Today, the Eleventh Circuit rejected the exceedingly common law enforcement practice of warrantlessly tracking suspects’ physical location using cell phone tower data. The opinion, United States v. Davis, is both welcome and overdue. Defendants who have and will be physically tracked without a warrant have new legal support to challenge that surveillance. Additionally, because the case involved stored cell site data, Davis undermines the government’s legal arguments that other warrantless “metadata” collection practices are constitutional.
Mobile phones leave digital footprints of their owner’s travels through space and time. Whenever these handsets are turned on, even when they are not in use, they send signals to the nearest cell towers so that the communications network system knows where to route a call should one come in. These inevitable “pings” can be used to identify the tower or towers closest to the targeted handset. Since people usually do not share handsets, and since most individuals carry their cell phones with them at all times, tracking pings is a highly effective way of tracking any particular individual’s physical location around the clock.
Not only is cell tracking an effective methodology for determining where an individual suspect or suspects have been or is going, it is also useful for determining the identity of unknown individuals who are near a particular place at a particular time. The wireless provider can identify to law enforcement all the handsets that were connected to any particular tower at any time. This technique is called a “tower dump” and providers charge about $75 for the information, which can disclose the location of any number of individuals. Law enforcement can discover who is or was connected to the towers nearest to a crime or political protest, for example.
How common is cell phone tracking? We don’t have exact numbers, but it is exceedingly common. For example, in 2011, cell phone companies reportedly received over 1.3 million requests for text messages, caller locations and other information.
One of the reasons we don’t have better numbers is that, even in the law enforcement context, the Department of Justice, sometimes with the unwitting cooperation of magistrate judges and sometimes with the active collusion of the judiciary, takes steps that hide the numbers from the public.
Cell tracking applications, and the orders approving those applications, are sealed, and often never get unsealed. In 2009, U.S. Magistrate Judge Stephen Smith of the Southern District of Texas analyzed the 4,234 electronic surveillance orders issued in his Houston courthouse between 1995 and 2007. Judge Smith found that 91.8% of them remain sealed as of 2009. In his paper “Kudzu in the Courthouse: Judgments Made in the Shade”, Judge Smith argued that the secrecy of electronic surveillance orders means that “each magistrate judge has effectively become a law unto himself.”
Against this backdrop, Brian Owsley, an outgoing United States magistrate judge, from the same Southern District of Texas decided that one of his final judicial acts would be to unseal more than 100 of his own judicial orders involving digital surveillance that he himself had sealed at the government’s request. Despite Owsley’s effort to clear his kudzu from the courthouse, a US district judge vacated Owsley’s order and resealed all Owsley’s orders and then sealed his own order. In response to this Kafkaesque turn of events, the media company Dow Jones has filed a motion with the US District Court in the Southern District of Texas asking it to unseal all such documents and to make them available publicly online.
More transparency is important because Americans’ privacy rights are regularly decided in secret, and those secret decisions, as the Eleventh Circuit held today, are violating peoples’ Fourth Amendment rights.
The issue of cell tracking came to the forefront in August 2005. That’s when Magistrate Judge Orenstein in the Eastern District of New York published the first decision on cell tracking. He publicly denied a government request to use cell phones as tracking devices without probable cause and explained his reasoning. In doing so, Judge Orenstein revealed that the Justice Department had routinely been using a baseless legal argument to get secret location tracking authorizations from courts, probably for many years. The government had been using provisions of the Stored Communications Act to get real time cell tower data as an end run around the Rule 41 warrant requirement.
Magistrate Judge Smith followed up on Judge Orenstein’s work with cell tracking opinions of his own, as well as an opinion denying the government access the numbers you dial after your call has been connected—for example, the digits you dial to use to communicate with your bank’s automated service, order a prescription, or vote on American Idol—without a warrant.
Since 2005, the majority of magistrate judges writing opinions on the issue have rejected the DOJ view and require a search warrant for real time cell tracking. Still, there are many districts where magistrates accept less than a warrant for cell tracking, especially when the government gets data from only a single cell tower, rather than the more accurate triangulation. That appears to be because two 1970s era cases involving primitive beepers that assisted officers following suspects distinguished between following the conspirators on a public road (no warrant needed) and determining that a barrel of precursor chemicals was located in a private home (warrant required).
As of today, no judge in the Eleventh Circuit will be doing that anymore. The Davis case explicitly rejects the government’s argument that single tower data, because it is “less precise”, may be obtained without probable cause.
We are not sure why this should be significant. We do not doubt that there may be a difference in precision, but that is not to say that the difference in precision has constitutional significance. While it is perhaps possible that information could be sufficiently vague as to escape the zone of reasonable expectation of privacy, that does not appear to be the case here. The prosecutor at trial stressed how the cell phone use of the defendant established that he was near each of six crime scenes. While committing a crime is certainly not within a legitimate expectation of privacy, if the cell site location data could place him near those scenes, it could place him near any other scene. There is a reasonable privacy interest in being near the home of a lover, or a dispensary of medication, or a place of worship, or a house of ill repute. Again, we do not see the factual distinction as taking Davis’s location outside his expectation of privacy. That information obtained by an invasion of privacy may not be entirely precise does not change the calculus as to whether obtaining it was in fact an invasion of privacy.
