusion, the Second Circuit, I believe, decided as a matter of law that 25 of the works were fair use.
Steve Schindler: Correct.
Katie Wilson-Milne: So no need to do any fact finding. Just a matter of law, they’re fair use.
Steve Schindler: Right. Which that was really extraordinary in a way, because for a circuit court to sort of overrule the trial court on a factual matter, I think they just did not care for the District Court’s decision, but then they did send I think 10 of the works back to be—
Katie Wilson-Milne: Five or 10, I don’t remember. Could be five, yeah.
Steve Schindler: —to be reviewed again by the District Court and left open the possibility if they were infringing that there could be liability both for Prince and for the Gagosian Gallery and Larry Gagosian.
Katie Wilson-Milne: Yeah. To your point. So it left open that possibility. The case then was resolved without further decision. But yeah, it was unclear, and I think it’s worth looking up some of these artworks, because the 25 that were as a matter of law fair use versus the five that weren’t, to a lay observer, it’s not entirely clear what legal distinction the court was drawing.
So, I mean, back to you Matthias and Alexander, but just to sort of tee up your paper, which is all about the confusion that this decision created, even though the higher court decision, as you said, found in favor of the appropriation artist, it also created a standard that was nearly impossible to apply precisely. And so for people in the art market who paid any attention, they didn’t know.
Steve Schindler: Right. And just to add one more point before we turn it back over to you, is that the Second Circuit kind of became an outlier in a way, in the sort of permissive use of fair use because a year later there was a case out of our Seventh Circuit by a well-known judge, a circuit judge, Judge Easterbrook, which basically took issue with the entire approach that the Second Circuit was taking and said, transformative is not a test that exists under the copyright law.
Katie Wilson-Milne: Yeah.
Steve Schindler: And that is something that’s entirely been made up, and the most important thing to look at is the four factor fair use test, which is—
Katie Wilson-Milne: Which is in the Copyright Act.
Steve Schindler: —which is in the Copyright Act. And it basically says, did this work usurp the market for the other work? Did it impose itself on the value of the work, as opposed to this kind of vague transformativeness test? And so by the time, as you were looking at the data, we have the Second Circuit going in one direction, which is a very vague, hard to apply standard, and then we have even the Seventh Circuit saying that’s not even the standard.
Katie Wilson-Milne: Right, you’ve totally just departed from the Copyright Act and created a standard that isn’t grounded in how we understand copyright to be. Yeah, that’s the situation shortly after Prince. And there have been some further developments which we can talk about later.
Steve Schindler: Sure.
Katie Wilson-Milne: But let’s turn back to your study. So clearly you did some research into that state of legal affairs and noticed scholars and commentators remarking that this created some uncertainty in the market, even though it was favorable in this case to an appropriation artist. So how did you go about measuring the impact of this court decision?
Alexander Cuntz: Your discussion of factors really highlights our starting point at the end of the day in the sense that it puts spotlight on the question of uncertainty, which for us as economists, a market is a major thing. The idea that things are unclear and difficult to manage from you as a practitioner’s perspective, but maybe as well for other trade intermediaries, this is really, really a moment in time where basically there’s potential that people in the market will respond to that. So if the legal framework or the way the legal framework is shifting is introducing uncertainty, that can actually trickle down to the people that are operating on the market. That’s basically the generic background. Let me maybe even go one step further back and give another argument why we think a contributory liability is important. It is because in many of the discussions we are in, and most of the research we are doing, or most of the research that has been done so far is actually informing discussions on how revenues are shared between stakeholders, which comes natural, right? It’s kind of the moment when new things are created and people talk about how— who has contributed and who should be entitled to certain revenues. Why we think liability is important, or what this misses out on is basically a dynamic perspective, as we would call it, as economists, where we are interested how the legal framework plays out over time. And this brings in this important perspective which crystallizes in a sense around appropriation artists, how well does the legal framework balance the interest between current artists and future artists, and how well can the two exchange, right, and operate together. And that’s kind of an area that is, I think, very much understudied and very important if you are interested as a government and policymaker to set up a system that is over time helping the flourishing of creativity and business of course.
