We’ll never know what Philando Castile was feeling when the police lights first flashed across his rearview mirror on a balmy night in the summer of 2016. But we can be reasonably certain of what he wasn’t feeling: surprise. The traffic stop—ostensibly for a broken tail light—that precipitated his tragic death, and captured the nation’s attention, was nothing out of the ordinary for Castile. It was in fact the 46th time he had been pulled over. And while this figure may seem shocking to some, there is sadly nothing aberrational about it.
Our traffic laws are designed to function this way. The offenses that populate our traffic codes are so numerous and sprawling that the average driver can scarcely hope to spend a few minutes behind the wheel without committing an infraction. Simply hanging an air freshener from the rear-view mirror or failing to illuminate your license plate can get you pulled over and ticketed in some places. It’s only a matter a time before any driver falls prey to the minefield of traffic infractions that awaits them every time they fire up the engine.
As a practical matter, this gives police nearly unbounded discretion to pull over any driver, any time they choose. And one way that they put this discretion to use is by engaging in what’s known as pretextual traffic stops. These are when police have a hunch that a driver has committed a crime (or perhaps the driver just looks “suspicious”), but not enough evidence to stop and interrogate them about it. So instead, police just pull the driver over for a petty traffic offense. Once the motorist is stopped, officers can start questioning her about the non-traffic crime that they’re really interested in. The traffic stop is just a pretext for their true motivation.
This is exactly what happened to Philando Castile. As dispatch audio recordings would later reveal, the officer believed that Castile’s “wide-set nose” fit the description of someone who had committed a robbery. The officer, as the world would later discover, was mistaken. (The officer who shot Castile was acquitted of the charges he faced related to the shooting.)
So, what’s wrong with pretextual stops? For starters, they don’t make us safer. Rigorous studies have shown that pretext stops turn up evidence of non-traffic crimes at abysmally low rates, and that they have no effect on crime rates. These same studies confirm that when we invite officers to be led by their gut instincts and other unchecked heuristics, it is people of color who are disproportionately affected. Racial disparities in who gets pulled over erode trust in the police, and deepen the perception that police use race as a proxy for criminality.
There are roughly 20 million traffic stops every year, many for petty infractions. For drivers like Castile and Daunte Wright, these stops can quickly turn lethal. But the consequences need not be deadly to have lasting effects. Even citations for minor traffic offenses can lead to hefty fines, and for those unable to pay, late fees and even license suspensions. And this is to say nothing of the fright, humiliation, and other indignities that these stops visit upon the innocent.
Read more: She Wrote a Book About Bias. Here’s How She Thinks Police Departments Should Approach Reform
Policymakers have begun to take notice. Philadelphia recently became the first city to implement a comprehensive policy to curtail pretextual traffic stops. I applaud its efforts. Policymakers owe it to their constituents to divert taxpayer dollars away from the failed regime of pretext stops and invest that money in programs that will keep officers and communities safe.
That is why, on Wednesday, in my role as a Police Commissioner for San Francisco, I introduced a new regulation to do just that in our city. If enacted into law, here’s what the new regulation will do.
First, it prohibits stops for low-level traffic offenses, like failure to illuminate your license plate. If you can’t stop someone for this type of offense, you can’t leverage it as a pretext for further investigation. To be clear, officers can still make traffic stops for things that pose a real threat to public safety; drunk drivers and drag racers are still fair game. They can even still enforce the low-level violations if they wish. All they need to do is take down the driver’s license plate number and send them a ticket in the mail.
Second, it limits consent searches. Prohibiting officers from asking for consent to search cars once they’ve pulled the driver over—except under certain narrow circumstances—will reduce the incentive to make pretextual stops in the first place, as well as the temptation to go searching for a needle in a haystack.
Finally, it will improve data collection. By letting the numbers tell the story, policymakers and the public can draw evidenced-based conclusions about police traffic enforcement and cast a light on any racial disparities.
My hope is that, after soliciting and incorporating the public’s input, the Police Commission will be able to vote on a final version of the regulation this fall.
Now here’s why—regardless of your politics or feelings about police reform—you should be all-in for limiting pretext stops.
For starters, the data out of major metropolitan areas like Nashville show that pretext stops are not making anyone safer and don’t have any impact on crime rates. Yet these stops do put officers in a potentially dangerous situation. While most traffic stops will be routine, it’s impossible to know when a driver pulled over for a broken taillight will be a criminal with a murder warrant who doesn’t want to be taken in.
The substantial amount of money and officer time spent carrying out these ineffectual stops could all be redirected to strategies proven to stop and prevent crime — including greater responsiveness to emergency calls.
While commentators often reflexively assume that there is an inherent tradeoff between police reform and public safety, that is generally not the case. And curtailing the use of pretext stops is a prime example. It doesn’t matter whether you’re chiefly concerned with officer safety, or better police response times, or racial disparities in the criminal justice system: everyone comes out a winner when pretext stops are phased out. Seldom is there such a confluence of diverse stakeholders that stand to gain from a single policy proposal.
More Must-Read Stories From TIME