The Davis court didn’t even both to cite those beeper cases. Instead, it relied on the more recent Supreme Court case of United States v. Jones (2012), where accuracy was not at issue. In Jones, the Supreme Court considered long-term recording and aggregation of location information from a GPS device police warrantlessly installed on a suspect’s car. The government had argued that use of the device was not a search because it only revealed information the defendant already disclosed to others—the location of his vehicle on the public roads.
The Supreme Court unanimously rejected this argument, though judges employed different reasoning. The majority held that attaching a GPS-tracking device to a vehicle and using the device to monitor the car’s movements over a period of twenty-eight days was a Fourth Amendment “search” because it interfered with the defendant’s property interest in the car. In other words, it used a trespass test to determine constitutionality. Justice Sotomayor joined that majority opinion. However, in two concurring opinions, Sotomayor and four other Justices also concluded that the surveillance “impinge[d] on expectations of privacy.”
As Sotomayor explained:
GPS monitoring generates a precise, comprehensive record of a person’s public movements that reflects a wealth of detail about her familial, political, professional, religious, and sexual associations. The Government can store such records and efficiently mine them for information years into the future. And because GPS monitoring is cheap in comparison to conventional surveillance techniques and, by design, proceeds surreptitiously, it evades the ordinary checks that constrain abusive law enforcement practices: limited police resources and community hostility.
Based on this reasoning, the five concurring justices would have found that the GPS tracking violated the “reasonable expectation of privacy” test established in Katz v. United States.
The Eleventh Circuit looked at the SCOTUS concurrences in Jones, found that the Katz expectation of privacy test applied, and said that the cell tracking violated that test for at least the same reasons as GPS tracking. In fact, Davis’ constitutional case was stronger than that in Jones because the car can only move on the public roads, while a cell phone regularly goes in private places likes homes and doctors’ offices. Additionally, while the defendants in Jones may have known that police could follow their car, there’s no reason to think that Davis or anyone else expects they have surrendered their location privacy by exposing his cell site location to his service provider when he placed the call. For that proposition, the Eleventh Circuit cited a Third Circuit case regarding historical cell site data in which the Electronic Frontier Foundation served as amicus (In that capacity, I participated in the case, which was ably argued before the Third Circuit by Kevin Bankston).
The Third Circuit went on to observe that “a cell phone customer has not ‘voluntarily’ shared his location information with a cellular provider in any meaningful way.” That circuit further noted that “it is unlikely that cell phone customers are aware that their cell phone providers collect and store historical location information.” 620 F.3d at 317 (emphasis added). Therefore, as the Third Circuit concluded, “when a cell phone user makes a call, the only information that is voluntarily and knowingly conveyed to the phone company is the number that is dialed, and there is no indication to the user that making that call will also locate the caller.” Id. Even more persuasively, “when a cell phone user receives a call, he hasn’t voluntarily exposed anything at all.” Id. at 317–18.
In deciding Davis, the Eleventh Circuit (partially) took my side in the debate I had with Professor Orin Kerr here at Just Security. Kerr and I were discussing whether the NSA’s collection of all Americans’ phone call detail records violates the Fourth Amendment. I argued, among other things, that it did under the Jones concurrences. Kerr replied that the Jones concurrences were not law and therefore could not support my purely aspirational position:
[G]iven that we’re sticking to current law, we can reject arguments based on the concurring opinions for a simple reason: There was a clear majority opinion in Jones, and it expressly declined to adopt the novel theories suggested by the concurring opinions. Indeed, the majority opinion scoffed at the mosaic approach of Justice Alito’s concurrence on the ground that it was “novel” and its method derived from “no precedent.” So as interesting as the mosaic theory is, presently it is not the law. See, e.g., State v. LeMasters, 2013 WL 3463219 (Ohio App. July 8, 2013) (“While [the appellant] spends a great amount of time in his brief quoting and referencing the concurring opinions in Jones that suggest that the Fourth Amendment should be stretched to include other privacy rights, we are bound only by the majority opinion of the court, rather than questions raised and suggestions made within the dicta of concurring opinions.”).
To the contrary, the Eleventh Circuit said today in Davis that the Jones concurrences—while not conclusive—confirmed that the reasonable expectation of privacy test was alive and well, and that it applied to the transmission of this digital information. Indeed, Scalia writing for the Jones majority expressly noted that “[s]ituations involving merely the transmission of electronic signals without trespass would remain subject to [the] Katz [privacy] analysis.”