Katie Wilson-Milne: Yeah, I mean, in many ways the concept of licensing tries to solve that, right? This relationship between one creative person to another where they can voluntarily loan or give limited rights to use of their material. And that’s in many industries like music, publishing, very routinized.
In the fine art world we don’t see a robust licensing exchange between creative parties. We see it in brand partnerships and the monetization of artworks, but between artists using each other’s work there’s no culture of that. So that maybe is a particularly underdeveloped area of study because there’s no mechanism in place to negotiate those factors.
Steve Schindler: Right. And I think as a practical matter, when we talk about appropriation art where one artist is sort of taking apart, or recontextualizing, whatever you want to call it, the work of another artist, artists are not often very receptive to having their photographs doctored up or picked apart and kind of put together in a different way. That’s not something that they would—
Katie Wilson-Milne: So they might not license it.
Steve Schindler: I think that we see that in—I mean, even in the first, you know, the sea change here was in the Campbell v. Acuff-Rose case in the 1980s with the parody of 2 Live Crew of Roy Orbison’s “Pretty Woman.”
Katie Wilson-Milne: Yeah, in the music contest.
Steve Schindler: And they tried to get a license. 2 Live Crew actually approached Orbison’s copyright holders and said, can we have a license? And of course they said no, because you’re going to make fun of my song.
Katie Wilson-Milne: My song. Don’t make fun of my song.
Steve Schindler: So we’re not going to license this. So I think that’s the same thing with a lot of these appropriation artists, is that the original artist is not—Patrick Cariou is not going to want to have his works painted over and things like that. So it’s not really practical to think that that’s going to happen.
Katie Wilson-Milne: Yeah. Right. The incentives are not there and the culture isn’t there. So you are interested in this sort of economic market reaction to these kind of shifts in the law that create uncertainty around what standard will be applied. How then specifically are you measuring that? I mean, what do you do to test that premise and see what the result is? You can get into the weeds here.
Matthias Sahli: Yeah, so I’ve mentioned already that we were exploiting this rich dataset from Artsy that we’re categorizing artists in not only obvious distinctions, like Modern Art or Pop Art, for example, but with this genetic information, and I could give you like an overview of tags that we were using, but basically all of them were nested information around this appropriation genetic information.
Steve Schindler: Sorry, could I maybe just ask a quick question, because you keep referring to this genetic information.—
Katie Wilson-Milne: That’s going to be misconstrued.
Steve Schindler: I’m not sure it’s going to be understood by everybody. Because obviously—
Katie Wilson-Milne: Not genetic.
Steve Schindler: —we’re not talking about actual genetics. So maybe you could just describe a little bit more what you’re meaning when you say you’re looking at genetic information of artists or artworks.
Alexander Cuntz: So the starting point was for us the genome database, which started off as a project by art historians that tried to look at systematically categorized artworks and artists. And they basically, to give you an example, they basically looked at an art artwork by Picasso and they said like, this is blue and this is painted on wood.
So they started to roll out those tags, and I think they ended up with at least 1,000 different type of tags in order to classify and distinguish the work. And basically that project was taken up by Artsy as a service, because they had in mind to develop something like if you go on Amazon, you get recommendations, this is what other people bought, or this is what you might like. And such algorithms are actually building on tech data that makes different products comparable and makes them—
Steve Schindler: It’s a marketing ploy.
Alexander Cuntz: Yeah, so they tried to do this for the visual arts market and for artworks at the end of the day. So they took over and they further developed basically this tagging classification, and that’s very data heavy and very rich. And we basically took all that knowledge they developed around artists and artwork to define what appropriation artists are.
That’s important for our research design to come up with a group of similar artists that would just differ in the appropriation genome as they call it, in just one tack. Otherwise we try to set up a group of similar artists that was otherwise very similar. Matthias, am I getting it right or would you like to add something?
