The en banc Eleventh Circuit has ruled that historical cell-site records are not protected by the Fourth Amendment under the third-party doctrine. The case, United States v. Davis, also adds an alternative holding that is even more important: Even if cell-site records were protected, the en banc court holds, accessing them would trigger only minimal Fourth Amendment concerns and would not require a warrant or probable cause. My bottom line: I agree with court’s ruling that the third-party doctrine applies and there was no search, but I think the alternative holding is puzzling, inconsistent with precedent, and unnecessary. But stay tuned. It’s a long shot, but that second alternative holding might end up drawing Supreme Court review of both holdings.

This post will go through the majority’s reasoning and then offer my thoughts. There’s also lots of interesting stuff in the concurring and dissenting opinions, but in the interests of space and time I’ll stick with the majority opinion.

I. The Facts

I summarized the facts in an earlier post as follows:

Davis participated in seven armed robberies. He helped rob a Wendy’s restaurant, a Walgreen’s pharmacy, a beauty salon, and a bunch of other places. The evidence against Davis was pretty strong. At trial, eyewitnesses testified that Davis was a participant in some of the armed robberies. At one, armed with a gun, he pointed his gun at a victim’s head, pushed a 77-year old woman to the ground, and robbed them. At another, Davis was seen firing shots at an eyewitness who tried to write down the license plate number of the getaway car. Davis’s DNA was recovered from the getaway car. Video surveillance footage at some of the victim stores showed him committing the crime, as well. So far, it’s a pretty old-fashioned robbery case.
As part of its evidence, however, the government also obtained cell-site records of suspected members of the groups that committed the robberies. They obtained the records under the federal Stored Communications Act, which requires the government to get court orders based on “specific and articulable facts” but not probable cause. The records existed because cellular service providers record the cell tower that a phone was connected to when a cell call was placed and when it was ended. The records generally tell you the neighborhood the phone was in when the call was placed and ended, as the cell tower used is usually (but not always) the one closes to the phone.
The cell-site records obtained in Davis showed that during six of the seven robberies, the phones belonging to members of the alleged conspiracy (Davis included) placed and received calls in the neighborhoods where the robberies occurred at around the time of the crimes. At trial, the prosecutor used the fact that Davis’s cell phone had placed/received calls from co-conspirators at the time of the robberies in the area around the victimized stores to indicate that he was one of the robbers.

The panel opinion had held that the cell site records were protected by the Fourth Amendment.

II. The En Banc Ruling

The en banc Court disagreed. The opinion divides into two parts. In the first part, Judge Hull, writing for the majority, concludes that historical cell site records are not protected because they are governed by the third-party doctrine of Smith v. Maryland:

