Free-Riding on a Dream
By coincidence, the SDNY’s rejection of ReDigi’s business model happened at almost the same time as the Second Circuit’s seeming affirmation of Aereo’s business model. This coincidence led to a certain amount of bewilderment. How could one court rule to strengthen copyright at the same time another court ruled to weaken copyright? The answer, of course, is that courts don’t—or shouldn’t—worry about the relatively weakness or strength of copyrights. They’re in the business of implementing the Copyright Act—a task that just gets harder and harder. The main lesson here is that, regardless of the copyright law’s purpose and policies, it is (outside of fair use and a couple of other things) often a highly technical law that can have counterintuitive results.
I believe this images shows part of Aereo’s array of TV antennae, each the size of a dime.
Copyright as Economic Policy
The bewilderment had two sources. First, there are those with an extra-legal interest in the strength or weakness of copyrights. For both content providers, who prefer stronger copyright (but have mixed feelings about fair use), and information providers, who see strong copyright as a nuisance, the courts went 1-for-2. Either ReDigi and Aereo should both have lost or had won, depending on the rooting interest.
For those who are used to the normal world of commerce, the cases seemed to have gone 0-for-2 because they were both counterintuitive results. ReDigi was just trying to let you do with digital “goods” what you can do with normal goods: sell them despite copyright. And Aereo was pretty clearly trying to sell a product—TV programs—that it had never paid for: a case of classic free riding. If you expect copyrighted goods to behave like normal goods, these two cases left you wondering what in the world is wrong with copyright law.
It is not too much for folks to ask that copyright law behave in an economically efficient manner. The ultimate policy behind copyright law is to incentivize creation, and incentive is an economic concept if there ever was one. We accept that the incentives can be very subtle, and we hope that the incentives operate the way markets normally operate. It is also not a good thing if copyright law is so complex and counterintuitive that ordinary people just give up trying to understand it because that just discourages businesses from getting involved, or increases the transaction costs in the form of legal fees.
So we are correct to feel a bit disturbed when copyright cases depart from our expectations of how markets should behave.
Economic Value of Secondary Markets
ReDigi was counterintuitive because secondary markets are a normal and necessary part of our economy, but the court there held that there can’t realistically be a secondary market in digital “goods.” We have already discussed the importance of secondary markets, and this recent Planet Money story does a better job of explaining their benefits. They get goods out of the hands of those who don’t need them and into the hands of those who do. A bicycle that sits unused in your garage is economically inefficient because, if you don’t ride it, no one rides it. Also, secondary markets support primary markets by reducing the risk of buying new goods*. How many fewer people would buy new cars or houses if it were illegal to re-sell a car or a house**? Also, not everyone who can afford a used car or house would be willing or able to buy a new one, so secondary sales don’t necessary displace new ones.
* Ill just come out and say that I’d probably buy a lot more music if I knew I could re-sell songs I ended up no liking. iTunes is always trying to get me to try unfamiliar music, but 90-second previews aren’t enough. I often take iTunes up on its offer, but only after a lot of study. I’ve never regretted a decision, but I’m very conservative, precisely because I don’t want buy something I can’t re-sell. The question is: would I buy enough new music to more than compensate for all of the sales I displace by re-selling the music I don’t like? To answer that question, you’d need to know how good the underlying product is, and how good iTunes is at figuring out what I’d like.
** Interestingly, one of the most important developments in Anglo-American law is the victory of the right to sell land (“alienation of realty”) over the duty to preserve land for the family, during the 17th century. It used to be that you couldn’t sell your land because you were, in a sense, just holding it for the next generation. Until you had the right to sell land to people who would do a better job of using it, there wasn’t going to be any “green revolution.” Imagine if the English gentry were still in charge of deciding what was to be done with land…
ReDigi’s secondary market is a bit different from an ordinary one. Its re-sold goods are shiny and new: they’re just as good as new goods, so there’s bound to be more displacement of new sales. The companies that set the price did so on the expectation that there wouldn’t be a secondary market, and now they’re stuck with price-points that don’t build in the effect of secondary markets. A robust secondary market should imply a higher price point, but sale displacement should drive prices down.*
* The way around that is to add valuable products or services that only the primary retailer can offer, such as free upgrades. For music, I’m not sure what those add-ons would be, but I’m sure an smart, imaginative person could up with something.
