Fox v. Dish Revisited or: How I Learned to Stop Worrying About Aereo and Love the Cloud

On January 15th, the news broke that Fox and Dish had settled their carriage dispute, and Dish will now carry Fox’s news and business channels as part of its satellite service. That same day, the parties jointly requested that District Judge Dolly M. Gee put their copyright and contract battle on hold. They told Judge Gee that it’s “highly likely” that they will be able to reach an agreement later this year. Just three days earlier, Judge Gee had issued her decision on the parties’ cross-motions for summary judgment. That opinion, in redacted form, was made public last Tuesday. On the copyright front, Dish was found to have infringed Fox’s reproduction rights with its “Quality Assurance” copies. These were internal copies Dish made to manually mark the beginning and end of programming segments for its AutoHop commercial-skipping feature. Judge Gee brushed aside Dish’s argument that fair use excused its copying: “The QA copies are not transformative.”

On the rest of the copyright claims, however, Dish came out smelling like roses. Judge Gee held that Dish’s Hopper Transfers feature, which let users download recordings from their set-top boxes, its PrimeTime Anytime feature, which let users record the entire primetime lineup of the four major broadcasters, and its AutoHop feature, which let users skip commercials when watching their PrimeTime Anytime recordings, were all noninfringing. Judge Gee also held that Dish’s Hopper with Sling set-top box and Dish Anywhere feature, which let users stream live and recorded content from their set-top boxes over the internet to their personal computers and devices, did not infringe Fox’s public performance rights. For this latter determination, Judge Gee relied heavily on the Supreme Court’s opinion in Aereo from last year.

I wrote about Fox’s public performance claim in an earlier post, and I argued that Dish’s Hopper with Sling combined with the Dish Anywhere feature is noninfringing under the Supreme Court’s reasoning in Aereo. After the Court granted certiorari in Aereo, there was a lot of concern about how a decision against Aereo would affect the public performance analysis, particularly with respect to the cloud. Many of Aereo’s supporters maintained that Aereo’s service was indistinguishable from other cloud computing services, such as Dropbox. During oral argument before the Supreme Court, the cloud once again took center stage as the jurists opined about the broader repercussions of a decision against Aereo. In the Aereo opinion itself, Justice Scalia in dissent complained that the majority’s test “will sow confusion for years to come.” Despite these doomsayers and naysayers, Judge Gee quite easily applied Aereo to find that Dish did not infringe Fox’s public performance rights. Moreover, she did so in a way that provides robust support for the cloud.

The public performance test from Aereo has two elements: (1) whether the service performs, and (2) whether that performance is public. Unless both elements are met, the service does not publicly perform. The test is particularly simple to apply because of the second element. In a typical case, such as here with Fox and Dish, it’s easy to determine whether a performance is public or private—even if we don’t know who is doing the performing. Performances are private, Aereo teaches, when they are transmitted “to individuals in their capacities as owners or possessors” of copies of the underlying works. Performances are public, by contrast, when they are transmitted to those “who lack any prior relationship to the works.” If the performances are private, it doesn’t matter who performs them—whether the service or its users—as either way they are noninfringing.

While the second element is easily applied, the first element proves a bit trickier. The Supreme Court in Aereo reasoned that Aereo performed by analogizing its service to a cable system: Aereo received and retransmitted broadcasts, just like a cable system, and since a cable system performed, Aereo did too. Here, the Court’s analogy to a cable system works exceedingly well for Dish’s initial satellite transmissions to its users. There’s no doubt that Dish publicly performs as it beams down content to its users in the first instance. Dish has licenses for these transmissions, of course, and consequently they were not at issue here. Instead, Fox’s claims were directed at transmissions that occurred after the users had already obtained the content from the earlier satellite transmissions. The identity of the performer in this context turned on the volitional conduct test, which looks at who caused the transmissions to occur.

In his Aereo dissent, Justice Scalia argued that the majority’s two-factored test really had but one element: Once the Supreme Court determined that Aereo performed, that performance was necessarily public. On his view, the second factor collapsed into the first. If anything, however, I think the reverse is likely true—especially in close cases. Whether the service causes the transmissions to occur turns, in no small part, on whether the service is transmitting content to which the user has a “prior relationship.” In other words, if the performances are public, they’re presumably performances caused by the service provider—it’s the one supplying content to members of the public in the first place. Moreover, this holds true no matter who pushes any button.

