Vermont Supreme Court Concludes that Magistrates Have the Discretion to Add at Least Some Kinds of Ex Ante Search Restrictions to Computer Warrants

I have blogged often about the ex ante search restrictions on computer search warrants. In such cases, the government applies for a warrant to search a computer, and the magistrate judge concludes the warrant application satisfies the Fourth Amendment’s requirements of probable cause and particularity. Nonetheless, the judge is worried that the subsequent execution of the warrant may be too broad, so the judge devises a set of conditions for executing the search and then conditions the issuance of the warrant on complying with those conditions. In my view, judges lack the authority to impose such conditions, and those conditions are undesirable because they will impede the necessary development of Fourth Amendent law. I made that argument in an article published two years ago, Ex Ante Regulation of Computer Search and Seizure, 96 Va. L. Rev. 1241 (2010). Until today, however, no appellate court had squarely addressed the question.

Today the Vermont Supreme Court finally issued a decision in its long-awaited pending case on the issue. I have uploaded the court’s decision here: In re Appeal of Application for Search Warrant. The court’s basic conclusion is that magistrates can “sometimes” add ex ante search restrictions to try to protect privacy, at their discretion, although not all restrictions are allowable. Further, restrictions are usually binding on the police, but not if the magistrate’s guess as to what would happen that was the basis of the condition later proves to be inaccurate. A significant part of the opinion responds to my article, so I hope readers won’t mind an unusually long post explaining the court’s reasoning and offering my brief commentary.

In this case, Vermont police applied for a warrant to enter a home and search the suspect’s computer for evidence of identity fraud. The judge agreed that the government established probable cause, and he signed the warrant. However, the judge imposed ten individually-numbered restrictions detailing how the police had to execute the warrant and what rights they retained going forward. Several but not all of the restrictions were taken from Judge Kozinski’s concurrence in a recent computer search and seizure case, United States v. Comprehensive Drug Testing. For example, the magistrate forbade the police from relying on the plain view exception to use any evidence outside the scope of the warrant that the police might found. The judge also ruled that the police could not use “sophisticated hashing tools” and “similar tools” to search the computer without first obtaining special permission from the magistrate judge. The police searched the home and seized the suspect’s computer, imaged it, and returned. But it has not searched the image. Instead, the state filed a motion for relief in the Vermont Supreme Court challenging the restrictions. The state’s goal is to have the state Supreme Court strike down the ex ante restrictions so the state can search the computer and leave the constitutionality of its search to ex post challenges.

In the decision handed down today, the Vermont Supreme Court reviewed the magistrate’s conditions for abuse of discretion rather than de novo, as required by Vermont state law that gave the state Supreme Court jurisdiction. Under that standard, the Court concluded that ex ante search restrictions are “sometimes acceptable” because they can be ways of ensuring that searches are executed in a narrow way.

The Court’s argument is based on two premises. First, in the court’s view, the warrant clause of the Fourth Amendment is not just about establishing probable cause and particularity. Rather, warrants are designed to minimize privacy invasions by making sure that “investigatory promise” of a search justifies the resulting “collateral exposure” in that particular case. The court seems to have in mind a kind of balancing scale that each magistrate must weight with each warrant application. The more invasive a warrant seems, the more the magistrate has the authority (although not the requirement) to devise and impose restrictions designed to ratchet back the overall invasiveness of the resulting search. As a result, ex ante restrictions are “sometimes” permitted as ways of regulating the process to try to protect privacy in light of the “interplay” between probable cause, particularity, and reasonableness.

The Court’s second premise is that particular procedures can substitute for particular descriptions under the Fourth Amendment. Instead of particularly describing what the police can seize, the warrant can describe how the police can search for items. Imposing specific procedures acts a functional substitute for a particular description of the thing to be seized: It ensures a narrow search by expressly regulating what the police can and can’t do. As a result, imposing search restrictions is a way of satisfying the requirement that warrants must “particularly describ[e] . . . the persons or things to be seized.”

The court then runs through the four kinds of restrictions imposed by the magistrate judge in this case. One of the conditions violated the magistrate’s discretion, the court concludes; the remaining restrictions did not.

1) The first restriction was requiring the government to forswear reliance on the plain view exception. The court rejects this condition as “unnecessary for privacy protection and inappropriate” for two reasons. First, the other restrictions in the warrant make it unlikely that this would be a problem: There won’t be any information outside of plain view, so there is no need to add this restriction. Second, a magistrate doesn’t have the authority to overturn U.S. Supreme Court precedent, so it’s improper to make the government give up reliance on that precedent.