I’m Free, Free-Ridin’
Turning to Aereo, isn’t economic free-riding self-defeating? If you were thinking about building an amusement park, but you learned that you couldn’t legally control the concessions within the park, would it still be economically feasible to build the park? You can’t stop all free-riders: folks can set up concession stands just outside your gates, and competitors build their own parks nearby and take advantage of the “draw” you created. But to economists, these are bugs in the system, not features, because they discourage creation.
Aereo basically free-rides. There really isn’t a polite way to put it. It sells “free” over-the-air television programs to internet subscribers. Here’s how it works. Aereo maintains thousands of tiny TV antennae, which are attached to transcoders, which are in turn attached to a centralized DVR system. Subscribers essentially rent one or two of these antennae, which are tuned by the subscriber to a particular channel. Only the subscriber has access to the signal received by that antennae or by the attached transcoder, even though hundred of other antennae might be receiving the exact same signal. The subscriber can then view the TV channel in real time, or have it copied on the DVR system for later.
The subscriber pays for this service. Aereo doesn’t pay for the TV signals. It doesn’t pay for the TV programs. It has to pay for its own equipment, marketing, etc., but so do the free-riding concessionaires who invade your amusement park.
Larry the Cablevision Guy
Copyright law won’t tolerate this, right? Well… It does—because of a previous opinion that happens to be controlling here: Cablevision. In Cablevision, the Second Circuit held (among other things) that, once a signal or data stream is allocated to an identifiable viewer, the performance of that signal or stream is not public and, therefore, not an infringement of the public performance right—provided that all decisions about whether and when to watch or copy the program is made by that identified viewer—even though all the necessary equipment are provided by, and the necessary actions are performed by, the cable company.
It’s a matter of perspective. If you’re a cable company, you receive a signal from the program provider, then split that signal among your subscribers. You are in the business of performing the TV programs publicly and, therefore, need a license—which cable companies naturally have.* From the subscriber’s point of view, however, the performance isn’t public at all. At some point along the line, the public performance becomes a private performance. The Cablevision court decided that point was when the viewer starts to exercise control (“volition”) over the program’s performance.**
* Indeed, Congress provided cable companies with a statutory license.
** To the district court judge whose decision was reversed in Cablevision (and coincidentally was elevated to the Second Circuit and dissented in Aereo), the performance never loses its “public” nature if the transmission is commercial in nature.
Cablevision makes a certain amount of economic sense. The cable company was adding incremental value to the service it provided its subscribers, and in doing so, it incidentally infringed the content creators’ copyright. The number of subscribers remains the same—they just get a more enjoyable viewing experience. Therefore, the incentives to create are essentially the same whether centralized DVR exists or not.
Poison Aereo
Aereo, by contrast, isn’t just providing a better viewing experience; it’s increasing the total number of viewers without increasing the revenue to the content creators. It’s true that many of these viewers would not watch the programs otherwise, so the content creators aren’t being hurt. But they also represent a market for the content creators to exploit. Imagine if you sold your book exclusively in the Nashville area, where it’s very popular, so I make copies of the book and sell it in the Memphis area. True, I’m not “hurting” you because I’m not displacing any sales. But what if you want to move into the Memphis market?
To be clear, Aereo is actually providing a valuable service. That’s not the problem. To be clear, the content creators aren’t being hurt, because they weren’t exploiting the market that Aereo was exploiting. The problem is that the content providers aren’t being paid for the expanded market.
