Stanford Law Review
Copyright (c) 2000 The Board of Trustees of Leland Stanford Junior University
May, 2000; 52 Stan. L. Rev. 1125
Privacy As Intellectual Property?
Pamela Samuelson
( Professor of Information Management and of Law, University
of California, Berkeley.)
I wish to thank Robert J. Glushko, Mark A. Lemley, Marc Rotenberg, Carol
Rose, Jason Schultz, Leah Theriault and Hal Varian for their insightful comments
on earlier drafts of the article, and Peter Huang, Jason Schultz and Leah
Theriault for excellent research assistance for the article. Research support
for this article was provided by NSF Grant No. SES 9979852. All Internet
citations were current as of May 22, 2000. Copyright © 2000 by
Pamela Samuelson and the Board of Trustees of the Leland Stanford Junior
University.
[*1126]
Introduction
Information privacy is a scarce commodity in cyberspace. n1 The technical
infrastructure of cyberspace makes it remarkably simple and inexpensive to
collect substantial amounts of information identifiable to particular individuals.
n2 Once these data have been collected, information technologies make it
very easy and cheap to process the data in any number of ways (for example,
to make profiles of particular users' interests). n3 Although some
privacy-enhancing technologies (PETs) are being developed and deployed, these
technologies have thus far done little to make cyberspace more privacy-friendly.
n4 The market incentives for firms to collect and process personal data are
very high. Data about users is not only useful in assessing how a firm might
improve service for its customers, n5 but it also has become a key
commercial asset which firms use both for internal marketing purposes and
for licensing to third parties. n6 Although the Clinton Administration
has worked very hard to persuade Internet economy firms to adopt privacy
policies and practices to make users more comfortable about engaging in e-commerce
transactions in cyberspace, n7 these efforts have done little to overcome
the inertia of the current technical and economic environment n8 that
is [*1127] generally hostile to privacy interests. n9 This
symposium has been convened to consider whether the law should play a greater
role in promoting greater information privacy in cyberspace.
A recent book succinctly stated the principal utilitarian argument for providing
greater protection to personal data in cyberspace and elsewhere:
Consider the incentives of a company that acquires private information. The
company gains the full benefit of using the information in its own marketing
efforts or in the fee it receives when it sells the information to third
parties. The company, however, does not suffer losses from the disclosure
of private information. Because customers often will not learn of the overdisclosure,
they may not be able to discipline the company effectively. In economic terms,
the company internalizes the gains from using the information but can externalize
some of the losses and so has a systematic incentive to overuse it.
This market failure is made worse by the costs of bargaining for the desired
level of privacy. It can be daunting for an individual consumer to bargain
with a distant Internet merchant ... about the desired level of privacy.
To be successful, bargaining might take time, effort, and considerable expertise
in privacy issues. n10
To overcome this market failure, some American commentators have proposed
that the law should grant individuals a property right in their personal
data which would enable individuals to bargain over which personal data to
reveal to which firms for what purposes. n11 Other American commentators
have recommended a contractual approach to protecting personal data in cyberspace
(or more generally). n12 Some suggest that the law should try to facilitate,
and perhaps to approximate, the "privacy agreement the two sides would reach
if they were both well informed and it was not expensive to reach an agreement."
n13 American commentators generally prefer market- [*1128] based
solutions to personal data protection over the strict comprehensive regulatory
regime adopted some years ago in Europe. n14
While utilitarian considerations weigh heavily in the minds of many Americans
who have written on information privacy issues, noneconomic considerations
provide an equally or more compelling rationale for legal protection of personal
data in cyberspace, according to other commentators. Those who conceive of
personal data protection as a fundamental civil liberty interest, essential
to individual autonomy, dignity, and freedom in a democratic civil society,
often view information privacy legislation as necessary to ensure protection
of this interest. n15 Others regard cognitive limitations on the ability
of individuals to comprehend and accurately assess the risks of revealing
personal data to others as a reason for the law to provide corrective measures.
n16 Still others argue for information privacy protection to guard against
identity theft, harassment, and other wrongful uses of personal information.
n17 Achieving consensus on the rationale for information privacy protection,
however, may be unnecessary if both economic and noneconomic considerations
favor greater protection for personal data in cyberspace. n18
Part I considers both the appeal and limitations of the property rights model
for protecting personal data. A property rights model offers two principal
benefits: First, it would establish a right in individuals to sell their
personal [*1129] data and thereby capture some of the value their
data have in the marketplace. Second, a property rights model would force
companies to internalize certain social costs of the widespread collection
and use of personal data now borne by others. By internalizing these costs,
firms may make better investment decisions about what data to collect and
what uses to make of the data. A property rights model for protecting personal
data nevertheless presents many problems. This approach to personal data
protection would, in essence, establish a new form of intellectual property
right in information. But it would be an intellectual property right of a
very different sort than existing regimes provide. Deep differences in the
purposes and mechanisms of traditional intellectual property rights regimes
and the proposed property rights regime in personal data raise serious doubts
about the viability of a property rights approach and about its prospects
of achieving information privacy goals.
Part II explores an alternative market-oriented legal regime for protecting
personal information. Such a regime need not ground itself in property law.
The law can establish a default rule providing individuals with certain rights
to control the collection or processing of personal information about them
while also providing individuals with the power to contract away this right
(e.g., when they receive compensation for doing so). Because market imperfections
may impede fair and effective licensing of personal data in cyberspace, the
law can supply some default terms for the licensing of personal data. Certain
trade secrecy licensing default rules may be adaptable to the licensing of
personal data. The Uniform Computer Information Transactions Act (UCITA)
may supply additional default rules for the licensing of personal data in
cyberspace. n19 Adoption of online privacy policies could facilitate
a market-based licensing approach to personal data protection. When Web sites
post notices saying personal data will not be collected, disclosed, or used
except for named purposes, users who supply data in reliance on those restrictions
may be able to enforce the restrictions. A market-based licensing approach
may also arise if technology evolves to allow "negotiated" agreements on
the collection, use, or disclosures of personal data.
Although this article endorses a licensing approach to the protection of
personal data, it recognizes that the law alone cannot solve information
privacy problems in cyberspace. Work must continue on evolving norms about
appropriate and inappropriate uses of personal data, on persuading firms
that the trust necessary for electronic commerce to flourish requires the
interests of individuals in information privacy to be given appropriate deference,
and on adapting the technological infrastructure of cyberspace so that information
privacy becomes easier to achieve. The principal challenge of these
[*1130] multifaceted endeavors is not to recreate in cyberspace some
preexisting zone of privacy from the physical world, n20 but to articulate
values inhering in information privacy that should constrain and structure
social, economic, technological, and legal relations. n21
I. The Appeal and Limitations of a Property Rights Approach To Protecting
Personal Information
A. The Appeal of a Property Rights Approach
It may seem natural for individuals to assume that they do or should own
data about themselves. n22 It is surely true that the law will enforce
the expectations of individuals that certain private information (for example,
a diary or journal) should remain secret. n23 Because individuals generally
have a legal right to exclude other people from access to their private data,
they may have a sense that they have a property right in the data as well
as a legal right to restrict access to it. Even when data about individuals
are in the hands of others (such as banks, doctors, and insurance companies),
individuals may perceive themselves to have a protectable interest in records
of their financial transactions or medical histories. n24 Because the
law will sometimes protect these and other types of data from unauthorized
uses and disclosures, n25 this too may reinforce a sense of ownership
in personal data.
Although the law often protects the interests of individuals against wrongful
uses or disclosures of personal data, n26 the rationale for these legal
protections has not historically been grounded on a perception that people
[*1131] have property rights in personal data as such. n27 Indeed,
the traditional view in American law has been that information as such cannot
be owned by any person. n28 The Fourth Amendment and real property
law may provide protection against certain unauthorized intrusions into one's
real or personal property for purposes of getting access to information that
might be stashed there, and the Fifth Amendment may provide protection against
compulsion to reveal certain information about oneself. But these results
are not grounded on a belief that people have property rights in information
about themselves, but on the recognition of legally protectable interests
of other sorts. n29 An individual, for example, may be able to obtain
relief if a doctor releases details of her medical history to a prospective
employer, but the individual's rights would arise under contract or privacy
law, not from the existence of any property rights in this information.
n30
Many examples illustrate that the law does not generally recognize the legal
right of individuals to control uses or disclosures of personal data.
n31 Individuals, for example, have no legal right to stop firms from marketing
[*1132] their personal data to other firms based on information that
the individuals disclosed on a product warranty card sent to manufacturers
of that product. n32 Nor can they stop state governments from selling
drivers' license data about themselves. n33 Thus, however intuitively
powerful the notion of property rights in one's data may be, it is clear
that in the United States the existence of some legally protectable interests
in personal data in certain circumstances is not equivalent to a legal rule
that a person has a property interest in one's personal data. n34
In recent years, a number of economists and legal commentators have argued
that the law ought now to grant individuals property rights in their personal
data. n35 Some favor propertizing personal data as a way to allow individuals
to make appropriate deals for selling their personal data and to receive
compensation for uses of their personal data so that markets in personal
information will work more fairly. n36 Others favor propertizing personal
data as a way of forcing companies to internalize more fully the costs associated
with the collection and processing of personal data, in the hope that this
will lead to greater privacy. n37
There is at the moment a "lively market" in personal data, but it is a market
in which individuals play at most a very small role. n38 Many firms
collect and process personal data because of its value and because information
technology makes the collection and use of such data so much easier and cheaper.
n39 They also do so because they are not forced to internalize the societal
[*1133] costs of private sector processing of personal data.
n40 Because they may have invested time, money and energy in compiling, organizing,
or processing the data, they may well think of themselves as owning the data
they have gathered or otherwise acquired. n41 Perhaps firms would collect
or process less personal data than they currently do if they had to pay individuals
for rights to do so. n42 If so, this would simultaneously achieve information
privacy goals and allow individuals who wish to sell their data to receive
some benefits from this market. In addition, a property rights regime might
enable firms to make fewer wasteful investments in personal data and to develop
higher quality databases, since individuals would presumably agree to release
personal data to firms from whom they would be willing to receive information,
and would have less incentive to lie as a way to protect their privacy.
n43
[*1134] Governments clearly have power to create property rights when
appropriate to cure or ameliorate market failure problems. n44 Creating
property rights in informational assets is, in fact, remarkably common. Intellectual
property law grants exclusive rights in information-based creations in order
to promote development of a thriving marketplace for them. n45 A number
of commentators have observed that, in an information economy, an increasing
commodification of information and a creation of new property rights seems
almost inevitable. n46 Granting individuals property rights in their
data would seem consistent with this general trend and with the emergence
of an "attention economy." n47
A property rights approach to solving the information privacy problem may
also be consistent with survey evidence suggesting that most Americans are
willing to disclose personal data to businesses and allow them to use these
data as long as the individuals obtain a discernible benefit from this disclosure
and use (for example, a discounted price for certain goods or services).
n48 If what upsets Americans most about the loss of control over their personal
data is that they are not receiving any benefits arising from private sector
reuses of the data, a property rights approach would arguably provide individuals
with a way to exercise meaningful control over the market in personal data
which they do not currently enjoy. This would arguably cure a market failure,
as well as halt the unjust enrichment that compilers of personal information
now enjoy.
A property rights approach may be especially useful to accommodate the varying
preferences of individuals about private sector uses of personal data.
n49 Although some individuals may value privacy so highly that they will
choose not to engage in market transactions about their personal data, others
may be quite willing to sell their personal data to firms A, B, and C (even
if [*1135] not to X, Y, or Z). Or they may be willing to sell
personal data about their recreational interests, but not about the associations
to which they belong. The market arguably provides an efficient device -
namely the price mechanism - with which individuals can express their preferences
about who should be able to use which of their personal data and to what
degree. n50 Private sector buyers would, of course, dicker on price
and other terms, but economists generally assume that the market is a good
way to achieve an efficient outcome satisfactory to both buyer and seller.
n51 If the market works well in enabling transactions in other commodities,
it would presumably work for transactions in personal data as well.
A property rights approach to the information privacy problem would involve
substantial transaction costs for individuals if they have to separately
negotiate with each prospective buyer of their personal data. n52 To
overcome such problems, some commentators have predicted the emergence of
new businesses to serve as intermediaries on behalf of individuals to represent
their interests and negotiate with buyers of these data ("infomediaries").
n53 Others anticipate the development of electronic agents to perform negotiations
and make deals to sell personal data in cyberspace. n54 Still others
expect individuals to be able eventually to program their browser software
to incorporate their privacy preferences. n55 Well-programmed browsers
might then avoid Web sites that do not conform to their masters' preferences
and only make automated deals with Web sites whose privacy terms are within
an acceptable range.
A property rights approach offers a further potential advantage over other
legal approaches to protecting privacy in that it could protect personal
data without requiring the establishment of a substantial government bureaucracy,
as some nations have done, to oversee regulation of personal data protection.
n56 Americans generally disfavor the substantial costs associated [*1136]
with direct government oversight of industry practices. They also tend to
bristle if the government requires firms to establish internal oversight
procedures and structures, as the European Directive requires. n57
To the extent that a property rights approach would avoid such costs, this
would seem to be another factor in its favor.
B. Limitations of a Property Rights Approach
Despite these appealing features, 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. A property rights
approach may have some unintended consequences that proponents of this approach
have not recognized.
To understand some possible disadvantages of the property rights approach,
it is necessary to think beyond the initial creation of a property right
in an individual's personal data. Proponents implicitly assume that the creation
of the property right is the only significant act necessary to enable the
growth of a functioning market in which individuals could engage in personal
data transactions. n58 Kenneth Laudon is one of the few commentators
to consider the infrastructure required to make a property rights system
work. n59
Laudon proposes the establishment of a regulated National Information Market
(NIM) to allow "personal information to be bought and sold, conferring on
the seller the right to determine how much information is divulged."
n60 Individuals would first "establish information accounts and deposit their
information assets and informational rights in a local information bank,
which could be any local financial institution interested in moving into
the information business." n61 The banks would then pool these information
assets and sell "baskets" of them in a National Information Exchange.
n62 Buyers would receive the right to make commercial uses of personal information
in those baskets for stated periods of time, in exchange for compensation
[*1137] paid to the seller-banks. The banks would then equitably allocate
this compensation among the individuals whose information was pooled in a
particular basket (less a service fee). n63 Laudon foresees assigning
every participant in the NIM a unique identifier and barcode symbol (to be
known as a National Information Account number) which "would help individuals
keep track of who is using their information by informing the account whenever
the individual's name is sold as part of a basket of information."
n64 Laudon proposes to make it a crime to use personal information without
permission. n65 He also foresees a substantial role for government
oversight of this market. n66
One need not agree with all the particulars of Laudon's vision in order to
agree with his basic insight that an institutional infrastructure would be
needed to make a new property rights market in personal information work.
Even if one "grandfathered" in private sector "rights" to continue using
personal data collected before the effective date of legislation establishing
a property right in personal data, the new property system would introduce
significant "friction" to a market that currently operates without it. This
friction may be justifiable as a way to force data compilers to internalize
certain costs they currently impose on others, n67 but it is fair to
say that the costs of establishing new procedures and implementing them would
be far from trivial for both companies and for individuals. n68 Collectors
of personal data would presumably have to pay individuals for rights to process
the data; this cost would unquestionably have to be passed on to others in
the form of higher prices for the firms' own products or services, and establishing
an enforcement system would also be costly. Property rights systems are not
costless. n69 Too little thought has been given as yet to how to move
from where we are today to a thriving market in personal data under a property
rights regime in which individuals would have a right to control market transactions
in data about themselves.
Achieving information privacy goals through a property rights system may
be difficult for reasons other than market complexities. Chief among
[*1138] them is the difficulty with alienability of personal information.
n70 It is a common, if not ubiquitous, characteristic of property rights
systems that when the owner of a property right sells her interest to another
person, that buyer can freely transfer to third parties whatever interest
the buyer acquired from her initial seller. n71 Free alienability works
very well in the market for automobiles and land, but it is far from clear
that it will work well for information privacy. An individual may be willing
to sell his data to company N for purpose S, but he may not wish to give
N rights to sell these data to M or P, or even to let N use the data for
purposes T or U. The individual may be able to make a reasonable estimate
of the value they should receive from N for a grant for S purpose, but may
at the time of transacting with N be unable to assess what value he should
receive for any transfer of the same data to M, P, or any other licensee
of N. Collectors of data may prefer a default rule allowing them to freely
transfer personal data to whomever they wish on whatever terms they can negotiate
with their future buyers. However, individuals concerned with information
privacy will generally want a default rule prohibiting retransfer of the
data unless separate permission is negotiated. They will also want any future
recipient to bind itself to the same constraints that the initial purchaser
of the data may have agreed to as a condition of sale. Information privacy
goals may not be achievable unless the default rule of the new property rights
regime limits transferability.
Consider also that the most common justification for granting property rights
- to enable market allocations of scarce resources - does not seem to apply
to personal data. n72 What is scarce is information privacy, not personal
data. If anything, personal data are being too plentifully distributed in
the marketplace right now. Indeed, a reason many people argue in favor of
granting individuals property rights in these data is, in essence, to make
the distribution of them scarcer. While there are other instances in which
property rights have been created in order to make a too plentiful a resource
more scarce - for example, the creation of property rights to allow emissions
of [*1139] pollutants up to certain levels as a way to achieve
environmental goals n73 - such a property rights system works because
of the free transferability of the property rights. The right to pollute
to a certain level is, by virtue of the property right grant, made into a
scarce resource that the market can then allocate efficiently. n74
The alienability of this property right is an essential part of what enables
the property regime to accomplish its objective of controlling pollution
levels. Yet, as noted above, the free alienability of property rights in
personal data may prove troublesome.
Consider also differences between the rationale for the proposed property
rights in personal information and the rationale for existing property rights
regimes that regulate markets for information-based products, namely, intellectual
property law. n75 The economic rationale for intellectual property
law arises from a public goods problem with information products that this
law strives to overcome. n76 In the absence of intellectual property
rights, there may be too little incentive to induce an optimal level of private
investments in the production and dissemination of intellectual products.