Based on the SCA and governing Supreme Court precedent, we too conclude the government’s obtaining a § 2703(d) court order for the production of MetroPCS’s business records did not violate the Fourth Amendment.
For starters, like the bank customer in Miller and the phone customer in Smith, Davis can assert neither ownership nor possession of the third-party’s business records he sought to suppress. Instead, those cell tower records were created by MetroPCS, stored on its own premises, and subject to its control. Cell tower location records do not contain private communications of the subscriber. This type of non-content evidence, lawfully created by a third-party telephone company for legitimate business purposes, does not belong to Davis, even if it concerns him. Like the security camera surveillance images introduced into evidence at his trial, MetroPCS’s cell tower records were not Davis’s to withhold. Those surveillance camera images show Davis’s location at the precise location of the robbery, which is far more than MetroPCS’s cell tower location records show.
More importantly, like the bank customer in Miller and the phone customer in Smith, Davis has no subjective or objective reasonable expectation of privacy[.] . . . Even if Davis had a subjective expectation of privacy, his expectation of privacy, viewed objectively, is not justifiable or reasonable under the particular circumstances of this case. The unreasonableness in society’s eyes dooms Davis’s position under Katz. In Smith, the Supreme Court presumed that phone users knew of uncontroverted and publicly available facts about technologies and practices that the phone company used to connect calls, document charges, and assist in legitimate law-enforcement investigations. See 442 U.S. at 742-43, 99 S.Ct. at 2581. Cell towers and related records are used for all three of those purposes. We find no reason to conclude that cell phone users lack facts about the functions of cell towers or about telephone providers’ recording cell tower usage.
Smith’s methodology should not be set aside just because cell tower records may also be used to decipher the approximate location of the user at the time of the call. Indeed, the toll records for the stationary telephones at issue in Smith included location data far more precise than the historical cell site location records here, because the phone lines at issue in Smith corresponded to stationary landlines at known physical addresses. At the time of Smith, telephone records necessarily showed exactly where the user was—his home—at the time of the call, as the user’s telephone number was tied to a precise address. And the number dialed was also tied to a precise address, revealing if the user called a friend, a business, a hotel, a doctor, or a gambling parlor.
Admittedly, the landscape of technology has changed in the years since these binding decisions in Miller and Smith were issued. But their holdings did not turn on assumptions about the absence of technological change. To the contrary, the dispute in Smith, for example, arose in large degree due to the technological advance from call connections by telephone operators to electronic switching, which enabled the electronic data collection of telephone numbers dialed from within a home. See 442 U.S. at 744-45, 99 S. Ct. at 2582-83. The advent of mobile phones introduced calls wirelessly connected through identified cell towers. This cell tower method of call connecting does not require “a different constitutional result” just “because the telephone company has decided to automate” wirelessly and to collect the location of the company’s own cell tower that connected the calls. See id. at 744-45, 99 S. Ct. at 2582. Further, MetroPCS’s cell tower location information was not continuous; it was generated only when Davis was making or receiving calls on his phone. The longstanding third-party doctrine plainly controls the disposition of this case.
The use of cell phones is ubiquitous now and some citizens may want to stop telephone companies from compiling cell tower location data or from producing it to the government. Davis and amici advance thoughtful arguments for changing the underlying and prevailing law; but these proposals should be directed to Congress and the state legislatures rather than to the federal courts.

According to Judge Hull, you can’t read the Jones concurrences as overturning the third-party doctrine, either. For example:

Nothing Justice Alito says [in his Jones concurrence] contravenes the third-party doctrine. His concurring opinion does not question, or even cite, Smith, Miller, or the third-party doctrine in any way. The opinion never uses the words “third party” or “third-party doctrine.” It would be a profound change in jurisprudence to say Justice Alito was questioning, much less casting aside, the third-party doctrine without even mentioning the doctrine.
. . . If anything, Justice Alito’s concurrence, joined by three others, suggests that a legislative solution is needed. Id. at __, 132 S. Ct. at 964 (“In circumstances involving dramatic technological change, the best solution to privacy concerns may be legislative. A legislative body is well situated to gauge changing public attitudes, to draw detailed lines, and to balance privacy and public safety in a comprehensive way.” (citation omitted)). At present, the SCA is that solution.

Nor does the time-period of the records collection matter under what seems to be a mosaic theory claim:

[R]easonable expectations of privacy under the Fourth Amendment do not turn on the quantity of non-content information MetroPCS collected in its historical cell tower location records. The § 2703(d) order covered 67 days of MetroPCS records. In his brief before this en banc Court, Davis argued that the length of the records covered by the order made the production an unconstitutional “search.” But at oral argument Davis’s counsel firmly contended that even one day of historical cell tower location information would require a search warrant supported by probable cause. Counsel’s response at oral argument is faithful to Davis’s broader claim, but misapprehends the governing law. Because Davis has no reasonable expectation of privacy in the type of non-content data collected in MetroPCS’s historical cell tower records, neither one day nor 67 days of such records, produced by court order, violate the Fourth Amendment.

Nor can the Supreme Court’s decision in Riley v. California enable a mosaic theory:

[T]he Supreme Court in Riley made a special point of stressing that the facts before it “do not implicate the question whether the collection or inspection of aggregated digital information amounts to a search under other circumstances.” Id. at __, 134 S. Ct. at 2489 n.1. It is not helpful to lump together doctrinally unrelated cases that happen to involve similar modern technology.