Matthias Sahli: Absolutely correct. It’s maybe easier to call them tags.
Katie Wilson-Milne: Tags, yeah.
Matthias Sahli: The treatment group, so the appropriation artists that we wanted to identify— like Richard Prince had 100 tags that we identified and some of there were appropriation or appropriation close tag information, and our control group of artists basically had the very same tags as Richard Prince, but excluding the appropriation close tags.
So that’s how we constructed a control group of artists that we identified. And also the Artsy service itself, based on their algorithms they—It’s also part of their business model. I always assume that they promote, like if you like this, then you’ll probably also like this artwork, for example. So we can rely on their algorithms as well there.
Katie Wilson-Milne: Just, I guess to be— this is probably somewhat of a limitation. You weren’t identifying appropriation art, you were looking at the market for artists and their whole body of work, which presumably you identified artists that mostly make what we call appropriation art or copying other people’s work as part of their work. But you weren’t actually identifying which artworks were appropriation, right?
Alexander Cuntz: That was our initial hope actually, that we get very fine grain data on the level of the artwork that would give us the original underlying photograph, let’s say, and then we could even look at the original artists, right, and how they would fare.
But at the end of the day we were only able to say this is, on the level of the artist, this is likely somebody who is deeply into appropriation art versus this is somebody who does very similar things, who loves to paint blue paintings and work with wood, but is just different in that aspect of appropriation practice.
Ideally you would have a situation where you can randomly assign appropriation practice to artists. So you would say you have artists that haven’t done anything and basically you would say like you go, please, you go forward and practice appropriation art. So look for works that you would like to transform and those things, and everyone else— starting from scratch artists— everyone else would just not be appropriating other people’s art. But that’s not how reality looks like, right?
Katie Wilson-Milne: You have to look backwards. You can’t create, yeah.
Alexander Cuntz: Exactly. We need to look back and we need to basically, well, identify a group of appropriation artists, yes, but then we would also need other people who are very similar in multiple dimensions so they are as successful in the market, let’s say, they produce in similar decades. So we look for everything that is observable to us to make it very comparable that other group. And that’s what we call natural experiment in a sense. And that’s one of the last times I’ll use terminology, but difference and difference design is actually what we chose to do.
Katie Wilson-Milne: Yeah, and what does that mean?
Alexander Cuntz: Yeah. That’s basically a comparison of market outcomes for the appropriation artist group versus market outcomes for the group of similar artists that we identified as well. And what it typically does, it goes back in time, as you said, and basically looks at, for example, how well did artists in a similar group sell before the court decision, and how well did appropriation artists sell before that decision?
And then it basically tracks how sales change once the decision comes in. And in that sense we say like the appropriation artist group is impacted, is treated as we call it, once the decision comes in versus the group of similar artists shouldn’t care or shouldn’t really care about the decision.
Katie Wilson-Milne: So the goal, I guess is, as with any control group, is you’re trying to create a group as close as possible to the appropriation data set, except the only difference being, which is impossible, but it’s close enough, the main only difference being that they don’t create appropriation art. But in other ways I guess I’m just guessing but, age, sex, style medium, that they’re comparable sets. Is that right?
Alexander Cuntz: Exactly. So everything that is literally observable to us based on the data sources we had. So we used auction data, that has a certain set of information attached, right? You know where typically the artist sells, in which auction houses. You would know the size of the artwork, you would know, as you said, characteristics, socio-demographics of the artists. And we basically compile all of that, and that’s a major assumption actually in the difference and difference design that before the court decision comes in, both groups are literally performing the same. That’s a major assumption. And we basically statistically test that to see that yes, they are really comparable.
So any change we might observe after the court decision maybe driven by the court decision, it tries to causally link the decision to changes in the treatment group. Did I get it right, Matthias, or am I simplifying too much?
Matthias Sahli: Yeah. One specific point is maybe important, if you say that they have to perform the same, it is worth noting that the assumption requires only a common pre-trend, like what you say. So the trends have to be the same so that whatever trend happens after our treatment that we defined is then the average change for the treatment, and before you have the same trend. So we can rule out the possibilities that noise comes in from pre-treatment periods.
