This is the second in a series of posts on the justiciability of Fourth Amendment disputes. Specifically, the issue is whether magistrate judges have the power, when asked to review a statutory court order for surveillance under the Stored Communications Act, to decline to sign an application that satisfies the statute but that the magistrate fears will be executed in ways that violate the Fourth Amendment. The question is currently pending in the Fifth Circuit, with oral argument scheduled for October 2nd. As I argue in my amicus brief, I don’t think judges have that power because such cases are not yet ripe for adjudication. The facts of how the government will obtain the records are unknown, so there are no facts yet on which to apply the law. Without actual facts, Article III jurisdiction does not yet exist, so the magistrate judge has no constitutional authority to opine on the legal issues and the Court of Appeals has no constitutional authority to rule on the issues in the appeal. The Court of Appeals therefore must reverse the district court with instructions to sign the order because the statutory threshold has been satisfied. In this post I want to run through some of the major cases that support this argument, and then consider the two major precedents that have been asserted against my position.
Perhaps the most analogous ripeness cases are those dealing with the limitation of the federal courts to impose injunctive relief in Fourth Amendment litigation. In these cases, a party who has been searched or seized — or who expects to be searched or seized soon — comes into court and seeks an injunction ordering the government not to engage in the practice again because the practice is unconstitutional. The issue is whether the federal court has the power to order the government not to violate the Fourth Amendment prospectively, and more specifically to rule on what the constitution means prospectively and then tell the government not to do that thing that the court finds violates the constitution. The order in the Fifth Circuit appeal raises a very similar issue, I think: The court’s adjudication of the issue and denial of the application to avoid a future constitutional violation operates as an injunction. Because the statute requires an order before collecting the evidence, denying the application for the order effectively enjoins the government from collecting the evidence (without the formality of a motion for injunctive relief). So the question is, what are the limits on the constitutional authority of federal courts to enjoin potentially unconstitutional conduct on Fourth Amendment grounds?
The Supreme Court doesn’t have much authority on this issue, although it’s worth flagging the somewhat analogous issue that arose in City of Los Angeles v. Lyons, 461 U.S. 95 (1983). This case dealt with the closely-related case or controversy requirement rather than ripeness doctrine, but it raises some similar issues. Lyons sued the City of Los Angeles after a city police officer had subjected him to a “chokehold” that rendered him unconscious. Lyons sought damages and also injunctive relief; the injunctive relief asked the Court to block the LA police force from using chokeholds “except in situations where the proposed victim of said control reasonably appears to be threatening the immediate use of deadly force.” The district court found that the LA police department had authorized its officers to use chokeholds “in situations where no one is threatened by death or grievous bodily harm,” and that the officers were insufficiently trained, and that the chokeholds were very dangerous. The district court enjoined the use of chokeholds “under circumstances which do not threaten death or serious bodily injury.”
The Supreme Court overturned the injunction on the ground that Lyons had not established a case or controversy. To establish a case or controversy permitting injunctive relief, the Court held, Lyons had to show “a real and immediate threat that he would again be stopped for a traffic violation, or for any other offense, by an officer or officers who would illegally choke him into unconsciousness without any provocation or resistance on his part.” It wasn’t enough that Lyons might be stopped, or that he might be subject to a chokehold. Nor was it enough that someone in Los Angeles might be subject to an illegal chokehold. To establish Article III standing, Lyons needed to show a real and immediate threat that he himself would be stopped and would be subject to a the same specific set of facts he alleged — a chokehold that rendered him unconscious without any provocation or resistance. Without proof of a real and immediate threat that the very specific set of facts he alleged would happen — down to the details that he would be choked in unconsciousness after being stopped for a traffic violation, and that he would not have resisted or provoke it — there was no constitutional authority for the federal courts to step in and provide prospective relief.
