Note this post was updated the same day it was posted around 9 p.m. my time to address issues raised in one of the comments. I had reversed the polarity on Teleprompter and forgotten that Cablevision pretermitted (i.e., punted) the performance issue on grounds that the performance of a copy of a program earmarked and segregated for the subscriber wasn’t a public performance.
Outlier or Game Changer?
So the Supreme Court handed down its decision in American Broadcasting Cos. v. Aereo, Inc. on the first day of my vacation because, of course it did. It was building up to be perhaps the most significant Supreme Court decision on copyright law in several years. Sure, last year’s Kirtsaeng case could have had far-reaching effects—the “parade of horribles”—but, in the end, all the Court was doing was unknotting a small, terribly-drafted part of the Copyright Act that related to imports. Boring! By contrast, Aereo forced the Court to examine some fundamental concepts of copyright law, could have had far-reaching effects (especially on “cloud computing”), and involved something everybody gets: watching television!
Oh, and once again, a decision I blogged about—and described as “so wrong (yet so right)”— has been reversed. But unlike last time, I feel no need to eat humble pie.
But I’ve been back from vacation for a while, and still I’m only just now writing about Aereo? Honestly, it’s taken me a while to digest it. It’s easy enough to understand on its face: Congress deliberately wrote the current Copyright Act of 1976 to make 60’s and 70’s era CATV systems illegal, and Aereo so closely resembles a CATV system that it, too, must be illegal. There are lots of reasons why that reasoning is or isn’t flawed, but that’s beside the point. A lawyer criticizing the Supreme Court (which can be fun sometimes) is bit like shouting at the wind.
Technical vs. Squishy
What’s been hard is fitting this decision into the overall structure of copyright law. Although Aereo involved a fundamental question of copyright law—what does it mean to cause an infringement?—the Court deliberately ran away from that question. Perhaps my problem is that, ever since I started teaching copyright law, I see copyright law as systematic, because that’s the way you teach it. And that’s the way you analyze it, when you are working on a copyright case. And Aereo just doesn’t fit. So the question becomes: is Aereo an outlier, or is it a harbinger of some broader change to the structure of copyright law?
For the most part, copyright law is a technical law, and like any technical law, it often leads to anomalous results. By “technical” law, I mean the law mostly consists of a series of discrete factual issues, what lawyers might call “bright lines.” On any given issue, you have a good idea on what side of the bright line you stand (assuming sufficient facts). The challenge is working through the thicket of rules in the correct way—and figuring out what those rules are in the first place, where there is disagreement among courts or just plain confusion. You also have to resist your intuition or some inherent sense of justice (in your analysis—not in your opening statement!). You have to trust where the analysis leads, which doesn’t always make sense.
For example, you don’t need a license to read or lend a book, but you do need a license to load software into RAM. So, while a bookstore can’t take back a book it sold you, Amazon can prevent you from accessing books on your Kindle that you bought (but don’t really “own”).1 This makes sense only to a copyright lawyer. A painstaking reproduction of an Old Master receives no copyright protection, but a chance photograph or film of an important historical event does. We put up with these counter-intuitive results because we desire consistency: that, once we work our way through the thicket of definitions and rules, we’ll all arrive at the same place.
The opposite of a technical law is a “rule of reason.” These are squishy concepts, often necessary because bright line rules just aren’t up to the task of arriving at fair results. You wouldn’t just be getting the odd anomalous result, but you are getting lots of weird and manifestly unfair results. Such rules are capable of taking into account many factors, weighing them, and hopefully arriving at a just result. We usually leave it to the jury to figure these things out. If, for example, you swerve to miss a cat crossing the road, hit a lamppost, which falls through somebody’s front window, the jury will have to figure out, first, whether it was reasonable for you to swerve out of the way of the cat and whether you did do reasonably. How fast were you going, were paying attention, what was the weather like, what precautions did you take because of the weather, how cute was the cat,
are jury members cat or dog lovers, etc.? Then the jury will have to determine whether you’re responsible for the damage to the window: is a broken window a reasonably foreseeable result of hitting a lamppost with such-and-such force? While it might be nice to have a rule stating that you cannot be liable for falling lampposts if you were traveling the speed limit or less, because that would be an easy rule to apply, such a rule wouldn’t be a very fair to the homeowner if, say, you were driving like a moron.