Katie Wilson-Milne: Basically the markets for the control group and the test group, the markets were aligned pre-2013, is what you’re saying, and you can control for that.
Matthias Sahli: Yeah, or specifically the outcomes that we look at. We generally can frame it as the market, but we are looking at specific outcomes and outcome variables. That is, in our case, for example, the price of the artwork, that is the number of artworks artists have, or the probability that an artwork results in sale success. So in those kind of outcomes it has to have the same trends.
Steve Schindler: Right, so that’s more data-driven. And we kind of impose the term “market” on it as lay people or participants in the market. So then can we ask what were your findings ultimately?
Alexander Cuntz: Well, the main finding we have is basically that the number of auctions among appropriation artists sees a slight drop after the decision compared to our baseline, which is the similar artist group. So yes, they continue to have auctions, but there’s a different trend, a different average number happening for the similar artists. That’s our first main result. But as Matthias mentioned, there’s a number of other outcome variables, for example— well, that’s an indication that maybe those potentially infringing artworks, auction houses maybe, were less willing to actually carry that litigation risk, right?
Katie Wilson-Milne: Meaning they were not as willing to sell through certain appropriation art. So finding one is— they’re just fewer pieces up for auction after the decision.
Alexander Cuntz: Exactly. Exactly. Mm-hmm.
Katie Wilson-Milne: But you have some other findings that are actually performance, sales performance and performance at auction too, which are interesting.
Alexander Cuntz: So basically if you make it to the auction, you see a decline in the probability that work will actually sell off. And so you have that, kind of the overall drop in the number of auctions, but once you are in you’re still less likely to sell actually.
Katie Wilson-Milne: And what about if you do sell, was there a price shift that you investigated or it was meaningful?
Alexander Cuntz: I think we looked at the prices as well, but there was no significant effect if I recall that correctly. Matthias, can you confirm or disagree?
Matthias Sahli: No, no, no. The answer there is that price formation— so we look in much detail at the price formation of those artworks. So again, our model does control, for example, for the size of the artwork or the time when the sale took place, for example, the year, or the auction house, most importantly. But we were able to replicate all those studies on basic auction house theories and sales, but what we didn’t find was actually a reaction of the sales price after the decision that we defined as our treatment period. There are a number of explanations why this is the case, mostly probably because price formation is something that’s a bit more stable. And since we were able to observe less auctions after the court decision, it could be that the one with the stable prices were like kept in the market while others were already taken out. So this is a limitation of the study that we cannot identify the same artwork over the time. We can only see the performances of artists over time.
Katie Wilson-Milne: Well, and it’s interesting that sort of fits in with the more narrow scope of your study, which is not about the performance or the market receptivity to appropriation art in general, but that intermediary behaviors, meaning not the creators but the sellers, the displayers of this work. That was obviously the question in your study, and so the sale price wouldn’t be as pertinent to that analysis, right? Because the intermediary decisions are, do I put this up for auction?
Alexander Cuntz: Exactly. I think it’s— any measurement of a price effect would be more blurred also by a decision of buyers, right, and kind of art investors? So I think this is what we try to be, as you say, like we try to be narrow around the supply question a bit. What is supply to the market? What is curated to the market even?
Katie Wilson-Milne: And buyers would be less concerned with fair use, because presumably they’re going to hang it in their home or put it in storage, and that does not raise copyright infringement concerns the way putting it in an auction catalog or displaying it in a showroom would. That makes sense to me. So you design this study, you pick— based on the Artsy data set available to you, you pick and categorize different appropriation artists, but it strikes me that a certain few very famous appropriation artists like Richard Prince, Warhol, Lichtenstein, I don’t know, I mean, you have a list, might skew the data set in a way that like maybe the market only for them changed or the market for them didn’t change at all. But for the vast majority of artists who aren’t that famous, whose work does not sell for that much, that the market does not pay that much attention to, there really wasn’t a shift. So I don’t if that’s something you parsed but we’d be curious to hear if there were outliers that skewed the results.
