2007 Symposium/Symposium Articles
The Olmsteadian Seizure Clause
The Fourth Amendment's Seizure clause is mired in the Eighteenth century. Its
counterpart, the Search clause, has evolved through a steady progression of Supreme
Court cases from Katz to Berger to Kyllo, no longer to be confined to the property-based
notions of privacy embodied in Olmstead v. United States. Instead it is sensitive to
modern privacy concerns by extending Constitutional protection to situations that satisfy
the reasonable expectation of privacy test. While imperfect, the evolved Search clause
has kept the protections of the Fourth Amendment relevant in an age of digital evidence,
ubiquitous communication networks, and increasingly sophisticated and invasive
surveillance capabilities.
In contrast, the Seizure clause is in an Olmsteadian holding pattern, consistently
interpreted to protect only physical property rights and to regulate only the deprivation of
tangible things. In particular, Courts interpreting the clause rarely consider what
"deprivation" means when we are talking about intangible property such as digital
evidence, and voice and data communications.
In this essay, Professor Ohm argues for a Twenty-First century definition of
Constitutionally-proscribed property deprivation. He argues that a Constitutionally
significant "Seizure" occurs whenever the State obtains the original or a copy of
personally-owned, non-public data. By copying the data, the State deprives the owner of
the property of the ability to delete or alter the State-possessed copy of the data. In
addition, modern Intellectual Property rules, and in particular Copyright, acknowledge
other harms caused by an unauthorized copy.
The Supreme Court has already opened the door to this definition, in Berger and Katz, by
holding in no uncertain terms that voice conversations are both searched and seized when
recorded by the police. Professor Ohm traces lower-court cases that came before and after
Berger and Katz and recognizes that most lower courts have ignored these Supreme
Court's Seizure holdings.
Embracing a modern interpretation of the Seizure clause is consistent with the Framers'
intent, because copying affects the property rights of owners of intangible property in
many of the same ways that physical dispossession deprived property owners at the time
the Fourth Amendment was adopted.
Reconceiving the seizure clause in light of modern concerns about intangible property
rights helps solve many vexing Fourth Amendment puzzles that arise if the sole test is the
reasonable expectation of privacy. For example, does a bit-by-bit copy of a computer's
hard drive implicate the Fourth Amendment, if the human operator does not "view" the
contents as they are copied? Could the government lawfully capture all of the
communications traversing a network without a warrant so long as they did not look at
the contents without a subsequent warrant? Do government-run network intrusion
detection systems implicate the Fourth Amendment?
Viewed as possible violations of the Search clause, these are frustrating, metaphysical
inquiries; if a bit falls in a packet sniffer, has it been searched? In contrast, under the new
definition of Seizure, these questions result in straightforward answers. In every one of
these situations, a seizure has occurred. The owner of the information has lost the ability
to delete, modify, secrete, or contextualize a copy of the information, even though he may
have retained his own copy. No less than when the police commandeer an automobile or
grab a box of records, the owner of the intangible property has lost dominion and control
over his property. A seizure has occurred, and the Fourth Amendment should proscribe
these acts absent warrant or exception.
Audio from Paul Ohm's presentation of his paper at the 2007 STLR Symposium can be accessed here
Comments
Greetings, all. Paul [first name ok? we've not met], is there more here than in your Harvard forum comment on Orin? More generally: Ironically? --apropriately?-- the best thing on the Fourth Amendment is not available online, at least not on Westlaw or anything I could find on google. Lexis might be different. Anyway, since it looks like it has to be summoned from the libes on (omg) photocopy, I take the liberty of posting the cite on a Saturday night: Anthony Amsterdam, Perspectives on the Fourth Amendment, 58 Minn. L. Rev. 349 (1974). Best--
--dd
Posted by: Don Dripps | January 20, 2007 7:09 AM
It's unlikely that Lexis has the Minn. L. Rev. piece. For those who have access to Hein Online, however, you can access it at any time at http://heinonline.org/HOL/Index?index=journals/mnlr&collection=top30
Posted by: Eric Chan | January 20, 2007 7:15 PM
The point about “nonrivalrous,” perfect copying (of formless but sensitive data) makes good sense. And a beefed up “seizure” prong, at first blush, seems like a good way bridge the 4th Amendment gap. However, I think the unconditional “right to delete” that Ohm proposes goes too far. That proposal replaces one bright line rule (the rule that property must be tangible in order to be “seized”) with another (the rule that any copying of data by law enforcement is a “seizure” because it results in the loss of an individual’s right to delete that data – even if no human has viewed that data.)
While the outcome from a personal privacy perspective is certainly desirable, recognizing such a right would fundamentally shift the power balance between the government and defendants. There may be very legitimate reasons for law enforcement to preserve data before “searching” it – for example, if a defendant might destroy the data, or if it is ephemeral data such as network traffic. For that reason alone, it might be very hard to implement such a strong seizure prong.
In addition, it is unclear whether a property-like right should be recognized in every bit of information that is generated about a person. The case for saying that an individual “owns” the data to be seized (and thus has the right to “delete” it) seems to vary with the kind of data – it’s relatively clear in the case of personal files that reside on a hard drive image, for example, but maybe less clear with ISP records, e-mail routing information, or Windows registry and log files that the user did not create. Does the amount of conscious control matter in granting a right to delete? Should an individual have a right to delete certain bits of data that he doesn’t know exists? What about the argument that another entity “owns” particular bits of data about a particular individual, and only that entity should have the right to delete?