The problem with squishy rules like these is that they are difficult to apply consistently and, thus, results of such rules are difficult to predict. You often don’t know whether what you did was right or wrong until the jury returns a verdict—which is a bit too late if you were wondering how to conduct yourself legally in the first place. You can see the trade-offs here: you can certain but at the cost of anomalies, or you can have flexible but at the cost of consistency.
Copyright law, although mostly a technical law, has its fair share of flexible rules. In fact, two of its most important concepts are very, very squishy, indeed. Fair use is notoriously flexible, inconsistently applied and difficult to predict the outcome of, but no set of technical rules would lead to consistently fair results. Substantial similarity, which is a requirement for copyright infringement, requires a jury to determine whether you took too much of what is too important to the underlying work (the so-called “quantitative/qualitative” analysis). How much is too much, and what is important to the underlying work? Ask the jury.
The question that Aereo leaves us with is: is there a third major squishy concept in copyright law?
By now, you probably know how Aereo works, so you can skip the next couple paragraphs if you want. Imagine if you had a friend in New York city whose apartment gets great TV reception, but you live
far enough out of town that you can’t get New York TV stations only crummy Albany ones2 somewhere you can’t get TV signals very well. You place a TV antenna in your friend’s apartment, run the signal through a digitizer, buffer the digital output on your friend’s computer, then stream the stored output over the internet through a little gate through your friend’s firewall, a gate that you and you alone have the credentials to access. Your friend doesn’t want her computer streaming your content all the time, so you rig it so the computer only receives and streams the content when you tell it to through the gateway. Oh, and to cut down on bandwidth, you don’t want to receive ALL the content all the time, only the content you’re watching, so you install a TV tuner on your friend’s computer that you can change the channel and output to the streaming device only the content on that channel.
Now let’s say that you’re friend is enterprising. She has purchased lot and lots of antennas, and because the signal strength is so strong, she discovers the antennas don’t have to be very big. (Maybe she moved somewhere with an even better signal.) She rents the antennas to you and other customers, for a fee. Each antenna has its own tuner and digitizer, and the output is kept carefully segregated so only the antenna’s “owner” (renter, really) can access it. [Edited to add: Somehow (maybe using IP geolocation?), she only allows customers who are in the New York City TV market.]3 She actually has more customers than antennas, but that’s OK because not all customers are using their antennas at the same time, so she sets up a system that assigns a random antenna to a customer when the customer demands access. A given customer might end up using multiple different antennas over time, only “renting” a particular antenna for the period in which the customer is seeking access.
From the customer’s point of view, the experience isn’t quite the same thing as just watching TV. For one thing, unless you have an Apple TV, Roku or some other interface between your computer and your TV, you can’t watch it on your TV, just on your computer. But, hey, in the YouTube age, the computer is the new TV, right? But, still, you get the same content that your friend gets, for just a few bucks a month. Your friend stands to get very rich. She just needs to invest in the equipment and keep it running. She doesn’t have to pay for the content. Who knew there was a market for “free” over the air TV?
From the perspective of rights holders, it’s clear as day that Aereo is illegal. Aereo is making money from content without compensating the rights holders, so it’s got to be illegal.4 Some of the analysis is roughly the equivalent of Calvin yelling at Hobbes, “You can’t do that!”, but more seriously, they argue that Aereo is violating the “public performance” right.