Matthias Sahli: This is an excellent question and absolutely, yes. So this is the short answer. Like Andy Warhol is definitely, or Picasso is definitely in almost all auction data sets like very much of an outlier. And this is something that our econometric models can address at first stage.
So we basically run regression, so like calculate the effect based on the average treatment of an artist. We compare then the averages across the artists to be a bit more outlier robust. In the next step, this is then what we call a robustness check of our models. We can exclude specific artists, we can only run the calculations based on the superstars. We can interact what you call our treatment with the superstars only, for example, in order to find out specific impact of them. So this is something that the econometrics, so the statistics for economists can address pretty well, I would say.
Katie Wilson-Milne: And what did you find when you did that?
Matthias Sahli: One of the interesting findings actually that we find so-called heterogeneity in the treatment. That means that, for example, that we run the calculations based on the pictures generation only what we identified as core appropriating artists, for example. And interestingly, we found in this study that artists from the Pictures Generation— the pictures generation from art historians, a defined group of artists, I think it was based on a exhibition in the 80s—
Katie Wilson-Milne: But I mean, you isolate them, why do you—you isolate them, because they might be more high profile or more famous or why isolate them?
Alexander Cuntz: No, I think they have a higher liability risk attached to them, in a sense the iconic practitioners of appropriation art. So for us, it was basically a test to see— the hypothesis we ran was actually there the effects must be most pronounced in that group, because it is crystal clear basically even for most outsiders, this is the core of the practice, right? So the test was to see if we can confirm the effects are most pronounced in that group. And we can confirm right, Mathias?
Matthias Sahli: We can confirm, yes.
Steve Schindler: Right. So I’m curious, and maybe I think I’m understanding what you’re saying, but I mean, if you look at, for example just anecdotally, not a data-driven analysis, but about a year after the Cariou v. Prince decision came down, there was another lawsuit against Prince by Donald Graham, a photographer, right, who had posted a photo on Instagram. And this was the project where Richard Prince basically downloaded— commented on other people’s Instagram posts, blew them up, and then sold them at an exhibition at the Gagosian Gallery for $90,000-ish a piece. And this was post-Cariou v. Prince. So certainly from one intermediary, Larry Gagosian and his gallery, they were not at all inhibited.
Katie Wilson-Milne: Well, they felt like they’d gotten permission from the Prince decision.
Steve Schindler: Yeah, so they were prepared to go ahead and have a whole exhibition. And I’m just wondering, is this one of the kind of outliers that you were trying to eliminate?
Alexander Cuntz: Yeah. Maybe one step back, if I may, it’s kind of— I constantly run into those issues when discussing the research, this economic research with lawyers, because I understand your practice is wholly based on assessing case by case and being very concise there. The whole statistical approach that we run in the research is really trying to tease out average effects, right, Matthias? So it’s kind of across the board. You may still have that assessment that we have a change in the number of auctions, includes the possibility that for an individual artist in that group, it is actually increasing after the court decision, right? But the majority of those people will see a slight drop in the number of auctions. So it’s similar like that, just to explain that a bit.
So excellent point, Steve, on the Gagosian Gallery. I fully agree this is anecdotal evidence that they actually didn’t care. However, as economists, we’re basically interested in those average effects and probably not everyone has those deep financial pockets as the Gagosian gallery. And our point is basically that the average effect may actually have led the average auction house or average intermediary to step down and sell less artworks by appropriation artists, so that’s basically part of our analysis.
Katie Wilson-Milne: And one of your other findings was that there was a relocation of auction activity for appropriation art. So maybe we didn’t make this clear, but your research is about the effect of this US, really one circuit in the US, decision on the US art market, specifically intermediary activity with respect to appropriation art, which you assume is related to litigation risk. But you do something interesting, which is you look at whether there are changes in other jurisdictions and you haven’t talked about that yet. So maybe you can summarize that finding as well.
