Pages

Tuesday, June 26, 2018

Carpenter's Curiosities (and its Potential to Unsettle Longstanding Fourth Amendment Doctrines)

Not surprisingly, there are already a slew of reactions to the Court's landmark decision on Friday in Carpenter v. United StatesMost observers understandably have focused on two major aspects of the ruling:

(i) The Court held that customers have at least some "reasonable expectation of privacy" in the cell-site location information (CSLI) records that their service providers maintain about them--a new "exception" to the so-called "third-party doctrine," and thus a repudiation of the principle the Court announced in Smith v. United States that “a person has no legitimate expectation of privacy in information he voluntarily turns over to third parties.”  (Such a categorical principle was obviously vulnerable and inadequate from the outset--indeed, it can't be reconciled with the landmark Katz holding itself, in which he Court found that Katz had at least some reasonable expectation of privacy in information that he "voluntarily" shared with a third party, namely, his bookie.)  After the Court's decision in Carpenter, the fact that one has shared information with a third party is relevant to the Fourth Amendment analysis but it does not, in and of itself, resolve the question of whether an expectation of privacy in that information is legitimate, and thus subject to some Fourth Amendment solicitude, including in cases where the state directs the third party to produce the records. 

(ii) Carpenter also confirms the Court's recent willingness to shape its Fourth Amendment doctrine to address the extraordinary new technological surveillance capabilities of the state--in particular, its ability to glean huge amounts of detailed information about individuals, without significant cost, using computer searches of now-ubiquitous databases that track our every transaction, communication, and movement.  (In this respect the best early take, emphasizing the possible implications of Carpenter for "bulk" foreign intelligence surveillance collections, is this piece by David Kris.  [UPDATE:  And this post by my colleague Paul Ohm is also must reading.])

Those aspects of the decision are certainly momentous, but they don't begin to tell the whole story.  Indeed, at least on my preliminary first reading, Carpenter appears to be even more extraordinary and groundbreaking than the initial reports have suggested, for at least three reasons.

1.  It's actually a 6-3 decision--and Justice Gorsuch's rationale might be even broader than the Chief Justice's


The decision has widely been described as a 5-4 split, with the Chief Justice joining the four more liberal Justices.  That's understandable, given that the other four Justices did purport to "dissent" from the disposition.  Justice Gorsuch's separate opinion, however, is in all practical respects even more protective of Fourth Amendment rights than the Chief Justice's (or it would be, anyway, if Gorsuch applies his rationale in future cases).

The first 11 pages or so of Justice Gorsuch's opinion effectively convey the confusion and frustration most newcomers sense when they first confront the Miller/Smith third-party doctrine and Katz's "reasonable expectation of privacy" principle.  Justice Gorsuch vividly describes how difficult it is to think though the problems in this area under the constraints of these undertheorized doctrines.  In particular, he does a nice job demonstrating how the usual justifications for the third-party doctrine are woefully inadequate and how they would lead to deeply counterintuitive results.  For a quick and useful introduction to this area of the law, one could do a lot worse than starting here.

Then, in the next eight pages of his opinion, Justice Gorsuch offers a tentative solution to his dilemma:  "There is a another way," he confidently declares.  Truth be told, however, Justice Gorsuch is less than crystal clear on just what that "other way" consists of (and admirably candid in acknowledging that he hasn't quite settled on his precise approach yet).  Here's my best shot at untangling it:  As I read his analysis of "third party" cases, he would hold that if Party A "entrusts" information about herself to Party B and has a reasonable expectation that Party B will keep it "private"--and a concomitant "positive law right," based upon a promise or contract or property right or statute, that Party B will honor the confidentiality of the information by not sharing it with others--then the state's effort to compel Party B to disclose that information triggers Fourth Amendment scrutiny (i.e., is a "search"), and at least presumptively requires a warrant based upon probable cause to believe that the records contain the fruits, instrumentalities, or evidence of a crime.  Moreover, he suspects that this rule might apply to the CSLI maintained by your cell-phone service provider--information that you convey to the provider as a condition of service.

