One of the major differences between traditional criminal investigations and network surveillance investigations is that network investigations generally require lots of lawyers and lots of court orders. In traditional criminal investigations, lawyers and judges play a relatively minor role. The cops run the show, and lawyers and judges mostly get involved only if a warrant is obtained. The police might watch a suspect, stop and detain him, frisk him for weapons, and make an arrest all without lawyers or judges. Review of the legality of the investigation is ex post: The defendant moves to suppress the evidence and argues that the past conduct was unconstitutional.
Criminal investigations involving network surveillance are quite different. When investigators want to obtain evidence over the telephone network or the Internet, Congress has imposed ex ante statutory requirements of court orders at many stages. For the most part, lawyers and judges are involved from the beginning of the case. Lawyers need to be involved at every stage, getting various kinds of court orders required by the Electronic Communications Privacy Act (ECPA), such as pen register orders, 2703(d) orders, and the like. Judges have to be involved from the beginnning, too, as they rule on the government’s applications. There’s a very good reason for the requirement of ex ante court orders, I think. Network surveillance investigations involve third party networks, and often the orders are executed by the third party network provider rather than directly by the government. The introduction of the neutral third party justifies ex ante procedural protections that provide notice to the provider and leave a record of what occurred. But it also leads to a very different role for lawyers and judges. While a typical magistrate judge might only hear about a traditional criminal case when the government has already arrested a suspect, a magistrate judge in a digital evidence case often has signed lots of court orders about the suspect long before an arrest is made.
So here’s the interesting question: As a matter of policy, when should targets of digital evidence investigations receive notice of the court orders? And when and how should they be allowed to challenge the orders as unlawful? In a traditional criminal case, suspects don’t receive notice that they are subjects of monitoring. When the government decides to “tail” a suspect around town, they don’t send them a letter letting them know. Suspects receive notice only in specific contexts, such as if their home is searched pursuant to a warrant. And they have to wait to bring challenges until late in the game. In the case of a warrant, for example, the defendant challenge the warrant until after it has been executed. The question is, if you were writing the statutory network surveillance laws, when would you impose a statutory notice requirement and when would you allow challenges to be brought? Would you try to match overall amount of notice in digital investigations to that of traditional physical investigations? Or would you aim for more or less notice in the electronic setting than in the physical setting? Would you allow challenges to surveillance practices as they were ongoing, or would you require challenges to wait until the order had been executed?
In an interesting new article in the Harvard Law & Policy Review, Magistrate Judge Stephen Smith argues that everyone who is the subject of an electronic surveillance order ordinarily should receive notice as well as a right to challenge and appeal any ongoing surveillance order that targets them. As best I can understand Smith’s proposal, the government would not only have to get a court order to take any surveillance steps, but every surveillance step would be ordinarily followed by notice and an opportunity for immediate and full litigation of the legality of the surveillance through at least the Court of Appeals and presumably also the Supreme Court. Magistrate Judge Smith does recognize that there should be cases when notice can be delayed in light of investigatory needs, but he suggests that the government should have to apply for delayed notice and provide a need for it — and that requests for repeated delay should be scrutinized carefully. Smith also argues that docket sheets of electronic surveillance should be open, so that the fact of the surveillance should be public. In one sense, Magistrate Judge Smith takes the standard that applies to warrants in the case of traditional physical investigations and would instead apply it to all court orders. But Smith’s standard goes beyond even warrants, creating a right to challenge ongoing surveillance and appeal any adverse decisions. Given that proceeding by way of court orders is the norm in digital evidence/surveillance cases, it seems to me that the result would be that most digital evidence/surveillance investigations would require a default of notice to each suspect about what the government is doing and a right to challenge surveillance at its inception.