Perhaps proving Justice Scalia’s broader point, Judge Gee determined that Dish did not perform because it did not “receive programs that have been released to the public and then carry them by private channels to additional viewers in the same sense that Aereo did.” The streaming enabled by Dish’s Hopper with Sling and Dish Anywhere feature “takes place after the user has validly received it,” Judge Gee noted, “whereas Aereo transmitted its programming to users directly, without a license to do so.” The license, of course, was irrelevant to whether Aereo performed, but Judge Gee rather astutely recognized that the Aereo majority applied its own version of the volitional conduct test when it analogized Aereo to a cable system:

The Aereo majority’s analysis can be reconciled with the volitional-conduct requirement for direct infringement. The Aereo Court distinguishes between an entity that “engages in activities like Aereo’s” and one that “merely supplies equipment that allows others to do so.” Id. at 2504. The Court held that a sufficient likeness to a cable company amounts to a presumption of direct performance, but the distinction between active and passive participation remains a central part of the analysis of an alleged infringement.

The Aereo Court cited three points of comparison that established Aereo’s “overwhelming likeness” to traditional cable providers: (1) Aereo sold a service that allowed subscribers to watch television programs almost as they were being broadcast; (2) Aereo used its own equipment, housed in a centralized warehouse, outside of its users’ homes; and (3) by means of its technology (antennas, transcoders, and servers), Aereo’s system received programs that had been released to the public and carried them by private channels to the additional viewers. 134 S. Ct. at 2506.

Judge Gee nailed it. The Supreme Court’s comparison of Aereo to a cable system was its way of telling which side of the active-passive line Aereo was situated, that is, whether Aereo’s volitional conduct was sufficient such that it was directly performing. Dish’s technologies weren’t transmitting content to “additional viewers,” Judge Gee reasoned, so Dish wasn’t another Aereo. Note, however, that these “additional viewers” were simply members of the public, and they had no “prior relationship” to the particular content being streamed. This is the second element seeping into the first: She knew that Aereo performed, while Dish did not, based in part on whether the transmissions were “to the public” under the Transmit Clause. The volitional conduct turned, at least in part, on the publicness of the transmissions. Whether Dish performed was a function of whether the performances were public.

As computing moves further into the cloud, distinctions drawn on whose-hardware-did-what and who-pushed-what-button become harder to discern. Figuring out who caused a transmission to occur when it involves hardware and software utilized by both the service and its users takes on a metaphysical dimension. This is especially true as courts are wont to say it’s one party and not the other. The Aereo majority at least had a known quantity—a cable system—to compare with Aereo. Usually, the volitional conduct analysis is far less certain, and courts end up analogizing to other fact-patterns gleaned from the decisional common law. This is exactly what Judge Gee did. After first applying the specific cable-system-lookalike version of the volitional conduct test from Aereo, she then turned to its more general form. This was necessary because the Aereo test is merely sufficient for determining whether a service directly performs, but it’s not necessary. Other services may directly perform as well, even if they’re not just like a cable system.

The trick is figuring out where to draw the line, and this is where Judge Gee’s analysis got a tad wonky. She first acknowledged that the “process depends to some extent” on Dish’s involvement. She then concluded that Dish didn’t engage in sufficient volitional conduct to be a direct performer, because “it is the user who initiates the process, selects the content, and receives the transmission.” Of course, the user initiated the process, selected the content, and received the transmission in Aereo, but that didn’t mean that Aereo didn’t directly perform. I initiate the process, select the content, and receive the transmission when I use YouTube, but that doesn’t mean that YouTube doesn’t directly perform. The difference between Dish and Aereo or YouTube is not so much who causes what to occur—with all three, it’s a combination of both the service and the user. The difference is that with Aereo and YouTube, the user obtains content in the first instance, but with Dish, the user does not. In other words, it turns on whether the user has a “prior relationship” to the particular content being streamed.