This Gorsuch "test," in turn, is based upon two key suppositions:  First, CSLI is, in his words, a "modern-day equivalent[] of an individual’s own ‘papers’ or ‘effects,’" regardless of whether the individual has a property law interest in the CSLI.  This is one of the principal differences between Gorsuch and the other three "dissenters":  Justices Alito, Thomas and Kennedy insist that CSLI records do not trigger the Fourth Amendment because they are not the modern-day equivalent of the customer's "papers or effects."  The Roberts opinion for the majority, by contrast, does not deal wth this question at all, because it follows the lesson of Katz that the Fourth Amendment is not literally limited to intrusions upon persons, houses, papers and effects.*

Second, Gorsuch recognizes that "the fact that a third party has access to or possession of your papers and effects does not necessarily eliminate your interest in them."  He uses common practices to illustrate his point:  "Ever hand a private document to a friend to be returned?  Toss your keys to a valet at a restaurant?  Ask your neighbor to look after your dog while you travel?  You would not expect the friend to share the document with others; the valet to lend your car to his buddy; or the neighbor to put Fido up for adoption."  He also rejects the reasoning of the Court in Smith that by conveying information to a third party one “assum[es] the risk” that the third party will reveal it to the police and therefore lack a reasonable expectation of privacy in it--particularly in the case where the third-party chooses to honor, rather than betray, your confidentiality:  "Suppose I entrust a friend with a letter and he promises to keep it secret until he delivers it to an intended recipient.  In what sense have I agreed to bear the risk that he will turn around, break his promise, and spill its contents to someone else?  More confusing still, what have I done to 'manifest my willingness to accept' the risk that the government will pry the document from my friend and read it without his consent?"

OK, so what's the upshot of Justice Gorsuch's "positive law right" theory?  Some readers have assumed Gorsuch is simply asking whether the defendant has a property-law interest in the information.  No doubt a property-law right would almost certainly be sufficient on Gorsuch's view.  As my description above suggests, however, I don't read his opinion as indicating that it's necessary to invoke property law, or that Party A's "positive legal right" to Party B's confidentiality might not arise from some other legal source, such as a statute or, most importantly, a confidentiality agreement.  Gorsuch notes, for example, that a bailee "normally owes a legal duty to keep [an entrusted] item safe, according to the terms of the parties’ contract if they have one," and he also writes that "[p]eople often do reasonably expect that information they entrust to third parties, especially information subject to confidentiality agreements, will be kept private."  To be sure, he notes that the protections Congress has established for CSLI in 47 U. S. C. § 222 "might even rise to the level of a property right"; it's noteworthy, however, that he does not suggest that Fourth Amendment protection depends upon whether they do so.  (It's possible, I suppose, that Gorsuch or others might consider all of this a form of "new property," no matter the source of the confidentiality obligation.  The label doesn't matter, however--the important point is that the customer can bargain for the right to have the service provider maintain confidentiality, at which point presumably the Gorsuch theory of constitutional protection would kick in if the state tries to compel or induce the provider to breach that confidentiality.)

Importantly, Gorsuch also emphasizes that it would not be fatal to the Fourth Amendment claim that the "positive legal right" includes a carve-out that allows or requires Party B to disclose the information to the government:  "Legislatures cannot pass laws declaring your house or papers to be your property except to the extent the police wish to search them without cause."  The important question for Gorsuch, it seems, is whether Party A generally has legal right to prevent Party B from breaching confidentiality, apart from whether that right extends to the government.  (In this respect, Justice Gorsuch's opinion owes an obvious debt to Will Baude and James Stern's The Positive Law Model of the Fourth Amendment, which he cites.  See especially pp. 1871-76 of the article, and page 1879 ("If the government forces a telecom provider to breach a contractual obligation to its subscriber, a Fourth Amendment search plainly occurs.").  As I read his opinion, though, Justice Gorsuch is less concerned about whether the law prohibits other private parties from interfering with the relationship between Parties A and B (the principal inquiry of the Baude/Stern proposal), and more concerned with whether Party B has a legal duty of confidentiality in the first instance.)

