Privacy as property - Part V: Democratic Process and Nonpublic Politics

Social Research, Spring, 2002 by Lawrence Lessig

A SOCIETY protects its values in different and overlapping ways. The values of free speech in America are protected by a constitution that guards against speech abridging regulation (U.S. Constitution, First Amendment). They are supported by copyright regulation intended as an "engine of free expression" to fuel a market of creativity (Harper & Row Publishers, Inc. v. Nation Enterprises, 471 U.S. 539 (1985)). They are supported by technologies such as the Internet that assure easy access to content (Lessig, 1999: 164-185). And they are supported by norms that encourage, or at least allow, dissenting views to be expressed (Lessig, 1999: 164-185; 235-239). These modalities together establish cultural resources that to some degree support the fight to speak freely within American society.

This is an essay about the cultural resources that support the values of privacy. My aim is to promote one such cultural resource--the norms associated with property talk--as a means of reinforcing privacy generally. In my view, we would better support privacy within American society if we spoke of privacy as a kind of property. Property talk, in other words, would strengthen the rhetorical force behind privacy. (1)

Such a view is not popular among privacy advocates and experts in the field of privacy law. It has been expressly rejected by some of the best in the field. (2) Thus it is my burden to demonstrate the value in this alternative conception.

This essay alone will not carry that burden. But my hope is that it will at least evince a benefit from our reconceiving privacy talk. I will address some of the objections to this form of speaking, but only after advancing the arguments in its favor. The essay moves in six parts. I first introduce two stories that will frame my argument (parts I and II). I then consider one account by Professor Jonathan Zittrain that might explain the tension that is revealed between these two stories (part III). Part IV offers a different account from the one proposed by Professor Zittrain. In the final part, I offer a brief response to some of the criticisms that have been made of the privacy as property model. I then conclude.

I

At a public debate about the increasing scope of patent protection granted by U.S. law to software and business method inventors, Jay Walker, the president of Priceline.com and Walker Digital and a holder of many of these new, and controversial, patents, was asked to justify them. Did we know, Walker was asked, whether increasing the scope of these new forms of regulation would actually increase innovation? Or would they, by requiring every developer to add to its team covens of lawyers, increase the costs so much as to chill or stifle innovation? ("Internet Society," 2000).

Walker did not have a strong answer. In the face of evidence that these patents harm innovation, he could cite no firm evidence to the contrary. Would it not make sense then, Walker was asked, to have a moratorium on this new form of patent protection until we learn something about its economic effects? Why not study the effect, and if it is not clear that they do any good, then why not halt this explosion of regulation?

Walker exploded in anger at the very suggestion. Grabbing the mic, he said to the questioner: "Does that mean if Microsoft takes my property, I can't bring a suit against them, is that what you're suggesting?" ("Internet Society," 2000; emphasis added).

II

Amazon.com is a collector of data. It sells books to collect that data. These data help it do an extraordinary job understanding its customers' wants. By monitoring their behavior, Amazon can build large and accurate profiles about its customers. And using powerful data-matching techniques, it can predict what books they are likely to want to buy.

Amazon had a privacy policy. The data it was collecting, Amazon said, would not be sold to others, at least if a customer sent Amazon an email asking that such data never be sold. The data was therefore collected only for Amazon's use. And for anyone who had used Amazon frequently, much of that use was obvious and familiar. Amazon watched its customers, and used the data from that watching to better serve its customers; relying on this policy, people gave Amazon years' worth of data from this watching.

At the end of 2000, however, Amazon announced a new policy. (3) From that point on, data from Amazon could be sold to or shared with people outside Amazon, regardless of a consumer's request that it not. The privacy policy would therefore no longer assure users that only Amazon would know what books they bought. Amazon, and anyone Amazon decided, could know their consumers' buying patterns.

The extraordinary feature of this announcement--and the one on which I want to focus here--was its retroactive effect. (4) Not only was Amazon announcing that from then on, data collected could be subject to sale. It also made that policy retroactive. Customers who had relied on Amazon's promise in the past and had indicated that they "never" wanted their information sold were now told their data could be sold. Amazon refused requests to delete earlier data; the consumers who had relied on its policy were told they had no right to remove the data they had given. Their data was now subject to sale. (5)

Amazon justified this change, and this retroactive effect, based on an escape clause built into the company's privacy policy. This policy, the clause explained, was always subject to change. Consumers were therefore on notice, Amazon explained, that the rules could change. They knew the risks they were taking.