Matthias Sahli: So I was referring to those kind of outcomes variables that we were looking at, and one of it was the relative percentage of auctions that were in the US compared to outside of US on the artist level. And we actually found that, at least temporarily, a pretty significant decrease after 2013 of US auctions compared to non-US auctions on average per artist. And this somehow makes sense, because the decision took place in the US, and it could be defining that you’re basically outsourcing the risk in other jurisdictions, but since we are confronted here with 100 countries maybe, it’s very complicated to look at all of the jurisdictions of course, and to find out where the trade probably relocates. But we were looking at was that shift outside the US, at least temporarily.
Alexander Cuntz: Exactly.
Katie Wilson-Milne: And you found a statistically significant movement of appropriation sales by intermediaries to outside of the US?
Matthias Sahli: Yeah, so a decrease in auctions in the United States, so a shift outside, because we were looking at relative numbers. Yeah.
Alexander Cuntz: It seems that some of the artworks by appropriation artists they went from, let’s say, New York houses to London auction houses. Some of them are operating in many jurisdictions. So imagine it may be fairly easy to actually relocate the sales to another place that has a more favorable jurisdiction set up.
Katie Wilson-Milne: I think to your analysis it’s just where are the rules clearer, and you touch on this in your paper, none of us are foreign law experts, but we are all under the impression that the US system for exceptions to infringement, especially in the fair use appropriation context, are much less specific and defined than they are in the exceptions to copyright infringement that exist let’s say the UK and Europe, where there absolutely are exceptions to infringement. For example, in the UK, parody and commentary, which we have here, which are about free speech concerns, but they’re not as amorphous, I guess, in their illegal application or judicial interpretation, and that may create a market effect.
Steve Schindler: Right, and obviously when we talk about auctions in the United States, we’re really talking about auctions in New York City, and auctions in New York City—
Katie Wilson-Milne: Of expensive art.
Steve Schindler: —of expensive art in New York City, whether it be at Sotheby’s or Christie’s or Phillips, and those are all covered by within the jurisdiction of the Second Circuit, which is the decision that we’re talking about.
Katie Wilson-Milne: Yeah, so that makes sense that the Second Circuit would have kind of nationwide impact on these major sales, because they occur in New York. I was interested in your paper that you also found no impact in a couple of areas you studied. One you already mentioned, which was artist production, that there didn’t appear to be artist’s sensitivity to this particular court case. And maybe we can extrapolate to say litigation risk in general. That would ring true to me just based on my observation of how much or little certain artists are concerned with precise legal rules around how they can create their art. So that was interesting.
And also, correct me if I’m wrong, you found that in terms of sort of museum-type intermediaries, who are not selling but they’re just exhibiting work, that there was no statistically significant change in the exhibition of appropriation art, so access by the public to see this work. Is that right?
Matthias Sahli: It’s— only partly. We had the idea that it’s an incomplete picture to only look at auctions. Of course it would be great to also look at like the primary market, galleries, for example, or museums. And we’ve done an initial and very preliminary analysis on one museum, the Metropolitan Museum of Art to see whether the appropriation artists were identified and also changes in exhibitions in this specific museum, for example.
We had to look at a couple of museum datas that we were able to find, but in the end we didn’t find an effect there. But this would definitely be a part of like a follow up study, for example, to look in great detail actually. Because if I may mention that case in France, I think it started in 2015 and now got confirmed by— an artwork by Jeff Koons is no longer able— the Pompidou is no longer able to exhibit this piece of artwork. So it seems to be the case that also there copyright infringement and liability risk actually really are taking artworks out of a museum. So this is definitely— it is anecdotal evidence here, so I would assume you would see reaction as well in the museum world, but yeah.
Katie Wilson-Milne: Oh, that’s very interesting. I mean, that’s obviously moving in the complete opposite direction that the Second Circuit moved in, the decision you analyzed.