Just as significantly, Justice Gorsuch--like the Chief Justice (see Point 3, below)--notes that the Fourth Amendment protection does not disappear simply because the state has used a subpoena to try to compel Party B to disclose:  "No one thinks the government can evade [Ex parte Jackson’s] prohibition on opening sealed letters without a warrant simply by issuing a subpoena to a postmaster for 'all letters sent by John Smith' or, worse, 'all letters sent by John Smith concerning a particular transaction.'”  (For what it's worth, this is a fairly bold and confident claim by Justice Gorsuch about what "no one thinks."  To be sure, "no one" should think such a thing after Carpenter, but my understanding is that many people did, in fact, think exactly that about the breadth of the subpoena power until last week--indeed, that's the upshot of much of Justice Alito's dissent.)

Justice Gorsuch's opinion, therefore, appears to be even broader in terms of Fourth Amendment protection than the Chief Justice's--indeed, it includes as comprehensive a repudiation of the third-party doctrine as any I've seen from a Justice.

Moreover, the Chief Justice's majority opinion identifies several other contexts that he specifically insists the Court's resolution of Carpenter does not resolve:  a order to compel production of "real-time CSLI"; “tower dump” downloads "of information on all the devices that connected to a particular cell site during a particular interval"; "conventional (?!) surveillance techniques and tools, such as security cameras"; "other business records that might incidentally reveal location information"; "other collection techniques involving foreign affairs or national security"; and possibly even CSLI records for a period shorter than seven days (something that's suggested in footnote 3, although the Court's actual holding ("The Government’s acquisition of the cell-site records here"--which included only two days of records from SPRINT--"was a search under [the Fourth] Amendment.")) is not so limited).

Justice Gorsuch's rationale, by contrast, is not so limited--it would appear to apply to most, even if not all, of the cases on which the Chief Justice reserves judgment.

On the other hand, Justice Gorsuch himself signals some trepidation about the implications of his own reasoning.  For example, he writes that "we must be wary of returning to the doctrine of Boyd v. United States, " which restricted the use of subpoenas "even for ordinary business records."  Gorsuch does not say why his theory doesn't suggest just such a "return" to Boyd--unless, perhaps, he intends to provide greater Fourth Amendment protections to individuals than to business entities.  He also, oddly, styles his opinion as a dissent rather than as a concurrence, even though he appears to acknowledge that customers do have a "positive legal right" under 47 U.S.C. § 222(c)(1) to prevent their service providers from "disclos[ing], or permit[ting] access to, individually identifiable” CPNI without the customer’s consent.  Why wasn't that fact--the existence of that "positive law" right of confidentality--enough to compel Gorsuch to rule for Carpenter?  He writes that he is "reluctantly" compelled to conclude that Carpenter "forfeited" the "positive law" argument--in a way that would preclude Justices from relying upon it!--by not raising it below and by only giving it "cursory" treatment in the Supreme Court.  Carpenter's lawyers, however, afforded it much more than cursory treatment (see pages 21-23 and 31-34 of their opening brief, and their reply brief at pages 9-10, 14-15) and, in any event, Justice Gorsuch, like other Justices, often relies upon rationales that were not, or were barely, raised by the parties.  The "forfeiture" argument, then, appears to be a makeweight.

The better explanation--even if it doesn't quite explain why Gorsuch "dissents"--is that Justice Gorsuch has not quite fully worked out his theory to the point where he's comfortable relying upon it.  Which is fine, of course:  he evidently shares Roberts's view that the Court should "tread carefully . . . to ensure that we do not 'embarrass the future.'”  There's not much point in debating just what the "true" implications of Gorsuch's opinion are:  Because he hasn't yet himself worked them out, only time will tell.  The important point going forward, however, is that even if one of the Justices who joined the Roberts opinion is wary about extending the holding or reasoning of Carpenter to one or more of the other contexts specifically reserved by that opinion, that doesn't necessarily mean that there aren't five Justices willing to do so, because Justice Gorsuch appears to be amenable to even broader protections than those identified in the Roberts opinion in at least some cases where the defendant has a "positive legal right" to the confidentiality of information and, in the context where she has "entrusted" such information to a third party, where she has a "positive law right" to compel the third party to honor that confidentiality.