Everyone benefits if such investments are made, regardless of whether they
are in technological, [*1140] artistic or literary fields.
n77 However, without a legal protection system, creators will find it difficult
to exclude free-riders from appropriating the fruits of their labor and selling
identical or very similar products in the marketplace at a cheaper price.
n78 The prospect of being unable to recoup research and development costs
may deter such investments from being made in the first place. n79
A limited grant of property rights in intellectual productions gives creators
assurance that they can control the commercialization of their work and enjoy
some fruits of their labor, assuming the market finds the product attractive.
n80
The standard rationale for granting property rights in personal data is,
of course, quite different. n81 The personal data most likely to become
the subject matter of such a property right, for the most part, already exist.
Property rights are not needed to bring them into being, nor to achieve widespread
distribution of them. There are, in addition, no research and development
costs to recoup. It is, of course, possible that people might invest more
time, money and energy in the creation of additional personal data about
themselves (for example, hobbies the person would like to have or famous
people the person would want to meet) if they could assert property rights
in this new data, but there is some reason to think that people may be willing
to do this even without a grant of property rights in the data. n82
A further cause for concern about a property rights approach to protecting
personal data is the potential that such grant of intangible rights in intangible
information will lead to greater incoherence in intellectual property law.
A fundamental principle for Congressional grants of intellectual property
rights is that such legislation should "promote the Progress of Science and
[the] useful Arts ...." n83 It is difficult enough these days for Congress
to adhere to this principle: Expanding intellectual property law to protect
personal [*1141] data would only strain the coherence of this
body of law further. n84 This constitutional principle does apply to
personal data. The creation and dissemination of personal data does not generally
promote "science" in the constitutional sense (i.e., knowledge), n85
nor does it promote technological innovation. n86 Indeed, the purpose
of the proposed new personal data property right is almost the inverse of
traditional intellectual property law, for it would grant a property right
in order to restrict the flow of personal data to achieve privacy goals.
n87
It is also far from clear what constitutional authority Congress would have
to enact legislation creating a property right in personal data. Given the
mismatch between the purposes of personal data protection and of traditional
intellectual property rules, it would be difficult to justify such legislation
under the enabling clause for copyright and patent legislation. n88
Because of the interstate character of the Internet and web, it might be
possible to justify congressional legislation granting property rights to
personal data in cyberspace under the Commerce Clause. n89 However,
a more general grant of property rights in personal data might be constitutionally
troublesome. n90 [*1142] Grants of property rights are
generally the province of state law. n91 Indeed, the state law doctrine
out of which a property right regime in personal data would seem the most
natural extension is right of publicity law which gives individuals some
rights to control commercial exploitation of their names, likenesses, and
other indicia of the commercial value of their person. n92 Although
the right of publicity has often been characterized as a property interest,
n93 it is an interest that law has allowed celebrities, not ordinary folk.
n94
Creating a property right in personal data may, moreover, be objectionable
to those who consider information privacy to be a fundamental civil right.
n95 While the civil right conception of personal data protection is predominant
in Europe, n96 sometimes this conception is evident in U.S. decisions
[*1143] on privacy, n97 even in cases involving uses or disclosures
of personal data. n98 Other cases have been less deferential to information
privacy as a protectable civil liberty interest, n99 but this conception
of information privacy unquestionably has adherents in the United States.
n100 From a civil liberties perspective, propertizing personal information
as a way of achieving information privacy goals may seem an anathema.
n101 Not only might it be viewed as an unnecessary and possibly dangerous
way to achieve information privacy goals, it might be considered morally
obnoxious. If information privacy is a civil liberty, it may make no more
sense to propertize personal data than to commodify voting rights.
n102
Europeans have more of a civil libertarian perspective on personal data protection
in part because of certain historical experiences they have had. n103
One factor that enabled the Nazis to efficiently round up, transport, and
seize assets of Jews (and others they viewed as "undesirables") was the extensive
repositories of personal data available not only from public sector but also
from private sector sources. n104 Europeans may realize more than most
[*1144] Americans the abusive potential for reuses of personal data
that may initially have been provided to a particular entity for a specific,
limited purpose. If more Americans had an appreciation of the negative consequences
that might arise from commercial distributions of their personal data, they
might perceive personal data protection differently. n105
Congress has sometimes legislated information privacy protections out of
concern about cognitive difficulties in appreciating the risks of supplying
personal data to private sector firms, for example, in respect of gathering
information from children under the age of thirteen. n106 On occasion,
Congress has also recognized that adults, too, may not appreciate certain
risks in supplying personal data to private sector firms and has decided
that in those instances even the adults should be protected. When renting
certain video cassettes from a corner rental store, Robert Bork, for example,
surely did not anticipate that he was running the risk that the owner of
the video store might disclose his rental choices to the press while he was
a nominee to the U.S. Supreme Court. n107 The disclosure of his viewing
choices was not, under then existing law, illegal. It is illegal now. And
the Video Privacy Protection Act [*1145] is far from the only
law of this kind. n108 Congress has also acted to protect individuals
against public sector commercialization of drivers' license data in part
because of the involuntary nature of this particular kind of data collection
and in part because of negative consequences arising from the widespread
market availability of such data. n109
As difficult as it may be for the average person to judge the risks of personal
data misuse as a general matter, it may be even more difficult for the average
person to judge the risks of selling her property rights in personal data.
n110 Data collectors may well insist on broad transfers of all of a person's
right, title and interest in her personal data. n111 While such a broad
transfer works very well in a sale of a used car or a house, it may be troublesome
in the context of personal data. As a result of such a transfer, an individual
could potentially be foreclosed from any control over these data in the hands
of the transferee or in the hands of other firms to whom the data might have
been transferred. The individual could even be precluded from engaging in
further transactions to sell the same data to other firms because her rights
in the data now belong to a personal data aggregator. Other firms wanting
to access or use these data would have no choice but to go to the data aggregator
and license the data from that firm - on terms that would likely reflect
the interests of the aggregator rather than those of the individual whose
data has been licensed.
This cluster of problems could be mitigated if the individual makes a more
limited grant of rights to a data aggregator, n112 but this may suggest
that a different approach to protecting information privacy may be more satisfactory
than a property rights approach. It is unusual for a property rights regime
to establish a rule or strong presumption against alienability. n113
A [*1146] property approach may also thwart information privacy
goals unless the law makes it clear that a person does not abandon property
rights in personal data when visiting Web sites that collect personal data.
n114 The rhetoric of property law may also be unsuited to further elucidation
of normative understandings about acceptable and unacceptable uses of personal
data that is sorely needed in this era of rapid technological, economic,
and social change.
C.A Moral Right in Personal Data?
As vigorously as this subsection has argued against a property rights model
for protecting personal data, it has done so because the standard models
of property rights seem unsuitable to achieving information privacy goals.
There is, however, one rather unusual class of property right that protects
personhood interests of individuals, melding economic, reputational, and
autonomy interests at its core. In the spirit of providing exemplars from
the existing tool kit of property law, it may be worth mentioning "moral
rights" of authors as a model for a nontraditional property right that might
be adaptable to protecting personal data. n115
In Europe and many other nations, authors have "moral rights" in the works
they have created. n116 These rights are distinct from the purely economic
rights that European law, like American copyright law, grants to authors.
The moral rights regime derives from a conception of artistic and literary
creations as emanations of the author's personality in which he can
[*1147] and should retain an interest even after copies of the work
have entered the stream of commerce. n117 Among the commonly recognized
moral rights are the rights of attribution (i.e., the right to be identified
as the author of the work) and of integrity (i.e., the right to protect the
work from alterations that would be harmful to the author's reputation).
n118 In some jurisdictions, authors also have moral rights of "divulgation"
(i.e., the right to decide when and under what circumstances to divulge the
work) and sometimes even of withdrawal (i.e., the right to withdraw all published
copies of the work if the work no longer represents the author's views or
otherwise would be detrimental to the author's reputation). n119
Moral rights are generally waivable by contract, although some countries
- notably France - regard such rights as sufficiently important and vulnerable
to unfair contractual overrides that they have made them inalienable rights.
n120 An advantage of moral rights is that these rights can be exercised long
after the author has sold copies of her work to the public and can be exercised
against remote purchasers. If the owner of a sculpture, for example, alters
it in a way that the sculptor deems detrimental to his interests (for example,
by tying red ribbons around its neck), the sculptor can assert his moral
right of integrity in the work and can obtain injunctive relief requiring
restoration of the original. n121 While moral rights generally focus
on the personal, reputational interests of authors, an economic consideration
may partly [*1148] underlie moral rights. "Mutilation" of an
author's work can tarnish the author's reputation in ways that may be difficult
to measure, akin to the harm to goodwill when trademarks are disparaged or
tarnished. n122
A moral rights-like approach might be worth considering as to personal data.
As with the moral right of authors, the granting of a moral right to individuals
in their personal data might protect personality based interests that individuals
have in their own data. The admixture of personal and economic interests
could be reflected in the right. The integrity and divulgation interests
may be the closest analogous moral rights that might be adaptable to protect
personal data. An individual has an integrity interest in the accuracy and
other qualitative aspects of personal data, even when the data are in the
hands of third parties. n123 An individual also has an interest in
deciding what information to divulge, to whom and under what circumstances.
n124 An advantage of a moral rights-like approach is that this right can
be asserted against persons beyond those with whom one has contracted. Contract
law, in general, provides relief for breach as between the parties to a contract,
not rights against third parties. n125 Firms that collect and process
personal data are often not in privity with the individual whose data is
being used. n126
[*1149] A moral right-like approach would overcome a second important
limitation of a purely contractual approach which generally aims to compensate
the non-breaching party through an award of damages, not by granting injunctive
relief. n127 Property law, in contrast, generally allows the owner
of the right to exclude other people from engaging in certain activities,
and injunctive relief is consequently generally available. n128 A person
who has licensed a particular use of her personal data, but not another use,
would almost certainly want injunctive relief upon learning that her licensee
is using the data for more than the authorized purpose. n129 A property
right in her personal data could provide grounds for injunctive remedy.
[*1150] However, the idea of creating a moral right-like interest
in personal data presents many difficulties. For one thing, U.S. law has
generally been inhospitable to the idea of moral rights of authors,
n130 even though it has ratified a treaty that requires such protection.
n131 This augurs poorly for adaptation of the concept to protection of personal
data. It is also unclear what constitutional authority Congress would have
for enacting legislation of this sort. Moreover, even the Europeans might
balk at the idea of generalizing the moral right concept for personal data
because it undermines the special status of authorship that provides the
theoretical justification for existing moral rights law. n132
Two state law doctrines out of which a moral right-like interest might emerge
are right of publicity law and the appropriation branch of privacy law. Right
of publicity law, like moral rights law, has generally protected the interests
of special status individuals (in the case of publicity rights, the interests
of "celebrities"), n133 and like intellectual property laws, publicity
law largely concerns itself with providing appropriate incentives to induce
investments in creative efforts, not to protect personality based interests.
n134 The appropriation tort could be extended to provide individuals with
a protectable interest in personal data. n135 Even though the right
created would not be a "property right," n136 it could still allow
individuals to contract for allowable [*1151] uses of personal
data n137 and to police third party uses of personal data insofar as
these uses were unreasonable. n138 This tort protects dignity-, integrity-,
and autonomy-based interests of individuals by setting bounds on acceptable
behavior. n139 However, the appropriation privacy tort seems an unsuitable
way to establish a market-based system for enabling transactions in personal
data in which the individual participates, even though this is what many
Americans seem quite willing to do with personal data. n140 The next
section will explain why a licensing system built on modified trade secrecy
default principles might offer a useful model for licensing of personal data,
and will offer some suggestions about how such a system might be implemented
to facilitate greater protection for personal data in cyberspace.
II. Modified Trade Secrecy Default Rules for Promoting Information Privacy
The law can grant individuals a protectable interest in their personal data
without grounding that interest in property law. n141 It can do so
by setting a default rule forbidding certain activities with respect to these
data, such as unauthorized collection or uses of them unless the individual
has agreed to these activities. n142 Because market imperfections make
it difficult to negotiate effectively about terms of use as to personal data,
n143 it may make sense to establish some default terms for such agreements
which the parties could override if they so choose. Although trade secrecy
and information privacy [*1152] laws obviously differ in many
significant respects, these laws nonetheless have at least three important
interests in common: first, an interest in protecting the interest of the
claimant to restrict access to and unauthorized uses of secret/private information;
second, an interest in giving firms/individuals control over commercial exploitations
of secret/private information; and third, an interest in setting and enforcing
minimum standards of commercial morality. To achieve policy goals embodied
in these interests, trade secrecy law has evolved a set of default licensing
rules. Some of these default rules may be adaptable to the licensing of personal
information.
A. Rationale for Adapting Trade Secrecy Default Rules to Licensing of Personal
Data
Like the information privacy law contemplated in this article, trade secrecy
law facilitates license transactions in information, while at the same time
providing default rules to govern uses and disclosures of protected information,
and setting minimum standards of acceptable commercial practice. Information
privacy rights, like trade secrecy rights, can be based on three types of
interaction: first, on contractual agreements; second, on conduct between
the parties from which it is reasonable to infer that information was disclosed
in confidence, and use and disclosure beyond those purposes is wrongful;
and third, on the use of improper means to get the information. n144
Agreement-based trade secrecy typically occurs when A has nonpublic information
to which B wants access. A agrees to give B access to the information in
exchange for B's agreement to respect certain restrictions on its use, and
abide by other terms and conditions (for example, payment of a stated sum
or royalty). Because of the exchange value of such information, trade secret
information can be a highly valuable asset of the firm and provide it with
a substantial revenue stream. n145 The information, however, does not
become "public" simply because a number of firms possess it - as long as
each is under an implicit or explicit pledge to maintain the nonpublic status
of the information. n146
Confidential relationship-based trade secrecy may arise when A reveals certain
nonpublic information to B under circumstances in which B would have reason
to understand the limited purpose of the disclosure, and that use [*1153]
and disclosure for other purposes would be wrongful. n147 For example,
if a firm discloses certain nonpublic information about the firm's operations
to a consultant, the consultant will understand that he is entitled to use
this data only for purposes of analysis in order to advise the company about
how to improve its operations. The revealed information may have a commercial
value beyond its utility to aid the consultant in doing his job, but the
consultant understands that it would be inappropriate to sell or release
the information to another firm or to reveal it to stockbrokers so they could
make better decisions on whether to trade in that firm's securities. Both
the consultant and the firm would understand, even if they did not specifically
agree, that rights to control uses of the information reside in the firm,
not the consultant.
For similar reasons, individuals often regard the data that they reveal to
others - their accountants, doctors, and banks, to name a few examples -
as having been provided to those firms for limited purposes. Uses and disclosures
of the data, whether internally or to third parties, may be inappropriate
unless undertaken for purposes consistent with the initial disclosure. Just
as the consultant could not justify revealing information to a third party
on a theory that this disclosure would enable the other firm to provide new
or better service to the company, individuals may be skeptical of those who
argue that disclosure of their personal data to a third party is justifiable
because it enables that firm to offer service to them.
In trade secrecy law, as in the information privacy law contemplated in this
article, there is no need to say that a property right exists in the protected
information. n148 Although courts have sometimes loosely referred to
trade [*1154] secrets as the "property" of the firm that licensed
them and have on occasion held trade secrets to be property for certain purposes,
n149 the more appropriate way to characterize the firm's interest in a trade
secret is to say that the law protects the firm against breaches of contracts
and confidential understandings, n150 [*1155] as well as
against the use of improper means to obtain the secret. n151 Despite
its frequent presence in texts of intellectual property law, n152 trade
secrecy law remains firmly rooted in unfair competition law. n153 A
true intellectual property right provides the owner with rights to exclude
that are good against the world at large as to innovations that are generally
widely distributed to the public. n154 Trade secrecy law, by contrast,
remains a tort law that enforces minimum standards of commercial morality.
n155 Going through trash bins outside a firm's office may, for example, be
an acceptable way for the government to obtain information when investigating
a crime, n156 but the law of trade secrecy regards this means of obtaining
trade secrets to be improper and the trash searcher as a misappropriator
of trade secret information. n157
Trade secrecy law has a number of default rules that might be useful for
information privacy protection. The general rule of trade secrecy licensing
law is that if the licensor has provided data to another for a particular
purpose, the data cannot be used for other purposes without obtaining permission
for the new uses. n158 Licensing law generally accommodates the
[*1156] reasonable expectations of the parties. n159 If a licensor
has failed to specify a limitation on use, the limitation may still be enforced
so long as circumstances surrounding the agreement reasonably support an
implicit understanding about limitations on use. n160 Moreover, licensing
law generally permits revocation of the license for breach of material terms.
n161 Contract law, far more than property law, takes into account cognitive
difficulties individuals may have in assessing the risks of certain transactions
and provides protections to overcome these cognitive problems. n162
Some of these doctrines may be adaptable to licensing of personal data, particularly
in view of the cognitive difficulties people often have in assessing the
risks of permitting certain uses of personal data. n163
One of the most significant advantages of the licensing regime is that it
avoids the problems of a property rights approach deriving from its preference
for free alienation. The general default rule of trade secret licensing law
is that license rights are nontransferable unless the licensor grants a right
to sublicense. n164 Sublicenses, if permitted, generally oblige the
sublicensee [*1157] to abide by the same terms as the license
imposes on the now sublicensor. n165 Licenses are also nonexclusive
unless expressly provided otherwise. n166
Trade secrecy law also provides some rights against third party uses of protected
information. n167 If a third party has obtained the protected information
from one whom the party knows or has reason to know got the information by
improper means, or in breach of confidence, the trade secret can be enforced
against the third party. n168 If the third party got the information
innocently, the firm seeking to protect the information may nevertheless
be able to stop unauthorized use of the information after giving notice to
the third party about its rightful claim to control uses of the information.
n169
Adopting modified trade secrecy licensing default rules for protecting personal
data may also be less likely to interfere with or contribute to confusion
in the law in respect of intellectual property rights and the First Amendment
because it would focus on enforcing agreements and confidential relationships
and monitoring acceptable commercial practices. n170 In addition, such
an approach makes it unnecessary to engage in a quasi-religious [*1158]
war to resolve whether the nature of a person's interest in her personal
data is a fundamental civil liberty or commodity interest. n171 A licensing
approach to protecting personal data is consistent with the widespread use
of licenses in the digital networked environment. n172 If software
and Internet companies have devised licenses to cover virtually every Internet
transaction between them and their customers, it may seem only fair for the
customers to start insisting on contractual terms that serve their interests
as well.