At the circuit court level, the most analogous case is Warshak v. United States, 532 F.3d 521 (6th Cir. 2008) (en banc). Like the pending Fifth Circuit case, Warshak involved an effort to stop reliance on 18 U.S.C. 2703, parts of which allow the government to collect certain kinds of evidence from network service providers without a warrant. Mr. Warshak had the contents of his private e-mail accounts obtained from his ISP without a warrant pursuant to 18 U.S.C. 2703(b). The government investigators refused to say that they wouldn’t get more of his e-mail without a warrant in the future based on the same statute. So Warshak sought an injunction blocking the government from obtaining personal e-mails from ISPs without a warrant. Warshak had a very good case that his Fourth Amendment rights had been violated, and that obtaining e-mail without a warrant is often unconstitutional. Indeed, the Sixth Circuit ruled just a few years later that Warshak’s rights had been violated and 18 U.S.C. 2703(b) was generally unconstitutional for allowing the government to obtain e-mail without a warrant (at least in the usual case). Nonetheless, the en banc court ruled that Warshak’s effort to obtain an injunction must fail on ripeness grounds. The facts were speculative, so the legal claim was not ripe. The Warshak case is closely analogous to the Fifth Circuit case, and the whole opinion is worth reading, but here’s the heart of Judge Sutton’s reasoning:
Not only do “we have no idea whether or when” such a search will occur but we also “have no idea” what e-mail accounts, or what types of e-mail accounts, the government might investigate. Toilet Goods Ass’n, Inc. v. Gardner, 387 U.S. 158, 163, 87 S.Ct. 1520, 18 L.Ed.2d 697 (1967). That uncertainty looms large in a debate about the expectations of privacy in e-mail accounts. The underlying merits issue in the case is this: In permitting the government to search e-mails based on “reasonable grounds,” is § 2703(d) consistent with the Fourth Amendment, which generally requires “probable cause” and a warrant in the context of searches of individuals, homes and, perhaps most analogously, posted mail? The answer to that question will turn in part on the expectations of privacy that computer users have in their e-mails—an inquiry that may well shift over time, that assuredly shifts from internet-service agreement to internet-service agreement and that requires considerable knowledge about ever-evolving technologies[.]
Think of just one of these moving parts—the variety of internet-service agreements and the differing expectations of privacy that come with them. An agreement might say that a service provider will “not … read or disclose subscribers’ e-mail to anyone except authorized users.” United States v. Maxwell, 45 M.J. 406, 417 (C.A.A.F.1996) (describing testimony about AOL’s then-existing policy). An agreement might say that a service provider “will not intentionally monitor or disclose any private email message” but that it “reserve[s] the right” to do so in some cases. See Privacy Statement for Juno Members, http:// www. juno. com/ legal/ privacy. html (last visited July 7, 2008). An agreement might say that a service provider “may or may not pre-screen Content, but … shall have the right (but not the obligation) in [its] sole discretion to pre-screen, refuse or move any Content that is available via the Service”—as indeed Warshak’s Yahoo! account did. JA 89, 163 n. 3. An agreement might say that e-mails will be provided to the government on request—as indeed the same Yahoo! account did. An agreement might say that other individuals, besides the recipient of the e-mail, will have access to it and will be entitled to use the information in it. See, e.g., JA 208 (explaining that Gmail, a service provided by Google, gives users “an enormous amount of storage capacity … in exchange for … terms of service which say that Google is allowed … [to] take a look at the content of [users’] e-mail and … target advertising at [users] accordingly”). Or an agreement might say that the user has no expectation of privacy in any of her communications. See, e.g., JA 207 (government counsel explaining that “every day when we log into our e-mail account, we agree that we have no expectation of privacy in the account”).
Some of these service-provider agreements could cast doubt on the validity of § 2703(d) in a given case; others might not. Better, we think, to decide the validity of the statute in the context of a specific internet-service agreement and a specific search and seizure.
