Kozinski and Goldfoot on Cyberspace Exceptionalism and Internet Regulation
By Eric Goldman
Alex Kozinski & Josh Goldfoot, A Declaration of the Dependence of Cyberspace, 32 Colum. J.L. & Arts 365 (2009).
In early 1996, in response to Congress’ enactment of the Communications Decency Act (the first comprehensive attempt to regulate the Internet), John Perry Barlow published his cyberspace exceptionalist screed, “A Declaration of the Independence of Cyberspace.” The manifesto (naively, IMO) tells government regulators that they are outdated and should not—and cannot—regulate the Internet.
Judge Kozinski, chief judge of the Ninth Circuit, and Josh Goldfoot, a trial attorney in the DOJ’s CCIPs division, use Barlow’s article as an entry point to discuss Internet exceptionalism/regulation generally. Although the article expresses the authors’ personal views, the article amplifies some themes Judge Kozinski has been developing in his recent Internet jurisprudence, most notably the Roommates.com case and Perfect 10 v. Visa. Because Judge Kozinski plays a crucial role on the federal appellate court governing both Hollywood and the Silicon Valley, this article is worth a close look.
Judge Kozinski made his distaste for Internet exceptionalism clear in the Roommates.com opinion. In this article, the authors explain this view more thoroughly:
It is a mistake to fall into Barlow’s trap of believing that the set of human interactions that is conducted online can be neatly grouped together into a discrete “cyberspace” that operates under its own rules. Technological innovations give us new capabilities, but they don’t change the fundamental ways that humans deal with each other….[W]hen the internet is involved in a controversy only because the parties happened to use it to communicate, new legal rules will rarely be necessary. When the substance of the offense is that something was communicated, then the harm occurs regardless of the tools used to communicate….[T]he vast majority of internet cases that have reached the courts have not required new legal rules to solve them.
While I generally agree with this, I also think it’s an antiquated sentiment. Whether or not cyberspace exceptionalist law are logical or even appropriate, legislators have found them irresistible, resulting in dozens or hundreds of Internet-specific statutes. I explore this dynamic in my article “The Third Wave of Internet Exceptionalism.” So to the extent the authors are arguing that we don’t need new cyberspace-specific laws, that ship sailed a long time ago.
The authors conclude that “the internet is doing wonderfully. It has survived speculative booms and busts, made millionaires out of many and, unfortunately, rude bloggers out of more than a few. The lack of a special internet civil code has not hurt its development.”
I agree that the Internet is doing wonderfully, but I would assign causality differently. Legislatures in the 1990s passed a number of Internet-favorable laws, such as the Internet Tax Freedom Act, which kept taxing authorities from loving the Internet to death, and 47 USC 230, which provided a crucial immunity to online intermediaries. Reverse-engineering the Internet’s success is a tricky science, but my hypothesis is that the success is partially due to these “special Internet civil codes,” not due to their absence. For more on this with respect to 47 USC 230, see my talk notes from the Denver University Cyber Civil Rights event.
”Death of the Internet” and “Death of Innovation” Arguments
The authors address two common arguments that Internet defendants make to support favorable exceptionalist rulings, including that an adverse ruling (1) will end the Internet or (2) harm innovation.
They suggest that “end of the Internet” arguments can be powerful (specifically addressing Judge McKoewn’s doomsday concerns in her Roommates.com dissent):
The argument that a legal holding will bring the internet to a standstill makes most judges listen closely. Just think of the panic that was created when the Blackberry server went down for a few hours. No one in a black robe wants to be responsible for anything like that, and when intelligent, hard-working, thoughtful colleagues argue that this will be the effect of one of your rulings, you have to think long and hard about whether you want to go that way. It tests the courage of your convictions.
While end-of-the-Internet arguments can grab judges’ attention, I have to assume that the litigant loses credibility if the claim is overstated. So use the argument sparingly, like when your client’s loss will pry beloved Crackberries out of the judges’ hands.
The authors are less impressed with the “death of innovation” argument.