It is also noteworthy that virtually all of the advantages offered in support
of the property rights approach for the legal protection of personal data
would be achievable through a licensing regime. n173 A licensing model
would allow a market to exist in personal information insofar as individuals
wished to participate in that market. New infomediary businesses could also
arise under a licensing regime. Licensing also avoids the need for a government
bureaucracy to regulate information privacy practices. Like the property
model, the licensing model assumes that the marketplace can generally achieve
workable outcomes.
There are obviously significant differences between trade secrets and personal
information which may require each law to have different rules. n174
However, borrowing trade secrecy licensing default rules makes sense insofar
as a person and a firm have agreed that the person will reveal nonpublic
information to the firm in exchange for a stated sum and a willingness to
restrict uses of the information to stated purposes. It also makes sense
when a person reveals information to a firm in circumstances in which it
is fair to infer that the information has been disclosed in confidence and
for limited purposes. Borrowing from trade secrecy law's default rules may
even make sense if one can articulate some means of obtaining personal data
that the law should be considered improper. Consider, for example, the impropriety
in getting personal data by engaging in unauthorized surveillance, by fraud,
trickery, misrepresentation, or by hacking into a cryptographic envelope
in which the data are being stored. n175 The law of information privacy,
like the law of trade secrecy, could monitor commercial morality, adapt to
changing [*1159] circumstances, and at the same time accommodate
the interests of individuals who are quite willing to reveal or allow uses
of their personal information as long as they derive a benefit from it.
B. Developments That Might Cause Licensing To Emerge As a Viable Solution
to Cyberspace Information Privacy Problems
Societal consensus about appropriate and inappropriate uses of personal
information in cyberspace is forming in the United States, shaped in part
by news coverage about information privacy issues. Several times a week,
major news stories about information privacy issues appear. One day the story
may be about legislation forbidding states to sell drivers' license data
as a commercial product. n176 Another day someone will have discovered
that widely used software is sending surreptitious messages back to the firm
when a user is playing a sound recording. n177 Yet another day will
bring news that Congress has passed legislation to deregulate the financial
services industry will enable subsidiaries to share information about customers
(which the industry claims will promote better service to customers and which
privacy advocates say will bring harmful consequences, for example, the denial
of a person's application for a mortgage on the ground that the insurance
data about him suggests he will not live long). n178 In view of the
negative publicity that occurs when information privacy is not respected,
major Web sites now worry about whether the information sharing they do is,
in fact, fair or unfair. n179 This publicity has caused firms to back
down very publicly when they have acted in a privacy-unfriendly way.
n180 Internet companies know [*1160] that an installed base of
millions of users can quickly evaporate if customers do not trust the provider.
While fears of negative publicity is one inducement to attend to information
privacy concerns, companies have realized that the news can be favorable
as well, as when it publicizes private sector initiatives to further information
privacy goals. The Online Privacy Alliance has been particularly active in
taking a proactive stance on information privacy policy issues and getting
the word out about its initiatives. n181 Industry commentators also
frequently point out that information privacy is key to building trust among
consumers and trust is essential for the promise of e-commerce to be realized.
n182 In addition, American firms with substantial international market presence
are becoming more attentive to information privacy practices and policies
because of the need to comply with data protection rules in other jurisdictions.
n183
1. From self-regulation norms to licensing.
For e-commerce Web sites, having a privacy policy is no longer optional.
Federal legislation, Federal Trade Commission (FTC) enforcement, the European
Union Privacy Directive, economic coercion, and consumer demand have all
recently converged to create a new environment in which implementing a privacy
policy is a business necessity for most, and legally advisable for all.
n184
To give content to "self-regulation," the Clinton Administration has endorsed
privacy principles that it strongly recommends private sector firms should
adopt as part of a self-regulatory strategy. n185 The FTC announced
the following five pairs of principles as critical components of a true self-regulatory
regime:
. Notice/Awareness
. Choice/Consent
. Access/Participation
[*1161] . Integrity/Security
. Enforcement/Redress. n186
In 1998, the FTC conducted a survey of more than 1400 commercial Web sites
on privacy policy practices. The agency reported to Congress that a high
proportion of such sites (92%) collected personal information from visitors
to their sites, although nearly as substantial a proportion (86%) provided
no notice about their information privacy policies. A year later, the FTC
reported a substantial increase in the proportion of commercial Web sites
that provided some notice about their sites' privacy policies. n187
Based on this progress, the FTC indicated that self-regulation should be
given additional time to succeed. n188
While it is true that more online firms have privacy policies today, it is
also true that if the FTC had judged the adequacy of privacy policies based
on the criteria it set forth defining meaningful notice, the agency might
have perceived less progress than it reported. n189 And if it judged
progress based on private sector adherence to all five privacy principles,
it might well have concluded that self-regulation had a very long way to
go. Nevertheless, there is some evidence that American-based commercial Web
sites provide more notice about privacy policies now than they did a year
ago. n190 Some progress also continues in implementation of the other
principles, in part because of the well-publicized actions of major firms,
such as IBM Corp., that have announced they will not place advertising with
Web sites that do not meet certain privacy standards. n191
Providing users with meaningful notice about what information a site is collecting
about an individual, and what the site intends to do with this data, is definitely
a step in the right direction. Notice alone, particularly one that is vague
in content, may provide little basis for inferring that the site owner has
bound itself to collect only these data and use the data only for stated
purposes. However, misrepresentations in Web site privacy notices about
[*1162] the collection or use of personal data might be actionable.
n192 In addition, the FTC has authority to monitor sites to ensure that they
are not engaging in deceptive or other unfair trade practices with respect
to personal data they collect. n193 And the FTC, among other agencies
and groups, can be expected to press for greater adherence to the privacy
principles over time.
As firms adhere more fully to the FTC privacy principles, it may enable the
emergence of a contractual basis for holding firms to privacy representations.
The more notice a Web site gives about what data will be collected and for
what purposes, and the more the site seeks consent for collection and use
of personal data, the more robust the firm's representations about the integrity
of its data and the security with which it maintains the data. Also, the
more explicit a firm is about remedies available for failure to adhere to
stated privacy policies, the more reasonable is an inference that firms have
contracted with users about personal data practices. As one legal commentator
has observed:
As between the Web site and the user, a privacy policy bears all the earmarks
of a contract, but perhaps one enforceable only at the option of the user.
It is no stretch to regard the policy as an offer to treat information in
specified ways, inviting the user's acceptance by using the site or submitting
the information. [*1163] The Web site's promise is sufficient
consideration to support a contractual obligation, as is the user's use of
the site and submission of personal data. n194
The modified trade secrecy licensing default rule approach discussed above
might supply some terms for such contracts.
The evolution of a licensing approach to personal data protection may be
necessary because, unlike other fields in which self-regulation has been
accepted, n195 there is no Internet e-commerce industry organization
to serve as the overseer of self-regulatory practices to ensure that members
of the organization are abiding by self-regulatory norms. Private sector
firms are likely to prefer a licensing approach to having the government
establish a new privacy bureaucracy. The more enlightened among private sector
firms are coming to realize that fuller adherence to privacy principles will
promote consumer trust which will, in turn, promote commerce. But providing
consumer protection through implied or explicit licenses may ensure that
self-regulation will work.
2. Promulgation of Uniform Computer Information Transactions Act.
A recent development that might have implications for the licensing of personal
data is the promulgation of a model law, once known as Article 2B of the
Uniform Commercial Code and now known as the Uniform Computer Information
Transactions Act (UCITA). n196 The paradigmatic transaction of the
Information Age is, in its view, that of licensing. n197 In July 1999,
the [*1164] National Conference of Commissioners of Uniform State
Laws (NCCUSL) approved this model law for submission to state legislatures,
n198 and it is already being considered for enactment by some states.
n199 For a variety of reasons, this model law has been highly controversial.
n200 UCITA could pave the way for a licensing regime for protecting personal
information. n201
In considering the possible implications of UCITA for personal data protection,
it is appropriate to begin with the recognition that the personal data gathered
in cyberspace falls within UCITA's rather open-ended definition of "computer
information." n202 Interactive communications between an individual
and a commercial Web site, moreover, would seem to constitute a "transaction."
n203 Since the paradigmatic transaction of UCITA is a license, transactions
between an individual and a commercial Web site may be among the transactions
which UCITA could govern. n204
[*1165] For a license in computer information to be enforceable under
UCITA, a prospective licensee of personal data (in this case, the Web site
owner) must manifest assent, through conduct or otherwise, to the terms of
the license after an opportunity to review the terms and conditions of the
license. n205 A potential problem with using UCITA to protect personal
information in cyberspace is that individuals today do not generally articulate
terms and conditions to which the site must agree before the individuals
will supply the site with personal data; nor do they present such a license
to site owners for their review before using the site. n206 Site owners
could conclude from the absence of proffered terms that whatever information
individuals might provide to the site, wittingly or unwittingly, is being
provided without license restrictions. n207
However, it may be possible to establish restrictive licensing terms for
personal data by looking to the prospective licensee's privacy policy as
a statement of that party's willingness to restrict its uses of personal
data. After all, UCITA does not require restrictive license terms to be set
by the licensor; all it requires is a manifestation of assent to restrictive
terms. If users assent to the licensee's privacy policy restrictions by supplying
information to the site or using it otherwise in accordance with the site's
terms, a license agreement subject to these restrictions might be formed.
n208 This license might then be supplemented with the modified trade secrecy
licensing default rules proposed above to which site owners and the individuals
would agree unless expressly agreed otherwise. n209
[*1166] Future developments may also aid in the development of restrictive
personal data licenses for cyberspace transactions. Consumer protection organizations
could, for example, draft standard form restrictive licensing agreements
for individuals to use to protect their privacy interests when dealing with
Web sites. n210 Given the current technical infrastructure of the web,
individual users may not be in a position to present their standard form
contracts to the site owner in a meaningful way. However, the technical infrastructure
of the web may in time allow automated negotiations of privacy licenses that
will restrict uses that can be made of personal data (a matter to be considered
in the next subsection). n211
While much more could be said about the pros and cons of utilizing UCITA
for personal data protection, there is some reason to question whether UCITA
will be useful in achieving information privacy goals. UCITA was, after all,
drafted with very different kinds of licensing transactions in mind. From
the outset, the core subject matter of the UCITA/Article 2B project has been
computer programs. n212 Some years ago, the subject matter of this
model law was expanded to cover virtually all transactions in information.
n213 After several major information industries objected to this scope for
the law, in large part because the assumptions and default rules of UCITA/Article
2B did not match well with the licensing practices of those industries,
n214 the drafters eventually contracted the scope of the model law to computer
information. n215 Even with this contracted scope, major information
[*1167] industries continue to oppose UCITA in part because of the
"software-centric" nature of its rules. n216 If these industries are
correct in thinking that UCITA is not suitable for the licensing of such
computer information products as computer-processable motion pictures or
newspapers, it seems unlikely that UCITA would be suitable for protecting
personal data. After all, the commercial goals of the motion picture and
news industries would seem to be much closer to those of the software industry
than to the licensing of personal data. In view of this, it may be na<um
i>ve to think UCITA would provide a workable framework for achieving information
privacy goals.
Still, some believe that UCITA provides a licensing regime capable of providing
individuals with somewhat greater protection in transactions involving their
personal data than they might otherwise have. n217 To counteract concerns
about potential disparities in bargaining power of commercial Web site owners
and individuals about personal data matters, it might be worth considering
an adaptation of proposals made by Reichman and Franklin for public-interest
unconscionability default rules to achieve a better balance in non-negotiated
UCITA transactions. n218 Although Reichman and Franklin may have had
other public interests in mind, the concept of public interest unconscionability
default rules for licensing of personal data may provide a way to achieve
information privacy goals.
3. Privacy-enhancing technologies.
A number of privacy-enhancing technologies (PETs) have been developed in
recent years which are capable of masking personal identity in cyberspace
in order to achieve information privacy goals. n219 There is substantial
appeal in the idea of a technological solution to a problem that technology
itself seems to have created, in part because such technologies [*1168]
are self-enforcing and appear to reduce the need for regulatory interventions.
n220
One commentator has differentiated among four types of PETs: (1) subject-oriented
PETs (those aiming to limit the ability of others to discern the identity
of a particular person, for example, an anonymizing browser); (2) object-oriented
PETs (those aiming to protect identity through the use of a particular technology,
for example, anonymous e-cash); (3) transaction-oriented PETs (those aiming
to protect transactional data, for example, automated systems for destroying
transactional data); and (4) system-oriented PETs (those aiming to create
"zones of interaction where the identity of the subjects is ... hidden, where
the objects bear no traces of those handling them, and where no record of
the interaction is created or maintained," n221 for example, anonymous
remailer systems). n222
A fifth category of PETs capable of being programmed to interact with Web
sites about the privacy preferences of individuals potentially interested
in visiting the sites might be added as well. One well-publicized example
is the Platform for Privacy Preferences (P3P) effort underway at the World
Wide Web Consortium. n223 Some expect electronic agents to be programmed
to negotiate privacy and other user-preferred terms of contracts in cyberspace.
n224
If P3P's designers achieve the project's objectives, P3P would enable individuals
to program their browsers to identify classes of information that they are
willing and unwilling to disclose (for example, yes to zip code, but no to
street address) to Web site owners. n225 Individuals would then not
have [*1169] to haggle over terms and conditions with every site
they visit. Instead, their browsers could be set to avoid sites that do not
comport with the individuals' privacy preferences. n226 The prospect
of having fewer people visiting one's site if one's privacy policy does not
comport with common user preferences may create significant commercial pressure
for firms to offer more consumer-friendly privacy policies.
As promising as P3P and other PETs technologies may be, n227 it is
fair to say that they have yet to prove their worth in achieving information
privacy goals except in limited circumstances. n228 Other presenters
at this symposium are better able than I to assess the likelihood that such
technologies will provide greater privacy protection over time. n229
However, it is unlikely that technology alone can solve the problem.
As Professor Burkert has observed, the "main task ... [of] social scientists,
lawyers, regulators, and privacy practitioners ... [is] to accept the challenge
of information and communication technology as a challenge for social innovation."
n230 Information privacy is a social goal, not a technological one. To achieve
information privacy goals will require social innovations, including the
formation of new norms and perhaps new legal rules to establish boundary
lines between acceptable and unacceptable uses of personal data. It may be
easier for information technologists to embody such norms and legal rules
in code after society has configured what those rules should be, and they
will surely have greater incentives to do so if the law requires it.
n231
[*1170]
Conclusion
Europeans have realized that it is not just an information infrastructure
we are in the process of constructing, but an information society.
n232 They have identified information privacy as a fundamental value that
should be a keystone of the architecture for achieving an information society
in which people will want to live. n233 In addition, they have demonstrated
that political will can be found to utilize the law to ward off Scott McNealy's
vision for the information society ("you've got zero privacy now. Get over
it" n234). In these insights may lie some useful lessons for Americans
who also value information privacy. n235
Myriad reasons explain why the U.S. response to the challenges of information
privacy for an information society has been so much slower, more erratic,
and less comprehensive than in the E.U. n236 Among them are certainly
considerable differences in the regulatory cultures of the U.S. and the E.U.,
as well as dissimilar attitudes toward the private sector and toward technology.
n237 However, a serious impediment to a comprehensive approach in the U.S.
is the lack of clarity in this country about the nature of the interest that
individuals have in information about themselves: Is it a commodity interest,
a consumer protection interest, a personal dignity interest, a civil
[*1171] right interest, all of the above, or no interest at all?
n238 One of the strengths of the EU Directive is that the regulatory regime
it embodies is consistent with its underlying conception of information privacy
as a fundamental human right. Without a coherent conception about the nature
of a person's interest in personal data, it is difficult to design a legal
regime to protect this interest appropriately.
One of the virtues of the property rights approach to protecting personal
data discussed in Part I is that it would seem to solve the nature-of-the-interest
problem which, in turn, should simplify the task of constructing a legal
regime to protect the interest. However, as Part I has shown, a serious mismatch
exists between the traditional rationale for granting property protection
to an information resource and the rationale for granting individuals property
rights in personal data. n239 Also mismatched are traditional policies
of property law favoring free alienability and information privacy policy
preferences for restrictions on alienation. n240 If the goals and mechanisms
of property law are misaligned with information privacy policy objectives,
protecting privacy as intellectual property simply may not work.
Even though a one-dimensional conception of a person's interest in her information
makes crafting a legal regime easier, in truth, individuals may not have
just one interest in personal information, but many interests. Sometimes
a person's interest in personal data is a civil liberties interest (for example,
not being forced to disclose whether I am a member of the NAACP), n241
and sometimes it is not (for example, Amazon.com sending me an email to let
me know that an author whose books I have bought before has just released
a new novel). Sometimes it is a commodity interest (for example, I can get
a discount if I disclose my zip code) and sometimes it is not (for example,
I do not want software on my hard drive to scan what other software I have
installed there and report on this to its home base). Sometimes it is a dignity
interest (for example, whether I sweat profusely) and sometimes it is not
(for example, whether my eyes are blue).
The task of devising a workable legal framework for regulating private sector
uses of personal data is obviously more difficult if one takes a multi-dimensional
perspective on the nature of a person's interest in personal data. Yet it
is an advance to recognize that a person has more than one kind of interest
[*1172] in personal information. It is also an advance to realize that
the propriety of collecting or processing personal data depends in part on
context. n242 For my doctor to send information about my medical condition
to an insurance company so that it will cover the costs of treatment is appropriate,
but for the doctor to give the same information to a prospective employer
is inappropriate. It further advances understanding to realize that a major
factor in a contextual analysis about uses of personal information is whether
the person whose data is being collected or processed knows or has reason
to know that the data are being collected and what uses will be made of them.
n243 In addition, it may be important to realize that our concept of information
privacy, and in particular, our understanding of what is appropriate and
inappropriate to do with personal information, is evolving over time.
n244
One of the virtues of a contractual approach to protecting information privacy
is that it can accommodate the multiple interests people have in personal
information, the contextual nature of determinations about the appropriateness
of collection or use of personal data, the significance of consent as a factor
in determining appropriate uses, and the evolutionary nature of social understanding
about information privacy. It is a flexible, adaptable, market-oriented way
to allow individuals to control uses of personal data. Oddly enough, it may
more easily be achieved in cyberspace than in the physical world because
a Web site's privacy policy can become the basis of a contractual understanding
between the user and the Web site. n245 Although individuals and Web
site owners may sometimes reach express agreement on all relevant issues
pertaining to allowable uses of personal data, a set of default licensing
rules adapted from trade secrecy law might "fill in the gaps" of such agreement
(for example, restricting rights to sublicense the data to others if the
privacy policy is silent on this issue). Despite obvious differences between
trade secrecy and information privacy, there are some significant parallels
in the objectives of trade secret law and the information privacy law envisioned
in this article: protecting commodity and non-commodity interests of persons
in restricting others' uses of certain information; protecting information
disclosed in confidence; protecting information against the use of improper
means to obtain it; facilitating commercial transactions allowing the holder
of the interest to negotiate compensation for allowing uses of information;
enforcing agreements about nondisclosure or [*1173] limited use;
and establishing minimum standards of commercial morality that can evolve
over time.