III

In a recent essay, Harvard professor Jonathan Zittrain argued that there is no conceptual difference between the "privacy problem" and the "copyright problem" (Zittrain, 2000). Both, he argued, are examples of data getting out of control. In the context of privacy, it is personal data over which the individual loses control--medical records, for example, that find their way to a drug company's marketing department; or lists of videos rented, that before the protections granted by Congress, found their way to a Senate confirmation hearing. The problem with privacy is that private data flows too easily--that it too easily falls out of the control of the individual.

So understood, Zittrain argued, the problem of copyright is precisely the same. It too is a problem of data getting out of control. Music is recorded on a digital CD; that CD is "ripped" (that is, the audio extracted), and the contents are placed on a computer server; that content is then duplicated and placed on a thousand servers around the world. The data has thus gotten out of control. Copyright holders who would otherwise want to condition access to their data find the ability to condition access upon payment gone. Just as the individual concerned about privacy wants to control who gets access to what and when, the copyright holder wants to control who gets access to what and when. In both cases the presence of ubiquitous computing and saturating networks means that the control is increasingly lost.

Zittrain then considered the steps that have been taken to deal with each of these problems. These, not surprisingly, turn out to be quite different. In the context of copyright, technologists are developing many new technologies to re-empower the copyright holder--to assure that the use of his data is precisely as he wishes. Trusted systems, for example, will give the copyright holder the power to decide who can listen to what. A digital object with music inside can be released on the Internet, but it will only play if the possessor has a permitting key. It will then play just as the copyright holder wants--once, or ten times, or forever, depending on how the object is coded. Trusted systems promise to rebuild into the code an extraordinary system of control. It promises to use code to solve this problem of data getting out of control.

Changes in technology, however, are not the only changes that copyright owners have sought, and secured, to remedy the problem of data getting out of control. Copyright holders have also benefited from significant changes in law. In a series of significant acts of legislation, the U.S. Congress has increased the penalty for using copyrighted material without the permission of the copyright holder. It is now a felony to publish more than $1,000 of copyrights without the copyright holder's permission. (6) Congress has also added important rights and protections for copyrighted work online with the Digital Millennium Copyright Act (DMCA). (7) If a copyright holder uses technology to protect his copyright, then the DMCA protects that software with law by making it an offense to develop and distribute tools that circumvent that software. Software code thus increasingly protects copyrighted material. Legal code increasingly protects copyrighted material more strongly. And with the DMCA, law now protects code that protects copyright. These changes will thus balance the risk that copyrighted "data" will get out of control.

What about the response to the problem of privacy? Here the story is very different. We do not have a collection of new federal laws restoring control over their data to individuals. Law has been successfully resisted with a familiar rhetoric--that "we should let the market take care of itself." So far, these opponents have been quite successful. Congress has yet to pass an Internet privacy regulation, although it has passed regulation to protect financial privacy. (8)

Technology, too, has been slow to emerge. Not that there are not a host of creative technologies out there--technologies to anonymize transactions and presence, and technologies to facilitate control over the use of personal data. (9) But these technologies have not quite had the financial backing that copyright control technology has had. Germany, for example, requires that Internet service providers (ISPs) offer anonymous accounts. (10) That has spurred a market for providing such anonymity in Germany. But we have no such requirement in America, and hence we have not sufficiently spurred a market to provide this technology.

Zittrain puts these two stories together, and asks, What explains the difference? The same problem--losing control over our data--is raised in two very different contexts. In the one context, copyright, both law and technology (or we might say, "East Coast code" and "West Coast code") rally to defend the copyright holder against the users who would use it without control; in the context of privacy, both law and technology have been slow to respond. Same problem, two radically different responses. What explains the difference?

Zittrain offers a dark explanation: follow the money (Zittrain, 2000). If you want to understand why all controlled creativity in the world is allied on one side of this problem and only the public spirited, Marc Rotenberg-like are allied on the other, follow the money. When it pays to protect privacy--when it pays to build tools to protect privacy--you will see lots more privacy. But just now, it is Hollywood that pays, so it is copyright that gets protected.

IV

My aim in this essay is to explore a different account of why these two similar problems get fundamentally different treatment in Washington. This account focuses less on the dark motivations of dollars chasing policy. Instead, the difference I want to focus on is a difference in norms, or a difference in the ordinary understanding, or construction, or social meaning of the problem (see, e.g., Lessig, 1995: 943). This difference is seen in the two stories I began with at the start: the story of Jay Walker and Amazon.com. Recall Walker's response to the suggestion that we investigate a bit more whether patents do any good before we issue more patents to protect software or business methods. And allow Microsoft, he asked, to steal "my property?" Focus on this term, "property." Jay Walker framed the question of proper patent policy within a familiar, and deeply American, discourse about property. The issue, so framed, is whether one is for property or not; the punishment for those who would question patents is thus the same as the punishment for questioning property: Are you or have you ever been a communist?

