"It's 2:30 in the morning and my phone rings. My daughter says, 'Daddy, you need to come to the hospital,' " Michael Bell told me, of the moment in 2004 when he learned that his son had been shot by a police officer in their hometown, Kenosha, Wis.
Twenty-one-year-old Michael Bell Jr. died that night from a bullet wound to the head. In the nightmarish hours that followed, his father expected independent investigators to arrive on the scene and find out what had gone wrong. A former Air Force pilot, he knew that when an accident happened in the military, a forensic team performed an exhaustive review. Above all, he wanted to make sure that if a mistake had contributed to his son's death, it would be identified and fixed, so that nothing like it would happen again.
This investigative method is standard in aviation. When a plane crashes, experts pick through the wreckage to determine the cause and make recommendations to prevent the next accident. The process is so effective that for the last several years, the death rate from crashes of American commercial planes has been zero. But no comparable system exists in policing — and that may help explain why you are far more likely to die at the hands of a cop than to perish in an plane crash. Police officers in the United States now kill about 1,000 people and wound more than 50,000 every year.
Of course, no independent team arrived to perform a forensic analysis of the younger Bell's death. Instead, the Kenosha police department spent two days investigating its own officers before ruling that the shooting was justified. The police officers claimed that Michael had "failed to make a complete stop" (and tests later showed Michael had been drinking), so they followed him to his house and parked behind him. According to the police, the young man had lunged at them and tried to pull a gun out of an officer's holster.
Bell hired his own investigators. They contend that it all began with faulty equipment: Officer Erich Strausbaugh's holster caught on a cable dangling from one of the cars' side-view mirrors, so that when he tackled Michael, he felt a powerful tug on his belt. Assuming that the young man had grabbed for his weapon, he called out to his partners, "He's got my gun." Michael's mother and sister, who were watching nearby, yelled that Michael did not have the gun. But it was too late.
"My blond-haired boy was killed," Bell said, "and then blamed." He continued, "If that was how it was for my family, then I knew that the families of African-American, Hispanic or Asian boys didn't stand a chance. That was one of reasons I started raising a ruckus."
Police violence is tangled up with racism and systemic injustice. We desperately need to do more to address that, foremost by shoring up the criminal justice system so that it holds police officers accountable when they kill. But it's also true that deadly mistakes are going to happen when police officers engage in millions of potentially dangerous procedures a year. What aviation teaches us is that it should be possible to "accident-proof" police work, if only we are willing to admit when mistakes are made.
Bell, in fact, does not blame Strausbaugh, who committed suicide several years later. "The officer made an honest mistake," he said; the problem is that "the police department covered it up."
In 2010, the family received some vindication when the city of Kenosha agreed to pay $1.75 million to settle a wrongful death lawsuit. Afterward, Bell paid to erect billboards asking: "When police kill, should they judge themselves?" In 2014, Wisconsin passed a law requiring independent investigations of police actions that result in a civilian death.
Bell is still pushing for reform, touring Wisconsin with graphs and charts — think of him as the Al Gore of police shootings. In meetings in legislative offices, he explains that a proven method to improve safety already exists in the fields of medicine, nuclear power and aviation. Engineers call it an "external-learning system."
After an airplane plummets into a cornfield or a swamp, the National Transportation Safety Board sends a "go team" to interview survivors and pick through the debris for evidence of mechanical failures. Those investigations have led to revelations about how hidden problems can spin into disaster. For instance, in 1986, a single-engine airplane plowed into a jet in the air above Los Angeles County. That accident killed 82 people and led to new rules that made flying safer: Small aircraft flying close to major airports are now required to use transponders that indicate their position to controllers, and airliners are outfitted with collision avoidance systems.
Bell is calling for an agency like the National Transportation Safety Board to do the same thing for police procedures that end in violence. He has found an ally in Van Wanggaard, a former police officer and independent accident investigator who is now a Republican state senator in Wisconsin.
"We need a response team that doesn't have any horse in the race," Wanggaard told me. "We need safer ways of doing things. It's a lot like the medical world; you have the potential for a doctor to screw up and end someone's life." The challenge now, he said, is to teach lawmakers, citizens and police officers about the power of "accident-proofing" the police.
To that end, Wanggaard and Bell helped organize a meeting earlier this year that brought representatives from NASA, the transportation safety board and Harvard Medical School into the same room with law enforcement officials to discuss such an approach.
John Carli, the police chief of Vacaville, Calif., attended the meeting and said it opened his eyes. "If you could prevent just one fatal encounter, why wouldn't you?" he asked.
Carli also pointed out that police departments could overhaul the way that officers and citizens interact at close range, to make both feel safer. For instance, when an officer asks you to reach for your license and registration, that act can be mistaken for a grab for a gun — a chain of events that probably contributed to the death of Philando Castile last year. "That's a potential deadly threat that's got to be slowed down and re-engineered to create a higher degree of safety," he said.
The way it works now, civilians often feel like the ones responsible for keeping violence at bay. In May, an African-American journalist named Tonya Jameson was changing the license plates on the used SUV she had recently bought when an off-duty officer accused her of stealing the car — and pointed his gun at her. Ms. Jameson had proof of ownership in a bag that sat on the ground nearby, but she dared not move a muscle.