Americans may want information privacy, but they also want a strong information
economy. They appear to be willing to balance their interests in keeping
certain information about themselves private with their interests in getting
access to customized information and services that disclosure of their personal
data may enable firms to provide. n246 If information privacy goals
can be achieved without establishing a new government bureaucracy, as a modified
licensing regime should allow, Americans objectives for an information society
may more fully be realized.
FOOTNOTES:
n1. See Privacy Working Group, Information Infrastructure Task Force, Privacy
and the National Information Infrastructure: Principles for Providing and
Using Personal Information 1-3 (1995) <http://www.iitf.nist.gov/ipc/ipc/ipc-pubs/niiprivprin<uscore>final.html>
(defining information privacy and discussing risks to information privacy
in cyberspace).
n2. See, e.g., Fred H. Cate, Privacy in the Information Age 14-15 (1997)
(documenting the ease of collecting data); Jerry Kang, Information Privacy
in Cyberspace Transactions, 50 Stan. L. Rev. 1193, 1198-99 (1998) (providing
a concrete example of data collection).
n3. See, e.g., Joel R. Reidenberg, Setting Standards for Fair Information
Practice in the U.S. Private Sector, 80 Iowa L. Rev. 497, 516-18 (1995) (discussing
uses of personal data, including profiling).
n4. See, e.g., Herbert Burkert, Privacy-Enhancing Technologies: Typology,
Critique, Vision, in Technology and Privacy: The New Landscape 125 (Philip
E. Agre & Marc Rotenberg eds., 1997) [hereinafter Technology & Privacy].
n5. See, e.g., John Hagel III & Jeffrey F. Rayport, The Coming Battle
for Customer Information, Harv. Bus. Rev. Jan.-Feb. 1997, at 53, 53 (discussing
reasons companies want to collect personal information); Rohan Samarajiva,
Interactivity As Though Privacy Mattered, in Technology & Privacy, supra
note 4, at 277-79 (arguing that mass customization of the new economy requires
more surveillance and knowledge about customers).
n6. See, e.g., National Telecomm. & Info. Admin., U.S. Dep't. of Comm.,
Privacy and the NII: Safeguarding Telecommunications-Related Personal Information
18-22 app. A (1995) <http://www.ntia.doc.gov/ntiahome/privwhitepaper.htm>
(discussing the business of marketing profiles); see also Robert Pitofsky,
Opening Remarks at Public Workshop on Online Profiling (Nov. 8, 1999) <http://www.ftc.gov/opa/1999/9911/onlinepitofsky.htm>
(discussing online profiling).
n7. See, e.g., 1 Working Group on Elec. Comm. Ann. Rep. 16-18 (1998) (discussing
the Administration's efforts to promote information privacy as part of its
electronic commerce initiative).
n8. Lawrence Lessig has emphasized that law is only one of four principal
regulators of human behavior in cyberspace; norms, the market, and technology
also have regulatory functions. See, e.g., Lawrence Lessig, Code and Other
Laws of Cyberspace 85-90 (1999). I wish in this footnote to acknowledge this
influence on my perspective on information privacy issues.
n9. See, e.g., Kang, supra note 2, at 1255-67 (explaining why the market
is unlikely to provide a solution to information privacy problems in cyberspace);
Joel R. Reidenberg, Restoring Americans' Privacy in Electronic Commerce,
14 Berkeley Tech. L.J. 771, 771 (1999) (arguing that self-regulation has
been a failure).
n10. Peter P. Swire & Robert E. Litan, None of Your Business: World Data
Flows, Electronic Commerce, and the European Privacy Directive 8 (1998).
n11. See infra note 35 for sources.
n12. See, e.g., Steven A. Bibas, A Contractual Approach to Data Privacy,
17 Harv. J.L. & Pub. Pol'y 591, 592 (1994) (claiming that a contractual
solution most effectively protects privacy rights); Craig Martin, Mailing
Lists, Mailboxes, and the Invasion of Privacy: Finding a Contractual Solution
to a Transnational Problem, 35 Hous. L. Rev. 801, 850 (1998) (proposing an
expansion of existing legislation coupled with industry contracting); Scott
Shorr, Personal Information Contracts: How to Protect Privacy Without Violating
the First Amendment, 80 Cornell L. Rev. 1756, 1759 (1995) (suggesting a property
and contract law solution to protect privacy from credit bureau investigations).
But see, e.g., Reidenberg, supra note 3, at 546-47 (discussing limits of
contractual approaches to data protection).
n13. Swire & Litan, supra note 10, at 7.
n14. See Council Directive 95/46/EC, art. 1, 1995 O.J. (L 281) (protecting
individuals with regard to the processing of personal data and on the free
movement of such data) [hereinafter EU Directive]. See, e.g., Swire &
Litan, supra note 10, at 22-49 for a discussion of the main features of the
EU Directive. Although these authors agree with the EU Directive's underlying
premise about the need for greater protection for personal data, they are
among the Directive's strongest critics. See, e.g., id. at 7-21 (explaining
why the EU Directive is unworkable and overbroad).
n15. See, e.g., Simon G. Davies, Re-engineering the Right to Privacy: How
Privacy Has Been Transformed from a Right to a Commodity, in TECHNOLOGY &
PRIVACY, supra note 5, at 143-45 (noting a change in society's approach from
privacy protection to data protection); Reidenberg, supra note 3, at 497-98
(arguing that a citizen's right to participate in government depends "on
the ability to control the disclosure of personal information); Paul M. Schwartz,
Privacy and Democracy in Cyberspace, 52 Vand. L. Rev. 1609, 1611 (1999) (claiming
that the absence of privacy norms threatens democracy); see also Julie E.
Cohen, A Right To Read Anonymously: A Closer Look at Copyright Management
in Cyberspace, 28 Conn. L. Rev. 981, 982-83 (1996) (arguing that digital
copyright management technologies violate First Amendment rights protecting
speech and freedom of thought). The EU Directive is based on a conception
of personal data protection as a fundamental civil liberty interest. See
EU Directive, supra note 14, art. 1.1.
n16. See, e.g., Joseph D. Lasica, Your Past is Your Future, Wash. Post, Oct.
11, 1998, at C1; William Safire, Nosy Parker Lives, N.Y. Times, Sept. 23,
1999, at A29.
n17. See, e.g., Kang, supra note 2, at 1212-17.
n18. See notes 238-244 infra and accompanying text for further discussion
of the relationship between how one characterizes the nature of a person's
interest in personal data and the design of a regulatory regime to protect
this data. The underlying thesis of this article is that an individual has
one type of interest in her personal information: It is therefore both unnecessary
and counterproductive to choose between, e.g., the market-based and civil
liberty-based visions of privacy. This paper seeks to outline a workable
regime which would accommodate the full panoply of underlying interests.
n19. See Unif. Computer Info. Transactions Act 207 (1999) <http://www.law.upenn.
edu:80/library/ulc/ucita/cita10st.htm> [hereinafter UCITA]. See note 196
infra and accompanying text for a discussion of the implications of this
law.
n20. See, e.g., Frederick Schauer, Internet Privacy and the Public-Private
Distinction, 38 Jurimetrics J. 555, 560-61 (1998) (criticizing of this perspective).
n21. See, e.g., Lessig, supra note 8, at 142-63.
n22. See Kenneth C. Laudon, Markets and Privacy, Comm. ACM Sept. 1996, at
92 ("Why not let individuals own the information about themselves and decide
how the information is used?"); see also Catherine M. Valerio Barrad, Genetic
Information and Property Theory, 87 Nw. U. L. Rev. 1037, 1062-63 (1993) (discussing
natural rights theory for recognizing property protection in genetic information)
n23. See, e.g., Samuel D. Warren & Louis D. Brandeis, The Right to Privacy,
4 Harv. L. Rev. 193, 198-99 (1890)
n24. See, e.g., United States v. Miller, 425 U.S. 435, 441-45 (1976) (considering
arguments about the privacy expectations of individuals as to bank records).
n25. See, e.g., Right to Financial Privacy Act, 12 U.S.C. 3410 (1994); Horne
v. Patton, 287 So. 2d 824, 829-30 (1973) (holding that doctor's disclosure
of medical information to prospective employer was wrongful). But see, e.g.,
Paul M. Schwartz, Privacy and the Economics of Personal Health Care Information,
76 Tex. L. Rev. 1, 3 (1997) (indicating that little legal protection is available
for medical information).
n26. See, e.g., Fair Credit Reporting Act, 15 U.S.C. 1681 (1994). See generally
Paul M. Schwartz & Joel R. Reidenberg, Data Privacy Law: A Study of United
States Data Protection (1996) (providing an overview of state and federal
information privacy laws).
n27. See, e.g., Schwartz & reidenberg, supra note 26, ch. 5 (explaining
rationales of certain U.S. information privacy laws).
n28. See, e.g., Feist Publications, Inc. v. Rural Tel. Serv. Co., 499 U.S.
340, 399 (1991) (holding that copyright law does not confer exclusive rights
to information in order to achieve constitutional purpose of promoting knowledge).
Information can, however, sometimes be protected against unfair competition,
including breaches of confidential relationships. See International News
Serv. v. Associated Press, 248 U.S. 215, 235-36 (1918) (holding that news
may be protected against unfair competition); see also Yochai Benkler, Constitutional
Bounds of Database Protection: The Role of Judicial Review in the Creation
and Definition of Private Rights in Information, 15 Berkeley Tech. L.J. (forthcoming
2000) (arguing that the Intellectual Property Clause, the First Amendment
Speech Clause, and the Commerce Clause all constrain the power of Congress
to grant private rights in information); Yochai Benkler, Free as the Air
to Common Use: First Amendment Constraints On Enclosure of the Public Domain,
74 N.Y.U. L. Rev. 354, 354 (1999) (demonstrating that laws born of the conception
of information as an owned commodity are removing uses of information from
the public forum); Jessica Litman, The Public Domain, 39 Emory L.J. 965,
968 (1990) (examining the "gulf between what authors really do and the way
the law perceives them."); L. Ray Patterson & Judge Stanley F. Birch,
Jr., Copyright and Free Speech Rights, 4 J. Intell. Prop. L. 1, 4 (1996)
(discussing the need to balance the public's right of access to knowledge
and the entrepreneur's right to profit); Malla Pollack, The Right To Know?:
Delimiting Database Protection at the Juncture of the Commerce Clause, the
Intellectual Property Clause, and the First Amendment, 17 Cardozo Arts &
Ent. L.J. 47, 49-50 (1999) (criticizing the Collections of Information Antipiracy
Act); Pamela Samuelson, Information As Property: Do Ruckelshaus and Carpenter
Signal a Changing Direction in Intellectual Property Law?, 38 Cath. U. L.
Rev. 365, 366 (1989) (discussing two cases' treatment of information as private
property). See generally Diane Leenheer Zimmerman, Information as Speech,
Information as Goods: Some Thoughts On Marketplaces and the Bill of Rights,
33 Wm. & Mary L. Rev. 665 (1992) (examining the history of the friction
between information as a common resource and as a privately controlled good).
n29. See, e.g., Lessig, supra note 8, at 111-21 (explaining Bill of Rights
as a check on government power).
n30. See, e.g., Horne v. Patton, 287 So. 2d 824, 829-30 (1973).
n31. See, e.g., Polin v. Dun & Bradstreet, Inc., 768 F.2d 1204, 1207
(10th Cir. 1985) (rejecting privacy claim based on unauthorized release of
credit report information); Moore v. Regents of the Univ. of Cal., 793 P.2d
479, 487 (Cal. 1990), cert. denied, 499 U.S. 936 (1991) (rejecting individual's
claim of property right in his genetic information).
n32. But cf. EU Directive, supra note 14, art. 6.1 (maintaining a general
prohibition on the use of personal data collected to enable the customer
to qualify for warranty protection for marketing purposes).
n33. But cf. Reno v. Condon, 120 S. Ct. 666, 671 (2000) (holding that Congress
can, however, restrict the ability of states to release the personal information
of drivers without their consent).
n34. See, e.g., Randolph S. Sergent, A Fourth Amendment Model for Data Networks
and Computer Privacy, 81 Va. L. Rev. 1181, 1200 (1995) ("One might conclude
that an individual has no expectation of privacy in information kept by a
third party.").
n35. See, e.g., Developments in the Law - The Law of Cyberspace, 112 Harv.
L. Rev. 1574, 1634-49 (1999) [hereinafter Harvard Developments]; Laudon,
supra note 22, at 92; Lawrence Lessig, The Architecture of Privacy, 1 Vand.
J. Ent. L. & Prac. 56, 63-65 (1999); Lessig, supra note 8, at 122-35;
Patricia Mell, Seeking Shade in a Land of Perpetual Sunlight: Privacy as
Property in the Electronic Wilderness, 11 Berkeley Tech. L.J. 1, 26-41 (1996);
Richard S. Murphy, Property Rights in Personal Information: An Economic Defense
of Privacy, 84 Geo. L.J. 2381, 2383 (1996); Carl Shapiro & Hal R. Varian,
U.S. Government Information Policy 45 (July 30, 1997) <http:// www.sims.berkeley.edu/<diff>hal/Papers/policy.pdf>.
n36. See, e.g., Laudon, supra note 22, at 92-100; Shapiro & Varian, supra
note 35, at 29-30. Shapiro and Varian express concern that privacy protection
legislation may not promote consumer welfare because it will be too strong
and inflexible. Id. at 29.
n37. Email communication from Marc Rotenberg to Pamela Samuelson (Oct. 31,
1999) (on file with author) [hereinafter Rotenberg email].
n38. See Laudon, supra note 22, at 96.
n39. See, e.g., Kang, supra note 2, at 1199, 1220-30 (discussing the technical
infrastructure of cyberspace and how it enables collection of personal data).
n40. See, e.g., Laudon, supra note 22, at 98.
n41. See, e.g., Cate, supra note 2, at 74. The belief of data compilers in
their ownership rights in personal data compilations will be strengthened
if the U.S. Congress passes legislation to protect collections of data from
"piracy," as has been proposed numerous times in recent years. See Electronic
Commerce: The Current Status of Privacy Protection for Online Consumers,
106th Cong. 39 (1999). Notwithstanding their investment-based claim of rights
in data compilations, personal data compilers almost certainly recognize
significant limitations on their ability to use these data. A firm claiming
to "own" a list of ten thousand impotent men would surely recognize that
publication of the names of those men in a widely circulated newsletter would
be an invasion of privacy rights that these men would have in respect to
this information. A firm possessing such a list may feel justified in licensing
this information to the manufacturer of Viagra based on its belief that many
men suffering from this condition would be interested in or might otherwise
benefit from receiving information on this drug. Societal norms, then, already
limit to some degree what firms do with personal data. See, e.g., Robert
C. Post, The Social Foundations of Privacy: Community and Self in the Common
Law Tort, 77 Cal. L. Rev. 957, 958-59 (1989) (discussing normative concept
of privacy). When norms alone do not suffice, the law sometimes imposes community
norms on firms possessing personal data. Without resolving the question of
whether traffickers in personal data have "property rights" in these data,
it is easy to demonstrate that their rights (if any) in collections of personal
data do not extend as far as the "property rights" label might suggest. This
example illustrates that individuals have some residual legally protectable
interests in personal data in the hands of data compilers.
n42. Rotenberg email, supra note 37.
n43. See Laudon, supra note 22, at 104 (explaining that "the current situation
costs corporations billions of dollars in waste as they pour money into privacy-invading
marketing and authorization techniques." ); see also Scott Killingsworth,
Minding Your Own Business: Privacy Policies in Principle and in Practice,
J. Internet L., Oct. 1999, at 3-4 (discussing data quality issues); Mell,
supra note 35, at 78-81 (suggesting the creation, by statute, of an agency
relationship between an individual and an information holder, such that any
subsequent use or disclosure of the information becomes subject to a warranty
of authority to disclose and a warranty of accuracy). One consequence of
the property rights regime which most commentators have not explored is the
likelihood that individuals supplying false personal information under the
property model might themselves be subject to liability for the inaccuracy
of their information. See, e.g., id., at 80. While such liability would help
ensure the accuracy of the information that individuals provide about themselves,
it seems fair to say that most Internet users do not currently contemplate
personal liability when they provide information online.
n44. See, e.g., Margaret Jane Radin, Property Evolving in Cyberspace, 15
J.L. & Com. 509, 514-18 (1996) (discussing utilitarian criteria for creation
of property rights).
n45. See, e.g., Comm. on Physicial Sciences, Mathematice, and Applications,
The Digital Dilemma: Intellectual Property in the Information Age ES-1 (2000)
<http:// www.nap.edu/books/0309064996/html/>.
n46. See, e.g., Rochelle Cooper Dreyfuss, Information Products: A Challenge
to Intellectual Property Theory, 20 N.Y.U. J. Int'l L. & Pol. 897, 925-27
(1988); J.H. Reichman, Legal Hybrids Between the Patent and Copyright Paradigms,
94 Colum. L. Rev. 2432, 2434-42 (1994); Samuelson, supra note 28, at 366-68.
n47. See, e.g., Michael H. Goldhaber, Attention Shoppers!, Wired, Dec. 1997,
at 182-90; Radin, supra note 44, at 517 (commenting on the possible commodification
of attention).
n48. See, e.g., Web Privacy? Let's Make a Deal, Palm Beach Post, Aug. 26,
1999, at 4E (detailing a survey by Privacy and American Business and Opinion
Research Corporation found that 86% of web users would release personal information
as long as they received a direct benefit in return, such as money or free
products or services). But see Graphics, Visualization, & Usability (GVU)
Center, GVU's WWW User Survey (1998) <http://www.gvu.gatech.edu/ user<uscore>surveys/survey-1998-10/preview/privacy/q39.htm>
(reporting that between one-quarter and one-third of surveyed users would
be willing to reveal demographic data to get some benefit).
n49. See, e.g., Shapiro & Varian, supra note 35, at 30-31.
n50. See, e.g., id.
n51. See, e.g., id.; Laudon, supra note 22, at 102.
n52. See, e.g., Laudon, supra note 22, at 101.
n53. See, e.g., Hagel & Rayport, supra note 5, at 54.
n54. See, e.g., Lorrie Cranor, Internet Privacy, Comm. ACM, Feb. 1999, at
30.
n55. See, e.g., Harvard Developments, supra note 35, at 1646; Joseph Reagle
& Lorrie Faith Cranor, The Platform for Privacy Preferences (Nov. 6,
1998) <http://www.w3c.org/TR/1998/ NOTE-P3P-CACM>.
n56. See EU Directive, Supra note 14, art. 28 (requiring all member states
to establish "supervisory authorities" to ensure that the data protection
regulations are enforced). Many European countries already had established
data protection authorities. See generally, Colin J. Bennett, Regulating
Privacy: Data Protection and Public Policy in Europe and the United States
(1992). Americans tend to have reservations about the establishment of a
privacy bureaucracy as such, although privacy policy coordination can be
placed elsewhere in the government (e.g., in the electronic commerce group
at the Commerce Department). See, e.g., Swire & Litan, supra note 10,
at 17-18; see also Info. Pol'y Comm., Nat'l. Info. Infrastructure Task Force,
Options for Promoting Privacy on the Information Superhighway 24-28 (April
1997) <http://www.iitf.nist.gov/ipc/privacy.htm> (discussing possible
ways for the U.S. government to respond to challenges of information privacy,
but expressing reservations on the establishment of bureaucracy); Kang, supra
note 2, at 1285 (indicating that privacy bureaucracy is unlikely in the United
States). There is clearly a need for the U.S. to have expert negotiators
to participate in international discussions on information privacy issues.
