We can't fix policing without talking about race. This cartoon explains why.

Black and white Americans experience law enforcement in fundamentally different ways. We know this because of data.

Police hassle and question black people more frequently.

Police use force against black people more frequently .

And police are disproportionately likely to kill a black person than a white person.

Some Americans are uncomfortable talking about police reform in the context of race, and have urged reform so that policing can be better for everyone. But the problem is that, throughout American history, law enforcement has always been unfair to an underclass of people — and today that divide is most pronounced along racial lines.

And it’s caused this split in how black and white Americans feel about police:

These feelings are important because, in order for police to do their job, a community has to comply with them. And they will only willingly comply if they approve of police actions. This is what is called "legitimacy," and it's what gives modern police their authority in a democratic society.

But as the data shows, police are more likely to treat blacks poorly and, in turn, blacks are more likely have a legitimacy problem with police. Meanwhile, white Americans tend to be treated much better, so in turn they give more legitimacy to police.

So that's what this comic is about: The gap between how black and white Americans view police — how there has always been a group of Americans who believe the police to be a legitimate and rightful institution; how there has always been a group who have painful anecdotes to reject that notion; and how we can fix it.

So what does it take to earn legitimacy? Let's do an experiment.

If police need legitimacy to do their jobs, then they need to figure out how to earn it. But what does it take to earn it?

In 2008, Yale law professor Tracey Meares conducted an experiment to find out. She led a group of researchers who took about 1,400 people and showed them videos of police encounters.

But before each video, they split participants into three groups and gave them different information to frame the police actions as either lawful, unlawful, or ambiguous. The participants weren't necessarily told if the situations were legal or not, but it was enough that most lawyers would agree whether or not it was constitutional.

After researchers primed the participants with this information, they showed the three videos:

Afterward, the participants were asked various questions, including whether or not they wanted to punish the police officers in the video. The goal was to figure out what people considered rightful policing.

The researchers also asked about things that didn't have to do with legalities, like how fair, transparent, and respectful the officers were during the encounters — something called "procedural justice."

In short, this is the part of policing that isn’t in the law books. Rather, it’s about how police officers interact with people.

Then the researchers adjusted for a participant’s previous interaction with police, so it wouldn’t skew the results.

The results? People cared less about legality, and more about procedural justice.

It turns out that participants didn’t care that much about what is technically legal when deciding to punish an officer.

Rather, they cared about procedural justice — or lack thereof.

"Even when the public says that they care about legality, and they do say that, our research suggests that the way they construct legality actually has nothing or very little to do with how lawyers think about that," said Meares, the Yale researcher who also recently served on President Obama's police task force.

To model this point, she often includes this diagram in her work:

But it’s not just the way people are treated during encounters. The more you’re stopped, the less legitimate you feel they are.

In another study, Yale and Columbia University researchers surveyed about 1,300 men between the ages of 18 to 26 from diverse neighborhoods in New York. They asked them about their history with police, including how often they were stopped by cops.

And then they asked them to rate how legitimate they perceived police to be.

What they found was, after adjusting the responses for a respondent’s criminal history, people gave less legitimacy to police when they had been stopped by police more often.

This hints that when police hassle people excessively, they lose legitimacy.

In addition, their research backed up Meares’s study in finding that the most important factor for legitimacy was procedural justice — and that both black people and white people valued it greatly. Meanwhile, the thing that hurt legitimacy most was if a stop was intrusive, which is an example of low procedural justice.

This all gives us hints on how to think about police reform

The policing Americans want isn’t necessarily the kind that is written in law. Rather, it’s about the way people are treated during the encounters — and we generally agree what this "rightful" policing looks like.

But as shown at the top of this story, black people are far less likely to experience rightful policing. They are more likely to be stopped, pushed into a wall, and even killed.

That’s why the president's police task force, as well as activist groups like Campaign Zero, have tried to frame police reform around ways to encourage procedural justice. While this doesn't give us an exact roadmap, it does show us what is broken about the everyday encounters between police and their communities — and how fixing that goes a long way toward building trust.

But modern policing is partially responsible for a massive drop in violent crime. Why change something that’s working?