In the physical world, a defendant does not have “the right to delete” every physical trace that he or she may have left behind. Should that change just because digital copies are nonrivalrous and perfect? Perhaps there is a way to refine the idea of (and test for) “seizure” of such intangible, nonrivalrous information that addresses these problems. A more nuanced test might distinguish between different kinds of information, taking into account the role an individual played in generating that data, for example. Then again, any test short of a bright-line rule might lend itself to the sort of “Constitutional gamesmanship” that Professor Ohm claims that Professor Orin Kerr’s redefinition of “search” (search as any potential exposure of data to human observation) is vulnerable to. I am not entirely convinced of this.
In any case, I am looking forward to the final draft.
For reference – the Professor Ohm’s Harvard forum comment, referred to by Professor Dripps above, can be read here.
Posted by: Eric Chan | January 23, 2007 2:25 AM
I think Eric's [again, first name ok?] point about Paul's turn to the seizure clause as reaching an "outcome [that] from a personal privacy perspective is certainly desirable" nicely captures my doubts about Paul's thesis. Suppose that the outcome from a privacy vs. legitimate state interests concerns were deemed UNdesirable? That seems to be history's judgment on Boyd--but Paul flirts with history as a jurtification for his position, and Boyd might well be right as a matter of history. Note that under Boyd, following Entick and Wood, private papers could not be seized even pursuant to a warrant founded on probable cause. That would be a crazy rule in the digital age. At least with respect to these upper-class-privacy-interests vs. upper-class-national-security-concerns, I have some sympathy with Orin's willingness to leave the balance to the political branches.
Posted by: Don Dripps | January 23, 2007 7:29 AM
Don and Eric,
Thanks so much for the insightful comments. I don't have enough time to address everything you've said in this space, but let me respond briefly, with elaboration on Friday.
I am trying to present neither as sweeping nor as absolute a proposal as your comments (especially Eric's) suggest. Of course, I am to blame for not having written more already. The reason why I want to revisit what I have already said in the Harvard Forum (to address Don's first comment) is to flesh out these finer details.
In my two pieces, I have proposed this test: “Does the government’s copying of intangible property produce negative effects on par with the effects of physical dispossession? In particular, does the government’s copy prevent the owner from altering, destroying, or otherwise changing the state of his property?”
The innovation of my test, I think, is the idea that SOME intangible property should be protected by the seizure clause. I do not make nor do I even agree with the stronger claim (as I think Eric thinks I do) that ALL intangible property is thus protected. Eric says, “The case for saying that an individual 'owns' the data to be seized (and thus has the right to 'delete' it) seems to vary with the kind of data.” I agree completely with this statement.
Just as the tangible-property Fourth Amendment seizure rules exclude many categories of tangible property--things you’ve abandoned, things held by third parties, contraband in plain view--so too are there many, many categories of intangible property that are not covered under my test. (But, Eric, I don’t think I agree that “hidden” files such as cache and the registry fall into this category--after all they are on my computer--but I’m happy to debate this at the event.)
If I give a third party a copy of my data for his or her use (and not merely for safe-keeping), I shouldn’t be able to complain when the data ends up with the government. If my data is posted publicly--on my MySpace page, for example--I cannot protest that it has been copied by the police. If the police have probable cause to believe that data is about to be destroyed, exigent circumstances may justify a warrantless seizure. At the Symposium, I hope the panelists and audience will help me develop these examples and others.
My argument, then, is not as sweeping as Eric suggests, but I don’t think it a mere trifle, either. The examples I give in the paper--hard drive imaging, packet sniffing--are everyday occurrences, and I believe that confining these acts to protection under the search prong leaves gaps that are not merely regrettable, but also Unconstitutional.
Finally, in addition to these two examples, let me pose one more, ripped from last week’s headlines about the NSA wiretapping program: wiretap minimization. At some point in the past twenty years, the police switched from ex ante to ex post minimization of wiretaps. They used to be obligated (by the terms of their warrants, if not the reasoning of Berger) to cease recording calls that did not fall within the scope of their warrants. If Aunt Bea shared a phone line with her criminal-suspect nephew, most of her conversations would not have been taped.
Today, at least for data wiretaps, minimization is done after the fact, through the application of filters. In the meantime, the data belonging to Aunt Bea and many, many other people (in the case of the NSA program, maybe millions of other people) unrelated to the investigation are stored, sometimes indefinitely, on government hard drives. Again, when we confine our analysis to the search clause, this is probably no cause for alarm. Out of sight, out of Constitutional mind. But under my test, under the seizure clause, this may be a separate, cognizable harm.
I look forward to meeting you both (and you all) on Friday.
Posted by: Paul Ohm | January 24, 2007 6:16 AM
Professor Ohm's panel has generated some press. Declan McCullagh, a CNET reporter, attended and captured some comments between Ohm and Richard Downing (currently with the DOJ and a former colleague of Ohm's) during our event. The issue is whether the DOJ now uses new, more intrusive electronic monitoring techniques. This has prompted an official response from the DOJ:
http://news.com.com/FBI+turns+to+broad+new+wiretap+method/2100-7348_3-6154457.html?tag=st.num
http://news.com.com/2061-10796_3-6154934.html (DOJ's response; linked from first article)
http://www.wired.com/news/columns/0,72608-0.html?tw=wn_index_19 (article by Jennifer Granick)
This has prompted some demand for the audio recording of Ohm's panel. STLR is trying to secure release of the audio files for all panels, and should have them up soon. Please keep checking the website, and feel free to post comments.
Posted by: Henry Huang | February 7, 2007 7:26 PM
Die Privatsphäre eines Menschen ist sehr sehr wichtig. Hierbei geht es nicht um materielle Dinge, sondern um die persönliche Privatsphäre bei Comunikation und bei der freien Entfaltung!
Posted by: Michael | January 18, 2008 8:35 AM
thanx for nothing
Posted by: Telefon | January 18, 2008 4:39 PM
cep telefonu
Posted by: Telefon | January 19, 2008 10:14 PM