See Swire & Litan, supra note 10, at 17-18. For a discussion of institutional
infrastructure that might be required for a property rights regime, see notes
59-68 infra.
n57. EU Directive, supra note 14, arts. 17-18.
n58. See, e.g., Harvard Developments, supra note 35, at 1644-49; Lessig,
supra note 35, at 62-65; Murphy, supra note 35, at 2402. Other commentators
expect a "long and drawn out period of confusion" before this market becomes
stable, but expect standard contracts to solve the problem. See, e.g., Shapiro
& Varian, supra note 35, at 31.
n59. See Laudon, supra note 22, at 92.
n60. Laudon, supra note 22, at 92.
n61. Id. at 100.
n62. Id.
n63. See id. Private placements of personal data might also occur through
a National Information Accounts Clearinghouse which would be established
by Congress to permit individuals to collect fees for uses of their information.
See id.
n64. Id.
n65. See id. at 101.
n66. See id. at 103. Laudon recommends establishing a Federal Information
Commission to oversee the NIM and related activities. See id.
n67. Id. at 98.
n68. See id.
n69. See, e.g., Radin, supra note 44, at 516-17 (indicating that the costs
of enforcement must be included in the calculus of the costs and benefits
of establishing property rights).
n70. Some commentators have recognized the need for limitations on resale
rights. See, e.g., Hal R. Varian, Economic Aspects of Personal Privacy, in
U.S. Dept. of Com., Privacy and Self-Regulation in the Information Age 35-37
(1997) ("Information about an individual could not be resold, or provided
to third parties, without that individual's explicit agreement."); see also
Jessica Litman, Information Privacy/Information Property, 52 Stan. L. Rev.
1283 (2000) (raising concerns about alienation in connection with information
privacy protection).
n71. See, e.g., John P. Dwyer & Peter S. Menell, Property Law and Policy:
A Comparative Institutional Perspective 184-85 (1998) (discussing general
hostility to restraints on alienation in property law); Margaret Jane Radin,
Contested Commodities 18 (1996) (arguing that concept of inalienability "negates
a central element of traditional property rights."); see also Restatement
of Property 406 cmt. a (1944) (referencing rationale for disfavoring restraints
on alienation); Roger A. Cunningham, William B. Stoebuck & Dale A. Whitman,
The Law of Property 2.2, at 33-39 (1984) (tracing the public policy in favor
of free alienability back to Quia Emptores in 1290).
n72. See, e.g., Radin, supra note 44, at 514-16 (discussing scarcity rationale
for establishing property rights).
n73. See, e.g., Clean Air Act Amendment of 1990, 42 U.S.C. 7401, 7651-7651n
(1994); Carol M. Rose, The Several Futures of Property: Of Cyberspace and
Folk Tales, Emission Trades and Ecosystems, 83 Minn. L. Rev. 129, 164-80
(1998) (discussing regulatory property rights regimes).
n74. See, e.g., Rose, supra note 73, at 164-65. A further goal of this sort
of property rights regime is to ensure that firms will have incentives to
redirect investments toward nonpolluting or pollution-reducing equipment,
or otherwise to reduce production of the undesired substance. See id. at
166. Rose also emphasizes the critical importance of having the technological
means to set, monitor, and enforce emissions rights regimes. See id. at 167.
Of course, there are other differences between the right to pollute and the
information privacy rights contemplated here. Chief among them is that one
is a supplier's right and the other a buyer's right. In the environmental
context, the purpose of the property right is to limit the amount of pollution
any one supplier can distribute. In the personal data market, however, it
appears that we are not concerned with capping what suppliers want to do
with their information or with creating a property right to inhibit such
supplying. Instead, we want to cap what buyers do with the information they
purchase. By giving a property right to the suppliers, we make it harder
for the buyers to gather all the information they want. The Clean Air Act,
on the other hand, creates a market among producers of pollution to trade
among themselves, not a market between producers of pollution and buyers
of pollution. To achieve the goals of information privacy using a Clear Air
Act system for buyers, we would have to put a cap on the amount of information
any one company could own and then give companies limited rights in the ability
to own information, allowing them to trade those rights with other information
collectors in order to create a market in information collection that reflected
the value of amassing information.
n75. See, e.g., Rochelle Cooper Dreyfuss, Warren & Brandeis Redux: Finding
(More) Privacy Protection in Intellectual Property Lore, 1999 Stan. Tech.
L. Rev. VS.8, PP 5, 8, 32 <http://stlr.stanford.edu/STLR/Symposia/Privacy/99<uscore>VS<uscore>8/>
(contrasting incentives rationale for intellectual property protection with
rationale for privacy protection; Dreyfuss concludes that "the fit between
what intellectual property provides and what privacy advocates want is imperfect,
more apparent than real, and possibly evanescent").
n76. See, e.g., Robert P. Merges, Peter S. Menell, Mark A. Lemley & Thomas
M. Jorde, Intellectual Property in the New Technological Age 12-18 (1997).
n77. See id.
n78. See id.
n79. See id. The ability to recoup research and development expenses has
become increasingly difficult because so many of today's most commercially
valuable information products bear the know-how required to make them on
or near the surface of the product. See generally J.H. Reichman, Computer
Programs As Applied Scientific Know-How: Implications of Copyright Protection
for Commercialized University Research, 42 Vand. L. Rev. 639 (1989) (surveying
problems that university administrators face when seeking to exploit property
rights in new technologies).
n80. As with the other property rights considered thus far, alienability
of rights is a common feature of intellectual property rights systems. See,
e.g., 17 U.S.C. 201(d) (1994) (transfer of Copyright ownership rules).
n81. See notes 5-10 supra and accompanying text.
n82. See, e.g., Dreyfuss, supra note 75, at P 1 (bemoaning the willingness
of people to give away information about themselves).
n83. U.S. Const. art. I, 8, cl. 8. Some, of course, have considered alternative
rationales for grants of intellectual property rights. See, e.g., Wendy J.
Gordon, On Owning Information: Intellectual Property and the Restitutionary
Impulse, 78 Va. L. Rev. 149, 196-221 (1992) (discussing restitution-based
rationales for intellectual property law). This, of course, is closer to
the mark for information privacy concerns.
n84. See, e.g., Peter A. Jaszi, Goodbye To All That--A Reluctant (and Perhaps
Premature) Adieu To A Constitutionally-Grounded Discourse of Public Interest
in Copyright Law, 29 Vand. J. Transnat'l L. 595, 596 (1996) (explaining pressures
emanating from major copyright industry organizations on Congress to deviate
from constitutional and utilitarian purposes).
n85. See, e.g., Register of Copyrights, 87th Cong. Report on Copyright Law
Revision 5-6 (1961) (discussing constitutional purposes of copyright law);
see also L. Ray Patterson, Free Speech, Copyright, and Fair Use, 40 Vand.
L. Rev. 1, 6 (1987) (arguing that copyright functions should be recognized
as a regulatory concept).
n86. The U.S. Constitution speaks of promoting "Science" and the "useful
Arts" as the purposes for which Congress is empowered to enact intellectual
property legislation. U.S. Const. art. I, 8, cl. 8; see also Merges et al.,
supra note 76, at 12-15 (discussing constitutional purposes).
n87. See, e.g., Rosemont Enters., Inc. v. Random House, Inc., 366 F.2d 303,
311 (2d Cir. 1966) (vacating a preliminary injunction and holding that a
copyright owner could not suppress the biography of Howard Hughes). Of course,
if an author has chosen not to publish her work (or not to publish it yet),
copyright law will protect the work from unauthorized publication. See, e.g.,
Harper & Row, Pubs., Inc. v. Nation Enters., 471 U.S. 539, 561 (1985)
(holding that preemptive publication of excerpts from an unpublished book
was not fair use).
n88. See U.S. Const. art. I, 8, cl. 8. The Supreme Court repeatedly emphasized
constitutional limitations on the power of Congress to enact legislation
in explaining why copyright protection could not be extended to unoriginal
data compilations in Feist Publications, Inc. v. Rural Tel. Serv. Co., Inc.,
499 U.S. 340 (1991).
n89. See, e.g., Kang, supra note 2, at 1267 n.305.
n90. Of course, it might be possible to assert that Congress has constitutional
power to enact such legislation under Section 5 of the Fourteenth Amendment.
The Solicitor General relied on this constitutional provision in arguing
that Congress had power to enact the Drivers Privacy Protection Act, 18 U.S.C.
2721 (1994). The Fourth Circuit Court of Appeals rejected this argument in
Condon v. Reno, 155 F.3d 453 (4th Cir. 1998), on the ground that "neither
the Supreme Court nor this Court has ever found a constitutional right to
privacy with respect to the type of information .... to which individuals
do not have a reasonable expectation of privacy." Id. at 464. The Supreme
Court recently reversed this decision on other grounds. See Reno v. Condon,
120 S. Ct. 666, 669 (2000). Although it would have been clarifying for the
Supreme Court to decide whether an individual has a constitutional right
to privacy in her personal information, the Solicitor General expressly abandoned
the Fourteenth Amendment argument in the appeal to the Supreme Court. See
id. at 671 n.2. Legislation to create property rights in personal data might
also, unless narrowly drafted, run afoul of the First Amendment. See, e.g.,
Hicks v. Casablanca Records, 464 F. Supp. 426, 433 (S.D.N.Y. 1978) (rejecting
a right of publicity claim for commercial use of information about Agatha
Christie in a motion picture in part because of First Amendment considerations);
see also Eugene Volokh, Freedom of Speech and Information Privacy: The Troubling
Implications of a Right to Stop People From Speaking About You, 52 Stan.
L. Rev. 1049 (2000).
n91. See, e.g., Bd. of Regents of State Colleges v. Roth, 408 U.S. 564, 577
(1972) (holding that property interests are not created by U.S. Constitution,
but by state law); Pruneyard Shopping Ctr. v. Robins, 447 U.S. 74, 84 (1980)
(questioning the residual authority of the federal government to create property
rights).
n92. A number of states have enacted right of publicity statutes to protect
these interests. See, e.g., Cal. Civ. Code 3344 (West 1999). Other states
have recognized publicity rights through common law process. See, e.g., Zacchini
v. Scripps-Howard Broadcasting Co., 433 U.S. 562, 569 (1977) (holding that
surreptitious taping of human cannonball act at county fair violated common
law right of publicity).
n93. See, e.g., Midler v. Ford Motor Co., 849 F.2d 460, 463 (9th Cir. 1988)
(describing publicity rights as property rights).
n94. See, e.g., Sheldon W. Halpern, The Right of Publicity: Commercial Exploitation
of the Associative Value of Personality, 39 Vand. L. Rev. 1199, 1200 n.3
(1986) (characterizing publicity rights as "peculiarly celebrity based, arising
only in the case of an individual who has attained some degree of notoriety
or fame."); see also Hicks v. Casablanca Records, 464 F. Supp. 426, 429 (S.D.N.Y.
1978) (questioning whether Agatha Christie had publicity rights given the
paucity of evidence that she had made investments to promote the commercial
value of her persona as such); Pesina v. Midway Mfg. Co., 948 F. Supp. 40,
42 (N.D. Ill. 1996) (granting summary judgment to video game maker on publicity
claim by martial artist on the theory that before the video game, his name
and likeness had no commercial value). But see Dreyfuss, supra note 75, at
PP 17-18 (suggesting that there is no convincing basis for confining publicity
rights to celebrities).
n95. A person may have a civil liberty interest in voting or speaking freely
on issues of public importance in a public forum. These civil rights may
be legally enforceable, but they are not commodifiable interests akin to
property rights. If information privacy is a civil right, it may make no
more sense to propertize it than to propertize voting rights to protect the
franchise. See note 102 infra. See generally Radin, supra note 71, at 16-29
(discussing rationales for making certain rights inalienable).
n96. See, e.g., EU Directive, supra note 14, Recital 10 (referencing European
Convention for the Protection of Human Rights and Fundamental Freedoms as
well as general principles of European Community Law as recognizing data
protection as a fundamental civil liberty interest); see also id., art. 1.1
("Member States shall protect the fundamental rights and freedoms of natural
persons, and in particular their right to privacy with respect to the processing
of personal data.").
n97. See, e.g., Roe v. Wade, 410 U.S. 113, 153 (1973) (stating that, in the
Supreme Court's opinion, the right of privacy is "founded in the Fourteenth
Amendment's concept of personal liberty and restrictions on state action
....").
n98. See, e.g., Whalen v. Roe, 429 U.S. 589, 599-600 (1977) (recognizing
a constitutionally protected interest in information privacy, while upholding
a statute requiring the release of personal data in prescription drug records).
n99. See, e.g., American Fed. Of Gov't Employees, AFL-CIO v. Dep't of Hous.
and Urban Dev., 118 F.3d 786, 791 (D.C. Cir. 1997) (expressing "grave doubts
as to the existence of a constitutional right of privacy in the nondisclosure
of personal information"); U.S. West, Inc. v. Fed. Communications Com'n,
182 F.3d 1224, 1232 (10th Cir. 1999) (striking down a Federal Communications
Commission (FCC) rule aimed at protecting information privacy interests of
telephone subscribers on First Amendment grounds). The FCC, joined by amici,
has sought rehearing of this decision.
n100. See, e.g., Privacy Act of 1974, Pub. L. No. 93-579 a(4), reprinted
in The Privacy Law Sourcebook 1999 38, 38 (Marc Rotenberg ed. 1999) ("The
right to privacy is a personal and fundamental right protected by the Constitution
of the United States ...."); The Supreme Court on Privacy, N.Y. Times, Nov.
14, 1999, at 14 (endorsing the concept of information privacy as a fundamental
civil liberty interest); see also Schwartz & Reidenberg, supra note 26,
at 29-89 (discussing constitutional roots of privacy rights).
n101. See, e.g., Davies, supra note 15, at 161 ("The process of commodification
[of personal data] is inimical to privacy.").
n102. See generally Pamela S. Karlan, Not By Money But By Virtue Won? Vote
Trafficking and the Voting Rights System, 80 Va. L. Rev. 1455 (1994) (explaining
rationale for public policies against vote trafficking).
n103. See, e.g., Laura Lee Mall, The Right to Privacy in Great Britain: Will
Renewed Anti-Media Sentiment Compel Great Britain To Create a Right to Be
Let Alone?, 4 ILSA J. Int'l and Compar. L. 785, 805 (1998); Nora M. Rubin,
A Convergence of 1996 and 1997 Global Efforts To Curb Corruption and Bribery
in International Business Transactions: The Legal Implications of the OECD
Recommendations and Convention for the United States, Germany, and Switzerland,
14 Am. U. Int'l L. Rev. 257, 298 (1998) (discussing the historical context
of privacy protection in Europe).
n104. See, e.g., David H. Flaherty, Protecting Privacy in Surveillance Societies
306, 373-74 (1989):
European data protection laws include the hidden agenda of discouraging a
recurrence of the Nazi and Gestapo efforts to control the population, and
so seek to prevent the reappearance of an oppressive bureaucracy that might
use existing data for nefarious purposes. This concern is such a vital foundation
of current legislation that it is rarely expressed in formal discussions.
This helps to explain the general European preference for strict licensing
systems of data protection .... Thus European legislators have reflected
a real fear of Big Brother based on common experience of the potential destructiveness
of surveillance through record keeping. None wish to repeat the experiences
endured under the Nazis during the Second World War.
