In the excitement over the Second Circuit’s ruling on the NSA’s bulk collection program, another very significant appellate decision that was issued last week has been largely overlooked: the Eleventh Circuit’s en banc decision in United States v. Davis. A majority of the eleven judge panel held that the government did not need a warrant to collect 67 days’ worth of cell site location information on Quartavious Davis, who was suspected of involvement in several armed robberies.

On first glance, the panel’s holding appears to answer in the negative the question that the Second Circuit punted: whether telephony metadata receives protection under the Fourth Amendment. On closer examination, however, the fractured ruling, with its many separate opinions, highlights a fundamental lack of consensus over the reach of the third party doctrine.

Majority Opinion and Pryor Concurrence

Writing for the court, Judge Hull concludes that the case is controlled by United States v. Miller (1976) and Smith v. Maryland (1979), which together stand for the proposition that a person has no reasonable expectation of privacy in information that he or she voluntarily conveys to a third party. Below, Davis attempted to distinguish Smith — which involved the government’s use of a pen register to acquire phone numbers — by arguing that cell phone users do not voluntarily disclose their location in the same way that they disclose numbers they dial. Judge Hull dismisses this notion, noting that cell phone users are surely aware that their calls are routed through cell phone towers and that phone companies keep records of these transmissions.

Privacy advocates, academics, lawmakers, and judges have questioned whether the third party doctrine makes sense in the digital age. People today are effectively forced to convey a broad range of highly personal information to third parties, such as internet service providers and mobile carriers, on a daily basis. Judge Hull’s opinion is striking in its refusal to pay even lip service to these concerns. She does not chafe at the restraints of Smith and Miller; she is happy to be bound by them, and heaps scorn on the dissent for questioning whether their rule is suited to modern life. 

Judge Hull’s dismissal of the arguments raised by Davis and the dissenting judges is much too facile. At the outset, she emphasizes that Davis did not own or possess the business records he sought to suppress. Whether a person has a “reasonable expectation of privacy,” however — thus triggering the Fourth Amendment’s protections — does not turn on ownership. The government may not freely search a rented apartment or tap a telephone wire the caller does not own. Judge Hull’s characterization of the cell site data as company-generated information that merely “concerns” Davis misses the mark. The information contained in the phone records is entirely a byproduct of Davis’s communications. Davis generated the information; the phone company merely recorded it.

Judge Hull is also too quick to find that Davis willingly sacrificed any privacy rights he might have. While she is probably correct that most cell phone users know they are disclosing some location information, a knowing disclosure is not the same thing as a voluntary disclosure. The advice Judge William Pryor gives in his concurring opinion — “If a telephone caller does not want to reveal [information] to the telephone company, he has another option: don’t place a call” — is wildly unrealistic. It was probably unrealistic even when Smith was decided, but the ability to limit third party disclosures and still carry on a normal life has declined dramatically since the 1970s.

Judge Hull’s opinion also fails to grapple with the fact that sophisticated computer algorithms today can extract a wealth of detailed personal information from long-term location tracking. Noting that “[h]istorical cell tower location does not identify the cell phone user’s location with pinpoint precision,” Judge Hull concludes that “[h]istorical cell site location data does not paint the ‘intimate portrait of personal, social, religious, medical, and other activities and interactions’ that Davis claims.” In fact, Davis’s own case proves that cell site location data can be quite precise: the prosecutors asserted that “Mr. Davis’s phone [was] literally right up against the America Gas Station immediately preceding [the robbery]” and pointed to “the presence of his cell phone literally … right next door to the Walgreen’s just before and after that store was robbed.” In any event, the rule embraced by Judge Hull’s opinion — i.e., there can be no privacy interest in location information shared with a third party — would apply to highly precise GPS tracking as well.

Jordan and Wilson Concurrence

Judges Jordan and Wilson take a markedly different approach. Their concurrence acknowledges that “Smith does not fit this case like a glove” because “cellphones and smartphones (and the vast amounts of information they contain and can generate) are qualitatively different from land-line phones.” They accordingly assume that Davis had a reasonable expectation of privacy without expressing a definitive opinion on the matter.

Judges Jordan and Wilson nonetheless conclude that the warrantless search was permissible because it met the Fourth Amendment’s “reasonableness” standard. They assess the legislative scheme for obtaining phone records under the Stored Communications Act and find that it reasonably balances the government’s interest in obtaining evidence in criminal investigations against the “diminished” privacy interest phone users have in non-content data. (This reasoning also appears in Judge Hull’s majority opinion as an alternative basis for upholding the search.)