Accordingly, here are a couple of obvious lessons for future customers and advocates:  (i) Secure comprehensive confidentiality agreements wherever possible; and (ii) do not forget to rely upon them, and other "positive legal rights," in legal challenges to government acquisitions of, and orders for production of, information databases, because your fifth vote might well depend upon it.

2.  Acquisition of a database as the "search" that requires a warrant supported by probable cause


The Fourth Amendment concern that Chief Justice Roberts repeatedly expresses in his majority opinion is that modern technologies--computer searches of databases, in particular--will allow the state to perfect "tireless and absolute surveillance" of the populace, i.e., to become Bentham's panopticon.  In the case of CSLI records, for example, the government could "achieve[] near perfect surveillance, as if it had attached an ankle monitor to the phone's user," and could produce "a detailed log" or "exhaustive chronicle" of "a person's movements over several years," which would in turn afford the state dramatically new capabilities of exercising "arbitrary government power."

One obvious response to this concern would be to read the Fourth Amendment to prohibit the state from using CSLI databases to create, and prohibit state officials from reviewing, such "exhaustive chronicles" and "detailed logs."  The Court could have held, for example, that government officials could obtain the CSLI records from service providers, and could run searches of those databases, but only in order to determine whether Carpenter was near the relevant crime scenes, and that the police could not otherwise use the databases in their possession to develop or learn of Carpenter's more detailed movements and locations that weren't relevant to the crimes being investigated.

The Carpenter Court, however, went much farther than that, in a way that I haven't yet seen anyone emphasize:  The Roberts opinion holds--at least twice--that the state cannot acquire the database in the first instance without a warrant based upon a showing of probable cause and, indeed, that such acquisition is itself "a search within the meaning of the Fourth Amendment" (p.17); see also id. at 22 ("The Government’s acquisition of the cell-site records here was a search under that Amendment.").

That is, I think, an extraordinary doctrinal innovation, and (probably) a fairly big deal.  

Before Carpenter, the Court had held that the state's acquisition of an item, without examining its contents, was at most a seizure rather than a search because it does not itself "compromise the interest in preserving the privacy of its contents.”  Horton v. California, 496 U.S. 128, 142 n.11 (1990); see also United States v. Van Leeuwen, 397 U.S. 249, 252-53 (1970) ("The significant Fourth Amendment interest was in the privacy of . . . first-class mail, and that privacy was not disturbed or invaded [by the government's seizure of the package] until the approval of the magistrate was obtained [to open and search the package].").  Moreover, the Court also had previously held--wrongly in my view, but that's now beside the point--that because the Fourth Amendment's ban on unreasonable "seizures" (in contrast to "searches") implicates only property-based concerns and not privacy interests, a seizure of property only occurs when the state's possession of the item results in “some meaningful interference with an individual’s possessory interests in that property,” Jacobsen v. Illinois, 466 U.S. 109, 113 (1984), something that typically does not occur merely upon the government's acquisition or creation of a database of information, which does not divest the subject of any possessory interest in that information.  (In Arizona v. Hicks, for example, the Court held that copying serial numbers from the bottom of a stereo was not a seizure because it didn't divest the owner of any possessory interests.)  And even when there is a "meaningful interference with an individual’s possessory interests," the Court had also held that a seizure is permissible without a warrant if the government in fact has probable cause to believe that it contains contraband or evidence of a crime (see, e.g., Chadwick v. United States (1977)). 