Next the court adopts an alternative holding. Even if the cell site records are protected under the Fourth Amendment, they are only minimally protected, and no warrant would be required to obtain such records. This is new and unusual ground, so I’ll reprint the entirety of the analysis:

[T]he reasonableness of a search or seizure is evaluated “under traditional standards of reasonableness by assessing, on the one hand, the degree to which it intrudes upon an individual’s privacy and, on the other, the degree to which it is needed for the promotion of legitimate governmental interests.” Wyoming v. Houghton, 526 U.S. 295, 300, 119 S. Ct. 1297, 1300 (1999). In addition, “there is a strong presumption of constitutionality due to an Act of Congress, especially when it turns on what is ‘reasonable’” within the meaning of the Fourth Amendment. United States v. Watson, 423 U.S. 411, 416, 96 S. Ct. 820, 824 (1976) (internal quotation marks omitted).
This traditional Fourth Amendment analysis supports the reasonableness of the § 2703(d) order in this particular case. As outlined above, Davis had no reasonable expectation of privacy in business records made, kept, and owned by MetroPCS. At most, Davis would be able to assert only a diminished expectation of privacy in MetroPCS’s records. See King, 569 U.S. at __, 133 S. Ct. at 1969 (identifying “diminished expectations of privacy” as one of the factors that “may render a warrantless search or seizure reasonable”) (quotation marks omitted).
Further, any intrusion on Davis’s alleged privacy expectation, arising out of MetroPCS’s production of its own records pursuant to a § 2703(d) order, was minimal for several reasons. First, there was no overhearing or recording of any conversations. Second, there is no GPS real-time tracking of precise movements of a person or vehicle. Even in an urban area, MetroPCS’s records do not show, and the examiner cannot pinpoint, the location of the cell user. Ironically, Davis was using old technology and not the new technology of a smartphone equipped with a GPS real-time, precise tracking device itself.
Third, a § 2703(d) court order functions as a judicial subpoena, but one which incorporates additional privacy protections that keep any intrusion minimal. The SCA guards against the improper acquisition or use of any personal information theoretically discoverable from such records. See King, 569 U.S. at __, 133 S. Ct. at 1979-80. Under § 2703(d), investigative authorities may not request such customer-related records merely to satisfy prurient or otherwise insubstantial governmental interests. Instead, a neutral and detached magistrate must find, based on “specific and articulable facts,” that there are “reasonable grounds to believe” that the requested records are “relevant and material to an ongoing criminal investigation.” Such protections are sufficient to satisfy “the primary purpose of the Fourth Amendment,” which is “to prevent arbitrary invasions of privacy.” Brock v. Emerson Elec. Co., Elec. & Space Div., 834 F.2d 994, 996 (11th Cir. 1987); see, e.g., Terry v. Ohio, 392 U.S. 1, 21 n.18, 88 S. Ct. 1868, 1880 n.18 (1968) (explaining that the “demand for specificity in the information upon which police action is predicated is the central teaching of this Court’s Fourth Amendment jurisprudence”).
The stored telephone records produced in this case, and in many other criminal cases, serve compelling governmental interests. Historical cell tower location records are routinely used to investigate the full gamut of state and federal crimes, including child abductions, bombings, kidnappings, murders, robberies, sex offenses, and terrorism-related offenses. See, e.g., United States v. Troya, 733 F.3d 1125, 1136 (11th Cir. 2013) (“quadruple homicide” involving the “ganglandstyle murder of two children”); United States v. Mondestin, 535 F. App’x 819, 821 (11th Cir. 2013) (unpublished) (per curiam) (armed robbery); United States v. Sanders, 708 F.3d 976, 982-83 (7th Cir. 2013) (kidnapping). Such evidence is particularly valuable during the early stages of an investigation, when the police lack probable cause and are confronted with multiple suspects. In such cases, § 2703(d) orders—like other forms of compulsory process not subject to the search warrant procedure—help to build probable cause against the guilty, deflect suspicion from the innocent, aid in the search for truth, and judiciously allocate scarce investigative resources.
The societal interest in promptly apprehending criminals and preventing them from committing future offenses is “compelling.” See United States v. Salerno, 481 U.S. 739, 750-51, 107 S. Ct. 2095, 2103 (1987). But so too is the societal interest in vindicating the rights of innocent suspects. See King, 569 U.S. at __, 133 S. Ct. at 1974. Both interests are heavily implicated when the government seeks to compel the production of evidence “relevant and material to an ongoing criminal investigation.” 18 U.S.C. § 2703(d). Cell tower location records have the capacity to tell the police investigators that an individual suspect was in the general vicinity of the crime scene or far away in another city or state.
In sum, a traditional balancing of interests amply supports the reasonableness of the § 2703(d) order at issue here. Davis had at most a diminished expectation of privacy in business records made, kept, and owned by MetroPCS; the production of those records did not entail a serious invasion of any such privacy interest, particularly in light of the privacy-protecting provisions of the SCA; the disclosure of such records pursuant to a court order authorized by Congress served substantial governmental interests; and, given the strong presumption of constitutionality applicable here, any residual doubts concerning the reasonableness of any arguable “search” should be resolved in favor of the government. Hence, the § 2703(d) order permitting government access to MetroPCS’s records comports with applicable Fourth Amendment principles and is not constitutionally unreasonable.