This move--as a matter of rhetoric--is brilliant, although the idea that "patents" should be spoken of as this sort of property would be strange to anyone with a sense of history. The framers of our constitution did not speak of patents as "property." Patents were understood as exclusive rights granted by the government for a public purpose, not natural rights recognized by the government as an aspect of natural justice. This was the import of Jefferson's famous description of patents, (11) and it explains in part why the term "intellectual property" did not enter our legal vocabulary until late into the nineteenth century. (12)

But ordinary people are not constrained by any sense of history. They instead are open to this more familiar way of speaking about patents. And Walker's use of the term "property" is an increasingly familiar way in which "intellectual property" is discussed. Thus, despite the dissonance with Jefferson, Walker can talk about "stealing" his "property" without anyone noticing anything funny in his speech. The debate then focuses on why people should be allowed to "steal" rather than on why a "patent" is "property." (13)

Yet had Walker been forced to use the language of the framers, his rhetorically powerful rhetorical question--"you want Microsoft to steal my property while you conduct your studies about what does what good?"--would have been the extremely weak rhetoric--"you want Microsoft to invade my monopoly while you conduct studies to determine whether the monopoly does any good?" In their language, the idea that Walker, or anyone, has a moral claim to a government-backed monopoly would seem odd. Had he been forced to express himself in just these terms, his question would have answered itself quite differently.

There is something in this story that we who would like stronger protection for privacy might learn. The story shows the different social resources that are available within our culture to claims that are grounded in property.

To see the point a bit more directly, think about the Amazon story with one fundamental change. Imagine that people spoke about their privacy as if privacy were a form of property. What constraints--social constraints--on what people can do to private data would exist then?

I know many want to resist this idea of speaking about privacy as property, and I certainly know that one must justify the usage if it is to be accepted (though I really do not recall reading the justification that transformed monopoly-speak in the context of patents and copyright into property-speak, but let us put that aside for a second). Before we get to justification, or even possibility, just imagine if we thought about our personal data the way we think about a car. And then think about this analogous case about a contract governing a car.

You drive into a parking lot, and the attendant hands you a ticket. The ticket lists a number of rules and promises on the back of the ticket. The lot is not responsible for damage to the car; the car must be picked up by midnight, etc. And then imagine, as with Amazon, that at the bottom of the ticket, the last condition is that this license can be modified at anytime by the management.

Imagine you walk up to the parking lot, hand him the ticket, and say, "I would like my car." And the attendant says to you, "Well, as you'll see at the bottom of your ticket, we reserved the right to change the conditions at any time. And in fact, I'm afraid we have changed the conditions. From here on out, we've decided to sell the cars we take in. And so we've decided to sell your very nice car. We therefore cannot return your car; the car is probably in New Jersey by now. We're sorry, but that's our policy."

Obviously, in ordinary property thought, this is an absurd idea. It would be crazy to interpret a condition in a license stating that the license could be changed to mean that the license might be changed to allow the parking lot to sell your car. One might well imagine a Jay Walker moment--"Hey, that's my property. You can't steal my property. You can't change the license and then run away with my car."

And yet, notice that these same intuitions are not excited when people hear the story about Amazon. I do not mean that there are not people who think what Amazon has done is awful--clearly there are such people, and perhaps most think what Amazon has done is awful. My claim is not that people agree with Amazon; my point is that because we do not think of privacy the way Hollywood has convinced us to think about copyright, we cannot easily invoke the rhetoric of property to defend incursions into privacy. If it were taken for granted that privacy was a form of property, then Amazon simply could not get away with announcing that this personal information was now theirs. That would be "theft," and this is my point: "theft" is positively un-American.

Property talk would give privacy rhetoric added support within American culture. If you could get people (in America, at this point in history) to see a certain resource as property, then you are 90 percent to your protective goal. If people see a resource as property, it will take a great deal of converting to convince them that companies like Amazon should be free to take it. Likewise, it will be hard for companies like Amazon to escape the label of thief. Just think about the rhetoric that surrounds Napster--where thousands are sharing music for no commercial gain. This practice is comfortably described by many--not me, but many others--as "theft." And more important, it is hard for the defendants to defend against this label of theft. The issue becomes whether the user has a right to steal--not the kind of case you would want to have to prove.