Isn’t that the content providers’ problem? Isn’t it their fault that they weren’t exploiting Aereo’s market? Isn’t it more economically efficient to let third parties exploit copyrights (for free) where the rights holders aren’t involved?
Gossamer Dreams as Incentive to Create
I don’t think so, and the reason I don’t think so has to do with the nature of the economic incentive to create. While there are some works that are valuable even before they are made—the next sequel to Star Wars comes to mind—most creative works have only indeterminate potential value. It’s depressing but essential to understand: most acts of creation fail* in economic terms. What matters is the chance it might not only succeed, but succeed beyond your wildest dreams.
* They might succeed in other ways, of course.
When you write a song in Nashville, it will likely be worth nothing, but it could be worth something, and it might be worth a lot. It’s the difference between a song that is never recorded, a song that’s recorded on a middling country records, and a song that ends up on a Garth Brooks record. The unrecorded song might be much better than the Brooks song, but luck has a lot to do with it. The dream of “hitting it big” is a big part of the incentive to create.
A better example might a novel. It might never be published. It might be modest-successul. Or it might be modestly-successful but adapted into a successful movie. The movie market might not be one you ever seriously contemplated. It certainly wasn’t your intention. But if someone wanted to make a movie, to exploit a different market that you hadn’t exploited, that someone would have to pay you right? Even Garth Brooks has to pay his songwriters….
If we let Garth Brooks or a Hollywood big shot freely exploit your creativity simply because you didn’t or couldn’t exploit a market they’re positioned to exploit, we’re cutting off a major component of the incentive to create. “It’ll probably gross $25,000, but you never know….”
And that’s what Aereo is doing, in a smaller and less dramatic way: it’s taking away an unexpected market. Ever so slightly, it’s depressing the dream of “hitting it big.”
No My Job (Unfortunately, it’s Congress’ Job)
But that’s not the same thing as saying that the ReDigi and Aereo courts got it “wrong.” It’s not the court’s job to make sure that every copyright decision make economic sense, or even follow the basic policies underlying copyright law. That job belongs to Congress. With the obvious exception of fair use*, all courts can do is implement Congress’ implementation of the basic policy underlying copyright law.
* With fair use, courts can bring in all kinds of policy arguments, including at least one very explicitly economic one.
It might be nice if courts were able to bring some economic rationality to some of these decisions, but then, smart people can disagree about what is most economically rational, and we’d end up several different “versions” of copyright law, depending on the judge. So maybe it’s better to just leave it up to Congress.
If only Congress would. Congress hasn’t substantially updated copyright law since 1999 (the Digital Millennium Copyright Act), and it hasn’t revamped the Copyright Act since 1976. The 1976 Act* actually took more than 10 years of hard work to hammer out, and Congress probably only acted because treaty obligations were threatening to leave the U.S. behind the rest of the world. The 1976 Act was nevertheless a huge achievement, forward-looking**, careful in its balancing of rights, responsive to certain interests groups, yet still cognizant of the average consumer.
* It actually was enacted in 1978.
** I think it’s really impressive that Congress knew that software would be a problem, set up a special committee (of honest-to-goodness experts) to deal with it, and later implemented *most* of the committee’s proposals.
Since 1978, the Internet happened. Congress addressed that in 1999. Since then, the digital economy happened, and Congress hasn’t really responded. So, don’t hold your breath, and keep expecting difficult, counterintuitive court decisions.
Thanks for reading!
The problem with these decisions is that they make the mistake of focusing on the public performance right, but that is only one of the exclusive rights of a copyright owner. A copyright owner also has the exclusive right to control distribution of copies—indeed, the distribution right is at the heart of copyright law. The ReDigi court got this right, while the Aero court got this wrong—and if you focus on the distribution right there is nothing counterintuitive here. The First Sale doctrine applies to physical copies of copyrighted works, but it obviously cannot apply to digital copies because a digital copy cannot be shared without physically making another copy. If the First Sale doctrine applied to digital works, the Secondary market would swallow up the primary market. But the market effect is not the key point—the key point is whether a copy must be made to share a digital work. By its very nature, a work in digital format can only be shared by allowing a copy to be downloaded. Thus, a copy is made. Just as the First Sale Doctrine does not allow the person who made a CD to make additional copies of the CD for distribution, the First Sale Doctrine clearly does not allow the purchaser of a digital work to make additional copies of the digital work. Since a digital work cannot be shared without making a copy, the First Sale Doctrine does not work in the context of digital works.