Judge Gee held that Dish didn’t publicly perform because its volitional conduct was somehow insufficient as a matter of law. She could have reached the same conclusion by simply looking at whether the performances-by-transmission at issue were public or private. If private, Dish couldn’t be directly liable for violating Fox’s public performance rights. Since Fox also had argued that Dish was secondarily liable, Judge Gee needed to determine if the performances she had held to be caused by the users were public or private anyway—Dish couldn’t be an indirect infringer unless the users were direct infringers. Judge Gee could have found Dish not liable, either directly or indirectly, by pointing out that the performances were private. There was no need to look at who caused them, much less to say it was the users and not Dish. The problem with her approach is that, down the road, other services can point to this opinion and argue that they aren’t directly performing either, even if the performances are public in that they are supplying content to the users in the first instance.

Judge Gee adopted one of the broader versions of the volitional conduct test, which essentially turns on who pushes the button. The idea is that, no matter what the service does, users cause the copying to occur because they initiate things with the click of a virtual button. This test makes sense, at least somewhat, when it’s the reproduction right at issue. I think the service and the user often both cause the copy to be made, but it’s reasonable to pin it solely on the user since the service can be on the hook for violating some other right. For example, if a user downloads content from a service, the user reproduces and the service distributes. There is little problem in saying that only the user causes the reproduction since the service is still responsible for the distribution—if it is public in that it supplies the content to the user in the first place.

It’s the same thing with Dish. Judge Gee held that the users made the reproductions on their Hopper with Sling set-top boxes. She thought Dish’s volitional conduct was insufficient for it to be the direct cause of the copying. I don’t agree with this approach—especially with the PrimeTime Anytime recordings, which were individually programmed by Dish and which could not be stopped by the users. In the end, though, I don’t think it matters, as Dish would still be potentially on the hook for its satellite transmissions that supplied the users with the content in the first place. These transmissions were public performances, and they were, of course, licensed. Furthermore, Judge Gee held that the reproductions made from these transmissions were themselves fair use, as were the copies of these recordings made with the Hopper Transfers feature. In this way, Judge Gee upheld both time-shifting and space-shifting of the copies that were legally created as a result of Dish’s licensed public performances.

The problem with the who-pushes-the-button flavor of the volitional conduct test comes in saying that only the user causes the distribution, display, or performance when that user obtains content from the service in the first instance. Judge Gee’s intuition that only the user, and not Dish, caused the transmissions to occur with the Dish Anywhere feature was reasonable, but it only failed to be problematic because the content had already been legally obtained by the user. In other words, there was no problem because the performances-by-transmission were not “to the public.” Here, the content was transmitted from the Hopper with Sling set-top box to the user’s personal computer or device. There was some dispute about the extent to which Dish had its hand in things—for example, whether the content ever touched Dish’s servers. But, for the most part, the transmissions were accomplished using equipment in the user’s possession—at least insofar as the sending and receiving devices.

Splitting such hairs, however, can only take us so far. Our intuition may also tell us that the location of the equipment should be of little relevance. Think Cablevision with its DVRs in the cloud. What if Dish’s Hopper with Sling were a cloud-based DVR? What if the playback signal indubitably originated from Dish’s servers? This is where I think the public-private distinction under Aereo can do the heavy lifting. There is no need to figure out who causes the distribution, display, or performance when the content is stored in the cloud. Instead of beginning with the first element of the Aereo test, which looks at whether there is sufficient volitional conduct on the defendant’s part, it’s better to start with the second element, which asks whether the distribution, display, or performance at issue is public or private. If it’s private, then it doesn’t matter who causes it since it’s noninfringing. If it’s public, however, then it implies that the service, which is supplying content to the user in the first instance, is directly causing that distribution, display, or performance to occur.

On its face, Judge Gee’s opinion said nothing about the cloud. Underneath, by contrast, it told us all that we need to know. It’s the first opinion post-Aereo—that I’m aware of—that properly applied the Supreme Court’s public-private distinction under the Transmit Clause. Though, there was some confusion in Judge Gee’s treatment of ownership vs. possession:

DISH subscribers are not “owners” of the copyrighted programming. DISH has expressly disclaimed any ownership rights in the underlying programming, and agreed to various restrictions on its use of the material as a condition of the license. DISH is a licensee, and therefore cannot transfer title or ownership to its subscribers.