We could strengthen the cultural resources supporting the protection of privacy if we could come to think of privacy as property, just as the cultural resources available to Jay Walker have been strengthened by the happenstance of a legal culture that has come to refer to "patents" as intellectual property. Norms about property would support restrictions on privacy, just as norms about property resist limitations on Walker's "property."

That is the affirmative claim, but before I address objections, there are some qualifications I must offer to assure that certain confusions do not detain the debate.

(1) To promote property talk is not to promote anarchism, or even libertarianism. "Property" is always and everywhere a creation of the state. It always requires regulation to secure it, and regulation requires state action. The DMCA is a law designed to protect what most think of as "property." That is regulation. Police are deployed to protect property. That too is regulation. Zoning laws regulate and control property. Rules regulating the market control how property is exchanged. And rules establishing privacy as property would govern when and how a privacy right can be traded. Property is inherently the construction of the state; and to confuse the promotion of property with the promotion of laissez-faire is to fall into the vision of the world that libertarians delude themselves with. There is no such thing as property without the state; and we live in a state where property and regulation are deeply and fundamentally intertwined.
(2) Although property is often resisted by liberals because of the inequality that property systems produce, privacy as property could create less extreme inequality. If the privacy as property system were properly constructed, it would be less troubling from this perspective than, say, copyright or ordinary property. For if the law limited the ability to alienate such property completely--by permitting contracts about, for example, secondary uses but not tertiary uses--the owner of this property would be less likely to vest it in others in ways that would exacerbate inequalities.
(3) Property talk is often resisted because it is thought to isolate individuals. It may well. But in the context of privacy, isolation is the aim. Privacy is all about empowering individuals to choose to be isolated. One might be against the choice to be isolated; but then one is against privacy. And we can argue long and hard about whether privacy is good or not, but we should not confuse that argument with the argument that property would better protect any privacy we have agreed should be protected.
(4) To view privacy as property is not to argue that one's rights to use that property should be absolute or unregulated. All property law limits, in certain contexts, the right to alienate; contract law restricts the contexts in which one can make enforceable contacts. The state has a valid and important role in deciding which kinds of exchanges should be permitted. And especially given the ignorance about the Internet that pervades the ordinary user's experience, I would be wildly in favor of regulation of what people were allowed to do with "their property." Indeed, I imagine there is no legislative recommendation of the Electronic Privacy Information Center (EPIC), for example, for regulating privacy that I would oppose (see (http://www.epic.org)).
Thus, to promote property talk is not to demote the role of regulation, or to believe that the "market will take care of itself," or to question the strong role the government should have to assure privacy. It is simply to recognize that the government is not the only, or often, most important protector of human rights. And that where norms can carry some of the water, my argument is that we should not be so quick to condemn these norms.

V

Privacy and property talk is resisted, however, by many in the privacy community. Their resistance is strong, and their arguments are good. I will not in this short essay be able to rebut these arguments in opposition. But I do hope to suggest how they might be resisted.

To make the criticism clearer and, obviously, to aid in the resistance of these criticisms, consider first a picture of what I imagine a property system protecting privacy to be.

In the world that I imagine, individuals interact online through machines that connect to other machines. My computer is a machine; it links across the Internet to a machine at Amazon.com. In this interaction, the machine could reveal to the other machine that a machine with a certain ID--for example, that a machine with an ID with characteristics A, B, and C purchased goods X, Y, and Z after looking at pages k, l, m. This ID need not be linked back to a particular person, although it could be. The property right that I am imagining governs the terms under which one machine can access certain data from the other machine. It says that the party who would collect these facts cannot do so without permission from the ID. The default is that the facts cannot be collected, but that default can be negotiated around.

The negotiation occurs through technology that sets the terms under which facts A, B, and C may be collected. These facts then get wrapped in a digital envelope that in turn carries the terms with them. At this stage the problem is directly the problem of trusted systems in copyright law. And as with trusted systems, a system that trades these facts can only do so when dealing with a secure system that trades the facts according to the rules in the wrappers.

The terms on the wrapper could be many. I do not know which would be critical. Perhaps it is a permission for primary and secondary use, but not tertiary use. If so, after the secondary use, the fact would digitally "disappear." But access might also be granted on very different terms--conditioned on the promise that the fact will not be related back to the human who has control over the ID. Whatever the term, I assume the trusted system could implement it through secure trading technologies.

