The car of Philando Castile is seen surrounded by police vehicles in an evidence photo taken after he was fatally shot by St. Anthony Police Department officer Jeronimo Yanez during a traffic stop in July 2016. Picture released June 20, 2017. Minnesota Bureau of Criminal Apprehension/Handout via REUTERS.

In their book “Suspect Citizens,” the political scientists Frank Baumgartner, Derek Epp and Kelsey Shoub dive deep into an extraordinary database on how police officers treat ordinary citizens. They kindly answered some questions via email. A lightly edited transcript is below.

The focus on this book is traffic stops. What is the value of studying them?

Traffic stops are far and away the most common interaction that people have with law enforcement. They therefore play a central role in forming our perceptions of the police and show us in turn how the police view us. Do we look like full citizens, or do we appear as suspects? Frequent stops for minor traffic violations, especially if followed by a request to search the vehicle, send an unmistakable signal to those who experience them that they are viewed more as potential suspects than as full citizens.

The book is based on data on 20 million traffic stops in North Carolina. Where did those data come from and what kinds of information do they contain?

In the late-1990s, the concept of “driving while black” began getting national attention. North Carolina became the first state to mandate the collection of traffic stops data in 1999, thanks in large part to efforts by black representatives in the state legislature.

The database includes information on why the driver was pulled over, the outcome of the stop and demographic information about the driver. It also has an anonymous identification number for each officer as well as the time of the stop and the police agency that conducted it.

The initial law focused only on the State Highway Patrol, but it was expanded two years later to cover almost every police agency in the state. As a result, we have a record of virtually every traffic stop in the state since 2002.

The idea was to settle once and for all if “driving while black” was a legitimate grievance. The bill passed with bipartisan support and with the blessing of the leadership of the Highway Patrol. Editorials suggested either it would disprove the ugly allegations of bias, or that police leaders would take immediate steps to resolve any issues the data might reveal. The law also mandated that the state issue periodic reports analyzing the data.

None of this has occurred, which is one of the reasons we wrote this book.

A key topic in the book is who is targeted for traffic stops. What did your findings reveal about racial disparities in both who gets stopped and what happens during the stop?

We found that, compared to their share in the population, blacks are almost twice as likely to be pulled over as whites — even though whites drive more on average, by the way. We also discovered that blacks are more likely to be searched following a stop. Just by getting in a car, a black driver has about twice the odds of being pulled over, and about four times the odds of being searched. Hispanic drivers, overall, are no more likely than whites to be pulled over, but much more likely to be searched.

These racial disparities are particularly pronounced among men rather than women, and younger men rather than older ones. So the numbers certainly validate the idea that young black and Hispanic men are commonly viewed as suspects, not as citizens, by the police.

You note in the book that these racial disparities would be less noteworthy if they were simply due to underlying differences in criminality. But that’s not the case, correct?

It certainly does not appear to be the case. African Americans are much more likely to be searched after a stop than white drivers, but less likely to be found with drugs, guns, alcohol or other forms of contraband after discretionary searches. Hispanic drivers in particular are much less likely to be found with contraband after a search, compared to whites.

So blacks and Hispanics are more likely than whites to be stopped and searched, but whites who get searched are more likely to have committed a crime?

Contraband hit rates are 36, 33 and 22 percent for whites, blacks and Hispanics, respectively. So, yes, whites are more likely to be found with contraband. Note that all contraband hits do not lead to arrest, as many “hits” are very small amounts. But our general point is that the stereotyping that seems to be widely used — based on race, age and gender — puts young men of color at a distinct disadvantage. They are over-targeted, statistically speaking.

If it’s not criminality, what does help explain these racial disparities?

Stereotyping based on demographics. The way you look. The time of day or night. The location. All these factors have a significant impact on the likelihood of a search. Young men, especially minority men, are searched much more often. When we control for every factor included in the database, we cannot explain that disparity with any nonracial variable, and we do not find that it is justified by contraband hit rates or even arrests. Many police leaders have said it relates to location: those in high-crime areas are searched more. But when we were able to control for this, we still found that blacks were searched at higher rates than whites in the same area. So, as far as we can tell, the police use visible cues to determine the likelihood of criminality, and those tend to cause an overly great focus on young men of color.

You have a quote from defenders of this kind of policing: “You have to kiss a lot of frogs before you find your prince” — meaning, you have to stop lots of people who aren’t really criminals in order to find a criminal. But you argue that the costs of this kind of policing outweigh the benefits.

Using traffic stops to crack down on violent crime and drug trafficking has always been an inefficient strategy. It would be one thing if the police could tell by looking who might be involved in crime. But we find that just about 30 percent of searches lead to contraband hits, but only about 10 percent of these are what one would think of as a large hit. Most of the contraband discovered is very small amounts. For example, if it is ounces of drugs, it is less than one ounce, with fewer than 1 percent of the seizures being more than 10 ounces. Typically, contraband seizures are so small they do not lead to arrest.

And since searches occur only in about 3 percent of traffic stops, our conclusion is that we would all be better off if traffic stops were focused simply on unsafe drivers and if the police used other means to investigate crimes. Making 10,000 traffic stops in order to find three criminals is not a good strategy.

When Philando Castile was killed in 2016, research later showed he had been pulled over at least 46 times previously. Naturally, someone surveilled so commonly by law enforcement understands that they are more a suspect than a citizen, in the eyes of the police. We need to recognize the costs of these police strategies and compare them with the benefits.

Is there any kind of solution that might reduce racial disparities while perhaps actually improving the effectiveness of policing?

Absolutely. Emphasize traffic safety rather than regulatory enforcement. And reduce the search rate, particularly in cases without probable cause. These two simple reforms can reduce racial disparities dramatically. After such reforms were initiated in Fayetteville, N.C., community trust in the police appears to have increased, with no increase in the crime rate. In fact, the reforms helped reduce crime by enhancing community trust and cooperation with the police.

How has this research been received by police officers or leaders that you’ve interacted with?

After we issued our first 12-page report, the North Carolina Association of Chiefs of Police commissioned a 100-plus page report dismissing our findings as “deeply flawed.” A bill was introduced to stop collecting the data. So, it’s probably fair to say the initial response was hostile.

But, like the stages of grief, this response has shifted from anger and denial to something more positive. Chiefs, sheriffs and other law enforcement leaders have engaged with us very constructively and taken their own steps to attempt to reduce these disparities, to monitor their officers, and to take better account of just how they want to use the vehicle code.

Most recognize that drunk drivers, excessive speeders and those who ignore stop signs are true threats to community safety, and focusing on them will save more lives than questioning poor and minority community members about a crack on their windshield. So, we have seen significant engagement, and we look forward to more.

Find more details about “Suspect Citizens” here.