The mistake of the Aero court’s majority was in concluding that no additional copies of the work are made when they are transferred through the individual mini-Antennas. That is technically wrong. Even these transmissions through these mini-Antennas involve making a distributing a copy. Thus, regardless of whether the public performance right was violated, the distribution right was violated.
DVR’s in contrast, do not distribute new copies—they simply change the time of distribution. Reasonable people can disagree whether changing the time of distribution interferes with the right of distribution, but if someone already has a license to distribute (i.e., cable company), it is not unreasonable to conclude that this license includes the right to change the time of distribution.
Maurice–
Thanks for reading, and thanks for the great post!
The distribution right is defined as distributing “copies or phonorecords,” and those terms, as defined, require fixation in a tangible medium. Even if we accept the idea that the distribution of a digital good doesn’t literally have to be a distribution of the “copy or phonorecord”—an intuitive result, I’ll be the first to admit, even if it’s not a literal reading of § 106(3)—the first-sale doctrine is pretty clearly an exception to any distribution as defined by § 106(3). It says so. Thus, I can’t agree with a blanket proposition that the distribution of a digital item via a network CAN be an infringing distribution but CAN’T EVER come under the first-sale doctrine. As I point out in my ReDigi post, the court didn’t have to go this far because the statute has its own exception that takes care of the problem: the copy/phonorecord to be distributed has to be a legal copy. Once the court held that the uploaded copy was not a fair use, the first-sale doctrine ceases to be applicable. ReDigi/the customer is no longer selling the legally acquired copy but an illegal copy.
My point wasn’t that these decisions are counterintuitive to copyright lawyers. I’m not sure copyright lawyers even know the term, or what it feels like. My point is that consumers, lawmakers and other important non-specialists find these decisions counterintuitive, and that’s concerning because, ultimately, they (as voters) will eventually influence the law (and must comply with a law that they don’t understand very well). I would argue that, for consumers, this is still a counterintuitive result for two reasons. First, consumers just don’t think all reproductions are created equal. If you have to make a temporary copy of something in order to exercise your ancient right to alienate chattel, so be it, from this point of view. Second, not everyone realizes that what happening is a reproduction. Unlike copying your vinyl records to tape so you can listen to the music in your car (as we used to do in the 1980’s), the mechanism of uploading a file to a server is hidden from the user. So, I still maintain that ReDigi is a counterintuitive—if arguably correct (at least, not obviously wrong)—result.
I’m not sure I understand the CD analogy. The ReDigi court assumed for purposes of summary judgment that the customer could not retain a copy of the uploaded file or upload a file that had originated from a CD. Thus, this isn’t the same situation as someone who makes copies of a CD and distributes them to friends, which plainly violates the distribution right because the copies are not legally acquired. A better analogy would be if I made a backup copy (technically “phonorecord”) of my CD (in case my original one got busted—admittedly, not too many people will be obsessed enough to do this), the original phonorecord did, in fact, get busted, and I gave away my backup copy. That might be OK—but as the ReDigi court was at pains to point out, that’s not what happened. What happened was that I made a phonorecord of my CD, gave that phonorecord away, then deliberately broke my original CD (I can’t imagine why anyone would do that).