Nor can we rely on previous government searches of Warshak’s e-mails to hypothesize the factual context of the next search. Even if the record contained the full text of the NuVox and Yahoo! service-provider agreements (it does not; it contains just part of the Yahoo! agreement), we would run into a similar conjecture problem. Just as there is little basis for assuming the government will conduct another ex parte search of Warshak’s e-mails, there is little basis for assuming any future search will concern e-mails facilitated by these service providers, as opposed to e-mails facilitated by other service providers. In view of Warshak’s knowledge of the investigation and knowledge of his pre-existing service-provider agreements before he filed the complaint, it is surely possible that he might switch to a service provider that gave him a greater expectation of privacy—say, by contracting with a paid-subscription service provider that promises not to screen e-mail. Or he might decide that the convenience of free, web-based e-mail is ultimately worth the tradeoff of allowing the service provider to *528 screen his e-mail. We simply have no way of knowing what kind of accounts Warshak is likely to possess in the future and no basis for assuming that future e-mails will be controlled by the same type of service-provider agreement he used in the past. “The operation of the statute,” in short, will be “better grasped when viewed in light of a particular application.” Texas, 523 U.S. at 301, 118 S.Ct. 1257.
Concerns about the premature resolution of legal disputes have particular resonance in the context of Fourth Amendment disputes. In determining the “reasonableness” of searches under the Fourth Amendment and the legitimacy of citizens’ expectations of privacy, courts typically look at the “totality of the circumstances,” Samson v. California, 547 U.S. 843, 848, 126 S.Ct. 2193, 165 L.Ed.2d 250 (2006) (internal quotation marks omitted), reaching case-by-case determinations that turn on the concrete, not the general, and offering incremental, not sweeping, pronouncements of law, see O’Connor v. Ortega, 480 U.S. 709, 718, 107 S.Ct. 1492, 94 L.Ed.2d 714 (1987). Courts thus generally review such challenges in two discrete, post-enforcement settings: (1) a motion to suppress in a criminal case or (2) a damages claim under § 1983 or under Bivens v. Six Unknown Named Agents of Federal Bureau of Narcotics, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971), against the officers who conducted the search. In both settings, the reviewing court looks at the claim in the context of an actual, not a hypothetical, search and in the context of a developed factual record of the reasons for and the nature of the search. A pre-enforcement challenge to future e-mail searches, by contrast, provides no such factual context. The Fourth Amendment is designed to account for an unpredictable and limitless range of factual circumstances, and accordingly it generally should be applied after those circumstances unfold, not before.
That is why Warshak’s rejoinder — that this case presents a “purely legal question,” Toilet Goods, 387 U.S. at 163, 87 S.Ct. 1520—carries little weight. In addition to the fact that this “purely legal question” remains a purely speculative legal question, this case presents a legal question that may be answered differently in different settings and a legal question that “depend [s] … on an understanding of” complex factual issues. Id. In such cases, “judicial appraisal … is likely to stand on a much surer footing in the context of a specific application of [the law] than could be the case in the framework of [a] generalized challenge.” Id. at 164, 87 S.Ct. 1520.
Of course, this appeal has been filed in the Fifth Circuit. Where is Fifth Circuit caselaw on this? Perhaps the most relevant Fifth Circuit precedent is United Transportation Union v. Foster, 205 F.3d 851 (5th Cir. 2000). Two unions filed a preenforcement challenge to a newly enacted Louisiana railroad safety law, Act 81. The unions asserted that Act 81 violated the Fourth Amendment by allowing drug testing of railroad employees without probable cause in the event of a railway accident. They argued that such testing would be unconstitutional; the state countered that it was constitutionally permitted. The parties did not raise any ripeness concerns, but the Court raised ripeness sua sponte:
No one has challenged the ripeness of this case for adjudication. However, we must consider possible objections to our Article III jurisdiction sua sponte. See Lang v. French, 154 F.3d 217, 222 (5th Cir. 1998). “Every federal appellate court has a special obligation to ‘satisfy itself not only of its own jurisdiction, but also that of the lower courts in a cause under review,’ even though the parties are prepared to concede it.” Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 118 S. Ct. 1003, 1013, 140 L. Ed. 2d 210 (1998) (quoting Mitchell v. Maurer, 293 U.S. 237, 244, 79 L. Ed. 338, 55 S. Ct. 162 (1934)).