Finally, in a footnote, the court adopts yet another alternative theory — sort of an alternative to the alternative. Even if everything the court said up to now was wrong, the government would still win because the exclusionary rule doesn’t apply. Here’s the entirety of what the court says:

In the alternative, we hold that the prosecutors and officers here acted in good faith and therefore, under the well-established Leon exception, the district court’s denial of the motion to suppress did not constitute reversible error. See United States v. Leon, 468 U.S. 897, 919-21, 104 S. Ct. 3405, 3418-19 (1984).

III. My Take: A Mixed Bag

Her are a few thoughts:

1. I agree with the Eleventh Circuit on the first question, what is a ‘search.’ For a lower court bound by precedent, Smith governs. And I agree that the mosaic arguments shouldn’t go anywhere. Granted, I want there to be a circuit split to get the case up to the Supremes. That leaves me in an odd position: Although I think a judge should follow Smith, I also kinda want a lower court to not follow precedent in order to tee up the issue for the Supreme Court. Surely there are some judges who will be happy to oblige me at some point. Looking at you, Ninth Circuit. But alas, not yet. So far the Fifth and Eleventh Circuits are just following precedent. (The Eleventh Circuit does briefly mention the Florida Supreme Court’s Tracey decision, although Davis just says it’s about prospective cell site records rather than historical cell site records and otherwise ignores it. As I’ve blogged, I don’t think this is a real difference if the records are the same, but so far it is allowing the courts to disagree in practice without creating a formal split.)

2. On the other hand, I am really troubled by the en banc court’s alternative holding, which is by far more important. It’s a novel development of the law that cuts against a lot of practice and precedent. And it does so unnecessarily: It is not just an alternative holding, it is one of two alternative holdings, the latter of which the en banc court finds so obvious it gets only one sentence. So there was no reason to reach the reasonableness issue. And I think the Court was wrong to reach it, because it tries to answer a largely unanswerable question: In a world without the third party doctrine, what would happen to the reasonableness requirement? By saying that the third party doctrine applies but that alternatively the search was reasonable, the en banc court has to get into a sort of predictive exercise about what the Fourth Amendment reasonableness inquiry would look like in an alternative world where there was no third-party doctrine. I don’t think it was wise to reach that issue. It’s way beyond this case, waters down Fourth Amendment protections, and is only likely to confuse other courts.