What are the problems with a regime so constituted?

"It would be unconstitutional." The strongest argument against the privacy as property position--if true--is that it would be unconstitutional for the government to grant property rights in privacy. This, the argument goes, would be just like granting property rights in facts, and facts, the Supreme Court has come very close to holding, cannot be secured by at least one particular form of property right--namely copyright (Feist Publications v. Rural Telephone Service Co., 499 U.S. 340 (1991)). Thus, because copyright cannot protect facts, Congress cannot protect facts, and a property regime for privacy would therefore be unconstitutional.

But although I am a very strong believer in constitutional limitations on the intellectual property power, this argument moves too quickly. No doubt you could not grant an intellectual property right in private facts. It does not follow, however, that you cannot, through law, control the right to use or disseminate facts. Trade secret law protects certain "facts" from disclosure (Merges, Menell, and Lemley, 2000). Contract law can punish individuals for disclosing facts they have promised to keep secret (C0hen v. Cowles Media Co., 501 U.S. 663 (1991)). These limitations on the right to distribute facts stem from regimes that are grounded on a constitutional authority other than the Copyright and Patent Clause.

For this privacy as property claim to be constitutional, it would require a constitutional source of authority other than the Copyright and Patent Clause. The most natural alternative source would be the Commerce Clause (U.S. Constitution, Art. 1, Sec. 8, Clause 3). There is a substantial body of evidence to suggest that fear about privacy inhibits online commerce. The same fear was said to undermine the use of wireless communications generally. Congress was able to regulate that communication under the Electronic Communications Privacy Act and the Wiretapping and Electronic Surveillance Act. (14) The same authority should support the construction of a stronger right through property.

No doubt there are limits on the ability to protect facts through property, especially in light of the important First Amendment interests that are involved. Those limits were well described in the most recent case considering the intersection between privacy and the First Amendment: Bartnicki v. Vopper. (15) But Bartnicki does not establish that private facts cannot be protected through law. It establishes only that the use of those facts illegally obtained cannot, in certain circumstances, be constitutionally restrained. The same limits would restrict privacy as property. They would not, however, render it unconstitutional per se.

Of course, other constitutional powers cannot, or should not, be relied on to support a privacy as property right if that right is of the kind that the Copyright and Patent Clause protects. But if trademarks are not the sort of right that the Copyright and Patent Clause protects (see, e.g., Trade-Mark Cases, 100 U.S. 82 (1879)), I do not believe privacy would be either. Thus, whether or not privacy as property is a good idea, it would not, I believe, be unconstitutional because of the Copyright and Patent Clause.

"It would queer intellectual property law. "A second criticism is related to the first. Some fear that thinking of privacy as property would further strengthen property talk about intellectual property. Thus, increasing protection for privacy would perversely increase the already too expansive (in my view and the view of these critics) protection for intellectual property (see, e.g., Samuelson, 2000: 1140).

I share the concern about overprotecting intellectual property, so here again I think that if the criticism were correct, it would be strong. But I think the fear is overstated. Given the plethora of property talk that shoots through our society, it is fanciful at best to imagine that one more dimension of rights talk would tip any fundamental balance with intellectual property. If the power itself were grounded in the Copyright and Patent Clause, I would agree that it would create dangerous pressure on intellectual property law. I do not think that would be the source of any such power, and I therefore do not think that would be its effect.

"It would tend to promote the alienation of privacy, by encouraging a better, or more efficient, market in its trading. "A third criticism is more pragmatic than the previous two. Its concern is with consequences. The fear is that increasing property talk would tend to increase the alienation of privacy. Property is associated with markets; markets associated with trade; trade is the alienation of this for the acquisition of that. (16)

It is certainly true that by thinking of privacy as property, one makes it easier to think about trading privacy within a market. But equally, if the essence of a "property right" is that the person who wants it must negotiate with its holder before he can take it, propertizing privacy would also reinforce the power of an individual to refuse to trade or alienate his privacy. Whether he or she alienates the property depends on what he or she wants. And while people who have pork bellies may well prefer cash for their property (and hence property facilitates the trade in pork bellies), it would not follow that family heirlooms would be better protected if we denied the current owner a "property right" in those heirlooms. Property defends the right of the farmer to alienate pork bellies as much as it defends my right to keep you from getting a mirror my grandmother gave me.

