Published by The Lawfare Institute
in Cooperation With
In a recent post here at Lawfare, April Doss argues that the Ninth Circuit’s decision in United States v. Mohamud “got it right.” In her view, the critics of the decision—myself included—are just wrong. I disagree. Here’s why I think the reasoning of the decision is hard to defend.
First, an important caveat: My argument is about reasoning and not result. A lot of discussion on Mohamud focuses on the bottom line of which side won. That’s understandable, but it’s not my concern. My concern is whether the reasoning of the case makes sense as a matter of Fourth Amendment law.
When viewed from that perspective, Mohamud strikes me as awkward and unconvincing. Its core holding is weak and hard to square with Supreme Court doctrine. Other parts of it border on the incoherent. And it could have reached its result much more easily without the conceptual wrong turns.
I. There Is No “Targeting” Doctrine in Fourth Amendment Law
In Mohamud, the government obtained e-mails from a monitoring point inside the United States that collected e-mails between Mohamud inside the United States and another person who was a foreigner outside the United States. Mohamud had Fourth Amendment rights; the other person didn’t. How does the Fourth Amendment apply to a communication in transit between them? Here’s the Ninth Circuit’s first key Fourth Amendment holding:
[When] a search was not directed at a U.S. person’s communications, though some were incidentally swept up in it— [it] does not require a warrant, because the search was targeted at a non-U.S. person with no Fourth Amendment right.
As I read this, how the Fourth Amendment applies depends on who is the “target” of the monitoring. If Mohamud had been the target, hypothetically, then a warrant would normally have been required under United States v. Warshak because he was a party to those communications. But because the “target” was the person who had no Fourth Amendment rights, the rule for accessing the communications dropped to a lower general reasonableness requirement without a warrant.
The most obvious problem with this conclusion is that there is no “targeting” doctrine in Fourth Amendment law. Targeting is a core concept in foreign intelligence collection. The relevant statutes are person-based. They create different rules based on who is the person being targeted for monitoring.
The Mohamud case is about the Fourth Amendment, however, not statutes. In Fourth Amendment law, the concept of “targeting” doesn’t exist. According to the Supreme Court, the Fourth Amendment uses objective rules. Fourth Amendment law focuses what the government does, not what the government is thinking when it does it. As the Supreme Court has emphasized in literally dozens of cases, the subjective state of mind of the government is simply irrelevant.
This is a big problem for the reasoning of Mohamud, I think. The idea of the government’s attention being “directed” at a particular person, and therefore that another person’s communications are only “incidentally” obtained, appears to be a largely if not primarily a subjective inquiry of the type that the Fourth Amendment consistently prohibits. It’s a conceptual framework that fits the foreign intelligence statutes. But it is foreign to Fourth Amendment law.
II. The Absence of Support for the Ninth Circuit’s Rule
In defense of its holding, Mohamud relies on cases saying that “incidental collections occurring as a result of constitutionally permissible acquisitions do not render those acquisitions unlawful.” But I don’t see how that is helpful. First, the question in Mohamud is what counts as a constitutionally permissible acquisition. If A and B are communicating, the government will acquire the communication between A and B. The question is whether that acquisition is lawful.
Second, the authorities that the court relies on for that sentence were cases concluding that incidental collection doesn’t violate the federal Wiretap Act when the government has a wiretap order. In those cases, there is a warrant (a Wiretap order) that says that certain things can be collected and certain things can’t. Evidence can be collected “incidentally” when it is outside what the wiretap order authorizes to be collected. The cases interpreting the wiretap statute say that such collection is not a statutory wiretap act violation. But how is that helpful? This is a Fourth Amendment case, not a case about the criminal wiretap statute.
So where in Fourth Amendment law might the court get the idea of having special rules for “targets”? As the Wiretap Act cases suggest, caselaw on Fourth Amendment warrants distinguishes between evidence that is responsive to a warrant and evidence that is non-responsive to a warrant. When the government has a valid warrant, and the warrant authorizes the search for and seizure of particular communication, the law treats their seizure differently: The seizure of the communications described in the warrant is justified by the warrant, while the seizure of other communications outside the warrant has to be justified (if it can be) by some other doctrine such as a variant on the “plain view” exception.
That feels at least a little like a “targeting” doctrine, right? But that line of cases doesn’t offer support for the Ninth Circuit’s holding for three reasons.
The first reason is that the relevant caselaw is about how doctrines and orders can authorize different searches and seizures. The warrant authorizes the seizures of some communications but not others, so you have different rules for the different seizures of different communications. In contrast, Mohamud is talking about a single seizure of a communication between Mohamud and a target. The distinction drawn elsewhere to justify different seizures in different ways doesn’t work when applied to justify a single seizure. You can’t readily split the atom of a single e-mail.
The second reason the warrant caselaw doesn’t help is that it’s specifically about the scope of warrants. Warrants are special under the Fourth Amendment. They’re special because they turn unreasonable searches and seizures into reasonable searches and seizures. A section 702 collection order is obviously not a warrant. A non-warrant judicial order has no Fourth Amendment significance, at least outside the exclusionary rule context. A warrantless search is still a warrantless search, even if a judge issued a (non-warrant) order permitting it to satisfy a statute. Given that, the caselaw on the scope of warrants has no obvious relevance.
Finally, the idea of using plain-view-like principles to create a targeting/incidental distinction is in considerable tension with the Supreme Court’s guidance. In Horton v. California, the Court held that whether collection is inadvertent or intentional is irrelevant to whether the plain view exception applies. In Horton, the government had a warrant for some evidence but the police actually were looking for other evidence not named in the warrant. According to the decision, whether the government had a secret plan to get the evidence outside the warrant didn’t matter. All that mattered was whether the seizure was objectively reasonable.