Here’s the context. The usual rule in Fourth Amendment law is that a “search” in a criminal investigation requires a warrant or an exception to the warrant requirement. Recall the famous decision of Katz v. United States, where the Court first held that listening in on a phone booth was a search. The second holding of Katz was that the search required a warrant:

The Government . . . urges the creation of a new exception [to the warrant requirement] to cover this case. It argues that surveillance of a telephone booth should be exempted from the usual requirement of advance authorization by a magistrate upon a showing of probable cause. We cannot agree. Omission of such authorization “bypasses the safeguards provided by an objective predetermination of probable cause, and substitutes instead the far less reliable procedure of an after-the-event justification for the . . . search, too likely to be subtly influenced by the familiar shortcomings of hindsight judgment.” Beck v. Ohio, 379 U. S. 89, 96. And bypassing a neutral predetermination of the scope of a search leaves individuals secure from Fourth Amendment 359*359 violations “only in the discretion of the police.” Id., at 97.
These considerations do not vanish when the search in question is transferred from the setting of a home, an office, or a hotel room to that of a telephone booth. Wherever a man may be, he is entitled to know that he will remain free from unreasonable searches and seizures. The government agents here ignored “the procedure of antecedent justification . . . that is central to the Fourth Amendment,” a procedure that we hold to be a constitutional precondition of the kind of electronic surveillance involved in this case. Because the surveillance here failed to meet that condition, and because it led to the petitioner’s conviction, the judgment must be reversed.

The modern Supreme Court’s strongest pronouncement on this is probably Arizona v. Hicks, in which an officer was inside an apartment and picked up a turntable to see its serial number. In a dissent, Justice O’Connor argued that this was a small search that should be deemed reasonable without probable cause or an exception to the warrant requirement. The Supreme Court disagreed in an opinion by Justice Scalia:

JUSTICE O’CONNOR’S dissent suggests that we uphold the action here on the ground that it was a “cursory inspection” rather than a “full-blown search,” and could therefore be justified by reasonable suspicion instead of probable cause. As already noted, a truly cursory inspection – one that involves merely looking at what is already exposed to view, without disturbing it – is not a “search” for Fourth Amendment purposes, and therefore does not even require reasonable suspicion. We are unwilling to send police and judges into a new thicket of Fourth Amendment law, to seek a creature of uncertain description that is neither a “plain view” inspection nor yet a “full-blown search.” Nothing in the prior opinions of this Court supports such a distinction, not even the dictum from Justice Stewart’s concurrence in Stanley v. Georgia, 394 U.S. 557, 571 (1969), whose reference to a “mere inspection” describes, in our view, close observation of what lies in plain sight.

The Supreme Court has offered specific contexts in which this usual rule doesn’t apply. A basic summary of the Supreme Court’s cases might run something like this: When the search involves some kind of non-criminal investigation or purpose, the warrant requirement is often suspended. In that non-criminal context, reasonableness instead becomes a general balancing of interests. The Court has been expanding the general balancing cases, most recently in Maryland v. King. But the Katz rule of a warrant by default is still the Supreme Court’s blackletter law for a traditional criminal investigation search.

In this case, the Eleventh Circuit appears to take a different approach. It begins with the Supreme Court’s non-criminal cases and then applies them to the context of a classic criminal investigation. Instead of the Katz rule of a warrant, the court begins with general balancing. It’s important to catch criminals, the court reasons, and the statute has some good protections given that this wasn’t such an invasive practice. So on the whole the government’s conduct based on reasonable suspicion seems reasonable and therefore constitutional.

This alternative holding is a major development, I think. It’s at odds with the usual rule that a criminal search requires a warrant, and instead replaces it with a totality of the circumstances inquiry into whether the criminal search was the kind of thing that we would generally say is good or would generally say is bad. There’s not only no warrant requirement, there’s no probable cause requirement: It’s just a free-floating reasonableness inquiry.