What property does do affirmatively is to allow individuals to differently value their privacy. It is the consequence of a property system that by protecting the right of an individual to hold his property until he or she chooses to alienate it, different individuals get to value different bits of privacy differently. I may be a freak about people knowing my birthday, and so would never "sell" access to that fact for any price, but someone else might be willing to sell access in exchange for 100 frequent flyer miles. The advantage of a property system is that both of our wishes get respected, even though the wishers are so different.

There is nonetheless a legitimate and residual concern that propertizing privacy will tend to facilitate "too much" alienation of privacy, and that we should therefore resist the move to propertize.

This criticism, however, must be divided into two distinct parts. For the criticism begs the question of how much alienation is too much? And also, from what perspective is the judgment of "too much" made?

For some critics, the only legitimate perspective is the individual. Under this view, the criticism must be that propertizing privacy would create market pressures for people to alienate privacy beyond what they otherwise would individually prefer. Or they may be pressured into alienating a kind of privacy they would not otherwise prefer. But if this is the complaint, then there is no reason it could not be met by specific, or targeted legislation. If there is some indignity in alienating a particular bit of privacy, then a law could make trading in that privacy illegal (as it does, for example, with facts about children.) (17)

For others, the concern about "over alienation" is a kind of paternalism, though by "paternalism" I do not mean anything derogatory. I am all for paternalism in its proper place. It would be the proper place here if we could rightly conclude that people would, if given the chance, alienate more privacy than they should.

The problem is knowing whether the amount they would alienate is more than they "should." I am skeptical about whether we can know that yet. If we narrow the focus of the privacy that we are concerned about to the "stuff" revealed between machines in the exchanges with others online, then at this stage in the life of the Internet, I believe we know very little about the harms or dangers this exchange will create. And in a context where we know little, my bias is for a technology that would encourage diversity.

That is precisely the aim of a market. A market would help us find a mix that reflected people's wants, and if, as I believe, people's wants were very different, then the range of different wants could be respected.

It may be that people have a very similar set of wants, and that the property system is an unnecessary expense in finding this common set of privacy preferences. This relates to the final criticism that I will survey here, but if true, then once we discover it is true, we may well choose to substitute a rule for a market. I am just not as convinced we know enough now to know what that rule would be.

"A property system for privacy would be difficult to implement. "Property systems are not costless. They require real resources to make them work. A final line of criticism objects that the costs of making privacy property outweigh any benefit.

This last criticism may well be true, but we should be clear about the costs. The cost of a property system depends on the architecture that implements it. My assumptions about the value of a property system assume that the negotiations and preferences about privacy would be expressed and negotiated in the background automatically. This was the aspiration of the technology Platform for Privacy Protection (P3P) in its first description. It is a fair criticism of my position that the technology it depends on has yet to be developed. (18)

Yet technologies do not get invented in a vacuum. It was only after pollution regulations were adopted that innovative technologies for abating pollution were developed. And likewise here: establishing a strong property right in privacy and punishing the taking of that right without proper consent would induce technologies designed to lower the costs of that consent.

It may again in the end not lower the costs enough. The costs of a property rule may well exceed any benefits. And if that were so, then the alternative of a liability rule, which is pushed by privacy advocates who oppose privacy as property, would make most sense. But this, too, is not something we can know in the abstract.

I do not expect these responses to the criticism about privacy as property will convince. To answer the substantial criticism would require much more attention than I am permitted in these pages. But I do hope this is enough to identify the contours of a reply. My claim is not that the property view is right; it is the much less bold (much more balanced) claim that it could do some good, and that there is no obvious reason it is wrong.

VI

The law of privacy in America was born in a debate about property. In their seminal article giving birth to a "right to privacy," Brandeis and Warren related the need for that right to a change in how property was distributed (Brandeis and Warren, 1890: 193, 198-199). They described an earlier time when property functioned as the protector of privacy: since one could not trespass, one could not invade the sanctity of the right to be left alone. That world had changed, Warren and Brandeis argued. Because property had become so unequal, protecting privacy through protecting property would no longer equally protect privacy. The landed may have had privacy; the person living in a tenement did not. Privacy should therefore be separated from property, they argued, so that privacy could be better protected.

Privacy advocates embrace this argument as a way of resisting the argument in favor of propertizing privacy. But notice an important conflation. The complaints Brandeis and Warren made about physical property would not necessarily apply to the intangible property I am describing here. Indeed, given the difference in the nature of such "property," very different conclusions should follow.

In my view, we should be as pragmatic about property as Brandeis and Warren were. But such pragmatism will sometimes mean that we embrace property to protect privacy, just as it sometimes means that we should resist property to protect privacy. Whether we should depends on the contingencies of the technologies for establishing and protecting property. And those, I suggest, cannot be known in the abstract.