III. What’s the Harm, and the Path Not Taken
At this point you might be thinking: Okay, maybe the reasoning of Mohamud is hard to square with Supreme Court doctrine. But if the rule isn’t a problem, what’s the real harm?
I think the harm comes when you try to apply Mohamud to a criminal case. Criminal investigations do not use the concept of a “target,” meaning that the holding of Mohamud has to be mapped on to a context where targeting would presumably mean only subjective intent.
This creates a risk of all sorts of problems. Imagine the government is interested in collecting evidence about a criminal conspiracy. All but one of the members of the conspiracy are U.S. citizens inside the U.S., while one member of the conspiracy is a foreigner located abroad.
Under Mohamud, it appears that the government can just declare the foreigner the “target” of the investigation and conduct warrantless searches for the group’s communications. That might include physical searches, such as breaking into a suspect’s home without a warrant. The possible involvement of one member of a conspiracy being a non-US person abroad might mean that there is no longer a warrant requirement protecting anyone else. That kind of “reverse targeting” is prohibited by statute in the specific context under Section 702. But Mohamud isn’t a statutory case; it’s a Fourth Amendment case. And there is no such general prohibition in Fourth Amendment law, or in the criminal context where the Mohamud case will be used next.
One counterargument might be that Mohamud only means targeting in the sense that is used in the national security investigations. Mere subjective intent, as in criminal cases, should be excluded from consideration. But if that’s right, the court should have rooted its holding explicitly in a national security doctrine rather than announcing a new general Fourth Amendment principle.
Specifically, the court could have easily just said that a national security or foreign intelligence exception to the warrant requirement recognized a targeting concept, and that Mohamud’s holding was limited to that context. There’s authority for the view that you can look to subjective intent to decide whether the special needs exception applies. If the court had placed the Secton 702 Fourth Amendment analysis in that doctrinal box, it could have crafted a narrower rule that didn’t create headaches elsewhere. But the court didn’t do that, and that means its holding presumably applies to the criminal context.
IV. The Strange Doctrine of Diminished Expectations
The other strange turn in Mohamud is the court’s conclusion that Mohamud had only a “diminished” expectation of privacy in e-mails because of the third party doctrine. This part of the opinion strikes me as particularly puzzling.
Here’s some context to understand why. There are two basic ways to look at whether there is Fourth Amendment protection in e-mail. The first way is to focus on e-mail as a messaging system, like phone calls or postal letters, and to say that that e-mail retains full Fourth Amendment protection until delivery just it does with other messaging technologies to maintain Fourth Amendment protection across different technologies. Under this view, accessing e-mail is a search, for which the party retains Fourth Amendment protection until delivery, because the third-party doctrine doesn’t apply to e-mail.
The second approach is to say that the third-party doctrine does apply to e-mail and that there is no Fourth Amendment protection in e-mail once it leaves the sender because it is disclosed to the network. If you like to think by analogy, you might say that e-mail is like a postcard instead of a sealed letter. Under that view, accessing an e-mail after it has been sent is not a search and the Fourth Amendment does not limit it.
The Mohamud decision appears to get to its view that there is a diminished expectation of privacy in e-mails by suggesting that these two irreconcilable views are both correct. On one hand, the discussion starts with the first view: E-mail is like a letter, the court says, and therefore it receives full Fourth Amendment protection. But then the opinion shifts to the second view, invoking the third party doctrine cases which would say that there is no Fourth Amendment protection at all, and saying that they apply, too. The court accepts these two views at the same time and concludes that there is only a diminished expectation of privacy.
There are two ways to read that passage. The first is that this is the Fourth Amendment as quantum physics. Accessing e-mail is a search and a non-search simultaneously, so the result is apparently a kind of averaging where you get a “diminished” expectation of privacy.
The second way to read that passage is that the court just misunderstands the import of the third-party doctrine cases. The court says that under the third-party doctrine, Fourth Amendment rights are “diminished.” But that’s just wrong. Where the third-party doctrine applies, Fourth Amendment rights are completely extinguished. That’s why you can’t say that the Fourth Amendment applies but that the third-party doctrine applies at the same time. You have to choose.
Granted, some believe that the Fourth Amendment shouldn’t protect e-mail under the third-party doctrine. That’s a possible view, although one I don’t think is correct. But if the court had that view, and if they could square it with the Ninth Circuit’s recent decision in Kitzhaber, then they could have so held and created a clear split with the Sixth Circuit in Warshak to prompt Supreme Court review. That would have been a plausible path. But the court’s simultaneous embrace of the third party doctrine and cases that reject the third party doctrine doesn’t seem to make sense.
Once again, my problem is with the reasoning of the decision—and the implications for other cases—and not the result in this one case. The Ninth Circuit could have easily said in this case that e-mail is very private but that the access was still reasonable under the broad reasonableness balancing test. For an example of that kind of reasoning, see the Second Circuit’s analogous reasoning in a telephone wiretapping case, In re Terrorist Bombing.
The problem with saying that there is only a diminished expectation of privacy in e-mail in this case is that it gives the government an argument in other cases—cases that have nothing to do with Section 702 or national security—that privacy rights are generally diminished online. The passage doesn’t do much work in this opinion, and it doesn’t seem to change the outcome. But I would guess that DOJ will bring up that passage in other Internet privacy cases far removed from the context of this case.