In a legal opinion, judges are supposed to back every assertion with a citation to a source, like a statute or a previous case. Providing detailed references to existing authority helps show that an outcome is based on law, not just vibes.
Sometimes, however, it is really just turtles all the way down, and tracing each citation backward reveals that the original principle rests on nothing but air. So it is with federal courts discussing the ostensibly unique dangers that police officers face during traffic stops—a danger that judges have long used to justify allowing the cops conducting those traffic stops to do whatever they want.
A recent Fourth Circuit case, Sharpe v. Winterville, shows how these house-of-cards decisions unfold. Sharpe is about whether people can film the police making traffic stops, and the holding—that a blanket prohibition on live-streaming police interactions could violate the First Amendment—isn’t terrible. Along the way, though, the three-judge panel spends no small amount of time discussing the issue of the danger of traffic stops to cops. When that happens, the opinion unravels quickly.
Dijon Sharpe was a passenger in a car stopped by police in Winterville, North Carolina. During the stop, he started live-streaming on Facebook. One of the officers tried to reach through the window and grab Sharpe’s phone; “We ain’t gonna do Facebook Live because that’s an officer safety issue,” he said. Later, the same officer explained, “That lets everybody y’all follow on Facebook know that we’re out here. There might be just one me next time.” The officers threatened to arrest Sharpe if he streamed to Facebook Live again.
Ultimately, the Fourth Circuit held that Sharpe had credibly alleged a First Amendment violation against the town of Winterville, to the extent that it had an informal policy prohibiting livestreaming. However, the court also decided the officer was entitled to qualified immunity, as there was no “controlling authority” in North Carolina to let the officer know that stopping someone from livestreaming violated the First Amendment. As a result, Sharpe can pursue his claims against the town, but not the cop.
But this isn’t really a story about qualified immunity. This is a story about how judges help transform police mythology into federal law that prevents victims of police abuse from holding their abusers accountable.
There’s a long history of cases in which judges reiterate that stopping a car is dangerous for police officers. The Sharpe panel, for example, relied on Michigan v. Long, a 1984 Supreme Court case, to conclude that “officers often face increased risk during traffic stops from passengers in the stopped vehicles.” But the facts of Long—the facts that are supposed to support the case’s conclusion—do not reflect any particular danger to officers. In Long, police had stopped David Long for driving erratically and asked for his license and registration. On the floor of the car, police saw a closed folding knife. After confiscating the knife, the cops patted him down, flashed a light in the car, and found some weed, which is how Long ended up in court. But there’s no evidence in the record that he grabbed for the knife, or in any way moved to endanger the officers.
The Long Court relied on Pennsylvania v. Mimms, a Supreme Court case from several years earlier, for the proposition that there is an “inordinate risk confronting an officer as he approaches a person seated in an automobile.” Mimms, unlike Long, involved a stop of someone who had a gun, but the officer noticed it, frisked the man, and arrested him. Again, there’s nothing in the case that says Mimms lunged for the weapon or in any way threatened the officers; the fact that he had it, apparently, was enough.
Mimms, in turn, relied upon United States v. Robinson, yet another Supreme Court case from 1973. There, Robinson was stopped for driving with a revoked license, of which the cops were aware of based on a previous investigation. After ordering Robinson out of the car, the cops searched him and found a crumpled cigarette pack that felt heavier than usual, and turned out to contain capsules of heroin.
Again, there’s nothing particularly dangerous about the facts of the case—no evidence of a weapon, or that Robinson was behaving aggressively or erratically. But in a footnote, the Court finally provides some sort of factual support for its claims about the dangers of traffic stops: a study from 1963, and a brief from the government noting that, according to the FBI’s Uniform Crime Reports, 35 police officers had been murdered between January 1973 and March 1973, including 11 during traffic stops.
This is bad. It is also wildly out of date. Line of duty deaths are significantly lower than they were in the early 1970s. Although 458 officers died in the line of duty in 20201, 301 were attributable to COVID-19-related illnesses, according to an analysis by the National Law Enforcement Memorial and Museum. Of the 58 traffic-related deaths, 31 were attributable to accidents, and 27 occurred when an officer making a traffic stop was struck by another vehicle. The National Law Enforcement Memorial and Museum’s analysis lists only three firearms-related officer fatalities attributable to traffic stops—fewer than one percent of officer fatalities. Yet this data point forms the basis for the notion that police face a “special danger” that allows them to shred the Constitution whenever they pull someone over.
Chief Justice John Roberts, at the very tip-top of the federal court system, once dismissed attempts to quantify the effects of gerrymandering as “sociological gobbledygook.” It’s that sense of contemptuous intellectual superiority that allows judges to get away with ignoring data whenever it doesn’t conform with their priors. Instead of exploring the real dangers of traffic stops, which disproportionately target people of color, the court cites a 39-year-old Supreme Court case that cites a 46-year-old Supreme Court case that cites a 50-year-old Supreme Court case that is borderline irrelevant in 2023. Sharpe is the worst kind of legal decisionmaking: no factual support, just judges patting each other on the back and calling it a day.