Most AEI scholars were hatched with perfect issue knowledge. If you see Robert Doar in a white coat, he’s probably been creating a new scholar in the laboratory—a lair, really—far below the D’Aniello building. A rare human among the group, I’m still learning my issues. Below I’ll share some learning gained by producing testimony on Flock camera systems for the Pennsylvania House Republican Policy Committee.

Sure, I’d like people to think I was hatched. Preparing for this hearing about Flock cameras, I found that I imperfectly predicted this issue 25 years ago. In 2001 testimony to the U.S. House Transportation Committee, I said:

Red light cameras are only the first installation of the Big Brother infrastructure. There’s very much more to come, and very soon, too. Optical character recognition, for example, may already allow red light cameras to read license plates and scan for specific cars. Networked cameras will be able to track cars as they travel through a city and on the highways. Database technology will make it possible to record the movements of cars and make a permanent record of them….

Ten years ago I identified the privacy problem with the license plate. Technology has integrated these mundane metal rectangles into driver tracking.

But I am still learning. I was delighted recently by the case the Pacific Legal Foundation’s Amy Peikoff makes that mandating a license plate creates an easement on our cars. The easement allows for occasional observation and identification, which is reasonable. But using it to archive drivers’ movements without suspicion of wrongdoing abuses the easement. Watch this space, because a lot of work is making the Fourth Amendment amenable to textual interpretation using traditional legal concepts such as property. We don’t need the airy “reasonable expectation of privacy” test.

The Supreme Court used seizure and search articulately in 1968’s Terry v. Ohio. Cleveland Police Detective Martin McFadden espied Mr. Terry casing a store. Officer McFadden “‘seized’ petitioner and subjected him to a ‘search’ when he took hold of him and patted down the outer surfaces of his clothing.” It was the year after the Katz decision, which gave birth to befuddling “reasonable expectations” doctrine.

Writing my testimony made me firmer in my belief that indiscriminately using these systems searches for all cars, and by close inference all drivers, all the time. That is unreasonable, as it does not directly lead to evidence or fruits of crime. Watching all of us in case we do something wrong is not how the American system works.

Reconciling visual public surveillance with Fourth Amendment concepts also strengthened my belief that processing of imagery gathered in public is the search. Generally, anyone can record any visual or audible material that is publicly exposed. But when the government processes such imagery to bring out particular content—license plate numbers, bumper stickers, the faces of drivers—that, I believe, is searching. It has the “purpose of finding something” that the Supreme Court brought to the “search” concept from Noah Webster in 2001’s Kyllo decision.

My 2017 article for the National Constitution Center discusses the two conceptual parts of searching, and how technology can reverse them. Information technology can “canvass” the real or digital area where an object may be found before the object of the search is determined. That article also helps show how the “purpose of finding something” can help Fourth Amendment administration of “search.”

In the hours before my presentation to Pennsylvania’s Republicans, I watched part of a joint hearing on Flock cameras in the Pennsylvania House Transportation and Judiciary Committees. It brought me to thinking about Terry and small seizures and searches that are allowed without a warrant.

Before giving any time to warrantless vehicle tracking, I want to emphasize how small the Terry aperture is. Terry is essentially a flavor of exigency. When life or limb are in danger, or when evidence will soon abscond, it is reasonable to seize or search minimally without a warrant.

The case gave rise to the phrase “reasonable suspicion based on articulable facts.” I would tighten the standard to articulated facts. Investigating officers should state their basis for searching and seizing on body cameras, radios, or otherwise, not just have the theoretical ability to do so.

Let image collection happen. To nest Flock cameras with Terry, processing should await when articulated facts make a search reasonable. Often, that will mean getting a warrant, but exigencies will allow processing without. Non-suspects will have their cars searched. But rules for very fast disposal of irrelevant results may bring such searches within reasonable limits.

I say all this with great certainty! … of the fact that I will continue to learn about these fascinating problems. Unlike my colleagues, I was not hatched.