Note that this approach draws the court into a quintessentially legislative inquiry, just the kind of “new thicket of Fourth Amendment law” that the Supreme Court warned about in Hicks. Instead of looking at the reasonableness of a search of Davis himself, the court appears to take on the task of looking programmatically at whether Congress’s decisions in drafting the Stored Communications Act as it did was a reasonable decision. In effect, it’s a facial review of the Fourth Amendment adequacy of 18 U.S.C. 2703(d), with the Eleventh Circuit concluding that 2703(d) was good enough. (Or at least that’s how it seems to me; parts of the decision are not clear on whether it is answering the programmatic question.) Note the difference between this and the Florida Supreme Court’s decision in Tracey, which held that prospective cell site records are protected. The Tracey court then applied the usual Katz rule without a second thought: “Because probable cause did not support the search in this case, and no warrant based on probable cause authorized the use of Tracey’s real time cell site location information to track him, the evidence obtained as a result of that search was subject to suppression.”

To be sure, there are normative arguments for expanding the reasonableness framework for cutting back on the Katz rule. But beyond that being more a Supreme Court issue than a lower court issue, there’s an underlying problem: As I see it, the arguments for expanding the reasonableness standard hinge on whether there is a third-party doctrine. That is, the search and the reasonableness arguments are interrelated. As I have written, the case for the third-party doctrine depends on the Katz rule that searches generally require warrants:

My argument [for the third party doctrine] assumes the standard all-or-nothing options of Fourth Amendment law: if government conduct is a search, then it is a search that ordinarily requires a warrant based on probable cause to be constitutional. Epstein eschews these options in favor of a more flexible Fourth Amendment. In his view, courts should have the option of saying that a search is either a full probable cause search or merely a reasonable suspicion search. As a result, he imagines a world in which the issue is not whether to apply the third-party doctrine, but rather what degree of privacy is optimal given the range of tools that can be imagined.
That assumption changes everything, as it means that Epstein ends up answering a very different set of questions. Consider the role of substitution effects. In my article, I argue that the Fourth Amendment corrects for the substitution effects of how individuals use third parties. The use of the third party allows individuals to replace an outside transaction that is unprotected by the Fourth Amendment with an inside transaction that receives a full warrant protection. This argument largely depends on the all-or-nothing framework of existing law: If the choices are between no protection and full protection, a third-party doctrine that results in no protection is better than an alternative world in which crimes can be protected in their entirety by a full warrant requirement. By assuming away the all-or-nothing framework, Epstein dramatically changes the costs and benefits of the third-party doctrine.

The Eleventh Circuit court innovates by adopting the Epstein view of the Fourth Amendment that criminal searches are subject to general balancing. And in so doing it essentially adopts the interesting approach the New Jersey Supreme Court has taken to its state constitution. The New Jersey Supreme Court has rejected the third party doctrine under the state constitution, but it then does all the work under the reasonableness prong. Because every third party access is a search under the New Jersey Constitution, the New Jersey court then engages in the legislative task of saying how much privacy each search needs to have. In this book review, I argued that if you want to overturn the third-party doctrine and adopt an expansive version of the Fourth Amendment, something like that is probably the doctrinal next step.

But the troubling thing about the Eleventh Circuit’s ruling is that it reaches this issue as an alternative holding in a world that still has the third-party doctrine. That means that the third-party doctrine is Eleventh Circuit law and that the watered-down reasonableness analysis premised on a rejection of the third-party doctrine is also part of Eleventh Circuit law. It’s a confusing combination. Given these complications, I don’t see why the court reached the issue.

3. Looking ahead, one interesting implication of all of this is that the Eleventh Circuit’s alternative holding may make the case more certworthy than before. Maybe the Supreme Court will stay out this case because of the good-faith issue flagged in Footnote 20: This case involves an obvious application of Illinois v. Krull, which may make the case uncertworthy based on that alone. But if we take the Court at its word that the good-faith exception doesn’t interfere with Supreme Court review, and it’s willing to engage in advisory opinions (because the Court says they’re not advisory opinions), the alternative holding becomes a big juicy target for review. And you can’t review the second question without also reviewing the first question, as they are linked for the reasons explained above. So the alternative holding might be the way that Davis’s counsel can get the Court interested in resolving the first issue, too, even though there is no clear split on it.

UPDATE: I have fiddled a bit with the post shortly after putting it up, mostly to rewrite my Point 1.