Over the last few years, I’ve dedicated more and more effort to righting the Fourth Amendment, which has been weakened over decades by doctrines that don’t measure up to the times.

You can see my efforts and their evolution in my American University Law Review article, “Reforming Fourth Amendment Privacy Doctrine” (2008); Cato’s brief to the Supreme Court in U.S. v. Jones (Oct. 2011), Cato’s brief to the Supreme Court in Florida v. Jardines (July 2012); my Cato Supreme Court Review article, “Escaping Fourth Amendment Doctrine After Jones: Physics, Law, and Privacy Protection (Sept. 2012); my Cato Policy Report article, “U.S. v. Jones: Fourth Amendment Law at a Crossroads” (Sept./Oct. 2012); and, most recently, Cato’s brief to the Supreme Court in In re: EPIC (August 2013).

Today, I had the opportunity to expound on my thinking at a National Press Club event hosted by the Electronic Privacy Information Center to discuss their challenge to the National Security Agency’s bulk telephone data collection. Moderator Jeffrey Rosen, recently named President and CEO of the National Constitution Center, alloted me a good deal of time, and we discussed things a little more after the session. I’m ever-sharpening my thinking about how the Fourth Amendment should operate, and how to talk about it.

The starting point is this: The “reasonable expectation of privacy” doctrine, which grew out of Katz v. United States (1967), is a failure. Courts almost never actually investigate whether a subjective “expectation of privacy” is objectively reasonable, and they’re in no position to make broad societal pronouncements on the latter question anyway. The doctrine is not a product of the Katz majority, it’s worth noting, which focused on the steps Katz had taken to conceal the sound of his voice—steps upended by government agents’ placement of a bug in a phone booth without a warrant.

The Fourth Amendment should be administered as a law once again. To administer a law protecting “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures,” you’d ask four questions:

1. Was there a Seizure? (Many searches begin with seizure, so it makes sense to ask about seizures first.)

This goes to the question whether government agents invaded a property right. Unconsented and unwarranted entry into one’s home (or onto its nearby surroundings, the “curtilege”) is the classic Fourth Amendment violation. Given the directed looking that accompanies that property violation, courts often refer to it only as a “search.” But they don’t always collapse search and seizure. In Terry v. Ohio, for example, the court recognized the neat distinction between grabbing a suspect and turning him around (a seizure) and then patting him down (a search). In the Jones case, Justice Scalia clearly fixed on the property invasion in affixing a GPS device to a car. His majority opinion collapsed that small seizure with the search it facilitated and talked about it as a “search.” But it started with the small seizure of property that converted Jones’s car to the purposes of the government in transporting their tracking device.

Seizure is in play with digital data reflecting our communications. In our In re: EPIC brief, Randy Barnett and I wrote about the seizure that occurs when the government requires telecommunications providers to turn over data. It doesn’t matter what the government does next. The seizure must satisfy the Fourth Amendment.

Property and contract principles can and should protect communications under the Fourth Amendment. In the 1929 Olmstead decision, which wrongly approved warrantless wiretaps, Justice Butler dissented, pointing out that “contracts between telephone companies and users contemplate the private use” of telephone facilities. “The communications belong to the parties between whom they pass,” Justice Butler said. That’s true of modern communications as well. Some overlook this, treating contract law as a sort of legal also-ran. Explicit and implied contract terms promise us privacy. Courts and commentators should recognize that.

2. Was there a Search?

As I noted, seizures and searches often go hand in hand. But sometimes there are free-standing searches. These, too, must comport with the Fourth Amendment.

The best example is from Kyllo v. United States (2001), in which government agents used a thermal imager to observe abnormally high heat on the side of a home that they suspected of housing a marijuana growing operation. They did not enter onto the property but used their high-tech device to convert invisible radiation into visible images of the home’s emanations. They were “looking for or seeking out that which is otherwise concealed from view”—the definition of “search” from Black’s Law Dictionary.

Think of “search” as the exact opposite of “plain view,” the doctrine holding that law enforcement are entitled to note what they observe from lawful vantage points. It’s okay to look at stuff. And it’s a search when you use special efforts or outré technology to observe things law and physics otherwise didn’t permit you to see.

3. Was the thing seized or searched a person, house, paper, or effect?

There are a few things law enforcement might seize or search that aren’t protected by the Fourth Amendment—one’s farmland or unimproved vacation property, for example. But digitization is often just the rendering of constitutional papers and effects in a different format, which should not strip them of protection.

4. Was the Seizure or Search Reasonable?

The “reasonable expectation of privacy” test collapses many questions together, reasoning backwards, actually, from expectations to the existence of searches. Under that doctrine, a “search” is almost always unreasonable because it defeats a reasonable expectation.  But the right way to look at things is to follow the terms of the Fourth Amendment: See if there was a seizure. See if there was a search. And if there was either, decide if it was reasonable. This requires the contemplation of the reasonable seizure or search.

Posit a law enforcement officer walking down the street. She trips on crack in the sidewalk and reaches out to steady herself on a nearby automobile, leaving a noticeable smudge. The car was not hers to use this way. But for an instant, she converted the car to her purposes. It’s a small but real seizure of another’s private property.

Our instinct in such a case is to say, “that was no seizure.” The better way to think of it is as an entirely reasonable seizure. Were she to have converted the car to her purposes in a different way—attaching a GPS device to it, for example, so as to track its movements—this is not reasonable without a warrant supported by probable cause.

The same goes for reasonable searching. Say our law enforcement officer is at the beach. Espying odd behavior—maybe an incipient fist-fight—off in the direction of the wharf, she raises her binoculars to her eyes and looks at what is happening there. That’s the kind of directed looking that qualifies as a search, but the existence of something “odd” and her use of a relatively ordinary technology place the search well within the bounds of reasonableness. In a second scenario where a government agent sets up on a bluff and uses a military-grade instrument to read over a random beachgoer’s shoulder, that’s unreasonable searching.

There is certainly judging to be done in close Fourth Amendment cases. There’s no escape from that. But there’s a better way to administer the Fourth Amendment than the “reasonable expectation of privacy” test. That is to use the ordinary meanings of the words in the Fourth Amendment, to employ relatively familiar and settled property and contract concepts, and to ask judges to strike balances based on the facts in individual cases rather than make sweeping pronouncements about privacy.