These holdings allowed the government to argue, not without force, that if the state merely acquires or holds information, particularly digital information such as metadata, there is no Fourth Amendment issue unless and until government officials put their eyes on the information, at which point a search might occur--and even then, if the computer search through the database is tailored to disclose (to human officials) only wrongdoing, and those officials do not view "exhaustive chronicles" or "detailed logs" unrelated to the wrongful conduct, there'd be no search then, either.  (This last point was based on the doctrine of United States v. Place, 462 U.S. 696, 707 (1983), in which the Court held that a canine sniff of luggage was not a search where it “disclose[d] only the presence or absence of narcotics” without “exposing noncontraband items.”)  Indeed, the government even argued that until officials rummage through the database and cull information from it, the person whose data is at issue lacks standing to complain about the government's collection of it.  (See, e.g., pp. 2-4 of this brief in a case challenging the constitutionality of the government's Section 215 collection of telephony metadata.)

Ten years ago, my colleague Paul Ohm wrote a sharp little article (The Olmsteadian Seizure Clause: The Fourth Amendment and the Seizure of Intangible Property, 2008 Stan. Tech. L. Rev. 2) warning of this doctrinal development in an era in which the government was increasingly acquiring vast databases that were then available to be searched by computers rather than by the roving eyes of state officials.  On Friday, however, the Court apparently effected a fairly radical shift in the doctrine--and implicitly called into question the scope of precedents such as Horton and Jacobsen--by holding that the Government’s mere acquisition of a database of CSLI records was itself a Fourth Amendment search that requires a warrant based upon probable cause to believe that the fruits, instrumentalities, or evidence of a crime will be found in those records.  

If this reading is correct, then it appears the Chief Justice is suggesting that the Fourth Amendment put limits on the government's very development or acquisition of certain tools by virtue of their capability of being exploited to achieve the sort of "near perfect surveillance" that his opinion warns of.  (A colleague of mine suggests that such a view might even be tied to the textual guarantee of the people to be “secure” in their persons, houses, papers and effects:  One might read Carpenter to hold that such "security" can't be assured if the government has certain materials within its control--at least not if there aren't adequate safeguards in place to make sure the materials (the databases) are not exploited to create "a detailed log" or "exhaustive chronicle" of a person's conduct over an extended period of time.)  

To be sure, in Carpenter the government obtained the records in question for the specific purpose of running computer searches of them right away (rather than simply to hold them for a possible future perusal).  But even if the holding could somehow be recharacterized such that the "search" occurs when the government uses a computer to cull through the database rather than at the time of acquisition--which would be in direct conflict with the language of the Court's holding itself--that would still be a fairly big deal, because it would mean that a warrant is required whether or not any state agents ever see what's in the database (apart from evidence of wrongdoing), and regardless of the terms and scope of the computer search.  It's noteworthy that in Carpenter itself, the government did not create or (more to the point) examine a "detailed log" or "exhaustive chronicle" of Carpenter's "movements over several years," or of his whereabouts unrelated to the bank robberies--and yet the Court nevertheless held that, at a minimum, i.e., even apart from mere acquisition, the Fourth Amendment foreclosed even such a tailored search absent a warrant based upon probable cause. 

This (thus far) underappreciated aspect of the Carpenter holding could have dramatic practical effects on the practice of database collection and acquisition.  (Again, see David Kris's recent post.)  Of course, the Court might in a future case cut back on the broad language of its holdings in Carpenter and, at a minimum, permit the government to obtain databases where sufficient prophylactics are in place to ensure that computer searches of those databases are tailored and do not threaten to permit the sort of "tireless and absolute surveillance" that obviously most concerns the Court.  The Carpenter holding, however, doesn't suggest any such distinctions.  It'll be interesting to see how the Court navigates them in future cases.  [UPDATE:  Orin Kerr has followed up with a very detailed and interesting post on this topic. Recommended.]