How might the Davis case impact the broader question that Kerr and I were discussing, the constitutionality of the NSA’s phone metadata collection? To recap, in June, Americans learned that the United States National Security Agency (NSA) collects records and maintains a database of Americans international and domestic phone call records. The Bush and Obama Administrations have asserted that the Fourth Amendment does not protect Americans from mass collection and/or analysis of these phone records, and the Foreign Intelligence Surveillance Court (FISC), which has authorized the phone records collection program since 2006, agrees.
The program’s constitutionality depends on the 1979 case of Smith v. Maryland, where the Supreme Court found no constitutionally-protected reasonable expectation of privacy in the phone numbers a suspect dialed. Key to Smith was the argument that the suspect had knowingly disclosed that information to the phone company and thus lacked a reasonable expectation of privacy in it. Additionally, Smith built on the 1976 case of United States v. Miller, which held that a person does not have Fourth Amendment rights in their bank records because that information is the bank’s records of how the account was used and not the customer’s data.
In Davis, the historical cell site information kept by the cellular provider is analogous to the bank records in Miller. Yet, the Eleventh Circuit did not follow Miller. Rather, it found the defendant had an expectation of privacy despite the fact that the cell data was also the company’s business record. “While the government’s position is not without persuasive force, it does not ultimately prevail” for the reasons set forth by the Third Circuit above. Further, as I argued in my debate with Orin, people do not know that merely by using a phone, the provider can collect detailed calling records, and even if they do know that, they do not know or assume the risk that the government will amalgamate that information. As Justice Sotomayor said regarding Smith and Miller in Jones:
It may be necessary to reconsider the premise that an individual has no reasonable expectation of privacy in information voluntarily disclosed to third parties. E.g., Smith, 442 U. S., at 742; United States v. Miller, 425 U. S. 435, 443 (1976). This approach is ill suited to the digital age, in which people reveal a great deal of information about themselves to third parties in the course of carrying out mundane tasks.
Other bulk “metadata” collection programs have included Internet transactional records, financial transactions and more. The similarities between these categories of information, consumer expectations and knowledge, and the phone numbers dialed in Smith is even more attenuated. Davis, by refusing to apply Smith and Miller in a stored phone records case, has taken a giant step away from the legal justification propping up many of the government’s targeted and bulk metadata collection practices.
Once again, we see courts pressed into protecting constitutional rights where Congress has abrogated its responsibility to do so. Location privacy is a notable example. For some time now, the United States Congress has considered reforming the Electronic Communications Privacy Act (ECPA), the law under which the DOJ has been demanding warrantless access to cell data. One of the proposals on the table is to require a warrant for cell site data that can be used for location tracking. A counterproposal is to accept a warrant requirement for prospective tracking, but to allow historical tracking with simply a 2703(d) specific and articulable facts order. Neither of these proposals has moved forward in any meaningful way.
It is still important for Congress to act, even after the Eleventh Circuit has weighed in. The public won’t readily know whether that decision makes a difference in any of the other Circuits since, as I said before, those applications are orders are routinely sealed and rarely unsealed, apparently even when the magistrate who issued them thinks that would be appropriate.
However, the Eleventh Circuit opinion can and should inform and improve the congressional debate about ECPA reform. That is because cell tracking proposals which treat single tower or historical information as less invasive than real time or multiple tower information extend a privacy misconception well past its Sell By date. Single tower and historical information are seriously privacy invasive. As the Third and Eleventh Circuit have realized, as a general matter treating historical data differently than prospective surveillance does not make sense.
Moreover, the bases for disparate treatment are increasingly obsolute. In the context of location tracking, if historical information is in a particular case less revealing than prospective surveillance, that is only because today cellular companies generally keep only cell site initiation and termination information. These companies could readily start keeping tower pings and sector information that would make historical data just as accurate as prospective tracking for any, and every, cell phone customer. There is no time limit for historical surveillance other than that imposed by the providers’ business needs, which increasingly, given the interest in targeted marketing and the price of storage, trend towards forever.
Further, location accuracy based on mobile phone data depends on the density of cell towers and the type of information collected. The trend is towards more towers, and thus more specific location data. Providers are increasingly using tiny cell towers — called microcells, femtocells, picocells or the like – to improve their coverage area. These microcells often broadcast a signal no further than a subscriber’s home. So, the accuracy of even single cell site location data is increasing depending on the kind of tower with which the phone connects. Indeed, as early as 2006, the Department of Justice used single tower data to prove in criminal cases that a particular individual was inside a home. Today, seven years later, with the proliferation of towers, including microcells, single tower data can be far more revealing more of the time.
In sum, the cell site location privacy has been almost nine years in the making, from Judge Orenstein and Judge Smith’s initiation of the “Magistrate’s Revolt” in 2005 through today’s opinion in United States v. Davis. And we haven’t fully arrived yet at the right answer. It is, as Nelson Mandela once said, a long road to freedom.