Alexander Cuntz: Yeah, let me maybe clarify further on the issue of production of artists. So this is also rather anecdotal evidence, and research done by other people when it comes to the potential changes in creative artistic practice, right? This comes out of survey evidence and other places. So we haven’t inspected really— we weren’t able to track the production and the supply, all the supply and the new oeuvre that artists would create and if that would respond— that would be lovely to study actually as well, but we couldn’t cover that in this study.
Steve Schindler: Can I ask a question? Just going back to the sort of shifting of more auctions of works by appropriation artists to auction houses outside of the US, were you able to sort of consider or control for other factors? So for example, I mean whether more works in general were auctioned, say, in London versus the US in a particular period of time, or whether other factors, currency fluctuations and the like, could be a cause of the effect that you’ve observed?
Katie Wilson-Milne: Or like a major retrospective in London that coincided with the sales or just—
Steve Schindler: Right.
Alexander Cuntz: Both excellent points, I think. But we didn’t look specifically into exchange rates turbulences or something like that. That may arguably could cause that. The general point that Matthias raised before that we couldn’t systematically cover all foreign jurisdictions to see if they in this period somehow changed massively. But I think we are comparing the US to global, and I think then it makes maybe a still a useful benchmark. We also did other tests where we basically said like, let’s not compare similar artists to appropriation artist. Let’s just look at the appropriation artist and their global auction sales, limit that down.
And we basically then compared how this actually also showed this shift basically from the US to auction houses and other places. And maybe if I may add one more point in general, I think I would fully agree that the US with its fair use system may actually have been conducive for something like the appropriation artist movement to emerge in the very beginning. But that would probably be a different study, right? That’s something we couldn’t address. But I think it’s an excellent, an interesting point that in principle that’s our reading of the legal scholar literature, fair use is something conducive to this type of experimenting practice.
Still, the main point then that comes out of our results is that a fair use setup vis-a-vis a setup that is based on a list of exemptions, for example, is still not a perfect alternative in a sense. It has its imperfections. It seems that sometimes those flexibilities are actually increasing the uncertainty inside the market and that may actually change the behavior of the people that operate in that market.
Katie Wilson-Milne: Yeah. I agree that would be a fascinating kind of legal history, comparative legal history that I would love to read. I guess just on this sort of what other factors might be at play, I was thinking reading your paper again, that, you know, are there other reasons for a decrease in appropriation art activity, at least at auction in the US that might be related to these pieces being appropriation pieces, but aren’t necessarily related to legal uncertainty or fear of liability related to the prince decision?
For example, like tastes and artwork change over time, right? The market excitement about contemporary appropriation art, we wouldn’t expect it to be static forever, right? We would expect some real interest just like interest in sort of contemporary figurative art right now or portraiture and then a decline. And I mean I wondered if there was just sort of a change in market interest or taste or curation around appropriation art in this time in general that caused that decrease in interest. And it wasn’t because of the litigation risk associated with this decision. It might even have been because of this decision in terms of people’s taste changing or finding it distasteful or, you know, having a moral reaction to sort of how fair use was being expanded. But it wasn’t about the litigation risk, it was sort of a taste change. And I don’t know if you can control for that, but it just struck me that, you know, that is a plausible reason for the results you found that is not a litigation risk concern.
Matthias Sahli: Excellent point, and the short answer is no, we cannot rule out all those possibilities that you’ve mentioned. We can do probably two things that I’m thinking about. The first is that, like looking at the museum exhibitions for example, it’s an indication of that actually the people were still interested in appropriation art.
And you can do Google analytics, for example, trend search, whatever, that those kind of tools help you to understand whether people change taste. And secondly, I was thinking about the fact that we can also smaller the timeframes of our analysis. So we did various robustness check of only including some periods before and some periods afterwards. And then you can like make sure that at least not too many things change at the same time in your observation period. But your general point that actually the decision itself changed something else, but that is not the litigation risk is in the end something we cannot rule out. Yeah.