3.  The Court's unexpected, groundbreaking holding on warrants and subpoenas

As momentous as the Court's holding may have been on the question of whether an order for a third party to produce CSLI records is a Fourth Amendment "search," in the long run that holding might turn out to be far less important than two other, related holdings in Carpenter--namely, (i) that such a search generally will be unconstitutional unless done pursuant to a judicial warrant secured by a showing of "probable cause" to believe that fruits, instrumentalities, or evidence of a crime will be found in the records; and (ii) that the warrant requirement applies even in the context of a subpoena, including when a grand jury wishes to issue a subpoena duces tecum.

Carpenter presented two constitutional questions.  "Is it a search?" was the only the first.  The second question presented was this:  If the government's acquisition of a service provider's CSLI records is a search, does the standard prescribed by Congress in Stored Communications Act (SCA) for compelling production of those records--a magistrate's finding that the Government has offered "specific and articulable facts showing that there are reasonable grounds to believe that . . . records or other information sought, are relevant and material to an ongoing criminal investigation”--nevertheless satisfy the Fourth Amendment?

Most observers, myself included, thought that although the Court would probably rule for Carpenter on the "search" question, the Justices would either choose not to decide the second question, or they reach the question and a majority would hold that the SCA standard satisfies the Fourth Amendment.  The major reason I thought Carpenter would have a much harder time on the second question was because--as the government stressed in its brief (see pp. 44-50)--the statutory standard Congress has prescribed for compelled CSLI disclosure is more demanding than the showing that would be needed to permit a grand jury to subpoena those same records from any party.  In past cases, the Court had held that a grand jury subpoena duces tecum need only be "relevant" to the grand jury's inquiry” and that the “specification of the documents to be produced [be] adequate, but not excessive, for the purposes of the relevant inquiry.”  Congress's SCA standard would plainly satisfy that low bar--indeed, it would require the certification of a judicial magistrate, whereas a subpoena typically does not require any judicial involvement.  Therefore it would be very difficult to explain why what's sufficient for the grand jury is not good enough for a magistrate acting pursuant to the SCA.

The Court, however, not only reached the second question but decided it in Carpenter's favor--holding that the Fourth Amendment requires a judge to issue a warrant supported by a showing of probable cause before law enforcement can compel a service provider to produce CSLI records (or more than seven days of such records, at a minimum). 

But what about the government's grand jury subpoena argument?  How did the Court parlay that?  Remarkably, the Court held that the subpoena analogy did not help the government because the warrant requirement would also apply in a case where a grand jury or other government entity (e.g., Congress or an administrative agency) would wish to issue a subpoena for production of the same records.  Although "[t]he Government will be able to use subpoenas to acquire records in the overwhelming majority of investigations," wrote Chief Justice Roberts, a warrant supported by probable cause "is required in the rare case where the suspect has a legitimate privacy interest in records held by a third party."

To get a sense of just how extraordinary this "secondary" holding is, note that Justice Alito spends 17 pages of his dissenting opinion railing against it.  He argues that the original understanding of the Fourth Amendment did not extend its protections to orders requiring production of records (including subpoenas) at all (because in such a case no "private area becomes exposed to . . . officers’ eyes as they rummage through the owner’s property in their hunt for the object or objects of the search"), and that even after the Court departed from that understanding, its more modern cases have consistently held that as long as a subpoena is not "excessive" for purposes of a legitimate inquiry, it raises no Fourth Amendment concerns.  

In response, Chief Justice Roberts does not take issue with Alito's historical account; instead, he suggests that that history is simply inapposite because "[t]his Court has never held that the Government may subpoena third parties for records in which the suspect has a reasonable expectation of privacy."

Of course, as Justice Alito explains, the Court had never issued such a holding because the Katz "reasonable expectation of privacy" test had never been thought to be germane to the subpoena question in the first instance.  Indeed, as Justice Kennedy notes in his dissent, the "longstanding rule that the Government may use compulsory process," such as a subpoena, "to compel persons to disclose documents and other evidence within their possession and control" was one of the principles that supported the Miller/Smith third-party doctrine in the first instance, according to which the Court had held that there is no "reasonable expectation" that the government will not obtain records of your one's private conduct from third parties.  For Justice Alito, then, it gets things exactly backwards--the cart before the horse, as it were--for Justice Roberts to eschew the permissive subpoena doctrine in cases where the Court has independently found a defendant's reasonable expectation of privacy in nondisclosure of information from a third party, because there simply is no reasonable expectation that a third party will not comply with a subpoena, at least so long as that subpoena is not "excessive." 