Put another way, when invoking Brandeis and Warren's argument, it is important not to make Jay Walker's mistake. Real property is different from intangible property. Facts about real property do not necessarily carry over into the realm of intangible property. The physics of intangible property are different, and hence, so should be our analysis.

My claim is that in addition to the resources of law that Rotenberg and Catlett rightly advance (see (http://www.junkbusters. com>), and the resources of technology that Zeroknowledge et al. provide, and the mix of law and technology that Zittrain describes, we need to think, in a less politically charged and politically correct way, about adding to the arsenal in support of privacy the favored weapon of Disney and Jay Walker: the ability to rely on the rhetorical force of--"you mean you want to steal my property?"

* I am grateful to Jason Catlett for correction of and advice about parts of this essay. That is not to say he agrees.

Notes

(1) There are many others who have pushed the view that privacy be seen as a kind of property. Paul Sholtz (2001) has offered a transaction costs justification that closely tracks my own. His work, like mine, trades fundamentally on the distinction drawn in Guido Calabresi and Douglas Melamed's foundational work between property rules and liability rules (1972: 1089). See also Sailer (2000: 6); Schwartz (2000: 743, 771-776). Other property-based arguments about privacy include "Developments in the Law" (1999); Kang (1998: 1193, 1246-94); Shapiro and Varian (1997). The view that law should push to property over liability rules is not limited to privacy (whether or not it should be). For a strong push, see Epstein (1997: 2091).
(2) Stanford Law Review's recent "Symposium on Privacy" has a strong collection of property's opponents. See, for example, Cohen (2000: 1373); Lemley (2000: 1545); Litman (2000: 1283); Samuelson (2000: 1125-1126) (there are "some reasons to doubt that a property rights approach to protecting personal data would actually achieve the desired effect of achieving more information privacy"). See also Rotenberg (2001).
(3) For a summary of the facts surrounding these events, see .
(4) I do not mean it changed the rules that existed before. My claim is about expectations.
(5) Under the policy before the change, Amazon permitted customers to send a message to an Amazon email address and be automatically removed from the list of customers who would permit their personal data to be disclosed.
(6) No Electronic Theft Act, Public Law 105-147, 105th Cong., 1st sess. (16 December 1997): 111, 2678. This law, also known as the "NET Act," amended 17 U.S.C. [section] 506(a).
(7) The anticircumvention provision of the DMCA was recently upheld in the Second Circuit. See The Digital Millennium Copyright Act, 17 U.S.C.S. [section] 1201; Universal City Studios, Inc. v. Corley, 2001 WL 1505495 (2nd Cir. 2001).
(8) Gramm-Leach-Bliley Act, Public Law 106-102, 15 U.S.C. [section] 6801-6810 (1999).
(9) See, for example, .
(10) See .
(11) As Jefferson wrote: If nature has made any one thing less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of everyone, and the receiver cannot dispossess himself of it. Its peculiar character, too, is that no one possesses the less, because every other possess the whole of it. He who receives an idea from me, receives instruction himself without lessening mine; as he who lites his taper at mine, receives light without darkening me. That ideas should freely spread from one to another over the globe, for the moral and mutual instruction of man, and improvement of his condition, seems to have been peculiarly and benevolently designed by nature, when she made them, like fire, expansible over all space, without lessening their density at any point, and like the air in which we breathe, move, and have our physical being, incapable of confinement, or exclusive appropriation. Inventions then cannot, in nature, be a subject of property. Letter from Thomas Jefferson to Isaac McPherson, 13 August 1813 (Jefferson, 1861: 175, 180).
(12) Professor Fisher traces its origins to the late nineteenth century; see Fisher (1999: 2, 8).
(13) But among lawyers, however, this does not mean patents cannot be considered "property." All property is held subject to public necessity; any property right is defined in relation to conceptions of the public good. What makes a right a "property right" is that the holder has the right to alienate that right, and the person wanting the right must negotiate with the holder before he or she can use that right. Among lawyers, what defines a right as a property right is that the law protects an individual's right to dispose of that right as he or she chooses.
(14) Electronic Communications Privacy Act, 100 Stat. 1848 (1986), 108 Stat. 4279 (1994). See reference in Bartnicki v. Vopper, 121 S. Ct. 1753 (2001).
(15) See Omnibus Crime Control and Safe Streets Act, Title III, 82 Stat. 211 (1968); Bartnicki v. Vopper, supra. See also U.S. West v. FCC, 182 F.3d 1224 (10th Cir. 1999) (striking certain privacy regulations of businesses on First Amendment grounds). Bartnicki is a centrally important case defining the future balance between privacy and First Amendment interests. In my view, the key distinction that would enable privacy regulations to survive is that they regulate a kind of access, and not the use of the facts accessed. I believe a property right could be so structured. For a careful and balanced (if skeptical) view toward the other side, see Singleton (2000: 97).
(16) This view is well developed in Rotenberg (2001, [paragraph] 92).
(17) Rotenberg rightly criticized my overly condensed treatment of his position in Code (Lessig, 1999: 161). His and EPIC's view is a far more subtle mix of policies that builds directly and strongly on an important tradition in privacy law that balances privacy interests that are to be kept out of the market with interests that can, properly, be within the market. I have little to criticize about that balance in the context of these traditional, and critical, privacy concerns. My focus here is on the emerging issue of informational privacy, and the particular issues the expanded capacity to manipulate those data creates.
(18) For an extraordinary website that summarizes the debate on P3P exceptionally well, see "P3P Viewpoints" (http://www.stanford. edu/~ruchika/P3P/>. See also the World Wide Web Consortium .

