The Broad Reach of Carpenter v. United States
"The Broad Reach of Carpenter v. United States" by Paul Ohm was originally published June 27, 2018, on Just Security
Carpenter v. United States is an inflection point in the history of the Fourth Amendment. From now on, we’ll be talking about what the Fourth Amendment means in pre-Carpenter and post-Carpenter terms. It will be seen as being as important as Olmstead and Katz in the overall arc of technological privacy.
I hope to develop the argument that Carpenter is as important as Katz across two or three articles at Just Security, but let me begin with my overall big picture: The holding and reasoning of Carpenter is breathtakingly broad and will be applied far beyond the facts of this case. (For a detailed overview of the facts and five opinions of this case, see this blog post written by my star student and recent Georgetown graduate, Sabrina McCubbin.)
Carpenter holds that the police may not collect historical cell-site location information (CSLI) from a cellphone provider without a warrant. Footnote three (get used to a lot of scholarly commentary about Carpenter, footnote three!) restricts the holding, for now, to seven days of collection. I’ll write later about why I’m not as worried about the seven-day limit as other commentators.
This is the opinion most privacy law scholars and privacy advocates have been awaiting for decades. Oceans of ink have been spilled by those worried about how the dramatic expansion of technologically fueled corporate surveillance of our private lives automatically expands police surveillance, too, thanks to the third-party doctrine (a moment of silence, please, for the nearly departed). The third-party doctrine is the rule that provides that some information about an individual held by third parties falls entirely outside the protections of the Fourth Amendment.
With Carpenter, the third-party doctrine is almost dead. The majority opinion–which commanded the votes of five justices, none of whom muddied their full participation in the holding or reasoning with a separate concurrence–”decline[d] to extend” the third-party doctrine to the collection by the FBI from cellphone providers of seven-days of CSLI. “Given the unique nature of cell phone location information, the fact that the Government obtained the information from a third party does not overcome Carpenter’s claim to Fourth Amendment protection.” (p.17) Even on its own terms, this has sweeping consequences for privacy and law enforcement. But it’s the way Chief Justice Roberts reasoned his way to the new holding that assures that this opinion will be applied far beyond the facts of this case.
First, as described in the majority and dissenting opinions, the CSLI that has just been protected isn’t terribly precise. If the majority had placed an exaggerated gloss on the precision of CSLI at issue in this case, it would have given the government a way in future cases to distinguish other types of location information: “the data at issue in this case is not controlled by Carpenter,” the government could have argued, “because it is far less precise than CSLI.”
It’ll be difficult to make this argument because the majority opinion informs us that the CSLI records in this case “placed [Carpenter] within a wedge-shaped sector ranging from one-eighth to four square miles.” (p.14). In his dissent, Justice Kennedy characterizes these dimensions as “covering between a dozen and several hundred city blocks” in cities and “up to 40 times more imprecise” in rural areas. (Kennedy, p.4). GPS this certainly ain’t. The Chief Justice waves this away, in part, because “the rule this Court adopts ‘must take account of more sophisticated systems that are already in use or in development'”. (p.14) The source of this quote is Kyllo v. United States, the 2001 case requiring a warrant to aim a thermal imager at a house. I’ve always wondered if this methodology – rule on the facts but keep an eye on the future – represented merely a one time flourish from Justice Scalia. Carpenter firmly ensconces it in Fourth Amendment jurisprudence.
Second, the majority opinion isn’t restricted to CSLI, even on its own terms. Instead, this is an opinion about information that can locate people generally, not CSLI specifically. Part III of the opinion is all about the privacy interests individuals have in “the whole of their physical movements.” (p.12) This is a meditation on the nature of location information, whatever form it takes. Geolocation information, when there is enough of it, “provides an intimate window into a person’s life”, quoting Justice Sotomayor’s celebrated opinion from Jones, revealing “familial, political, professional, religious, and sexual associations.” (p.12) This case is “not about ‘using a phone’ . . . [i]t is about a detailed chronicle of a person’s physical presence compiled every day, every moment, over several years.” (p.16-17). It is about “a trail of location data.” (p.17)
By focusing on the nature of the information rather than on the telecommunications nitty-gritty used to gather the information or the structure of the database in which the information was held, this opinion clearly signals that its holding will apply to other massive collections of historical geolocation information, of which there are many. Many smartphone apps collect precise GPS information, including apps that have no need for this kind of information except to sell to advertisers. It’s not just your smartphone, as GPS information is gathered by the companies that provide fitness trackers, connected cars, and smart watches. Internet of Things gizmos place location trackers in our clothes, bags, and even our bodies. It might not be that every database of location information generated by every technology I list will fall within the Carpenter reasoning, but I think the police should think twice before trying to collect any of it without a warrant.
Third, the majority opinion will probably even apply to information that isn’t expressly about location but from which location may be inferred. “[T]he Court has already rejected the proposition that ‘inference insulates a search'” (p.14), quoting Kyllo once again. The opinion highlights how the government could use CSLI “in combination with other information, [to] deduce a detailed log of Carpenter’s movements.” (p.14) Many databases that don’t store location information directly can be used to infer location information. Credit card records, automatic toll transponder records, automated license-plate records, etc, can all generate inferences about a person’s location that are far more precise than CSLI. Any time the government accesses a privately assembled database in order to track location without a warrant, it risks suppression under Carpenter.
Fourth, the true test of the substantive sweep of Carpenter will be whether courts apply its reasoning to government access to databases full of sensitive and intimate information that isn’t directly related to location. I have no doubt that Carpenter will be extended to apply other massive databases, but I’m a little more uncertain about what the precise outer boundaries of this expansion will cover. David Kris has already written an interesting post exploring this question in national security cases, and I’ll try to write a little more about this soon.
Fifth, as my colleague Marty Lederman persuasively argues, Justice Gorsuch’s interesting “dissent” makes him a potential sixth vote for future cases involving some of the types of information I discuss above. Justice Gorsuch has signaled that he is no fan of the third-party doctrine, and scholars and defense lawyers will be studying his opinion, trying to unlock the formulation that will win his vote. Most importantly, this means that the death of the third-party doctrine will probably survive a change of opinion or retirement of one member of the Carpenter majority.
In sum, criminal defendants will test the outer boundaries of Carpenter‘s reasoning whenever the police use massive databases assembled by private parties that reveal location information, directly or by inference. Other defendants will challenge the collection of data unrelated to location. The broad reasoning of the majority’s opinion will give all of them plenty to work with. Anticipating this, risk-averse police departments will err on the side of caution, getting a warrant for data whenever they can, and turning promising leads into dead ends whenever they can’t. It’s a powerful reminder of the ability the Supreme Court has to protect civil liberties and reshape the contours of our relationship with the state. This opinion does no less, finally, at long last, giving us a tool to disrupt–at least for a moment–the steady march to a surveillance state.
About the author: Paul Ohm is Professor of Law at Georgetown University Law Center. Follow him on Twitter (@paulohm).
This article originally appeared June 27, 2018, on justsecurity.org and is reprinted with permission. Copyright, 2018 Just Security