DISH subscribers are, however, valid “possessors” of the copyrighted works that are stored in the STB in their home. See Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 456, 104 S. Ct. 774, 795, 78 L. Ed. 2d 574 (1984); see also Vernor v. Autodesk, Inc., 621 F.3d 1102, 1112 (9th Cir. 2010) (noting that some users “rightfully possess, but do not own, a copy of copyrighted [material].”).

The first paragraph discussed ownership of the copyright, not the copies. Dish couldn’t “transfer title” to the copyright because it was a nonexclusive licensee—it had no title to transfer. The second paragraph, by contrast, discussed possession of the copies, not the copyright. Judge Gee inadvertently slipped from the copyright to the copies as she switched her focus from ownership to possession. Even if the users had no license under the copyright to make the copies, they ultimately didn’t need one. They were legal possessors of the copies because fair use excused the time-shifting or place-shifting whereby the copies had been made in the first place. The users had neither ownership nor possession of the copyright itself—nor did Dish for that matter. But the issue here was ownership or possession of the copies, not the copyright.

The Supreme Court made the same mistake in Aereo: “Neither the record nor Aereo suggests that Aereo’s users receive performances in their capacities as owners or possessors of the underlying works. This is relevant because when an entity performs to a set of people, whether they constitute ‘the public’ often depends upon their relationship to the underlying work.” The phrase “underlying work” is used in copyright law to refer to the copyrighted work itself—that is, the intangible work that exists the moment it’s fixed in a tangible medium. Neither Aereo nor its users—nor Dish nor its users—had any ownership or possessory interest in the “underlying work.” That was never the issue in either case.

What the Supreme Court really meant to discuss was ownership or possession of copies of the “underlying work.” As Professor Jane C. Ginsburg noted:

The court’s reference to “owners or possessors” is very imprecise; the service’s customer is unlikely to be an owner of “the work” because “the work” is the incorporeal object whose “owner” is the author or other copyright owner. Presumably, the court was positing the request by a customer of a remote storage service to play back a digital copy that she was entitled, by express or implied license, or under the fair use doctrine, to deposit in a digital storage locker.

This is exactly what the Supreme Court meant in Aereo, and it’s exactly what Judge Gee meant here with Dish: The users’ preexisting possessory interests in the copies were what made the subsequent performances private. As Judge Gee elaborated:

When an individual DISH subscriber transmits programming rightfully in her possession to another device, that transmission does not travel to “a large number of people who are unknown to each other.” The transmission travels either to the subscriber herself or to someone in her household using an authenticated device. This is simply not a “public” performance within the meaning of the Transmit Clause.

The users’ possessory interests in the copies were easy to ascertain with Dish’s Hopper with Sling set-top boxes. This was because the users each had actual possession of a copy once that copy had been created for the first time. Under the Supreme Court’s test in Aereo, however, even if these copies were stored remotely, such as with a cloud storage service, the users would still have possessory interests in the copies such that subsequent distributions, displays, or performances would be private. And this is why I think Aereo makes the cloud safer than ever: We can tell whether a distribution, display, or performance is “to the public” based on whether the user has a “prior relationship” to the source copy. Judge Gee got it right because she realized that Aereo controls the public-private distinction analysis.

This realization was lost on the district court in the MP3tunes decision that I wrote about last October. The judge there thought that the number of source copies is what matters. It doesn’t. What matters is the “prior relationship” of the user to the content at issue. The number of source copies is irrelevant, and a service can even utilize one master copy that is served to multiple users. Such deduplication is permissible so long as each user has a preexisting possessory interest in the content. There will no doubt be hard cases where the limits of these possessory interests will be tested, but the doctrine post-Aereo is nevertheless clearer than ever. So long as the user has a “prior relationship” to the content at issue, its subsequent distribution, display, or performance back to that user will be private—whether the service directly causes it or not. The cloud has never been safer, thanks to Aereo.

Follow me on Twitter: @devlinhartline

© 2015 Devlin Hartline. Licensed under the Law Theories Public License 1.0.