References

Brandeis, Louis, and Samuel Warren. "The Right to Privacy." Harvard Law Review 4 (1890).
Calabresi, Guido, and A. Douglas Melamed. "Property Rules, Liability Rules and Inalienability: One View of the Cathedral." Harvard Law Review 85 (1972).
Cohen, Julie. "Examined Lives: Informational Privacy and the Subject as Object." Stanford Law Review 52 (May 2000). "Developments in the Law--The Law of Cyberspace." Harvard Law Review 112 (May 1999): 1647-48.
Epstein, Richard A. "A Clear View of the Cathedral: The Dominance of Property Rules." Yale Law Journal 106 (1997): 2091.
Fisher III, William W. The Growth of Intellectual Property: A History of the Ownership of Ideas in the United States (1999) .
Internet Society Panel on Business Method Patents <http://www.oreillynet.com/lpt/a/434>.
Jefferson, Thomas. The Writings of Thomas Jefferson. Vol. 6. Ed. H. A. Washington. 1861.
Kang, Jerry. "Information Privacy in Cyberspace Transactions." Stanford Law Review 50 (1998): 1193, 1246-94.
Lemley, Mark A. "Private Property." Stanford Law Review 52 (May 2000).
Lessig, Lawrence. "The Regulation of Social Meaning." University of Chicago Law Review 62 (1995).
--. Code and Other Laws of Cyberspace. New York: Basic Books, 1999.
Litman, Jessica. "Information Privacy/Information Property." Stanford Law Review 52 (May 2000).
Merges, Robert P., Peter S. Menell, and Mark A. Lemley. Intellectual-Property in the New Technological Age. New York: Aspen Law and Business, 2000: 557-794.
Rotenberg, Marc. "Fair Information Practices and the Architecture of Privacy: (What Larry Doesn't Get)." Stanford Technology Law Review 1 (2001) .
Safier, Seth. "Between Big Brother and the Bottom Line: Privacy in Cyberspace." Virginia Journal of Law and Technology 5 (Spring 2000).
Samuelson, Pamela. "Privacy as Intellectual Property?" Stanford Law Review 52 (May 2000).
Schwartz, Paul. "Beyond Lessig's Code for Internet Privacy: Cyberspace Filters, Privacy Control, and Fair Information Practices." Wisconsin Law Review (2000).
Shapiro, Carl, and Hal R. Varian. "U.S. Government Information Policy." School of Information Management and Systems, University of California, Berkeley. 30 July 1997 .
Sholtz, Paul. "Transaction Costs and the Social Costs of Online Privacy." First Monday 6:5 (May 2001) .
Singleton, Solveig. "Privacy Versus the First Amendment: A Skeptical Approach." Fordham Intellectual Property, Media and Entertainment Law Journal 11 (2000).
Zittrain, Jonathan. "What the Publisher Can Teach the Patient: Intellectual Property and Privacy in an Era of Trusted Privication." Stanford Law Review 52 (May 2000): 1201-1250.

Lawrence Lessig is Professor of Law at Stanford Law School. His publications include The Future of Ideas (2001) and Code And Other Laws of Cyberspace (1999) as well as articles on cyberspace regulation. COPYRIGHT 2002 New School for Social Research COPYRIGHT 2002 Gale Group