2) The second set of restrictions was requiring the computer to be searched by non-investigative personnel who would then set up a “wall” and not give any evidence to the investigators that was not, in their view, related to the crime under investigation. The court allows this because it was an effort to try to restore particularity: “In lieu of a particular description of the files” that would ordinarily be requried under the Fourth Amendment, this provision allows a substitute of “a procedure for identifying the relevant files and exposing them only to police investigators.” Exposure of “embarrassing information to a detached third party constitutes a lesser injury” to privacy interest than does exposure to the police, so this procedure minimizes the invasiveness of a computer search and is therefore “not so wholly without basis as to constitute an abuse of discretion.” (Two Justices dissented on this part of the opinion: They argued that these search restrictions were a procedure designed to frustrate the plain view exception, and thus were not permitted for the same reason that it was impermissible to force the government to forswear reliance on the plain view exception.)

3) The third set of restrictions was a requirement of special search protocols and a ban on using “sophisticated hashing tools” and “similar tools” without special permission. This was permitted because it was a way to limit the scope of the search that could limit the privacy invasion. And if the government searches the computer and thinks that there may be more on the machine, the government can always apply for a second warrant to use “sophisticated” tools (which the magistrate might reject, but at least the government can apply).

4) The fourth set of restrictions required only responsive data to be copied, non-responsive data to be destroyed, and the search to be completed in a particular period of time. The court permitted these restrictions because they were similar to the kind of restrictions traditionally imposed by statutory rules governing warrants.

The Court also indicates that the search restrictions “generally” are binding on law enforcement when issued, although the government can try to argue that circumstances changed and that therefore the restrictions need not be followed. The opinion is especially murky on this issue: It relies on a single outlier district court case, the Burnette case from Maine, for the view that search restrictions are binding. But Burnette provided no analysis for this point, nor even any citation, in support of its view. And the Vermont court limits to its facts the contrary U.S. Supreme Court case, Richards v. Wisconsin, based on grounds not articulated in that opinion. The court also suggests that the fact that the restrictions are always optional and not always binding when issued means that there can still be ex post litigation to figure out what kind of search rules are constitutional as a matter of law. But I’m not sure how that works: How can the police know if circumstances changed? Changed from what? The magistrate who imposed the order doesn’t tell the police what circumstances she has in mind, so I don’t know how the police are supposed to know that they changed or the facts turned out to be different from what the magistrate was thinking.

Anyway, that’s the Vermont Supreme Court’s decision. It’s fascinating. As one might guess from reading the opinion, though, I also think it’s wrong. I certainly see the appeal of trying to regulate computer searches ex ante: There’s a hard problem there, and this is at least one kind of attempt at an answer. It’s certainly well meaning. With that said, the Vermont Supreme Court’s view of warrants is very similar to the Ninth Circuit’s view that the U.S. Supreme Court unanimously rejected in United States v. Grubbs, 547 U.S. 90 (2006). Unfortunately, the Vermont Supreme Court doesn’t even discuss Grubbs. It is mentioned only in passing in a footnote, after which the court says that it reads Grubbs and similar cases only to say that “hard and fast rules” on warrants are “frowned upon.” It’s not entirely clear to me what the Court is thinking with that description, although it may be that the Vermont court reads Grubbs as saying that appellate courts can’t require magistrates to take these special steps but that it leaves open that appellate courts also can’t stop them if they choose to at their discretion. From that perspective, magistrates regulate searches, not appellate courts: Appellate courts shouldn’t interfere with magistrates in their effort to apply the Fourth Amendment ex ante. I think that’s backwards. It’s up to appellate courts and the Supreme Court to say what the Fourth Amendment requires, not individual magistrates. More broadly, the right way to protect privacy in computer warrant cases is to eliminate the plain view exception and amend statutory rules governing computer warrants, not to give individual magistrate judges the discretion to impose (or not impose) their own preferred set of rules in each case.

With that said, I greatly appreciate the Vermont Supreme Court’s extensive analysis of the issues as well as its engagement with the relevant scholarship, including my own. This is just the first of many appellate decisions that courts will issue on this question, and I have long thought that the question was ultimately destined for the U.S. Supreme Court. Trying to figure out how to regulate computer searches is a really tough Fourth Amendment problem, and courts are destined to divide over it. With the Vermont decision, the game is on: Other courts can now weigh in and agree or disagree. As always, stay tuned.

Powered by WordPress. Designed by Woo Themes