In his latest Sidebar column in the New York Times, Adam Liptak presents Smith’s view as a matter of common sense, and he doesn’t suggest any basis for criticism of it. But while I find Smith’s proposal interesting, it strikes me as unwise. The default of notice and opportunity to litigate a court order seems inconsistent with the court order requirement itself: If the point of the court order is to add an ex ante check, presumably the statutory requirement and review by the judge achieves that goal. And while I can see encouraging litigation when the investigation is done — more on this in a minute — tipping off suspects and giving them a way to slow down or stop court-ordered surveillance as the investigation is ongoing would make surveillance too costly and ineffective. Perhaps Judge Smith is assuming that delayed notice would be obtained as a matter of course: That way the government could monitor the suspect without tipping him off. So the requirement of notice in theory would become delayed notice in practice. But if so, I’m not sure why there would be a default of immediate notice with only an exception if the government makes the required showing.
A related concern with Smith’s proposal is the ex ante impact on the police of a notice requirement. Imagine the rule is that anyone whose records have been collected has a right to notice within 30 days, and that the government can renew the delayed period of notice for additional 30 days only if they can prove a specific need to do so. The government is investigating a narcotics trafficking conspiracy, and to identify members of the conspiracy the government obtains pen/trap non-content telephone records to link the phone numbers and see who is in the group. After obtaining the records on the low pen/trap standard, though, the government finds that it has significant cause but that getting a wiretap order would be prohibitively expensive and difficult. The investigation is stuck until the investigators can come up with another way to break open the conspiracy: It might take weeks, or months, or longer, and they may never have the evidence needed to move forward. Under Smith’s proposal, though, there is a default notice requirement to all the suspects whose records were obtained. Everyone in the conspiracy gets to know that the investigators had obtained their phone records, allowing members of the group to get together and know who was under suspicion. And if delayed notice is permitted, the government has to come back and renew the application of delayed notice every however-many days — and courts will strictly review the applications for delayed and renewed notice, meaning that courts may not grant the application. In that setting, I would think that the risk of requiring notice would have a significant ex ante deterrent effect on obtaining the orders in the first place. To the extent we want to limit the orders, though, I would think that the answer is raising the statutory threshold of obtaining the orders rather than creating a risk of tipping off suspects that might prevent evidence collection entirely.
Another problem with Judge Smith’s proposal is that it would appear to require judicial review to occur with no facts, as the challenges would occur before or as the surveillance was ongoing. This gets to the issue of what Judge Smith hopes courts will rule on when they review the lawfulness of the court order. Are the courts only supposed to review whether the application complied with the statute, or are they supposed to also consider whether the surveillance would comply with the Fourth Amendment? It sounds to me like Judge Smith has both roles in mind. If so, I don’t think that’s permissible: As regular readers know, I don’t think judges are permitted to make constitutional rulings in such settings because no ripe dispute exists, an issue that is currently on appeal before the Fifth Circuit. Basing an entire surveillance regime on this kind of challenge therefore strikes me as not only unwise but inconsistent with ripeness doctrine. (Presumably Judge Smith disagrees with me on the ripeness question currently before the Fifth Circuit — indeed, Judge Smith is the author the Fourth Amendment ruling that is on appeal before the Fifth Circuit that led me to file my amicus brief – although he has never written on the issue.)
So what’s the right answer? I think Judge Smith has recognized a significant issue. The electronic surveillance statutes are very murky, and the law is often unclear, making it difficult for the feedback mechanism of judicial and congressional oversight to work. But I think the better lever of statutory reform is remedies instead of notice. In a 2003 article in the Hastings Law Journal, I argued that the solution is to create a statutory suppression remedy for the electronic surveillance statutes. A statutory surveillance remedy would shine a line on the surveillance practices by encouraging defendants to litigate the surveillance statutes just as they do Fourth Amendment issues. That way criminal defendants can litigate the lawfulness of surveillance orders ex post with real facts and real stakes, leading to the needed precedents on the proper interpretation of the surveillance laws. A statutory suppression remedy would effectively match the remedies scheme of the statutory surveillance laws with that of the Fourth Amendment, harnessing the same mechanisms and leading to the same judicial oversight in the context of real cases and controversies. In my view, that’s the better way to satisfy the competing goals of investigatory needs and the protection of civil liberties in the surveillance arena.
But I’m sure many readers will disagree, so I’ll return to the question: In your view, when should the electronic surveillance statutes require notice and opportunity to challenge surveillance orders?
[UPDATE: I amended this a bit as I lost Internet connectivity as I was editing an earlier draft.]