That is not, however, how the majority of the Court (including Justice Gorsuch) sees things.  For them, the first question to ask is whether the defendant (the person whose conduct will be disclosed in the compelled records) has a reasonable expectation of privacy in the nondisclosure of evidence the third party is directed to produce--pursuant to a SCA order or a subpoena.  If the answer to that question is "yes," then, six Justices would say, a full-scale warrant is required, even in the subpoena context.  (Justice Kennedy intriguingly adopts a middle ground:  He would preserve the subpoena doctrine for the mine run of cases, but he acknowledges that a subpoena cannot be used to obtain certain information held by a third party, such as the content of emails held by a service provider--apparently because, in his view, such documents are analogous to the first party's "papers" in a way that CSLI is not.  In this respect, Justice Kennedy may be reflecting a lurking residuum of the Court's subpoena doctrine itself:  In dicta in Fisher v. United States, the Court wrote that “[s]pecial problems of privacy which might be presented by subpoena of a personal diary."  425 U.S. at 401 n.7.)

Justice Alito laments that the Court's treatment of compulsory production as constitutionally equivalent to a "real search" is “revolutionary,” ignores "a century's worth of precedent," and will cause “upheaval.”  In particular, he warns that the Court's “holding that subpoenas must meet the same standard as conventional searches will seriously damage, if not destroy, their utility.”  Likewise, Justice Kennedy writes that "by invalidating the Government’s use of court-approved compulsory process in this case, the Court calls into question the subpoena practices of federal and state grand juries, legislatures, and other investigative bodies."

Whether and to what extent the Carpenter ruling does work such a fundamental transformation of national subpoena practices (and other compulsory process practices) remains to be seen.  (There might even be a way to reconcile Carpenter with the Court's modern subpoena doctrine and fundamental Fourth Amendment objectives, even if not, perhaps, with the ways in which lower courts have permitted very broad subpoenas.**)  The practical impact will likely depend, in large measure, upon at least two things--not only how the Court extends its substantive limitations on the third-party doctrine, beyond the discrete context discussed in Carpenter (recall that the Chief Justice's opinion is at pains to insist that it does not settle a whole slew of analogous questions), but also what the Court decides about the manner in which those limitations will be enforced--what procedures it will insist upon--in the contexts in which the government would, but for Carpenter, resort to use of subpoenas.  

For example, in cases where the "Carpenter" doctrine clearly applies, and where law enforcement officials believe that the "probable cause" standard is satisfied, I assume that the warrant requirement will as a practical matter displace the subpoena process:  When a grand jury or administrative agency wishes to obtain the relevant records, the government will seek a warrant from a judge supported by a showing of probable cause to believe that fruits, instrumentalities, or evidence of a crime will be found in the records, and thereafter provide to the grand jury or to the administrative agency the records obtained.  

As the government stressed in its Carpenter brief, however, it is much more common in those contexts that probable cause has not yet been established--indeed, that the investigating entity, such as the grand jury, is trying to determine if there is probable cause.  What then?  Well, in cases where the records are clearly covered by Carpenter, I assume the government simply won't seek their production.  But what about in the countless contexts that the Court reserved in Carpenter, where it is not yet established whether Carpenter applies and thus whether compelled process is permissible absent a warrant?  If the grand jury or administrative agency issues a subpoena to a third party in such a case--unsupported by probable cause, let alone a warrant--what recourse will the "first party" have to challenge the constitutionality of the subpoena?  The Court has held that, at least as a general matter, the Fourth Amendment does not require that the target of an investigation be provided notice of a subpoena issued to a third party to allow the target to challenge the constitutionality of the compelled production.  The "first party," therefore, will usually be able to challenge the subpoena only much later, if and when the records are introduced against him at trial.  Will the "third party" itself be able to invoke the first party's constitutional rights?  That's not clear, either.  For one thing, service providers will generally lack the incentive or knowledge to raise their customers’ Fourth Amendment rights.  Moreover, it is unsettled whether they are even legally authorized to do so. 