The Fifth Circuit standard for ripeness was the following:
A court should dismiss a case for lack of “ripeness” when the case is abstract or hypothetical. The key considerations are “the fitness of the issues for judicial decision and the hardship to the parties of withholding court consideration.” A case is generally ripe if any remaining questions are purely legal ones; conversely, a case is not ripe if further factual development is required.
The Fifth Circuit dismissed the case on ripeness grounds because the pre-enforcement challenge to Act 81 was based on “conjecture and speculation” about the facts:
Appellees’ challenge sits atop a mountain of conjecture and speculation. In order for Act 81 to run afoul of the Fourth Amendment, the following train of events would necessarily have to occur: First, a train must be involved in a collision at a Louisiana railroad crossing. Although the law of probability suggests such a collision may be inevitable, we cannot determine with any degree of certainty when such an event will occur. Indeed, the Louisiana legislature may amend Act 81’s challenged terminology, or repeal Act 81 it in its entirety, before another locomotive collision at a railroad crossing in Louisiana. Second, even assuming that such a collision occurs, Act 81 does not operate automatically in the event of a collision. Rather, a law enforcement officer must have “reasonable grounds to believe the person to have been operating or in physical control of the locomotive engine while under the influence” of alcohol or other illegal controlled substances. Clearly, there will be many cases where an officer’s suspicion does not rise to the level necessary to trigger Act 81’s application. Third, “reasonable grounds to believe” would have to be interpreted to mean something other than “probable cause.” The appellees ask this court to interpret “reasonable grounds to believe” to mean “reasonable suspicion,” a level of suspicion clearly below the “probable cause” generally needed to justify a search in a criminal investigation. Finally, a Louisiana officer would have to order such testing without actually having “probable cause.” In light of the extreme prematurity of this action, we refuse to allow appellees’ Fourth Amendment facial challenge to Act 81.
In my view, the Fifth Circuit cell-site case should be governed by the principles of Foster, Warshak, and Lyons. The government has come to the court with an application under the statute, and it has provided the magistrate judge with the required reason to believe that particular telephone numbers were used in crime that (under the statute) requires the magistrate judge to sign the order requiring the phone company to disclose the records. The magistrate judge has those facts in front of him, but he does not know what the future will look like if he signs the order. He doesn’t know what particular records have that the providers will turn over. He doesn’t know what the service agreement between the phone company and the user might be that might impact expectations of privacy. He doesn’t know whether the phone was “on” during the relevant period — or, if it was “on,” whether the phone was with the person who registered for it. He doesn’t know whether the phone was used in public, where there is no expectation of privacy under Knotts, or whether it was used in a home, where at least there is a plausible argument that there was a reasonable expectation of privacy under Karo (only plausible because if the court accepts the view that Smith v. Maryland governs, then the home/outside distinction doesn’t matter). In short, what the facts might turn to be — and thus whether the Fourth Amendment might be violated — requires “speculation and conjecture.” As a result, the Fourth Amendment question is not yet ripe for adjudication.
But what about the cases on the other side? Those who argue that the Fifth Circuit case is ripe rely on two cases: Chandler v. Miller, 520 U.S. 305 (1997), and In re United States for an Order Directing a Provider of Elec. Commc’n Serv. to Disclose Records, 620 F.3d 304 (3d Cir. 2010). Neither is a ripeness case. The word “ripeness” never appears in either decision. So how can these cases support the view that the Fifth Circuit case is ripe?