Id.; see also Peter P. Swire, Financial Privacy and the Theory of High-Tech
Government Surveillance, 77 Wash. U. L. Q. 461, 495 (1999) (weighing "the
tempting advantages of high-tech surveillance with its sobering disadvantages.").
n105. See, e.g., Pitofsky, supra note 6, at 1 (indicating that consumers
have "little, if any, knowledge" about online profiling currently being done);
Privacy Rights Clearinghouse, Second Annual Report 21 (1995), cited in Kang,
supra note 2, at 1253 n.255 (""[Our] major finding ... is that consumers
suffer from a serious lack of knowledge of privacy issues. Many consumers
are unaware of personal information collection and marketing practices. They
are misinformed about the scope of existing privacy law, and generally believe
there are far more safeguards than actually exist.'"); see also Bibas, supra
note 12, at 597-98; Harvard Developments, supra note 35, at 1644; R. Craig
Tolliver, Filling the Black Hole of Cyberspace: Legal Protections for Online
Privacy, 1 Vand. J. Ent. L. & Prac. 66, 70 (1999) (noting consumer ignorance
of private sector data collection and processing practices). Often consumers
do not know that firms are collecting data about them. See Pitofsky, supra
note 6, at 1; Tolliver, supra, at 67-68.
n106. See Children's Online Privacy Protection Act of 1998, Pub. L. No. 105-277,
112 Stat. 2681-728 (1998) (requiring parental consent before Web sites targeting
children can collect personal data from children under the age of 13). The
FTC had recommended legislation of this sort in part because "children generally
lack the developmental capacity and judgment to give meaningful consent to
the release of personal information to a third party." See Federal Trade
Commission, Privacy Online: A Report to Congress 5 (June 1998) <http://www.ftc.gov/reports/privacy3/
toc.htm> [hereinafter FTC Report).
n107. See Video Privacy Protection Act, 18 U.S.C. 2710 (1994); Schwartz &
Reidenberg, supra note 26, at 10 (discussing the circumstances leading up
to adoption of the "Bork Bill").
n108. See, e.g., Cable Communications Policy Act, 47 U.S.C. 551 (1994); Electronic
Communications Privacy Act of 1986, 18 U.S.C. 2701 (1994). See generally
Schwartz & Reidenberg, supra note 26 (giving examples of U.S. information
privacy laws).
n109. See Drivers' Privacy Protection Act of 1994, 18 U.S.C. 2721 (1994);
Driver's Privacy Protection Act: Hearings on H.R. 3365 Before the Subcomm.
on Civil and Const. Rights 103d Cong. (1994) (statement of Congressman James
P. Moran), available in Fed. Doc. Clearing House, Feb. 4, 1994 (explaining
rationale for this legislation). The constitutionality of this legislation
is currently under review by the U.S. Supreme Court. See note 33 supra and
accompanying text.
n110. Consider also that if someone loses her car, she can always get a new
one, but when she loses her privacy, it may well be gone forever.
n111. Data compilers would likely prefer broad transfers because these might
mean fewer contractual restrictions to negotiate and keep track of. Yet,
if the goal of legal protection is to achieve information privacy, these
concerns of compilers of personal data should not be paramount.
n112. See, e.g., Varian, supra note 70, at 39 (discussing the advisability
of restrictions on transfer of rights in personal data).
n113. See, e.g., Restatement of Property, supra note 71, 409 cmt. a (1944);
Cunningham et al., supra note 71, 2.15, at 82-84 (stating that prohibitions
against restraints on alienation are relaxed in the case of life estates,
primarily because life estates are not that marketable to begin with; however,
even so, certain conditions must be met for the restraint to be valid); see
also Radin, supra note 71, at 16-29 (discussing general policy favoring alienability
of property rights and arguments against making property rights inalienable
in the market). In the United States, however, there is a statutory scheme
specifically designed to prevent the alienation of certain types of information.
42 U.S.C. 2274 makes it a criminal act to communicate "restricted data" when
it is known that communication of such data might injure the U.S. or benefit
a foreign nation. See 42 U.S.C. 2274 (1994). Although there are procedures
for determining when information is "classified," 42 U.S.C. 2162 (1994),
some information is considered by the U.S. government to be "born classified."
See, e.g., Harold P. Green, Constitutional Implications of Federal Restrictions
on Scientific Research and Communication, 60 UMKC L. Rev. 619, 630 (1992);
Peter Swan, A Road Map to Understanding Export Controls: National Security
in a Changing Global Environment, 30 Am. Bus. L.J. 607, 613 n.37 (1993).
n114. Another issue with which a property rights regime would have to contend
is whether an individual could assert property rights against a party who
obtained her data from public records (for example, publicly accessible drivers
license data). See Varian, supra note 70, at 35-37 (discussing public policies
favoring access to and reuse of personal data).
n115. The term "moral right" is a rather rough translation of the French
term, "droit moral." At least one commentator has suggested the use of a
more exact terminology, namely that of the German term, "Urheberpersonlichkeitsrecht,"
meaning "right of the author's personality," see 1 Stephen P. Ladas, The
International Protection of Artistic and Literary Property 272 (1938).
n116. There are many countries that protect moral rights, but the two most
commonly discussed are France and Germany. See Law on the Intellectual property
Code, No. 92-597 of July 1, 1992, in World Intellectual Property Org., Copyright
and Neighboring Rights, Laws and Treaties (1996) [hereinafter French Act];
Urheberrechtgesetz (UrhG) IV.2, arts. 12-14 <http://iecl.iuscomp.org/gla/statutes/UrhG.htm>
[hereinafter German Act].
n117. See, e.g., Neil Netanel, Alienability Restrictions and the Enhancement
of Author Autonomy in United States and Continental Copyright Law, 12 Cardozo
Arts & Ent. L.J. 1, 7 (1994):
Although a work may be commercially exploited, it is not simply a commodity--and
many commentators would say that it is not a commodity at all. Instead, the
work is seen, partially or wholly, as an extension of the author's personality,
the means by which he seeks to communicate to the public. "When an artist
creates,... he does more than bring into the world a unique object having
only exploitive possibilities; he projects into the world part of his personality
and subjects it to the ravages of public use.
Id. (quoting Martin A. Roeder, The Doctrine of Moral Right: A Study in the
Law of Artists, Authors and Creators, 53 Harv. L. Rev. 554, 557 (1940));
see also Radin, supra note 71, at 20 (noting that some interests are "incompletely
commodified").
n118. See French Act, supra note 116, art. L. 121-1 (codifying the right
of attribution); German Act, supra note 116, art. 13 (codifying the right
of attribution); see also French Act, supra note 116, art. L. 121-1 (codifying
the right of integrity); German Act, supra note 116, art. 14 (codifying the
right of integrity).
n119. See French Act, supra note 116, art. L. 121-2 (codifying the right
of divulgation); German Act, supra note 116, art. 12 (codifying the right
of divulgation); see also French Act, supra note 116, art. L. 121-4 (codifying
the right of withdrawal).
n120. Continental authors may choose not to enforce their moral right out
of fear of reprisals from producers and publishers in a tightly knit creative
community, but this does not mean that they could not legally do so if they
chose. See, e.g., Judgment of Dec. 12, 1988 (Delorme v. Catena-France), Cour
d'appel, P.I.B.D. III, No. 454, 231, cited in Netanel, supra note 117, at
25 n.123 (holding that even a copyright assignment "for all purposes" requires
the author's permission to modify the work). For a general discussion of
the actual inalienability of the Continental right, see Netanel, supra note
117, at 47 nn.254-305 (arguing that core of moral rights are properly considered
to be inalienable under Continental law).
n121. See Snow v. Eaton Centre [1982] 70 C.P.R.2d 105 (Ont. H.C.J.).
n122. See, e.g., Soc. Le Chant du Monde v. Soc. Fox Europe, Jan. 13, 1953,
Cours d'appel, Paris, Dallez, Jurisprudence, [D. Jur.] 16, 80, cited in Roberta
Rosenthal Kwall, Copyright and the Moral Right: Is an American Marriage Possible?,
38 Vand. L. Rev. 1, 28, n.103 (1985) (holding that Russian composers could
prevent their music from being used in a film that had an anti-Soviet theme,
because of the "moral damage" that would result). Section 43(a) of the Lanham
Act can sometimes be used to protect an artist's reputation in a manner similar
to protection available under moral rights law. See, e.g., Gilliam v. American
Broad. Cos., 538 F.2d 14, 18-19 (2d Cir. 1976) (invoking Lanham Act to prevent
the television broadcast of a modified version of a Monty Python movie).
n123. See, e.g., Rose Aguilar, Research Service Raises Privacy Fears, Cnet
News.com, (June 10, 1996) <http://news.cnet.com/news/0-1005-200-311506.html?tag=>.
After gathering personal data (such as social security numbers, addresses,
names and aliases) from various sources, Lexis-Nexis offers a centralized
searching service to government or anyone else seeking such information.
There is no oversight on who actually uses the service or how they use it.
The service is also marketed to individuals, journalists, etc., who might
want to find spouses who have missed support payments or engaged in criminal
behavior. The range of harms that could result from such a collection of
data appears obvious: think of a journalist working on a story about husbands
who skip their support payments, or investigators who pursue an individual
based on inaccurate information. Because the risk of data inaccuracy increases
with the number of times data changes hands, this type of service, which
involves at least four transfers, seems particularly prone to inaccuracy.
See the range of products offered at <http://www.lexisnexis.com/Incc/government/>.
n124. This interest was strikingly illustrated in the case of McVeigh v.
Cohen, 983 F. Supp. 215 (D.D.C. 1998). Like many users of America OnLine,
McVeigh took advantage of the opportunity to adopt a pseudonymous identity
for interacting with other people on AOL and create for this identity an
online profile which other users of AOL could see which included a reference
to his being "gay." U.S. Navy officials, after becoming suspicious that this
profile might be about McVeigh, sought confirmation from AOL, and after receiving
this confirmation, the Navy sought to expel him from service. Although this
case involved "public" information, McVeigh did not use his real name in
the profile, thus attempting to keep his identity undisclosed. See id.
n125. The requirement of privity is a foundational principle of contract
law (being an the inevitable consequence of bargain theory). In some cases
a third party is allowed to "step into the shoes" of one of the actual parties
to the contract. See, e.g., E. Allan Farnsworth & William F. Young, Contracts:
Cases and Materials (5th ed. 1995) at 863-70. Generally, this substitution
is allowed when the contract has been signed by A, but for the benefit of
a third party, B. In American jurisprudence such contracts are known as "third
party beneficiary contracts," and one of the most common examples is insurance
contracts. For a straightforward application of the privity doctrine, see,
e.g., Hanback v. Dutch Baker Boy, Inc., 107 F.2d 203, 203 (D.C. Cir. 1939)
(finding no suit in contract under the theory of implied warranty after a
child gets food poisoning from a chocolate eclair because child was not in
privity with the seller, even though her mother bought the eclair). The harshness
of the privity requirement has been recognized and relaxed in the case of
implied product warranties (which are imposed on contracts involving the
sale of goods). See, e.g., Henningsen v. Bloomfield Motors, Inc., 161 A.2d
69, 413 (N.J. 1960) (holding that strict privity was not required in cases
involving implied warranty of merchantability).
n126. This situation arises because companies who collect data from individuals
with whom they have a business relationship often sell this data to third
parties. These third parties are not in any contractual relationship with
the individual who originally supplied the data. This practice is especially
common in industries which generate "transactional information" such as the
banking industry. See, e.g., Julie Tripp, A Cause For the Masses: Banks Selling
Personal Data, The Oregonian, June 27, 1999 <http://www.oregonlive.com/business/99/06/bz062706.html>.
n127. The general unavailability of injunctions is witnessed by the fact
that we generally do not speak of injunctions per se in contracts, but instead
speak of awarding specific performance of the contract. Of course, in some
cases, specific performance requires that certain activities be enjoined.
A common example of this is the situation of non-competition agreements,
in which the courts will enjoin a former employee from competing with the
employer for a "reasonable" length of time, as long as the noncompete agreement
does not unreasonably prejudice the former employee's ability to earn a livelihood.
See Comprehensive Techs. Int'l., Inc. v. Software Artisans, Inc., 3 F.3d
730, 737-38 (4th Cir. 1993), vacated and dismissed on petition for rehearing.
For a general discussion of when specific performance will be ordered in
a contract, see, e.g., First Nat'l State Bank of N.J. v. Commonwealth Fed.
Sav. & Loan Ass'n of Norristown, Pa., 455 F. Supp. 464, 469 (D.N.J. 1978)
(holding that specific performance is only ordered when damages are otherwise
inadequate, or where they cannot be calculated accurately); see also Restatement
(Second) of Contracts, 357 (1982).
n128. See, e.g., Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d
1152, 1158 n.5 (6th Cir. 1978) (Markey, J., by designation), cited in Robert
P. Merges, Patent Law and Policy 973 (1997) (stating that injunctive relief
is the preferred remedy in property cases because "the right to exclude others
is the essence of the human right called "property.'")
Patents must by law be given "the attributes of personal property." 35 U.S.C.
261. The right to exclude others is the essence of the human right called
"property." The right to exclude others from free use of an invention protected
by a valid patent does not differ from the right to exclude others from free
use of one's automobile, crops, or other items of personal property.
Id.
n129. It is unlikely that McVeigh would license the use of certain personal
information for purposes that they would not, ex ante, have approved. An
injunction to prevent the use would therefore appear to be the desired remedy
in many cases involving personal data. See note 124, supra and accompanying
text.
n130. See, e.g., Kwall, supra note 122, at 57-72 (describing several entrenched
doctrines in American copyright law which prevent the wholesale adoption
of Continental-style moral rights); Netanel, supra note 117, at 3 n.12 (discussing
the vehement opposition to moral rights by Congress and American legal scholars).
n131. On March 1, 1989, the United States acceded to the Berne Convention,
which requires signatories to protect the moral rights of authors. See Bern
Convention for the Protection of Literary and Artistic Works, Sept. 9, 1886,
123 L.N.T.S. 233, art. 6[su' bis'].
n132. See, e.g., Netanel, supra note 117, at 2 (noting that the moral rights
idea is based on the notion of "literary and artistic works as inalienable
extensions of the author's personality.").
n133. See note 94 supra and accompanying text.
n134. The publicity right arises under the "commercial advantage" prong of
the invasion of privacy tort. See Restatement (Second) of Torts 652C (1976).
It therefore protects economic, rather than personality, interests. See,
e.g., White v. Samsung Elecs. Am., Inc. 989 F.2d 1512, 1512-13 (9th Cir.
1993) (discussing how property law must balance the protection of investments
in initial creation with the promotion of future creative efforts).
n135. See, e.g., Jonathan Kahn, Bringing Dignity Back to Light: Publicity
Rights and the Eclipse of the Tort of Appropriation of Identity, 17 Cardozo
Arts & Ent. L.J. 213, 213-21 (1999) (arguing that publicity and privacy
have the same appropriation tort origin and that the protection given to
publicity could also be given to privacy).
n136. Id. at 213; see also Robert C. Post, Rereading Warren & Brandeis:
Privacy, Property, and Appropriation, 41 Case W. Res. L. Rev. 647, 675 (1991)
(distinguishing between the right of publicity and the appropriation tort).
n137. See, e.g., N.Y. Civil Rights Law, 50-51 (McKinney 1976 & Supp.
1982) (proscribing use of a person's name or likeness unless written consent
has been obtained from that person).
n138. See, Kahn, supra note 135, at 215 (arguing that publicity and appropriation
call upon different authorities for enforcement, but basically use the same
criteria of virtue and vice).
n139. In addition to damages for mental anguish and injured feelings, injunctive
relief can be awarded in appropriation privacy cases. See, e.g., Stig Str<um
o>mholm, Right of Privacy and Rights of the Personality 151-64 (1967)
(discussing remedies generally); Kahn, supra note 135, at 265-266 (noting
that looking at the remedies sought can help determine the degree to which
plaintiffs feel their privacy has been compromised); Post, supra note 136,
at 667 (discussing damages).
n140. See note 16 supra and accompanying text.
n141. See Federal Elections Campaign Law, 2 U.S.C. 438 (a)(4) (1994) (limiting
commercial reuses of lists of campaign contributors); Privacy Act, 5 U.S.C.
552a(b) (1994) (prohibiting disclosure of information by Federal agencies);
Video Privacy Protection Act, 18 U.S.C. 2710 (1994) (prohibiting disclosures
of video rental records except under stated circumstances).
n142. See, e.g., Kang, supra note 2, at 1266-67 (considering an inalienability
rule for personal data, but concluding that "if the individual wants to exercise
... [her right of] control by disclosing information for various reasons
including monetary compensation, then the state should hesitate to proscribe
information flow on some paternalistic theory."); see also EU Directive,
supra note 14, art. 7(a) (stating that collection and processing of personal
data is lawful if collector/processor has the consent of individual).
n143. See note 10 supra and accompanying text. It may be sensible to consider
licensable personal data as "incompletely commodified," to borrow Professor
Radin's useful phrase. See Radin, supra note 71, at 20.
n144. See notes 148-157 infra and accompanying text.
n145. See James Pooley, Trade Secrets 9.01 (1997) (estimating that trade
secret theft costs the U.S. economy between 5 and 10 billion dollars annually).
n146. See, e.g., 1 Jay Dratler, Jr., Intellectual Property Law: Commercial,
Crea-tive, and Industrial Property 4.03[3][b] (1991). As Dratler points out,
the whole purpose of the law of trade secrets is to promote licensing and
exchange of nonpatented know-how between businesses and employees. The requirement
in trade secret law therefore is, not absolute secrecy, but rather "relative"
secrecy. See id.
n147. See, e.g., Smith v. Dravo Corp., 203 F.2d 369, 375 (7th Cir. 1953)
(holding that on implied confidential relationship arose from disclosure
of trade secret information to enable other firm to evaluate whether to negotiate
a proposed business deal).
n148. It must be noted here that although trade secret law does not rely
on property rights "as such," there is an ongoing debate about the exact
nature of the rights underlying this body of law. There are two main theories
behind trade secret law, generally referred to as the "property school" and
the "confidential relationship" school. See Pooley, supra note 145, 1.02[8].
The choice of characterization is more than semantic; it has a practical
impact on the legal consequences that courts will impose on parties. While
Pooley prefers to settle the debate by referring to the regime as "hybrid,"
id. 1.02[8][d], Milgrim gives the property theory more weight, pointing to
the fact that the owner of a trade secret can exclude the world from his
secret, and the fact that a trade secret can be assigned in the manner of
property, especially when a business is sold, etc. See 1 Roger M. Milgrim,
Milgrim on Trade Secrets 2.01 (1999). However, this author thinks that the
answer to the exclusion point is that the right is not "good against the
world," except insofar as the owner's power to control the behavior of those
he stands in confidential relations to (i.e., the exclusionary power is actually
just a by-product of the relational power that the owner has against those
in certain types of relationships with him). That this is so, is manifested
by the fact that the "exclusionary power" can only be maintained if accompanied
by the efforts of the owner to maintain actual secrecy. So, what appears
to be a right against the world is merely a functional product of actual
secrecy supplemented by enforced behavior on certain people who can destroy
that secrecy. Furthermore, the descendability of trade secrets does not really
support a characterization of the rights as a "property" regime, as the issue
here is not that the "owner" or the assignee can exclude the world. Instead,
the issue is who can exclude those in a confidential relationship or those
who would otherwise use improper means to obtain the trade secret. We should
not be confused in our characterization of the regime by the fact that courts
have adopted a legal fiction (i.e., that of calling a trade secret property)
for specific pragmatic reasons, such as to make the right descendable. An
even more convincing argument for the "property" characterization of trade
secrets is the development of the "improper means" branch of misappropriation.