These are the sorts of questions that courts will presumably have to confront in future cases.  [UPDATE:  Orin Kerr has published a post in which he argues that Carpenter is unlikely to have a profound impact on subpoena practices, in large measure because of the difficulties of challenging subpoenas on the ground that it violates the "first party"'s Fourth Amendment rights.]

At a minimum, however, it is fair to assume that questions of the reasonable expectations of privacy of the "first party" subject of the records sought by production orders will now play a much larger role in subpoena litigation and other challenges to compulsory process.  And it's also fair to assume that it'll be quite a while before the Court resolves many of those questions--before it settles on the ultimate breadth and impact of the "Carpenter doctrine." 

Aa Justice Gorsuch writes in his nominal "dissenting" opinion:  "stay tuned."

_____________________
* Justice Kennedy tries to re-write Katz as if it were a case involving "reliance on property-based concepts" involving domiciles such as the "home," in which the Court's holding depended upon viewing the phone booth in that case as analogous "to a friend’s apartment, a taxicab, and a hotel room," such that Katz "had a temporary interest in the space and a legitimate expectation that others would not intrude."  In fact, the Katz Court rejected that view, famously explaining that an "effort to decide whether or not a given 'area,' viewed in the abstract, is 'constitutionally protected' deflects attention from the problem presented by this case," because "the Fourth Amendment protects people, not places," and held that Katz had at least some constitutional interest in the confidentiality of the information he had shared with his bookie.

Even Justice Kennedy appears to concede that some information can be the modern-day equivalent of a person's "papers" even after he has conveyed them to a third party to hold (e.g., the contents of letters; emails held by an Internet service provider), in a way that could trigger Fourth Amendment scrutiny when the state tries to compel the third party to produce them.  Justice Alito expressly, and Justice Kennedy implicitly, tries to distinguish CSLI records from these examples on the ground that "Carpenter did not create the cell-site records."  But of course in every practical sense, he did:  Carpenter might not have compiled the CSLI in the form it was maintained by his service providers, but he conveyed to those providers the data themselves--that is, the information about where he and his cellphone were at all times.

** Although the Court has been very deferential to the state in the context of subpoenas, it did intimate, in one of its seminal modern decisions, that the Fourth Amendment "guards" (“at the most") "against abuse . . . by way of too much indefiniteness or breadth in the things required to be ‘particularly described.'"  Oklahoma Press, 327 U.S. at 208.  Perhaps Chief Justice Roberts's and Justice Gorsuch's opinions in Carpenter can be understood as holding, or at least implying, that a subpoena for vast databases such as the CSLI records in Carpenter are too "indefinite" or too "broad" in the sense that access to such databases empowers the government to engage in the modern-day equivalent of the sort of "general warrant" that allowed British officers to rummage through the homes of colonists "'in an unrestrained search for evidence of criminal activity'" (Carpenter, quoting Riley)--which was one of the principal evils that animated ratification of the Fourth Amendment.  If this is a plausible reading, then perhaps Carpenter might over time become the source of a reinvigorated doctrine about the permissible scope of subpoenas writ large, which would itself be a major development, seeing as how many lower courts are predisposed to approve subpoenas for vast reservoirs of records that can be mined by the state for evidence of wrongdoing (i.e., what some might characterize as "fishing expeditions").  To be clear, however:  The Carpenter decision as written is not, by its terms, about heightened standards for subpoenas themselves, but instead about when a judicial warrant supported by probable cause is required, i.e., when a subpoena just won't do.