Confused? Okay, so here’s the longer version of the answer. I’ll start with the history of the “search” test before Katz; then turn to the Katz test and cases interpreting it; and conclude by responding to criticisms of the Katz test.
I. The Meaning of Fourth Amendment Searches Before Katz
Let’s start with the early history, drawing on this recent article. Until the 20th century, there was very little caselaw on what is a Fourth Amendment “search.” The issue didn’t come up, in part because the Fourth Amendment only applied to the federal government (which did very little in the way of law enforcement) and in part because there was no exclusionary remedy. But at the time of the Framing, the debates on the Fourth Amendment and the cases that animated it mostly used the word “search” in the sense of breaking into someone’s house and searching it. That is, when folks talked about searches, they talked about acts of physical invasion and rummaging around.
Of course, just looking at the text, the word “search” has several possible meanings. As I have written:
On one hand, a search might refer merely to looking for an item. In that sense, you might gaze up at the night sky and search for the Big Dipper. On the other hand, a search might mean the act of observing an item closely. For example, you might search a page to find a particular word. Finally, a search might refer to the physical act of looking through a space in ways that expose its contents to plain view. In that sense, you might search a bedroom closet by rifling through the clothes it contains. The ambiguity of the word ensures that a wide range of concepts might plausibly define the meaning of searches.
This creates a problem. The fact that the Framing-era debates spoke about “searches” in the context of examples of the last type of search, that of breaking in and rummaging around, doesn’t tell you if a broader principle should apply to identify a search:
Devising a test from a set of examples raises a level-of-generality problem: Examples alone cannot identify how far beyond their facts the principle should extend. Clearly, physical entry of individuals inside the home to find evidence counts as a search. At the narrowest level, then, a search might be only a physical entry by government officials. A broader approach could focus on whether the officials interfered with property interests, such as whether a trespass occurred. Or perhaps the test should be whether the government interfered with privacy, with physical intrusion being just one example of government acts that violate privacy interests. Examples alone cannot identify which principle to use.
In the Court’s first major Fourth Amendment case, Boyd v. United States, 116 U.S. 616 (1886), the Court indicated that the Fourth Amendment should apply beyond physical entry and should instead “apply to all invasions on the part of the government and its employees of the sanctity of a man’s home and the privacies of life.” “It is not the breaking of his doors and the rummaging of his drawers that constitutes the essence of the offence,” the Court wrote, “but it is the invasion of his indefeasible right of personal security, personal liberty, and private property[.]”
That sounds like a pretty broad privacy-focused concept, but that broad approach was largely narrowed in the period from the 1920s to the 1960s. During that time, the Court eventually articulated the notion of physical intrusion into a protected space as the primary guide to what is a search. The basic idea at the time was that a physical intrusion was a search, but that conduct that was not a physical intrusion generally was not. The Court was never really clear about that being the test, but that’s the basic idea that emerged from the Court’s cases. That’s where the law stood in the early 1960s.
II. Katz and the “Reasonable Expectation of Privacy” Test
The Court ditched the physical intrusion idea in Katz v. United States in 1967, when it held that taping a microphone to the top of the phone booth and listening in on a call “searched” the phone booth even though there was no physical intrusion into the booth. The Katz opinion was very murky about the reason for its holding. The Court was relying on something more than physical intrusion. But what was it, exactly? Justice Harlan concurred with his two-step test that introduced the “reasonable expectation of privacy” idea. And within a decade after Katz, the Court eventually adopted the Justice Harlan’s formulation. In practice, the first part of the test (subjective expectations) is irrelevant, for reasons I’ll explain in a forthcoming article that I’ll post soon. But the key to applying Katz is the second part of the test. As phrased by Justice Harlan, the question is whether “society is prepared” to recognize an expectation of privacy as reasonable.
At first blush, Justice Harlan’s test only makes matters more confusing. Who is “society,” and how are Supreme Court Justices (of all people) supposed to know what it is “prepared” to recognize as reasonable? Don’t we have legislatures around to determine what “society” thinks?
I gather this is Justice Alito’s question: Just what is the Court supposed to do when it asks whether an expectation of privacy is “reasonable”? Maddeningly, Supreme Court decisions have avoided answering this question directly. See, e.g., O’Connor v. Ortega, 480 U.S. 709, 715 (1987) (O’Connor, J., plurality opinion) (“We have no talisman that determines in all cases those privacy expectations that society is prepared to accept as reasonable.”); Oliver v. United States, 466 U.S. 170, 177 (1984) (“No single factor determines whether an individual legitimately may claim under the Fourth Amendment that a place should be free of government intrusion not authorized by warrant.”).