As for Aereo, I take it your point is that the decision is simply legally wrong, and that if the court had applied the correct principles, the result would have gone the other way, and thus it would no longer be counterintuitive. The first question we have to ask is: why didn’t the networks raise the distribution right at all? They have pretty good copyright lawyers. You can read their brief here: there’s no mention of the distribution right (just “distribution channels”). If you’ve read my Online Music Series, you know that I’m not unsympathetic to your point of view. BUT, first things first. Even under London-Sire, there still has to be a copy/phonorecord resulting on the other end of the transmission. I don’t think it matters that an intermediate copy is made (and if it’s just a buffer copy, then it might even be fixed under Cablevision): what matters is what happens on the receiving end. Is that what happens, with streaming over the internet? I’m honestly not sure. If it’s a progressive download, then it seems to me the subscriber ends up with a big enough chunk of the work on her computer to possess a big enough chunk of the work to have a copy of it. If its true live streaming, maybe not. I’m betting it’s a progressive download because it’s supposed to have all the cool DVR-like features.
So to return to the main question: why don’t the rights holders pursue the distribution angle? Perhaps because it was a preliminary injunction, and they didn’t have enough time to investigate? Since all that’s happened is the affirmation of a refusal to grant a preliminary injunction, how the plaintiffs will have the chance. Do they simply not realize that it’s a possibility? You have to admit that, even if it ultimately leads to an intuitive results, it’s a pretty counterintuitive way to get there. We tend to think of distributions as resulting in a long-term possession (through sale or lease) of the copy/phonorecord, not a temporary cached copy that most users can’t even find on their computers. But if you accept the London-Sire view on digital distributions, then you have to accept that even temporary cached copies/phonorecords can be distributions….
Again, thanks for the thought-provoking post!
Oh, there might be another reason no one would want to interpret retransmission of a signal through a network as a violation of the distribution right: that basically describes a cable system. See § 111(f)(3). Cable systems rely on statutory licenses to retransmit signals through their cable networks, but those licenses are defenses only against violations of the public performance and public display rights. See § 111(c)(1). Either cable systems don’t “distribute” works by retransmitting them, or Congress has basically announced that retransmission isn’t distribution (without more). I seem to recall that the ivi case was about whether ivi was a cable system….
Note, this doesn’t apply perfectly to the Internet, because streaming isn’t the same thing as cable.
Rick,
Great piece. (I was also on the Planet Money piece.) I’ve been following the ReDigi saga for some time (see for example http://copyrightandtechnology.com/2013/04/01/capitol-records-prevails-in-redigi-case and http://copyrightandtechnology.com/2011/11/15/redigi-gets-riaa-nastygram/ from my blog). Their business model is itself somewhat counterintuitive: for all of the talk about leaving hardcopy business models behind instead of trying to shoehorn them into the digital world, ReDigi does exactly the latter. I think digital resale is going to happen, whether or not the copyright holder’s permissions are required. The question is, absent the possibility of getting the same product at a lower price, will anyone care.
Thanks, and thanks for reading! An honor to appear on the same program. Agreed that there is irony in ReDigi’s business model, how it spends so much energy trying to re-create a hardcopy mindset. Some of that is inevitable: everyone would agree that, under any digital first-sale doctrine, you can’t be left with a copy once you’ve exercised the re-sell right. Since that’s not (really) a problem in the hardcopy world, it makes sense for ReDigi’s model to mimic the hardcopy model. Nevertheless, it couldn’t get around the uniquely digital problem that transfers are also reproductions…
Any thoughts about whether the content industry would be better or worse off, on balance, with a robust (but properly regulated) secondary market for digital content? I mention in the post that I would certainly buy more new music if I knew I could re-sell it, but we’d need to balance that against the new music I wouldn’t buy simply because I bought it “used.” My *hunch* is the content industry would actually be better off. The fact that the “used” goods are as shiny-and-new as the new goods would (assuming a robust market) be reflected in higher prices for the used goods (i.e., ReDigi’s 69¢ wouldn’t last long), which would mitigate the undercutting, and there wouldn’t be enough “used” goods to make a significant difference, especially for really good content (because it would be re-sold less often).