Trade secrets were historically considered not to be property. Both the Supreme
Court's unequivocal statement in E.I. du Pont de Nemours Powder Co. v. Masland,
244 U.S. 100, 102 (1917) (holding that in this case involving trade secretes,
the issue is not property but the confidential relationship between the parties),
and the Restatement of Torts 757 (1938), make this abundantly clear. However,
courts and jurists soon saw that the "breach of confidential relationship"
ground of trade secret misappropriation was insufficient to police commercial
morality, and therefore interpreted the "improper means" branch of trade
secret misappropriation as completely separate from any relationship between
the parties. So, there are two, completely separate grounds for misappropriation:
"breach of confidence" and "improper means." One can see how the "improper
means," because it applies to everyone, looks like a property right. One
could even conceptualize the "breach of confidence" as one branch of improper
means--and this line of reasoning adds even more support to the property
school. And yet, even the "improper means" ground of trade secret misappropriation
does not transform the trade secret into a property right. We must remember
that the locus of the trade secret right is in the behavior of the nonowner
B, rather than the trade secret of the owner A. A does not have the right
to exclude B from the trade secret, he merely has the ability to prevent
B from taking certain actions to obtain it. An analogy might be useful in
drawing this distinction. If I drop my purse, I can still sue to get in back,
even though the person who finds it can not be charged with "theft." "Finders
keepers" rules are exceptional in the law, and usually are created for specific
purposes, such as to promote salvage on the high seas via pecuniary reward.
The fact that people do not sue people who find their purses because they
do not know who found the purse is an evidentiary, rather than a legal, issue.
On the other hand, if I "drop" my trade secret while walking down the street,
and my competitor discovers it, I cannot sue to get it back, even before
he has disclosed it to anyone else. I also could not get an injunction preventing
him from using or disclosing the trade secret, as he did not use improper
means to obtain it. So, we can view the obligation of the trade secret owner
to use reasonable efforts to maintain secrecy, as an implicit confirmation
of the nonproperty status of trade secrets because it is this obligation
which effectively destroys the property character of the right. Of course,
it should also be noted that the ALI has moved trade secrets closer to the
status of a property right by providing a right of action against third parties
who innocently discover the trade secret, once they receive notice of the
status of the trade secret. See notes 167-169 infra and accompanying text.
However, the author submits that this limited modification of the right seeks
to prevent misappropriators from carelessly "leaking" the trade secret to
"innocent" third parties, who can then claim that they did not misappropriate
the secret. Therefore, this rule is more about evidentiary issues involved
in policing business behavior, than about transforming the trade secret into
a property right. In the end, the overriding concern of the trade secret
regime is with policing the behavior of business entities. In addition, the
paucity of cases involving innocent third party "misappropriators," see note
169 infra, means that this modification to the right is more theoretical
than real.
n149. See, e.g., Ruckelshaus v. Monsanto Co., 467 U.S. 986, 987 (1984) (finding
trade secret information to be "property" within the meaning of the Fifth
Amendment for purposes of deciding whether the government's unauthorized
use or disclosure of the information should be subject to eminent domain
rules). See Samuelson, supra note 28, at 378-83 (criticizing the property
characterization for trade secrecy rights and of the Court's interpretation
of Missouri law in Ruckelshaus).
n150. The breach of confidential relationship and breach of contract grounds
are often closely related in trade secret law, but they are conceptually
distinct. Sometimes, a confidentiality agreement or other restrictive contract
will help to establish a confidential relationship, but courts will often
impose a confidential relationship without contractual restrictions on disclosure,
particularly in the case of employees. See, e.g., Milgrim, supra note 148,
at 4.02[1][b] and cases discussed therein. Restrictive licensing agreements
may also be used by the courts as evidence that sufficient efforts were made
to maintain secrecy. See Schalk v. State, 823 S.W.2d 633, 638-44 (Tex. Crim.
App. 1991), cert. denied, 118 L. Ed. 2d 425 (1992) (holding that a combination
of agreements including nondisclosure agreements "served to support trade
secrete status.").
n151. See, e.g., E.I. duPont de Nemours & Co. v. Christopher, 431 F.2d
1012, 1015-16 (5th Cir. 1970), cert. denied, 400 U.S. 1024 (1971) (holding
that improper means is a separate branch of trade secret misappropriation,
which neither requires a breach of a confidential relationship or illegal
conduct; industrial espionage, though not itself a criminal act, constitutes
improper means when the trade secret owner was using reasonable efforts to
maintain its secrecy.)
n152. See, e.g., Merges et al., supra note 76, at 29-120.
n153. See, e.g., Restatement (Third) of Unfair Competition, 39-45 (1988)
(noting that law "makes some rudimentary requirements of good faith" when
there is a confidential relationship between the parties).
n154. See, e.g., 17 U.S.C. 106 (1994) (setting forth exclusive rights of
copyright law); 35 U.S.C. 271 (1994) (setting forth exclusive rights of patentees).
n155. See, e.g, Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 497-98 (1974).
Although there are other policies implicated in trade secret laws, maintaining
commercial morality is a dominant interest. See 1 Melvin F. Jager, Trade
Secrets Law 1.03 (1999). Some of the other policies are: the promotion of
investment in research, exploitation of knowledge, privacy, mobility of labor
and free competition. For a thorough discussion of these alternate policies,
see Pooley, supra note 145, 1.02[3]-[7].
n156. See, e.g., California v. Greenwood, 486 U.S. 35, 40 (1988).
n157. See, e.g., Drill Parts & Serv. Co. v. Joy Mfg. Co., 223 U.S.P.Q.
521, 526 (Ala. 1983); Pooley, supra note 145, 6.02[2][e].
n158. Although this principle is illustrative of a more general contractual
rule of construing the actual agreement between the parties, this particular
default rule finds strong expression in the context of trade secrets. See,
e.g., Data Gen. Corp. v. Grumman Sys. Support Corp., 36 F.3d 1147, 1165-69,
(1st Cir. 1994), partial summary judgment granted, 32 U.S.P.Q.2d 1946 (D.
Mass. 1994).
n159. This is a general principle of contract interpretation. 1 Joseph M.
Perillo, Corbin on Contracts 1.1 (rev. ed. 1993) [hereinafter Corbin]; see
also, e.g., Darner Motor Sales, Inc. v. Universal Underwriters Ins. Co.,
682 P.2d 388, 396 (1984) (holding that unambiguous terms in standard-form
contracts will not be given their effect if they do not meet the reasonable
expectations of the parties).
n160. This is simply an application of the general contractual principle
that a court will seek to protect and enforce the reasonable expectations
of the parties. So, for example, a court may refuse to interpret a term in
a contract literally when the circumstances indicate that an alternate meaning
was intended. Tantleff v. Truscelli, 110 A.D.2d 240 (N.Y. App. Div. 1985),
discussed in Corbin, supra note 159, 1.1.
n161. See, e.g., Chameleon Dental Prod. Inc. v. Jackson, 18 U.S.P.Q.2d 1044,
1047 (7th Cir. 1991). The exact nature of what qualifies as material does,
however, differ between states and between judgments. See, e.g., Skil Corp.
v. Lucerne Prod. Inc., 206 U.S.P.Q. 792, 823 (N.D. Ohio 1980) (holding that
licensor is entitled to terminate a contract only if the licensee's behavior
indicated abandonment of the contract or caused irreparable injury).
n162. Numerous contract doctrines seek to prevent a weaker party from making
an improvident bargain. Whether this is conceived of as a cognitive dissonance
sufficient to negate a meeting of the minds, or whether it is viewed as judicial
undoing of the contract to prevent harm to a weaker party, the result is
the same. For a discussion of these doctrines, which include unconscionability,
inequality of bargaining power, and contracts of adhesion, see generally
Melvin Aron Eisenberg, The Limits of Cognition and the Limits of Contract,
47 Stan. L. Rev. 211 (1995); Anthony T. Kronman, Paternalism and the Law
of Contracts, 92 Yale L.J. 763 (1983).
n163. This expected increase in cognitive difficulties is a result of the
fact that transactions which transfer personal data most often involve, at
least at the initial point of data collection, an interaction between an
unsophisticated individual and a sophisticated business entity. See note
16 supra (citing sources pointing to cognitive difficulties in assessing
information privacy risks).
n164. It should be noted that this is somewhat of a simplification. There
are really two licensing issues which impact on alienability: sublicensing
and assignment. Sublicensing is more damaging from a privacy perspective
because it results in the creation of multiple right-holders. Assignment,
on the other hand, merely allows one right-holder to be substituted for another
(as when a business is sold, etc.). As a general matter, sublicensing of
nonexclusive licenses is not permitted unless such permission is express.
See, e.g., Noel Byrne, Licensing Technology 210-11 (2d. ed. 1998). On the
issue of assignment, which may or may not be permissible, depending on the
circumstances, see Terry B. McDaniel, Shop Rights, Rights in Copyrights,
Supersession of Prior Agreements, Modification of Agreement, Right of Assignment
and Other Contracts, 14 AIPLA Q.J. 35, 45-47 (1986) (discussing problems
that could arise with trade secret protection due to nonassignable employee
confidentiality agreements). In general, contracts which do not involve federally
granted intellectual property rights are assignable as a matter of state
law, except when the contract relies on the honesty, reputation, skill, character
or ability of one of the parties. 4 Arthur Linton Corbin, Corbin on Contracts
866 (1951 & Supp. 1971); see also Green v. Camlin, 92 S.E.2d 125, 127
(S.C. 1956) ("Rights arising out of a contract cannot be transferred if they
are coupled with liabilities, or if they involve a relationship of personal
credit and confidence [in the context of a franchise agreement]."); Restatement
(Second) of Contracts 317(2), 318(2) & 319(2) (1979) (discussing assignment
of rights, delegation of performance duty, and delegation of performance
of condition generally).
n165. See, e.g., Byrne, supra note 164, at 210.
n166. See id. at 23.
n167. See, e.g., Uniform Trade Secrets Act 1(2) (1985).
n168. See, e.g., id., 1(2)(ii)(B).
n169. See, e.g., id., 1(2)(ii)(C). However, if an innocent third party has
made substantial investments based on an understanding of its entitlement
to use the information, courts may withhold injunctive relief and provide
the trade secret claimant with a damages only remedy. See id.; see also Pooley,
supra note 145, 2.03[3][a]. There are very few cases involving innocent third
parties who thereafter receive notice; the author could find only one. See
Forest Laboratories, Inc. v. Pillsbury Co., 452 F.2d 621 (7th Cir. 1971).
n170. See discussion notes 83-92 supra and accompanying text for a discussion
regarding the constitutional problems with granting intellectual property-like
rights in personal data. A licensing regime would be less likely to interfere
with the First Amendment than a property regime would because, unlike property
rights, contract rights are not good against the world. See Robert P. Merges,
The End of Friction? Property Rights and Contract in the Newtonian World
of On-line Commerce, 12 Berkeley Tech. L.J. 115, 118-27 (1997); see also
Cohen v. Cowles Media Co., 501 U.S. 663, 665 (1991) (finding no First Amendment
right to breach negotiated agreement not to disclose identity of news source);
Kang, supra note 2, at 1277-82 (concluding that default rule providing protection
to personal data would not conflict with the First Amendment).
n171. See notes 95-105 supra and accompanying text. For a discussion of how
the regime proposed herein would accommodate both market-based and civil
liberty-based visions of information privacy, see notes 238-244 infra and
accompanying text.
n172. See, e.g., The Digital Dilemma, supra note 45, at ES 5-6.
n173. See notes 22-57 supra and accompanying text.
n174. For example, trade secrecy law aims to provide lead-time protection
to induce appropriate levels of investment in industrial innovations. See,
e.g., Reichman, supra note 46, at 2446-47. As a consequence, remedies for
trade secrecy protection will often be limited to those necessary to restore
adequate lead-time to the firm whose secret was misappropriated. See, e.g.,
Lamb-Weston, Inc. v. McCain Foods, Ltd., 941 F.2d 970, 975 (9th Cir. 1991)
(upholding eight-month injunction "to eliminate commercial advantage that
otherwise would be derived from the misappropriation").
n175. See Joseph Elford, Trafficking in Stolen Information: A Hierarchy of
Rights Approach to the Private Facts Tort, 105 Yale L.J. 727, 727-29 (1995)
(arguing that use of improper means to obtain personal information should
to be illegal).
n176. See, e.g., Bill Swindell, House Carries on Over Photo Sales, The Post
and Courier (Feb. 26, 1999) <http://www.charleston.net/news/imagedata/house0226.htm>
(discussing a South Carolina bill that would allow the state to continue
selling drivers' license photos); see also H.R. 1450, 106th Cong. (1999)
(citing the Personal Information Privacy Act of 1999, which would prevent
state departments of motor vehicles from transferring drivers' photos without
permission). Congress passed the Driver's Privacy Protection Act of 1994,
18 U.S.C. 2721(a) (1999), but several states have objected to it as an intrusion
on state prerogatives under the Tenth Amendment. See, e.g., Reno v. Condon,
120 S. Ct. 666, 671 (2000) (rejecting state claims that the Driver's Privacy
Protection Act is unconstitutional). A discussion of the case can be found
at Linda Greenhouse, States' Rights Adherents on Top Court Appear to Be Given
Pause, N.Y. Times, Nov. 11, 1999, at A20.
n177. See, e.g., Sara Robinson, CD Software Said to Gather Data on Users,
N.Y. Times, Nov. 1, 1999, at C1 (discussing how RealNetwork's software for
playing CDs gathers information about the user's activities in connection
with the software).
n178. See, e.g., Edmund Sanders & Robert A. Rosenblatt, New Era for Financial
Services, L.A. Times, Nov. 13, 1999, at C1.
n179. See David F. Gallagher, Amazon Moves to Ease Worry About Privacy of
Customers, N.Y. Times, Aug. 30, 1999, at C1 (discussing how Amazon.com changed
its policy of gathering personal information from its customers after critics
raised privacy concerns).
n180. See, e.g., Ted Bridis, RealNetworks Sorry for Tracking Data, The Chattanooga
Times, Nov. 3, 1999.
n181. See Steve Lohr, Seizing the Initiative on Privacy, N.Y. Times, Oct.
11, 1999, at C1.
n182. See Denise Caruso, Digital Commerce, N.Y. Times, Aug. 30, 1999, at
C5; see also Thomas P. Novak, Donna L. Hoffman and Marcos Peralta, Building
Consumer Trust in Online Environments: The Case for Information Privacy (Working
Paper of Vanderbilt University Project 2000, 1998) <http://ecommerce.vanderbilt.edu/papers/CACM.privacy98/CACM.
privacy98.htm> discussing barriers to establishing consumer trust in online
services).
n183. See Killingsworth, supra note 43, at 1. The need for private sector
firms to adopt privacy policies and practices to comply with the EU Directive
has also been recognized by the Clinton Administration, which has been working
on "safe harbor" guidelines. See U.S. Dept. of Commerce, Draft, International
Safe Harbor Privacy Principles (Nov. 15, 1999) <http://www.ita.doc.gov/td/ecom/Principles
1199.htm>.
n184. Killingsworth, supra note 43, at 1.
n185. See IITF Principles, supra note 1.
n186. FTC Report, supra note 106, at 7-14.
n187. See Self-Regulation and Privacy Online, Before the Subcomm. on Telecommunications,
Trade, and Consumer Protection of the House Comm. on Commerce (1999) (statement
of Robert Pitofsky) <http://www.ftc.gov/os/1999/9907/pt071399.htm>.
n188. See id. at 1-2.
n189. See, e.g., The Online Privacy Protection Act, 1999, Hearing on S. 809
Before the Subcomm. on Communications of the Senate Comm. on Commerce, Science
and Transp., 106th Cong. 4-5 (1999) (testimony of Marc Rotenberg, Director,
Electronic Privacy Information Center), <http://www.epic.org/privacy/internet/EPIC<uscore>testimony<uscore>799.pdf>
(noting that the FTC often fails to adequately evaluate whether companies
follow their privacy policies).
n190. See, e.g., Mary J. Culnan, Georgetown Internet Privacy Policy Survey
(June 8, 1999) <http://www.msb.edu/faculty/culnanm/gippshome.html>
(noting that 66% of all Web sites post their privacy policies); Online Privacy
Alliance Report on the Top 100 Web Sites (June 8, 1999) <http://www.msb.edu/faculty/culnanm/gippshome.html>.
n191. See Jeri Clausing, I.B.M. Vows to Pull Ads from Web Sites that Lack
Clear Policies on Protecting Consumer Privacy, N.Y. Times, Apr. 1, 1999,
at C4.
n192. So, for example, Real Networks had a privacy policy, but it did not
say that it was collecting data every time one used the software. See, e.g.,
RealNetworks Is Target of Suit in California Over Privacy Issue, N.Y. Times,
Nov. 9, 1999, at C16. For examples of bills pending in Congress which would
require web site owners to give consumers clear notice of the data being
gathered and of the uses being made of that data, see Consumer Internet Privacy
Protection Act of 1999, H.R. 313, 106th Cong. (1999) (prohibits disclosure
of personally identifiable information gathered online without consumer consent);
Internet Growth and Development Act of 1999, H.R. 1685, 106th Cong. 301 (1999)
(requiring notice); Online Privacy Protection Act of 1999, S. 809, 106th
Cong. (1999) (requiring notice).
n193. See, e.g., In re Geocities, FTC Docket No. C-3850 (Feb. 12, 1999) <http://www.ftc.gov/
os/1999/9902/9823015d&o.htm> (finding deceptive practices in the collection
of personal information from children deviating from stated privacy policy).
The FTC power "to prevent persons ... from using unfair methods of competition
... and unfair or deceptive acts or practices in ... commerce." 15 U.S.C.
45(a)(2) (1994). The legislative history of the FTC Act reflects a disinclination
to specify the unfair acts or practices because "there is no limit to human
inventiveness in this field." American Cyanamid Co. v. Fed. Trade Comm'n,
363 F.2d 757, 769 (6th Cir. 1966) (citing H.R. Rep. No. 1142, at 18-19 (1914)).
Before 1938, the FTC's jurisdiction was limited by the requirement that the
FTC show specific injury to competitors. See, e.g., Fed. Trade Comm'n v.