What Was Hip in the 1960’s and 70’s
The public performance right has a couple of strange wrinkles. First, it (along with the public display right) is limited to public uses of the copyrighted work (more on that later). Second, along with the display right, it’s the only right that can (potentially) be simultaneously violated directly by two entirely different parties. This is because, when a signal carrying a copyrighted work (like a TV show) is transmitted, both the transmitter and the recipient are said to be “performing” the work at the same time.5
It wasn’t always like this. It used to be that only the broadcaster was said to “perform” the transmitted work. How it got to be like this is actually central to the Supreme Court’s reasoning in Aereo, so let’s go back in time to the 1960’s and 70’s…
Back then, it was not uncommon for communities to have little or no access to television. What they would do is go some place where the signal strength was high (and preferably there were a lot of channels, like five or even six!), set up a big antenna there, run some cable back to the community, then amplify and divide the signal among subscribers, and run more cable to the subscribers’ homes. The subscribers would normally pay a modest subscription fee to make all this possible, and they’d have access to whatever the community’s antenna picked up. (This was all analog, obviously.) This was known as “community access television” or “CATV”. It was also the precursor of cable TV systems.
Twice—in a 1968 case called Fortnightly and again in a 1974 case called Teleprompter—broadcasters brought copyright suits against CATV operators all the way to the U.S. Supreme Court, on grounds that the operators were infringing their performance right in the broadcasts by re-broadcasting them. Both times, the Supreme Court held that broadcasters perform transmitted works, not viewers, but CATV systems were more like viewers than broadcasters because they did not choose what to broadcast6 but were mere conduits. It did not matter whether the CATV operators were really commercial enterprises, selling their content to people outside of the “community” or cross-selling their content to other CATV operators.
Removal of a Defense Does Not Automatically Create Liability
As luck would have it for the broadcasters, in the 1960’s and 70’s, Congress was waist deep in drafting a brand new copyright act—the Copyright Act of 1976, which is the one in effect today. Congress absolutely wanted to overturn Fortnightly and Teleprompter. At the same time, Congress looked with favor up on the nascent cable industry, which had been relying on Fortnightly and Teleprompter. So Congress crafted a compromise:
- The definition of “perform” would be written so that transmitted works are performed by both the broadcaster and the viewer, so it no longer mattered whether CATV operators were viewers or broadcasters, but
- A cable company won’t have to negotiate with every copyright owner, but instead will have a defense against claims of copyright infringement so long as it pays a statutorily-defined royalty (plus some other terms and conditions)7. This defense is known as § 111.
Here’s how the committee in charge with drafting the Copyright Act of 1976 put it:
In general, the Committee believes that cable systems are commercial enterprises whose basic retransmission operations are based on the carriage of copyrighted program material and that copyright royalties should be paid by cable operators to the creators of such programs. The Committee recognizes, however, that it would be impractical and unduly burdensome to require every cable system to negotiate with every copyright owner whose work was retransmitted by a cable system. Accordingly, the Committee has determined to maintain the basic principle of the Senate bill to establish a compulsory copyright license for the retransmission of those over-the-air broadcast signals that a cable system is authorized to carry pursuant to the rules and regulations of the FCC.
But not all performances are infringing. The performance must be public. This is why it’s not such a big deal when you turn on your TV set at home and watch with your family and friends. You’re “performing” the TV show, but you’re not doing so publicly. This is important.
Will-Power and Causation
So Aereo is a broadcaster of copyrighted TV shows, and therefore, must be publicly performing those TV shows, right? Well, not quite. All Congress did was deliberately overturn Supreme Court decisions that, in effect, immunized CATV systems from direct copyright liability. In other words, Congress effectively removed that immunity, but CATV systems aren’t copyright infringers simply because the broadcaster-viewer distinction had been removed. They must do something to the copyrighted work in question. And that something is “volitional conduct.”
All “torts” (bad acts that the law will redress, like copyright infringement or negligence) have some sort of causation element because you can’t really be at fault for causing harm if you didn’t somehow bring about the harm. It’s always been understood that copyright law requires a more stringent showing of cause than other torts, perhaps because direct infringement doesn’t have an intent or knowledge requirement. Under a normal causation standard, you’d be liable for just lending your camera to a friend who, unbeknownst to you, was planning to use it to pirate a movie. Since you can infringe copyright unwittingly, your ignorance of your friend’s intent would be no excuse.8 So what saves you is that you didn’t do enough to cause the eventual infringement.