To answer this, let me offer three points. First, as a practical matter, the Court only rarely strays from the basic physical intrusion approach to identifying what is a “search.” The adoption of the “reasonable expectation of privacy” test seemed like a big change, but 99% of Fourth Amendment fact patterns can still be answered by the old physical intrusion test. Surveillance that breaks inside a space is a search; surveillance that remains outside the space is not. That’s why the reintroduction of physical intrusion test in Jones is probably more about form than substance. Because the “reasonable expectation of privacy” test mostly tracks physical intrusion, you mostly end up with the same result under either test. The Katz test does do some things that the physical intrusion test doesn’t, to be sure. For example, pointing a thermal imaging device at a home is a “search” of the home under Katz even though technically there is no physical intrusion into the home. Similarly, taping a microphone to the top a phone booth is a Katz search but not a physical intrusion into it. But in both cases, the Katz test serves a pretty narrow role: It is used to cover a sort of “virtual entry” into the item that is not covered by a physical intrusion.
Second, the Court has interpreted the “reasonable expectation of privacy” test in ways that try to distinguish less invasive practices that don’t need regulation from more invasive practices that do. Here I’ll refer readers to this article, which looked at all the Supreme Court caselaw to try to figure out what makes an expectation of privacy “reasonable.” My basic finding was that the Supreme Court looks to a variety of methods (what I call “models”) to try to differentiate serious privacy invasions from less serious ones. The more serious privacy invasions are deemed to violate a reasonable expectation of privacy, and the less serious privacy invasions are not. And in each Supreme Court decision, there is an eye to the normative policy question of what should be a search. In some cases, it is the exclusive focus. For example, in Hudson v. Palmer, the Court ruled that there is no reasonable expectation of privacy in a prison cell because it would be a bad policy to allow privacy rights there.
Third, the Court has generally avoided resting the Katz test on the empirical question of what reasonable people would actually expect. That is, the Court normally doesn’t apply Katz by imagining what a reasonable person would anticipate occurring in a particular situation. And that makes sense: The question is what counts as a government search, not what counts as a government surprise. There are a few cases that focus on what reasonable people would anticipate, to be fair. For example, in Bond v. United States, the Court held it was a search when the government squeezed a bus passenger’s duffel bag to feel the brick of drugs inside on the theory that most bus riders don’t expect others to manipulate their baggage that way. But my sense is that in such cases, the question of what reasonable people expect acts as a stand-in for an implied consent argument like we saw last Term in Florida v. Jardines. The question of what reasonable people would actually expect is really about implied consent to a search that renders the search reasonable, not what is a search in the first place.
III. A Brief Comment on Alternatives to Katz
A good rule of thumb is that every discussion of the Katz test prompts several people to say that Katz is dumb and should be replaced by some better test. I’ll offer three brief responses. First, in my experience, most of the people who make this argument don’t appreciate how the Supreme Court actually interprets Katz. Critics usually imagine the test works pretty differently than it actually does.
Second, if you think some other test is better than Katz, I think it’s important to say what that better test is and how it should apply. The real challenge in this area is the level of generality problem I pointed out in Section I that first came up in the Boyd case back in the 19th century: How broadly beyond physical intrusion should we extend the notion of a “search”? One approach would be to stick only with physical intrusion, the core facts that motivated the enactment of the Fourth Amendment. Under that approach, non-intrusion cases like Kyllo and Katz are wrongly decided. That’s a coherent approach, I think, although I haven’t seen it advocated often: It leaves the virtual intrusions unregulated, which most seem to oppose. But once you want a level of generality broader than that, you need to say just how broad and how far it goes. What’s the formulation of the test, and how should that formulation apply? Some scholars argue that everything should be a search, and the rules for what is allowed should be generated under the reasonableness clause instead. But that’s just a punt, I think, as it’s just moving the hard question to a different doctrinal box rather than actually answering it.
Finally, I haven’t said anything yet about the mosaic theory of the Fourth Amendment. Re-reading Justice Alito’s concurring opinion in Jones in the wake of his question during the Wurie argument, it occurs to me that his mosaic approach is based on the assumption that the Katz test is empirical in nature rather than normative. He says that the Katz test is whether the government’s monitoring is what a reasonable person “would have anticipated.” He then draws a line based on an assumption about what would surprise people: in his view, it wouldn’t surprise people to learn that the police tracked a car for a short period but it would surprise people to know that the police tracked a car for a long period. This is an odd way to approach Katz, as I explained here. Some privacy law theorists have construed Justice Alito’s approach as signaling his adoption of novel and creative new theories of constitutional law. But in light of his question during the Wurie argument, I think it’s more apt to say that this brief passage in Alito’s Jones concurrence likely just reflected an assumption that the Katz test is empirical rather than normative. As I have argued in this post, the better view is that the test is normative rather than empirical, in which case the mosaic approach may be less likely to be adopted the next time the issue reaches the Court.