If the job of police is to keep communities safe, then it only makes sense to judge their performance on crime data.

And in the last 25 years, violent crime has been cut in half.

There are many theories as to why, but a Brennan Center study found that more police officers, paired with data-driven technology like CompStat, was the biggest factor. It allowed for more efficient and more focused policing.

But policing wasn’t always efficient, and it wasn’t always this professionalized. This era came about after generations of rich and powerful people used police for their own gain. In the early 1900s, police chiefs were appointed by politicians, and police officers often took payoffs. The institution was widely corrupt.

But in the 1950s, police leaders pushed back and implemented strategies that gave us the institution we know today. This is what's called the "professionalization" of police, which brought about military efficiency and organization of local police, as well as preventative tactics to reduce crime.

However, it turns out the point of policing isn’t just to reduce crime numbers.

In fact, only looking at crime numbers can be harmful

Meares, the Yale law professor, says that while these policing tactic may have reduced crime numbers, it also encouraged aggressive policing — and that destroyed the trust between communities and police, a trust that allows police to do their jobs.

"Research is clear that how people are treated is central to how they view police and other legal authorities — even more than whether police are effective at reducing crime," Meares wrote in the Washington Post. "That’s because people do not simply experience police interactions; they also learn from them."

We had inklings this was true a long time ago, which is partially why "community policing" exists. It encourages police to be a part of the neighborhoods they protect in order to rebuild that trust, instead of swooping in to enforce minor violations.

But despite huge investments from the federal government, a recent review found that many departments treated community policing as a "buzzword." And at police academies in the US, community policing strategies continue to be a small part of the curriculum.

But even knowing this, it's still hard for many Americans to support the movement in a racial context

When you ask Americans about specific police reforms — like putting body cameras on officers — most support the proposals.

But if you ask them about Black Lives Matter, only 14 percent of white people strongly support it, versus 41 percent of black people — even though this movement pushes for many of the same proposals Americans widely support.

The difference, however, is that it urges police reform in the context of race.

Some Americans are trying to decouple the issues of policing and race, often retorting by saying "All Lives Matter." But as the data shows, these two things are linked in such a way that it's silly to talk about police reform while ignoring race.

Meanwhile, some Americans believe this movement is a sharp critique of police officers themselves, and not the institution. But Ron Davis, who was director of President Obama’s police task force and served 28 years as an officer, says these are two separate things.

"Even if you have great cops, if the systems are bad, they're still going to have terrible outcomes," he said.

So why can't we get back to the days when police treated everyone well? Because we've never, ever had that in America.

It's tempting to think that we just need to restore the justice that existed before police brutality came to the limelight. And it was especially tempting after the recent spate of police shootings in Louisiana and Minnesota — and the subsequent targeting of police officers. It felt like this problem was spinning out of control.

It's a common sentiment, and one that Republican presidential nominee Donald Trump has embraced. "We have to give power back to the police because crime is rampant," Trump told NBC.

But what's crucial to understand is that we've never had this justice before. In fact, the very early days of policing were founded on targeting immigrants, minorities, and the poor.

In many cities, early police were often used to rein in labor strikes as work conditions deteriorated.

In the South, the root of early law enforcement was slave patrol, responsible for quelling revolts, pursuing runaway slaves, and disciplining slaves who broke plantation rules. Later on, after slavery was outlawed, police enforced Jim Crow laws.

In short, law enforcement as an institution has always favored the rich and powerful — and it has always given less value to the lives of the marginalized.

Throughout American history, this treatment has created an underclass of people who are criminalized, isolated, and labeled as inherently dangerous, says criminologist Gary Potter.

And while overt expressions of racism and classism have mostly faded, there are still enough unspoken hints that create this underclass. Meares writes, "On the streets, the hidden curriculum of policing can be seen in how people are treated in interactions with law enforcement. Too often this hidden curriculum sends certain citizens signals that they are members of a special, dangerous, and undesirable class."

So when we talk about police reform, what we're trying to figure out is how fair and just law enforcement can exist in America, and how it can be a legitimate institution for everyone, and not a source of fear — all for the very first time.