[P]romoting innovation alone cannot be a sufficient justification for exempting innovators from the law. An unfortunate result of our complex legal system is that almost everyone is confused about what the law means, and everyone engaged in a business of any complexity at some point has to consult a lawyer. If the need to obey the law stifles innovation, that stifling is just another cost of having a society ruled by law. In this sense, the internet is no different than the pharmaceutical industry or the auto industry: they face formidable legal regulation, yet they continue to innovate.
There is an even more fundamental reason why it would be unwise to exempt the innovators who create the technology that will shape the course of our lives: granting them that exemption will yield a generation of technology that facilitates the behavior that our society has decided to prohibit. If the internet is still being developed, then we should do what we can to guide its development in a direction that promotes compliance with the law.
I’m sympathetic to this point. Personally, I feel like arguments that a ruling or law will harm “innovation” are often make-weight. “Innovation” is ill-defined and difficult to measure (i.e., some folks believe patent applications/issuances quantify innovation, but we know better), and it is politically incorrect to oppose “innovation” (you might as well oppose other incontrovertible ideals like freedom, Mother Teresa and puppies). Thus, the “harms innovation” argument automatically, and often unfairly, puts the opponent on the defensive—they can either try to debate what’s better for innovation or stand silently and look like they oppose innovation. But debates about what’s best for “innovation” are almost always irresolute because innovation can take many forms, and we do not know what precise mixture of government intervention and deregulation will foster socially optimal levels of innovation. For more on this, see, e.g., Niva Elkin-Koren and Eli Salzberger’s analysis of Coasean allocations on innovation.
At the same time, the authors’ arguments are a little disquieting because they imply that innovation can result in only one of two outcomes—legal or illegal, with nothing gray in between. From my perspective, much (most?) Internet entrepreneurship/“innovation” exists between the two endpoints of the legality continuum. For example, in 1996, I believe many legal experts would have said that unconsented spidering and indexing of a website was probably illegal (a question that has not been definitively resolved even today)—so if we wanted to avoid possibly illegal innovation, Google would not exist today. As a result, it might sound great to channel innovation towards only clearly legal activities, but I don’t really think that’s what we want.
Secondary Liability and Anonymity
The article also has some troubling remarks on secondary liability and anonymity:
If the legal rules change, and companies are held liable more often for what their users do, then the cost of anonymity would shift away from victims and toward the providers. In this world, providers will be more careful about identifying users. Perhaps online assertions of identity will be backed up with offline proof; providers will be more careful about providing potential scam artists in distant jurisdictions with the tools to practice their craft. All this would be expensive for service providers, but not as expensive as it is for injured parties today.
I would like to see some empirical support for the last sentence’s comparison of expenses. It’s not self-evident to me. Further, if we are going to do a cost accounting, we also need to consider what socially beneficial activity is dissuaded by service provider authentication of identity.
The authors continue:
Secondary liability should not reach every company that plays any hand in assisting the online wrong-doer, of course. Before secondary liability attaches, the plaintiff must show that the defendant provided a crucial service, knew of the illegal activity, and had a right and a cost-justified ability to control the infringer’s actions. This rule will in almost every case exclude electrical utilities, landlords, and others whose contributions to illegal activity are minuscule.
This argument is consistent with traditional tort principles (as well as Judge Kozinski’s dissent in Perfect 10 v. Visa regarding copyright liability). 47 USC 230’s immunity breaks these venerable principles. As I’ve noted before, bright judges imbued in the common law can have a tough time with Congress’ rejection of traditional tort principles (as well as the concomitant reduction in judicial discretion).
Meanwhile, I’m wondering about the qualifier in the last sentence (“in almost every case”). Unless specified in a statute, I can’t imagine *any* circumstances where it would be appropriate to hold people who make “minuscule contributions” responsible for third party torts—especially electrical utilities, who as regulated monopolies usually have no discretion about whether or not to provide power to their customers.
Although in general most Ninth Circuit Internet rulings have reached the right result, recent Ninth Circuit rulings have shown some hostility towards 47 USC 230 specifically and Internet defendants generally. I am concerned that the Ninth Circuit has become a dangerous circuit for Internet defendants, and this article does not dispel my fears. I think Internet defendants should carefully weigh the pros and cons before appealing a case to the Ninth Circuit. The wild card factor is high, and the likelihood of getting an incomprehensible legal standard is higher still.