There is a surprisingly basic flaw in this approach. While it is true that the touchstone of the Fourth Amendment is reasonableness, the Supreme Court has held that warrantless searches are per se unreasonable “except in a few ‘jealously and carefully drawn’ exceptional circumstances.” Examples of such exceptional circumstances include exigency, searches incident to arrest, and the “plain view” exception. The cases on which Judge Hull, Jordan, and Wilson primarily rely are the so-called “special needs” cases, in which the Supreme Court has allowed warrantless searches where a special need other than law enforcement would render a warrant impracticable. Clearly, however, the search of Davis’s phone records was conducted for law enforcement purposes, so this exception does not apply — nor do any of the other recognized exceptions.

Rosenbaum’s Concurrence

The most interesting opinion by far is Judge Rosenbaum’s concurrence. He begins by observing that the third party doctrine has never served as an unyielding rule. When Smith was decided, he points out, telephone users voluntarily exposed the contents of their calls to the operators who could (and often did) listen in. Similarly, occupants of hotel rooms voluntarily expose their personal objects to the view of hotel housekeepers. Despite these third party disclosures, the government clearly needs a warrant to install wiretaps or search hotel rooms.

Judge Rosenbaum’s approach to reconciling these seeming anomalies is original and thought-provoking. He summarizes it as follows:

I believe that Supreme Court precedent fairly may be read to suggest that the third-party doctrine must be subordinate to expectations of privacy that society has historically recognized as reasonable.


Historically, privacy in one’s communications and one’s dwelling place were nearly sacrosanct; indeed, they formed the basis for the Fourth Amendment. These privacy interests therefore survive limited incursions by the third parties who provide facilities for communications or habitation.

Judge Rosenbaum next posits that long-term location tracking may violate just such a core, historical expectation — “the expectation of privacy in not being subject to constant, longer-term surveillance.” In support, he cites Justice Alito’s concurrence in United States v. Jones, which notes the practical limitations that historically constrained surveillance in public spaces. Disclosure to a mobile service provider, he asserts, cannot vitiate this longstanding expectation. He nonetheless concludes that Davis’s privacy interest was not violated because the cell towers at issue had a mile-and-a-half radius. In his view, “that is not precise enough to rival the invasion of privacy that point-longer-term surveillance represents.”

Even though Judge Rosenbaum ultimately agrees with the warrantless search, the most significant and influential part of the panel’s 102-page opinion is likely to be this concluding paragraph of his concurrence:

For all of these reasons, I believe that Smith (and therefore, the third-party doctrine) inescapably governs the outcome of this case. But when we must necessarily expose information to third-party technological service providers in order to make use of everyday technology, and the technological service merely allows us to engage in an activity that historically enjoyed a constitutionally protected privacy interest, Supreme Court precedent can be viewed as supporting the notion that the historically protected privacy interest must trump the third-party doctrine for purposes of Fourth Amendment analysis. If the historically protected privacy interest does not, then with every new technology, we surrender more and more of our historically protected Fourth Amendment interests to unreasonable searches and seizures.

Martin and Pryor Dissent

After Judge Rosenbaum’s provocative concurrence, the familiar arguments put forward by Judges Martin and Jill Pryor in their dissent are almost a let-down — persuasive as they are.

The judges distinguish the transmission of cell site location data in Davis’s case from the voluntary disclosure of phone numbers in Smith on the ground that a cell phone transmits data even when the user is not making a call. Even more important, they argue, the amount and sensitivity of the information disclosed to third parties through technologies like cell phones “has increased by orders of magnitude since the Supreme Court decided Miller and Smith.” Like Judge Rosenbaum, the dissent highlights several cases in which limited exposures to third parties did not nullify privacy interests; Judges Martin and Pryor conclude that the third party doctrine “has its limits,” and that “[s]ixty-seven days of near-constant location tracking of a cell phone — a technological feat impossible to imagine when Miller and Smith were decided — is an application of the doctrine that goes too far.”

Having decided that the third party doctrine does not control the outcome of the case, the dissenting judges apply the “reasonable expectation of privacy” test. They find that Davis had such an expectation and that society was prepared to accept it as reasonable, largely because long-term location tracking can reveal “quintessentially private” information, as explicated by five justices in Jones.

Most people likely would agree that long-term location tracking and analysis, using modern technology, can reveal highly sensitive information; the harder question is whether Smith can be distinguished or whether the Supreme Court must overrule it. The dissent missed an opportunity to explain how the government’s ability to derive personal information from location data affects the third party doctrine’s applicability. Even if it were true that cell phone users “voluntarily” disclose their location, it strains credulity to argue that, simply by virtue of putting a cell phone in their pocket, they voluntarily disclose “a wealth of detail about their familial, political, professional, religious, and sexual associations.” Yet this is exactly the information the government can deduce, as characterized by Justice Sotomayor in her Jones concurrence. This disconnect between what the subject of the search may have voluntarily disclosed and what the search itself may reveal should call into question whether cell site data searches implicate the third party doctrine as articulated in Smith.