But how much is too much? Recently, in a lower-court decision called Cablevision9, volitional conduct was defined as actually carrying out the infringing act. In that case, the defendant was a cable company that provided its customers a service that essentially duplicated the functionality of a DVR system, except it was in-house (i.e., remote from the customer). The court held that the cable company didn’t engage in volitional conduct when its equipment copied the copyrighted programs10, even though the cable company set everything up for the customer, it was still the customer, not the cable company, who decided whether and when to use the service, what programs to copy, and when to watch the program (privately).
I call this the “Who pressed the button?” theory of volitional conduct, and while it is strict, it has the virtue of being easy to apply. Find the person who pressed the button (or the equivalent), and you’re done—there’s your direct infringer. It’s a really technical rule (and fits in with copyright’s many other highly technical rules), and like all technical rules, it will lead to anomalies—hopefully not too many or too bad (or too annoying to those with lots of pull with Congress). We have to expand this definition a little bit, though, because we also know that you’re a direct infringer if you command, direct or force someone else to perform the infringing act, e.g., a boss telling an employee to make illegal copies of a newsletter, say.
I Happened to Notice that the House Is Currently Unoccupied, the Back Door Is Unlocked, and There Is a Fortune in Jewelry Upstairs. I’m Not Saying You Should Rob the House, of Course…
But what about an invitation to infringe? Let’s say I have discovered a new song that I really like, purchase it (legally) and download it to my MP3 player (again, legally). I then offer my player to anyone I know, encouraging them to listen to the song I love. I don’t limit myself to just friends and family but offer the song to co-workers, neighbors, a guy on the bus—enough people that the performance can’t be regarded as private.11 I don’t force anyone to listen to it; I’m not even very insistent. I just tell them it’s a great song and they can listen to it on my player if they want.
Like me in this hypothetical, Aereo (and Cablevision) don’t “push the button”; they don’t command you to push the button; but they invite you to.12 Does that cross some kind of line? And if there’s a “line,” where do we draw it?
Perhaps volitional conduct just isn’t the law? After all, it has never been formally endorsed by the Supreme Court. Terry Hart, the Director for Legal Policy for the Copyright Alliance13, has argued for some time that volitional conduct is, in fact, just short-hand for a more fundamental concept called “proximate causation.” Remember my hypothetical about the falling lamppost? The question about whether you should be liable for the damage to the window is an example of proximate causation. Hart’s argument is that copyright law should be treated like any other kind of tort, and torts (especially those without intent or knowledge requirements, known as “strict liability” torts) are limited (or extend out) to “proximate causation,” which is basically defined as: whatever is a reasonably foreseeable14 consequence of a (bad) act.
The great advantage of using proximate causation is that lawyers are very familiar with the concept. We learn it in our first year of law school15, and we never stop using the concept. The great disadvantage of using proximate causation is that nobody really understands it. It’s a jury question, and juries are (1) black boxes (i.e., you don’t know how they arrive at their verdicts, what they though was significant, etc.), and (2) not required to be consistent, just not insane. Over time, you can make educated guesses, but only if you have enough data. There are so few jury verdicts in copyright cases that we’ll never really have enough data. It would become the third major squishy area of copyright law.16 And squishy areas tend to favor plaintiffs because they create uncertainty among potential defendants, who will have to be extra careful to avoid infringement.
Hart argues that proximate causation isn’t a hard question in Aereo. It should surprise exactly nobody that Aereo’s customers will use Aereo’s service in the manner that Aereo instructs.17
Now, normally, we wouldn’t be bending ourselves into knots trying to figure out whether what Aereo is doing is direct infringement, because, normally, supplying equipment and explaining people how to use it to infringe is also a form of infringement—contributory infringement. And contributory infringement is just as bad as direct infringement. Under this doctrine, you are liable for copyright infringement if (1) you contribute to another’s infringement, (2) knowing that the other person’s conduct constituted infringement.
