Published by The Lawfare Institute
in Cooperation With
As Wells noted Wednesday, a three-judge panel of the Fourth Circuit handed down an important decision in United States v. Graham. The panel split 2-1, with the majority holding that the warrantless, extended, accessing of the two defendants' cell-site data amounted to an unconstitutional search under the Fourth Amendment. However, the court refused to suppress the collected information because of the Fourth Amendment's "good faith" exception—and thus affirmed both the defendants' convictions of various charges associated with a series of armed robberies.
Below, we overview the Fourth Amendment features of the majority, concurring, and dissenting opinions in Graham. (To be clear: The ruling spans 134 pages and touches many issues central to the two defendants' appeals apart from the constitutional question—only the latter of which is addressed in the precis below.)
So far as the Fourth Amendment issue is concerned, the facts are straightforward. Defendant-Appellants Aaron Graham and Eric Jordan were charged as co-conspirators in a string of armed robberies committed in the Baltimore area over a three-week period in early 2011. Police caught Graham during the last robbery, and recognized certain similarities with earlier ones. As a result, the government obtained a search warrant for Graham’s residence; the search uncovered, among other things, two cell phones. Investigators then obtained two court orders directed at Sprint/Nextel (the provider of the two cell phones), which commanded the company to disclose cell site location information (CLSI) associated with the defendants' devices. The orders were issued pursuant to the Stored Communications Act (SCA), which, as the panel would later explain in its opinion, “provid[es] an avenue for law enforcement entities to compel a provider of electronic communication services to disclose the contents and records of electronic communications.”
The SCA orders sought quite a lot of geolocation data. The first implicated fourteen days’ worth of cell site records. But the second comprised a much broader timeframe: from July 1, 2010 through February 6, 2011. In total, this amounted to 221 days’ worth of CLSI, which itself yielded an impressive 29,659 location data points for Graham and 28,410 for Jordan, enough to provide a “reasonably detailed account of their movements” during the intervals covered by the disclosure orders.
After the district court rejected the defendants' motion to suppress, prosecutors ultimately used the CLSI at trial to establish the location of Graham and Jordan at various times before and after most of the robberies—and, along with other evidence, to suggest both defendants' culpability. A jury convicted both defendants.
Senior Judge Andre Davis' Majority Opinion
Senior Judge Andre Davis, joined by Judge Stephanie Thacker (who also wrote a separate concurrence), held that though the government’s warrantless procurement of extended CSLI was an unreasonable search, the evidence was properly admitted nevertheless, because the government had acted in good-faith reliance on the SCA and the two court orders.
Judge Davis began his opinion by explaining the technological foundations of CSLI. Cellphone network providers maintain cell sites or “base stations” that emit signals that cell phones connect to. When a cellphone is used, such as in the case of sending a text or making a call, the phone connects to the station that has the strongest signal, which is usually the closest station. Moreover, phones can jump and connect to other stations as the phone moves in and out of a particular station’s coverage. All of this means that CLSI can “approximate the whereabouts of the cell phone at the particular points in time in which transmissions are made” based on analyses of which station(s) the phone has connected to. Crucially, Judge Davis and Thacker regarded this information as allowing the government to “discover the private activities and personal habits of the user.”
Specifically, the majority emphasized that there is a “recognized privacy interest in the comprehensive accounts of one’s movements . . . and location” and that the extended nature of the CSLI collection and inspection, which allowed the government to account for the defendants’ movements for nearly two-thirds of a year, was unreasonable. For support, Judge Davis relied on three Supreme Court cases: United States v. Knotts, where the Supreme Court upheld a search that tracked a suspect’s location, but only because the tracking was of a limited nature and duration and stopped when the suspect entered his home; United States v. Karo, where the Court held that the tracking of a suspect inside an area where the suspect has a reasonable expectation of privacy, such as his home, is a violation of the Fourth Amendment; and Kyllo v. United States, where the government used technological means not available to the general public (a thermal imaging camera) to ascertain a suspect’s location inside his home. These three cases—as well as the majority and concuring opinions in the much-discussed GPS tracking case, United States v. Jones—form the backbone of the majority's holding. Because CLSI would allow the government retroactively to pinpoint a person at his or her "home and other private locations at specific points in time;" and because CLSI can only be collected and inspected through technological means not in use by the general public, the majority determined that the warrantless search here ran afoul of the Fourth Amendment.
Though it placed great weight on the extended nature of the CLSI collection and identification, the majority specifically refused to identify the tipping point for when tracking information becomes unlawfully "extended." Rather, the majority simply held only that “the government engages in a Fourth Amendment search when it seeks to examine historical CSLI pertaining to an extended time period like 14 or 221 days.”
Second, and related to the first point, the court rejected the application of the third-party doctrine, which holds that a person does not have a reasonable expectation of privacy in information voluntarily conveyed to a third-party. The majority stressed that the defendants did not voluntarily convey their CSLI to Sprint/Nextel. For that reason, well-known cases like United States v. Miller and Smith v. Maryland were inapplicable. The syllogism is simple, in Davis' view: Because the defendants did not know Spring/Nextel was collecting their CSLI, they could not voluntarily convey that information to the company.