As best I can tell, the argument is that if there were really a ripeness problem with these kinds of prospective Fourth Amendment decisions, the courts in these two cases should have dismissed them on ripeness grounds sua sponte. But the fact that the courts did not sua sponte dismiss the cases on ripeness grounds signals that the issues they decided were ripe for adjudication. And if those two cases were ripe for adjudication, then so is this case. Here’s how the ACLU/EFF/CDT/NACDL argues the point in its joint amicus brief:
Professor Kerr’s amicus brief questions whether this case is even ripe, suggesting that at the time the government applies for a § 2703(d) order, a judge is to either approve or deny the request without determining “the constitutionality of the future execution of the search” because “[a] court cannot apply the Fourth Amendment when no facts yet exist.” Kerr Amicus at 2, 4, 8. This sweeping argument should be rejected as contrary to how the Supreme Court has applied the Fourth Amendment in the past. In Chandler v. Miller, 520 U.S. 305 (1997), the Supreme Court struck down a Georgia law mandating drug testing of certain candidates for elective office. The Supreme Court did not require the candidates to wait until after they were tested to pursue a challenge. Nor did the Supreme Court enjoin the statute only as to them, on the off-chance that a future candidate might be, for example, a parolee with a reduced expectation of privacy. See Kerr Amicus at 9. The Supreme Court struck the statute down in its entirety. Chandler, 520 U.S. at 323.
Similarly, where the government files an application requesting access to specific data—in this case, cell phone location data for whenever a phone is turned on—magistrate judges need not sit idly by while individuals’ constitutional rights are violated.
I find this argument puzzling. First, I don’t know how the absence of a ruling when no party raised the issue can be treated as the equivalent of a ruling in favor of one side. Sometimes courts recognize ripeness problems not brought to their attention, and sometimes they don’t. But the failure to spot and discuss the issue doesn’t mean no issue existed. That’s especially important with ripenes doctrine because parties often concede ripeness either out of cluelessness or because they want a ruling on the merits. Recall what happened in Foster: The parties declined to raise the issue, and the Fifth Circuit judges realized the problem and dismissed on those grounds sua sponte. (If you’ll pardon my speculation, I wonder if something similar is happening here. DOJ is the only party to this litigation, and it was curiously silent in its reply brief on the ripeness issue. But that may just be a tactical decision. Because the Fifth Circuit is the most pro-law-enforcement circuit in the country, DOJ may prefer a ruling on the merits to a dismissal on ripeness grounds — a win instead of a draw, from a Fourth Amendment perspective.)
Second, even if you accept the rather odd idea that the absence of a ruling on an issue not argued is akin to a ruling on the merits, the Chandler case raises a very different procedural context than does this case. In Chandler, three Libertarian candidates for state office had to be drug tested for a specific set of drugs and approved as drug free by a state-approved office by a specific date. This satisfied the Lyons test for cases and controversies, as it was a very specific set of facts that both sides agreed were required in a specific window of time. Further, the state conceded that the drug tests were “searches” that violated a reasonable expectation of privacy. The only question in dispute was a question of law, whether the need to keep candidates drug free was a “special need” for Fourth Amendment purposes that could make the searches “reasonable.” Importantly, whether that was a special need was not something that hinged on any future facts. That brings it pretty far from this case, in which it is unknown what facts might unfold and for which there has never been a hearing about the facts or even any claims by any parties as to what the future facts might turn out to be. (And why is it relevant what the scope of the relief was in Chandler? Ripeness is about judicial power to decide, not about what the scope of the relief should be after asserting jurisdiction, adjudicating the issue, and finding a violation.)
Finally, although it’s true that the Third Circuit did not dismiss the case sua sponte as unripe, the court did decline to rule on the Fourth Amendment question because there were no facts yet. See, e.g., 620 F.3d at 312 (“The MJ proceeded from the premise that CSLI can track a cell phone user to his or her location, leading the MJ to conclude that CSLI could encroach upon what the MJ believed were citizens’ reasonable expectations of privacy regarding their physical movements and locations. The MJ regarded location information as “extraordinarily personal and potentially sensitive.” MJOp., 534 F.Supp.2d at 586. We see no need to decide that issue in this case without a factual record on which to ground the analysis.”)
Next up, I plan to post on how to reconcile the two parts of the Constitution at issue in the appeal — Article III and the Fourth Amendment.