But contributory infringement didn’t help the plaintiffs in Aereo 18 because you have to contribute to the other person’s infringement. In other words, what the other person is doing must itself be direct infringement. And while the viewers are certainly performing the TV shows under the current definition of “perform,” they aren’t doing so publicly. Mostly what they are doing is watching TV by themselves or with their friends and family. Unless they’re projecting the TV program for the whole neighborhood (like for a block party), they aren’t engaging in public performance.
Thus, what we have is: one party that doing something publicly, but that’s not a performance; and another party performing something but not publicly.
What Did the Court Say? C-C-CATV!
Well, why are we having an academic debate about volitional conduct, proximate causation, exploding fireworks and falling scales19, when we have a Supreme Court opinion that addresses these issues? Sure, let’s see what the majority said about these crucial issues.
Let’s run a few text searches…
You know what? The majority says nothing about volitional conduct. Well, it’s an esoteric concept, and maybe nobody brought up? No, the dissent is all about volitional conduct. The issue must have been laid before the Court. Weird. OK, but surely proximate causation—which is well known to every lawyer—is discussed? Nope.20
The majority does address the fact that Aereo’s customers, not Aereo itself, determine what program Aereo’s equipment transmits to the customer. But the majority doesn’t understand this to be an argument about volitional conduct or even causation. To the majority, this is an attempt to distinguish Aereo from the CATV systems upheld in Fortnightly and Teleprompter, which holdings Congress intended to overturn with its definition of “perform” in the current copyright act. From this point of view, it’s a distinction without a difference because the viewer experience is the same either way—everything else is just “behind the scenes.” This reasoning is circular, because Congress didn’t intend to make all cable systems illegal, only to remove what was in effect an immunity to liability that cable systems had enjoyed. From the point of view of causation, who “presses the button” might really matter.
Shouting at the Wind
The majority’s refusal to consider volitional conduct or causation is really gobsmacking. It’s not as though the concepts of causation and volitional conduct weren’t put before the Court—they were. And, without a discussion of causation, the majority’s analysis is missing a crucial step. Just because cable companies are no longer effectively immune to copyright liability doesn’t mean they are automatically liable—you still have to show that they did something that was the legal cause of copyright infringement. Maybe Aereo did, but you can’t tell from the Court’s opinion. Furthermore—
Oh, there I go. Shouting at the wind. It is a difficult temptation to resist, isn’t it? My job isn’t to tell the Supreme Court how to do its job, but to figure out whether and how Aereo has changed copyright law. There are three main possibilities that I currently see (and I’m happy to consider others!):
- Hart is correct, and the Court is secretly applying his theory of proximate causation. In other words, Aereo is a major game-changer. The problem is that: the majority had every opportunity to use this well-known phrase and failed to use it once. Also, there are some theoretical problems with this model.
- Volitional conduct survives intact, but we expand our definition of direct infringement enough to encompass Aereo’s holding, i.e., direct holding requires either (a) volitional conduct or (b) that your service behaves a lot like a cable company’s. In other words, Aereo is an outlier. The problem is that: this sounds really stupid.
- Volitional conduct has been applied too narrowly, but proximate causation is too broad for copyright. In other words, Aereo is a clarification of copyright law, not a major revision, but not an outlier either. The problem is: the majority doesn’t suggest anything like this and provides no clues for us to figure out what the real causation standard should be.
Which is it? Got me. I’m fairly certain it isn’t no. 1 because proximate causation is such a common concept that the Court would have said “proximate causation” if that’s what it meant. No. 2 is quite possible, since the Court emphasizes the narrowness of its holding (in a transparent attempt to protect cloud computing). When I’m wearing my law professor hat, No. 3 is most appealing because it would fit Aereo into the structure of copyright law, and it would give us lots of room to argue about how volitional conduct should be understood.
As a copyright practitioner, however: AARGH!
Oh, you wanted me to talk about cloud computing and some other stuff? Maybe next time.
Thanks for reading!