Third, in rejecting the defense's objections, the district court below had distinguished the three cases mentioned above, insisting that in each, the disputed surveillance was done in “real-time” and of “greater precision and continuity” than historical CSLI. But the majority thought those differences beside the point; had the defendants used their phones constantly and the government monitored the CSLI in real time, the government would have had “information approaching that of GPS.” Any differences between the precision and continuity of location information obtained in this case and the three Supreme Court cases were due to differences in the defendants’ conduct—not the government conduct constrained by the Fourth Amendment.
Having determined there was a search, the court nonetheless refused to apply the exclusionary rule—here because of the government’s good faith reliance on the SCA and the two court orders permitting the discovery of CSLI. Judge Davis explained that the so-called "good faith" exception applies if even an unlawful search was conducted pursuant to “an enacted statute, unless that statute is clearly unconstitutional,” or “a search warrant or other court order issued by a neutral magistrate, unless issuance of the order is clearly defective.” Here the statute in question (the SCA) wasn't clearly unconstitutional; nor were the orders directed at Sprint/Nextel obviously out of bounds. Accordingly the majority affirmed the district court's denial of the defendants' motion to suppress.
Judge Stephanie Thacker’s Concurrence
Judge Thacker’s brief opinion adds little in the way of additional legal analysis, but instead emphasizes “concern about the erosion of privacy in this era of rapid technological development.” Her three-and-a-half page concurrence thust mostly consists of quotations from the Supreme Court’s Riley decision, and highlights the Court’s acknowledgement that the government’s use of cellular data poses particular dangers to individual privacy. She concludes by touting the panel's decision, which in her view "continues a time-honored American tradition --- obtaining a warrant is the rule, not the exception.”
Judge Diana Gribbon Motz’s partial dissent and concurrence in the judgment
Judge Motz’s opinion is, at least formally, a concurrence in the case's outcome. Like the majority, she finds no reason to apply the exclusionary rule. But on the primary question of the case—namely the constitutionality of the warrantless use of a cell phone user’s cell-site data—Judge Motz breaks with her colleagues. Her opinion thus closely tracks the opinions of the 5th and 11th circuits, both of which applied the third-party doctrine and found no Fourth Amendment problem with warrantless government access to cell-site data.
First, Judge Motz addresses, and distinguishes, each of the majority’s central precedents. She points out that in Karo the Drug Enforcement Agency surveilled the defendant directly, by placing a beeper which emitted signal from inside the defendant’s home. Similarly, in Kyllo and United States v. Jones, it is again the Government itself that monitored the defendant (through the use of thermal imaging and an installed GPS tracker, respectively). Judge Motz readily acknowledges that these cases show that when the government conducts invasive searches, it may very well be invading a reasonable expectation of privacy. But she emphasizes that this conclusion, and thus these cases, say nothing at all about the prosecutions of Graham and Jordan. Here, after all, the question is whether a reasonable expectation of privacy remains when data voluntarily has been transmitted to Sprint/Nextel.
The latter question, argues Judge Motz, is definitively answered by the “third-party doctrine” cases beginning with Smith v. Maryland. These cases hold that once a defendant had disclosed information to a third party, any prior expectation of privacy in that information is no longer recognized as “reasonable” and no longer protected.
The majority, of course, attempts to distinguish Smith by arguing that the provision of cell data to companies was neither “voluntary” nor truly “conveyed” at all. Judge Motz rejects this distinction: Just as in Smith, defendants disclosed their information to a company’s equipment “in the ordinary course of business.” More importantly, the majority’s insistence the users aren’t truly “conveying” their information to companies—or aren’t doing so “voluntarily”—defies common sense. Reasonable users know that location and proximity to cell towers matters, and demonstrate this every time they move to get better service. While “in the moment,” many users may not consciously recognize that they are conveying locational information to their service providers, this is the obvious implication of using a cell phone. Users choose to purchase and use their phones anyway.
Furthermore, a lack of an “in the moment” active decision to convey specific information is a standard feature of many settled third-party doctrine cases, including government use of the time of credit card use or incoming phone calls. So the requirement that the provision of information be “voluntary” simply can’t mean “contemporaneous recognition of every detail an individual conveys.” The majority, Judge Motz argues, has invented a rigorous “active submittal” requirement that is well beyond the existing “voluntary conveyance" to a third-party and that directly contradicts decades of precedent.
Judge Motz further points out that the majority’s emphasis on the length of the period from which data was drawn—and related reliance on concurrences from Jones—is based on an analytic mistake: In Jones, the length of GPS surveillance mattered because the surveillance was taking place in public, and so the length of the surveillance was what transformed public information to private information. With locational cell data, there is no question that such data would have been private—if not for the fact that it was voluntarily conveyed to a third party. And from the beginning, the Supreme Court that decided Smith understood the “third party doctrine” to be unaffected by the quantity of data, no matter how revealing that quantity.
Judge Motz concludes by noting that “although the Court formulated the third party doctrine as an articulation of the reasonable-expectation-of-privacy inquiry, it increasingly feels like an exception.” This may be undesirable and unfortunate. Nevertheless the precedent remains binding on the lower courts, Judge Motz argues, and it is the job of the Supreme Court or Congress to restore what may ultimately be the “proper balance between technology and privacy.”