Raladam Co., 283 U.S. 643, 649 (1931). But Congress responded in 1938 with
the Wheeler-Lea Amendment which added to the language of Section 5 a prohibition
of "unfair or deceptive acts or practices." 15 U.S.C. 45(1) (1938) The announced
purpose of the amendment was to overcome the limitation on jurisdiction imposed
by the Supreme Court in the Raladam decision, and to make the consumer injured
by unfair trade practices of equal concern, under the law, with injured businesses.
See Pep Boys - Manny, Moe & Jack, Inc. v. Fed. Trade Comm'n, 122 F.2d
158, 161 (3d Cir. 1941) (stating that 1938 amendments intended to broaden
FTC jurisdiction over business practices). Given this broad mandate, it would
seem possible for the FTC to investigate and issue orders concerning commercial
businesses that were practicing unfair or deceptive acts involving personal
information dissemination. However, it is somewhat unclear if the FTC has
power, for example, to order Web sites to post privacy policies. See, e.g.,
Tolliver, supra note 105, at 69 (discussing limits to the FTC's jurisdiction
on information privacy issues).
n194. Killingsworth, supra note 43, at 12. This attorney recommended that
Web site owners prepare explicit privacy licensing agreements, rather than
allowing such agreements to be inferred from the existence of a privacy policy,
so that the firm could include terms of choice, such as clauses requiring
arbitration of disputes. See id. at 13.
n195. Securities dealers, for example, have formed nonprofit organizations
to oversee and evolve self-regulatory activities in that business. Although
some cyberspace privacy self-regulatory enforcement mechanisms do exist,
such as the Truste privacy "seal" program, these have not proven particularly
effective. See, e.g., Courtney Macavinta, Truste Reports on RealNetworks
as FTC Examines Net Privacy, CNET News.com, Nov. 8, 1999 <http://news.cnet.com/news/0-1005-200-1431844.html>
(reporting that Truste had taken no action against several firms that violated
seal requirements).
n196. See U.C.C. art. 2B (Proposed Draft, Feb. 1, 1999) <http://www.law.upenn.edu/:80/
library/ulc/ucc2b/2b299.htm> [hereinafter U.C.C. 2B]; see generally UCITA
supra note 19. There were several reasons why this model law was removed
from the U.C.C. and promulgated as a stand-alone model law. For one thing,
the licensing paradigm did not fit well with the sales of goods transactions
covered by U.C.C. Article 2, and the U.C.C. is normally reserved for codification
of well-established commercial practices - which have not developed in the
area of information transactions. The American Law Institute (ALI) also had
significant reservations about U.C.C. 2B which might have made it difficult
for this model law to become part of the U.C.C. See Joint Press Release by
ALI and NCCUSL, NCCUSL to Promulgate Freestanding Uniform Computer Information
Transactions Act, (Apr. 7, 1999) <http://www.2bguide.com/docs/040799pr.html>
(stating only that "it has become apparent that this area does not presently
allow the sort of codification that is represented by the Uniform Commercial
Code.").
n197. See, e.g., Raymond T. Nimmer, Breaking Barriers: The Relation Between
Contract and Intellectual Property Law, 13 Berkeley Tech. L.J. 827, 829 (1998).
n198. See, e.g., Brenda Sandburg, E-Commerce Plan Faces Tough Fight, Cal.
Law., Aug. 4, 1999 <http://www.callaw.com/stories/edt0804e.html>.
n199. The Virginia Legislative Assembly, upon recommendation of the Joint
Commission on Science and Technology, has created its own version of UCITA.
See HB1 Computer Information Transactions Act <http://leg1.state.va.us/cgi-bin/legp504.exe?001+sum+HB561>;
see also Sandburg, supra note 198 (providing information on the adoption
process).
n200. See generally Symposium, Intellectual Property and Contract Law in
the Information Age: The Impact of Article 2B of the Uniform Commercial Code
on the Future of Transactions in Information and Electronic Commerce, 13
Berkeley Tech. L.J. 809 (1998).
n201. See, e.g., Lorin Brennan, The Public Policy of Information Licensing,
36 Hous. L. Rev. 61, 111 (1999) (anticipating the use of UCITA in consumer
licensing of personal data to private sector firms); Martin, supra note 12,
at 849 n.344 (same); see also Pamela Samuelson, A New Kind of Privacy? Regulating
Uses of Personal Data in the Global Information Economy, 87 Cal. L. Rev.
751, 776 (1999) (expressing doubts about the suitability of UCITA for information
privacy protection).
n202. UCITA, supra note 19, 102(a)(10) (""Computer information' means information
in electronic form which is obtained from or through the use of a computer
or which is in a form capable of being processed by a computer. The term
includes a copy of the information and any documentation or packaging associated
with the copy.").
n203. See id. 102(a)(11).
"Computer information transaction" means an agreement or the performance
of it to create, modify, transfer, or license computer information or informational
rights in computer information. The term includes a support agreement under
Section 612. The term does not include a transaction merely because the parties'
agreement provides that their communications about the transaction will be
in the form of computer information.
Id.
n204. It is somewhat unclear under UCITA whether someone needs to have a
legally protectable interest in information in order to be entitled to license
it. See id. 102(a)(38).
"Informational rights" include all rights in information created under laws
governing patents, copyrights, mask works, trade secrets, trademarks, publicity
rights, or any other law that gives a person, independently of contract,
a right to control or preclude another person's use of or access to the information
on the basis of the rights holder's interest in the information.
Id. For a criticism of UCITA's failure to be clear on this issue, see Jessica
Litman, The Tales that Article 2B Tells, 13 Berkeley Tech. L.J. 931, 931
(1998) (discussing the confusion in U.C.C. 2B over the nature of the rights
an information licensor might have in information other than those supplied
by intellectual property law). If the law confers on individuals a legally
protectable interest in personal data, these would seem to be "informational
rights" that UCITA would cover. It is not, however, at all clear that under
existing law, individuals can reasonably be said to have such rights in personal
data. See notes 26-34 supra and accompanying text. But if they did, or if
the law came to recognize that they did, such rights would seem to be licensable
under UCITA.
n205. See UCITA, supra note 19, 112.
n206. But see notes 219-226 infra and accompanying text (discussing how the
technological infrastructure might evolve to enable consumers to offer terms
for use and disclosure of personal data).
n207. UCITA does, of course, provide an array of default rules to fill in
missing terms. See, e.g., UCITA, supra note 19, 307, 308 (providing default
rules for a license and for the duration of a contract). Some of these, such
as its narrow implied right provision, might bode well for protecting personal
data. See, e.g., id. 307(a), (b). But see Jane C. Ginsburg, Authors as Licensors
of Informational Rights Under U.C.C. Article 2B, 13 Berkeley Tech. L.J. 945,
953-65 (1998) (reporting that the narrow implied rights provision of Article
2B might be good news for writers, but anticipating that publishers would
respond to this model law by developing elaborate contracts to protect their
interests to which they would require authors agree).
n208. See notes 205-207 supra and accompanying text.
n209. Consider also that Web sites set their own terms and conditions for
use of their sites. Under UCITA, individual users could be said to have "assented"
to such terms and conditions, either by clicking "I agree" or by continuing
to use the site after having an opportunity (which they will typically not
take up) to examine the site's terms and conditions. Even a cursory review
of the terms of service at commonly visited Web sites reveals how one-sided
they typically are (for example, they disclaim warranty and other responsibilities
on the part of the site owner and impose responsibilities on users). See,
e.g., Yahoo GeoCities Terms of Service <http://docs.yahoo.com/info/ terms/geoterms.html>
(imposing registration obligations and indemnification agreements on users,
while disclaiming warranties). Given this, it might be reasonable to expect
that if UCITA becomes the law, site owners will add to existing terms of
service a waiver of their responsibilities toward personal data that users
reveal at the site or a broad release of informational rights. See UCITA,
supra note 19, 208. If this occurs, it would constitute a step backwards
for information privacy, not a step forward.
n210. Consumer Reports Online currently makes "e-Ratings," which include
an evaluation of online merchants' privacy and security policies, available
to its subscribers. See Consumer Reports, E-Ratings Online Shopping Guide
<http://www.consumerreports.org/Special/Samples/Reports/ 9910etip.htm>.
Such activities could be expanded to include drafting of model licensing
agreements.
n211. See notes 219-231 infra and accompanying text.
n212. See, e.g., Robert W. Gomulkiewicz, The License Is the Product: Comments
on the Promise of Article 2B for Software and Information Licensing, 13 Berkeley
Tech. L.J. 891, 894 (1998).
n213. Compare U.C.C. art. 2B (Proposed Draft, Dec. 1, 1995) <http://www.2bguide.com/
drafts.html> (focusing on transactions in digital information), with U.C.C.
art. 2B (Proposed Draft, Feb. 2, 1996) <http://www.2bguide.com/drafts.html>.
A rationale for the expansion of scope can be found in: Notes on the February
1, 1996 Draft <http://www.lawlib.uh.edu/ucc2b/febnotes.html>.
n214. See, e.g., Letter from Motion Picture Assoc. of America, Recording
Indus. Assoc. of America, Magazine Publishers of America, Nat'l Assoc. of
Broadcasters, Nat'l Cable Television Assoc. of America & Newspaper Assoc.
of America to NCCUSL (Dec. 7, 1998) <http:// www.2bguide.com/docs/1298mpaa.html>
(voicing opposition to scope and enactment of U.C.C. 2B).
n215. See Drafting Comm., NCCUSL, Report on the November 13-15, 1998 Drafting
Committee Meeting (1998) <http://www.2bguide.com/nov98rpt.html>.
n216. See, e.g., Letter from Motion Picture Assoc. of America, Recording
Indus. Assoc. of America, Magazine Publishers of America, Nat'l Assoc. of
Broadcasters, Nat'l Cable Television Assoc. of America & Newspaper Assoc.
of America to NCCUSL, May 10, 1999 <http:// www.2bguide.com/docs/coalit5.html>
(voicing continued opposition to UCITA).
n217. See note 201 supra.
n218. See J.H. Reichman & Jonathan A. Franklin, Privately Legislated
Intellectual Property Rights: Reconciling Freedom of Contract With Public
Good Uses of Information, 147 U. Pa. L. Rev. 875 (1999).
n219. See, e.g., Burkert, supra note 4, at 125-42; see also Ian Goldberg,
David Wagner & Eric Brewer, Privacy-Enhancing Technologies for the Internet
<http://www.cs.berkeley.edu/<diff>daw/papers/privacy-compcon97-www/privacy-html.
html> (reviewing existing privacy-enhancing technologies for the Internet,
and the potential for further developments). But see Susan Freiwald, Uncertain
Privacy: Communication Attributes After the Digital Telephony Act, 69 S.
Cal. L. Rev. 949, 1119 (1996) (suggesting that advances in technology have
rendered the existing legal methods of protecting communications privacy
ineffective); Jonathan Weinberg, Hardware-Based ID, Rights Management, and
Trusted Systems, 52 Stan. L. Rev. 1251 (2000) (explaining the potential of
hardware-based identification systems to have privacy-destructive consequences).
n220. See Philip Agre, Introduction, in Technology & Privacy, supra note
4, at 7.
n221. Burket, supra note 4, at 127.
n222. Id. at 125-28. An example of the latter is considered in Bernardo A.
Huberman, Matt Franklin & Tad Hogg, Enhancing Privacy and Trust in Electronic
Communities <http://www.parc. xerox.com/istl/groups/iea/www/privacy.html>
(aiming to "facilitate finding shared preferences, discovering communities
with shared values, removing disincentives posed by liabilities, and negotiating
on behalf of a group" by adapting cryptographic techniques).
n223. See Reagle & Cranor, supra note 55.
n224. The prospects for electronic agent technology for engaging in electronic
commerce are explored in Pattie Maes, Robert H. Guttman & Alexandros
G. Moukas, Agents That Buy and Sell, Comm. ACM, Mar. 1999, at 81; see also
Brennan, supra note 201, at 109-14 (discussing the use of electronic agents
to contract in cyberspace on privacy terms); A Killer App for Computer Chat,
Economist, Apr. 10, 1999, at 79 (bots can be programmed to ask about a web
sites' privacy policies). UCITA validates contracts made by electronic agents.
See UCITA, supra note 19, 107, 112 & 206. Some speak of P3P as though
it will serve as an electronic agent negotiating privacy terms. See, e.g.,
Chris Oakes, The Trouble With P3P, Wired News, June 25, 1998, at 1 <http://
www.wired.com/news/technology/0,1282,13242,00.html> (outlining benefits
and problems with P3P).
n225. See Reagle & Cranor, supra note 55; see also Harvard Developments,
supra note 35, at 1646-47 (expressing enthusiasm for P3P as a means to protect
information privacy). Privacy advocate Marc Rotenberg is skeptical about
how useful P3P will be in the protection of personal data. See, e.g., Hearings
on Communications Privacy Before the Subcomm. on Courts and Intellectual
Property of the House Judiciary Comm., 106th Cong. (March 26, 1998) (testimony
of Marc Rotenberg), available in <http://epic.org/privacy/internet/rotenberg-testimony-398.html>.
Where privacy techniques focused on the means to protect identity, now the
focus is on means to obtain information. Many of the techniques that are
put forward as "technical solutions" - such as the Open Profiling Standard,
the P3P and Truste - will make it easier, not more difficult, to obtain information
from individuals using the Internet.
Id.; see also Karen Coyle, P3P: Pretty Poor Privacy? A Social Analysis of
the Platform for Privacy Preferences (P3P), (June 1999) <http://www.kcoyle.net/p3p.html>
(stating that P3P cannot adequately protect privacy because it is designed
to facilitate the gathering of data by web sites); Oakes, supra note 224,
at 2 (explaining difficulties for humans in adequately programming browsers
with P3P instructions).
n226. See Reagle & Cranor, supra note 55.
n227. See, e.g., Harvard Developments, supra note 35, at 1645-46.
n228. Some e-cash systems have been implemented with anonymizing features.
However, not all e-cash systems have this feature. See, e.g., Bruno Giussani,
Feeding the Meter - With a Pocketful of Micropayments, N.Y. Times, Aug. 19,
1997 <http://www.nytimes.com/library/cyber/euro/ 081997euro.html>.
See generally A. Michael Froomkin, Flood Control on the Information Ocean:
Living With Anonymity, Digital Cash, and Distributed Databases, 15 J.L. &
Comm. 395 (1994) (discussing technical and policy reasons for doubting technology
will protect privacy).
n229. See, e.g., A. Michael Froomkin, The Death of Privacy?, 52 Stan. L.
Rev. 1461 (2000).
n230. Burkert, supra note 4, at 140.
n231. See generally R<um u>diger Grimm, User Control Over Personal
Web Data, EEMA (1999) (discussing technical means of implementing German
data protection rules); R<um u>diger Grimm, Nils L<um o>hndorf
& Philip Scholz, Data Protection in Teleservices (The DASIT Project)
(on file with the author) (describing research project on uses of technology
to implement the EU Directive in telecommunications services).
n232. Compare Information Infrastructure Task Force, National Information
Infrastructure, Agenda for Action <http://www.metalab.unc.edu/nii/NII-Agenda-for-Action.html>
(advocating a market-driven "hands-off" approach to promoting the information
infrastructure), with European Commission, European Council, Europe and the
Global Information Society: Recommendations to the European Council <http://www.echo.lu/
eudocs/en/report.html> (discussing the importance of protecting privacy,
pluralism and freedom of expression in the global information society).
n233. See, e.g., EU Directive, supra note 14, art. 1.1.
n234. This rather infamous quote has been reported in various places. See,
e.g., Robert Lemos, The Dark Side of the Digital Home, ZDNet, Feb. 7, 1999
<http://www.zdnet.com/zdnn/stories/ news/0,4586,2203898,00.html>.
n235. See, e.g., Kang, supra note 2, at 1196-97 (citing polls about privacy
concerns).
n236. One impediment to the development of American information privacy law
has been its unduly heavy focus on "reasonable expectations of privacy."
See, e.g., Schwartz & Reidenberg, supra note 26, at 60-73. This has two
serious drawbacks. First, it largely excludes consideration of normative
purposes for limiting the collection and use of personal data, thereby undermining
society's ability to evolve norms and rules to regulate these matters because
it tends to make the law concerned about places, not people. See, e.g., Olmstead
v. United States, 277 U.S. 438, 476 (1928) (holding that wire tapping is
not within the scope of the Fourth Amendment because, when the Constitution
was enacted, the Fourth Amendment was intended to limit tresspass on property);
Lawrence Lessig, Reading The Constitution In Cyberspace, 45 EMORY L.J. 869,
872-875 (1996) (discussing Olmstead). Second, it is conducive to an ongoing
erosion of privacy. The more intrusive surveillance technology becomes, the
less reasonable is any expectation that individuals will have privacy, and
as a consequence, the less privacy the law will recognize. See Schwartz &
Reidenberg, supra note 26, at 65
n237. See, e.g., Swire & Litan, supra note 10, at 153-59.
n238. The lack of consensus about the nature of a person's interest in personal
data may help to explain the wide range of solutions to the information privacy
problem that legal commentators have proposed. For further consideration
of the nature of the interest problem in respect of personal data, see Julie
E. Cohen, Examined Lives: Informational Privacy and the Subject as Object,
52 Stan. L. Rev. 1373 (2000) (discussing property rights, choice-based, knowledge-based,
and personal autonomy-based rationales for protecting personal data).
n239. See notes 72-82 supra and accompanying text.
n240. See notes 70-71 supra and accompanying text.
n241. See NAACP v. Alabama, 357 U.S. 449, 458-59 (1958) (holding that the
state interfered with First Amendment interests by requiring the NAACP to
disclose its membership).
n242. See, e.g., Samarajiva, supra note 5, at 283 (arguing that "privacy
... is situational and relation-specific ....").
n243. See, e.g., id. (emphasizing the importance of consent). It is inappropriate
for my doctor to test my blood to see if I have HIV when I have not agreed
to this - unless, of course, the law has required the doctor to do so.
n244. See, e.g., Schauer, supra note 20, at 557 ("The standard rhetoric of
Internet privacy challenges ironically understates the Internet revolution,
because it does not acknowledge the way in which the Internet and related
technologies have changed the concept of privacy itself.").
n245. See notes 192-194 supra and accompanying text.
n246. See note 48 supra and accompanying text.