Katie Wilson-Milne: Yeah. I mean, it strikes me that’s something you could investigate. You’re right that museum shows would be one indication of what current taste is but maybe there are other indices, too, that would inform whether tastes are changing for different reasons.
Alexander Cuntz: Yeah, maybe another counter argument is a bit that if you think for example of Cariou and others showing up in those court cases, you could think this is kind of a media coverage in fact as well. Like they’re more prominent in the press now, they’re more frequently discussed. But wouldn’t you think that this effect would maybe operate in the other direction that they’re actually growing more popular and that eventually should make auction house directors schedule more auctions?
Katie Wilson-Milne: Maybe. It may be that there was some market saturation unrelated to this decision, I don’t know. But some people, this is totally anecdotally, some people certainly photographers, right, fine art photographers, some artists, certainly some lawyers, did not find this court case and all the attention around it, did not make them more interested in a Richard Prince’s work, right? It made them feel ambivalent or sort of negative about that type of work. So I just don’t know, but we observed a variety of reactions.
Steve Schindler: Right, it’s very hard to map those onto sort of market changes of course, because trying to map the art market is not something that certainly I can do.
Katie Wilson-Milne: We have not been able to do it. It constantly surprises us.
Steve Schindler: Right, right.
Katie Wilson-Milne: So maybe we should end by asking you both where your research is going next, what you’re interested in and are working on.
Alexander Cuntz: In particular, Matthias has been doing very beautiful research in two directions and we continue to work on the visual art market and one of them is actually museums. And what we are interested in is the impact of digitization on museums and on how easy it is in times of digital change to actually exhibit and make available artworks online. And that’s another direction where we’re chasing after and I think hopefully that will be something that is of use for policymakers and eventually hopefully will improve their decision making.
Matthias Sahli: Another study that we are currently working on is actually the impact of the death of an artist on the exhibition patterns and the auction results an artist have. So current research on death impact focuses almost only on auction data. And what we do is that we combine auction and exhibition patterns together and try to understand the inner dependencies of those markets, how maybe exhibitions could potentially impact price formations or the other way around. And we take the event of the death as a starting point. And this is as well interesting because copyrights are granted postmortem, and like most of the studies look at what happens when the copyright expires and works move to public domain. But what is a bit understudied is the effect of what actually changes when someone dies, because those rights are granted postmortem and in economic—
Katie Wilson-Milne: Or they continue postmortem. They’re not granted.
Matthias Sahli: Yeah. Exactly. In economics we call that a transaction cost that can appear after someone dies because the rights shift and like other things might happen and we want to understand that a bit better in a future study.
Steve Schindler: Well, when you finish that study, let us know, because we’d love to have you to—
Katie Wilson-Milne: We’ll do this again. Yeah.
Steve Schindler: We’d love to talk about it with you. This was really fascinating.
Katie Wilson-Milne: Yeah. Thank you, and it’s very interesting to compare our anecdotal observations in our practice to your economic analysis, which these studies will— we’ll also have plenty to talk about.
Steve Schindler: Right, and we’ll post the article with the show notes and thank you again.
Matthias Sahli: That would be great, yeah. Thank you.
Steve Schindler: Thank you.
Alexander Cuntz: Great pleasure.
Steve Schindler: Alright. And that’s it for today’s podcast. Please subscribe to us wherever you get your podcasts and send us feedback at email@example.com, and if you like what you hear, give us a five star rating. We are also featuring the original music of Chris Thompson, and finally, we want to thank our fabulous producer Jackie Santos for making us sound so good.
Katie Wilson-Milne: Until next time, I’m Katie Wilson-Milne.
Steve Schindler: And I’m Steve Schindler, bringing you the Art Law podcast, a podcast exploring the places where art intersects with and interferes with the law.
Katie Wilson-Milne: The information provided in this podcast is not intended to be a source of legal advice. You should not consider the information provided to be an invitation for an attorney-client relationship, should not rely on the information as legal advice for any purpose, and should always seek the legal advice of competent counsel in the relevant jurisdiction.
Music by Chris Thompson. Produced by Jackie Santos.