Jameson told me that the words "black people can die like this," kept running through her head. "That was my mantra. I knew I had to be careful not to freak him out. If he killed me, no one would know what really happened."
She spoke in a soothing voice, explaining to the police officer that if he looked in the bag, he would find the bill of sale. Instead he kept his gun trained on her and called in reinforcements.
Later, writing in the Charlotte Observer, Jameson described just how much nerve it took not to dig into that bag herself. "I fought every impulse to do anything" that the officer could interpret as threatening, she wrote. "I'm the one thinking my life could end if he panics. Yet I'm the one expected to remain calm. It seems that the legal system is really asking civilians to de-escalate adrenaline-fueled cops."
Could a design fix prevent situations like these and keep the police from pointing guns at unarmed people?
The routine traffic stop, like the one that killed Bell's son, is especially in need of redesign because it contains so many potential failure points that cause confusion and violence. In the computer science department at the University of Florida, a team of students — all African-American women — have developed a technology that they hope might make these encounters far safer.
The students — Jessica Jones, Dekita Moon, Michelle Emamdie and Isabel Laurenceau, working under Juan Gilbert, the chairman of the department — began by digging through data on deaths that occurred during traffic stops. "We studied these violent tragedies, and then ran them backward in time," Jones said. "We asked ourselves where the trouble began. Where was the point at which everyone was still safe?"
They discovered that as the traffic stop progresses, the dangers increase for both civilians and police officers. For instance, when an officer walks along the flank of a parked car, he cannot see the driver's hands — just the kind of situation that leads to a deadly misunderstanding. To make matters worse, confrontations often take place next to a highway with cars whizzing by at 70 miles per hour.
The whole thing seems almost diabolically designed to go wrong. But what if the police officer could stay inside his car?
The students created a smartphone app, called Virtual Traffic Stop, that allows him to do just that. After the officer pulls a driver to the side of the road, he would use the app to start a video chat with the driver as a first point of contact, allowing him to observe whether the driver seemed mentally ill or dangerous, notice clues in the interior of the vehicle and review identification documents.
Of course, there are considerable hurdles. Millions of drivers would have to download and learn to use the app. And police departments are likely to resist any changes to the traffic stop, which gives them broad authority to search cars for drugs and guns.
Still, even if the app fizzles, it represents an intriguing use of 21st-century technology to solve longstanding problems in policing.
Their project inspired me to dig into the history of the traffic stop. How was it developed and when?
It goes back more than 90 years to the Jazz Age, when bootleggers defied Prohibition laws by piling cases of whiskey into cars and speeding off to speakeasies. In response, the Supreme Court ruled in 1925 in favor of the "automobile exception" that allowed the police to search cars without a warrant.
Today, the legal patchwork of exceptions around the car means that with minimal cause — for instance, you forgot to use your turn signal — a police officer can ask for your identification, eyeball your stuff and make a judgment call about your behavior. If he decides that you're uncooperative, he can perform a deeper search. In other words, when it comes to traffic stops, the law gives broad powers to conduct warrantless searches.
Is it any wonder that the procedure has become a favorite tool of law enforcement looking to seize guns and drugs before they hit the streets? According to the Bureau of Justice Statistics, about 26 million Americans were pulled over by police officers in 2011 alone — that's over 10 percent of the population aged 16 and older. Of course, because of racial bias, a disproportionate number of those people are minorities. In fact, you could argue that the way the traffic stop is designed is inherently racist, since it encourages stop-and-frisk methods that unfairly single out African-American drivers.
How can we fix this system that puts civilians and the police officers who stop them at risk? The obvious solution is to take the officers — and their guns — out of the picture whenever possible.
New technologies allow us to do just that. In some cities, when you roll through a stoplight, a camera catches you in the act, and a few weeks later you receive a ticket in the mail. Data suggests that this automatic system is far cheaper than "human" ticketing and reduces pedestrian deaths. And a camera can't kill people.
Of course, we do need state troopers to pull reckless drivers from the highway, just as we need to police drug and gun smuggling. But the highways aren't the only place to do that. Police officers should not be questioning people about minor infractions like a broken taillight, especially when we know that this procedure can end in death.
Even when no one is hurt, the confrontation causes a toxic distrust of the police and exacts a horrible mental toll on minorities. Before he was killed by a police officer's bullets, Castile had been stopped at least 49 times by officers. The stress of driving while black has poisoned the roadways for millions of Americans.
One of the most frustrating aspects of this problem is that we already have models for fixing it, whether it is a version of the National Transportation Safety Board, as Bell seeks, or an empowered citizen review board with strong investigative powers, which Jameson is calling for.
Michael Scott, a former police chief who is now a professor of criminology and criminal justice at Arizona State University, is a fan of aviation safety-proofing and told me that "we need a parallel system for the police." But, he said, law enforcement agencies have a long way to go because they lack the most basic tools for learning from their mistakes."
"We don't even know exactly how many officer-involved shootings happen every year," he said, because "we still do not have a single national reporting system that chronicles and documents every police-involved shooting in this country."
"Of course, it's important to have a criminal and an administrative investigation of any death" that involves a police officer, he added. But it's not enough to determine who is to blame; we also need to ask, "Why did this happen?"
Until we can answer that question, innocent people will continue dying at the hands of the police.
Pagan Kennedy is the author of "Inventology: How We Dream Up Things That Change the World."
© 2017 New York Times