George Floyd’s death under a Minneapolis police officer’s knee triggered protests all over the U.S. in 2020. In the incident’s aftermath, The Associated Press launched an investigation into deaths that occurred after police used tactics known as “less-lethal force.”

Documenting Police Use of Force, a new FRONTLINE/AP documentary, takes an in-depth look at the circumstances under which people died after encounters with police using tactics such as physical holds, Tasers or body blows that are meant to stop people without killing them. Made in collaboration with the Howard Center for Investigative Journalism programs at the University of Maryland and Arizona State University, the documentary draws on an investigation built from police records, autopsy reports, hours of police body-camera footage and a database that documented more than 1,000 such deaths over a decade. The film was directed by Serginho Roosblad, produced by Mike Shum and Roosblad with reporters Martha Bellisle, Ryan J. Foley, Kristin M. Hall, Aaron Morrison and Mitch Weiss.

Roosblad, as well as the AP’s reporter Weiss and editor Justin Pritchard, spoke with FRONTLINE about how the team sifted through thousands of documents, built an extensive database and navigated reporting obstacles to create a comprehensive visual narrative that traces why and how such deaths occurred in the wake of police using what is known as “less-lethal force.”

This interview has been edited for length and clarity.

After George Floyd’s death, your team wanted to know how many people died following non-shooting encounters with the police. How did you arrive at that idea and what inspired you to create this database?

Weiss: I had a source who said to me, “You see George Floyd’s death, but it’s more common than you think.” Medical examiners play a role in [how] these deaths [are categorized]. And it sounded interesting enough for me and my colleague, Holbrook Mohr, to start looking into it. That’s how it started.

We started finding right away that the role of a medical examiner in a police[-related] death plays a big role in whether or not there’s going to be accountability.

What went into creating that database, collecting the information, organizing it? How many people worked on it? And what challenges did you encounter?

Pritchard: We filed, by the time this is done, more than 7,000 requests for information to government entities — what are called public records act requests.

We were looking to identify tips and leads from many, many different sources, and then go and find the video, find the police report, find the autopsy, the investigative file through these public records act requests and use that primary documentation to establish what happened.

Explore: Lethal Restraint

So the challenges of building that were both technical and human resource, as well as resistance on the part of agencies that receive these records act requests. Some of them produced what we would consider to be a full set of records in a timely way. That would be the exception. Many took many months. Sometimes, you hear back a year or more later. And then you get records that have been blacked out in certain important respects or video that’s been blurred.

Serginho, with so much data available in the project, how did you decide what to include in the documentary?

Roosblad: It was massive. It was pretty daunting at the beginning. Where do we go? But slowly, being in these weekly meetings, talking with the reporters, there were emerging certain types of force that were used over and over again.

Chemical restraint, forced sedation was one of the first ones that really stood out to me. I have never heard of that. I heard about the case of Elijah McClain in Colorado but didn’t know how widespread it was.

Read more: Dozens of Deaths Reveal Risks of Injecting Sedatives Into People Restrained by Police

Another thing that really struck me once I got to the project was the amount of video that the reporters were able to collect. From there on, also being able to see certain patterns, and one of them was the fact that so many people said the words “I can’t breathe” before they lost consciousness and eventually died.

Mitch, you also spoke with family members of those who died after the police used “less-lethal force.” How did you prepare for those difficult conversations?

Weiss: The one thing you don’t want to do is go in right away and say, “Hey, I got a video. Let’s watch it.” It’s a slow process. One thing I found is if you’re honest, you show empathy, you keep your temper, you never get defensive and you are understanding, usually they’ll respond. You just have to place yourself in their shoes when you go to their door and when you talk to them.

Justin, you mentioned how the federal government wasn’t documenting these types of deaths in a systematic way. Since the investigation collected records from all over the U.S., and states and departments have different procedures, what did the team have to do to make sure the information in the database was accurate, and what kind of work went through verifying and fact checking all the cases the team found?

Pritchard: With the help of colleagues who are more computer savvy than me, we created a system. We had to adapt a way of entering the data through, essentially, a program we called Harvester. AP developed it. And [we] enter as many specific facts that we wanted to collect, within limits.

Now, we have a big set of information that is all tied back to a primary document or a video. We then used a second system, which is called Document Cloud. You upload documents there and you can annotate them. We effectively annotate every fact that is entered. And then we hand that whole thing over case-by-case to a team of fact checkers who then go back using a third system called GitLab.

So that was under-the-hood of how we did it. We used these three different systems, tailored them for these specific needs.

Serginho, what would you consider were the most difficult challenges your team faced when making the documentary?

Roosblad: I think just the sheer size of this project was challenging and daunting — the fact that [the database] covers all 50 states. We’ve seen some journalistic projects around some of the issues that we’re touching on but it’s relatively small. It’s either looking at just one particular state or it’s looking at one type [of use] of force. Here, it’s all these different types of forces that were used all over the country over 10 years. So I think just to be able to distill all of that in basically [in] 52 minutes, it has been pretty challenging.

And then, also, how much do you show of the body cam of someone dying? [When] showing this to a general audience, how do you make sure that they are not getting bombarded with just people dying from the first minute to the to the last minute of the film? But then, at the same time, you’re still showing the [use of] force, you’re still showing the issue at hand. I think that has been a challenge.

How does this data help us understand the larger picture of police use of force and deaths after police encounters? Were there any findings that surprised the team?

Pritchard: There certainly are patterns that emerge. It’s interesting to highlight those and then get the feedback post-publication. So we had people who are in the policing world who research these issues. That was one of the things that I think they appreciated about what we’ve done.

This is a database that allows for that type of analysis. It has limitations too because of the [obstacles] we talked about. You can’t do comparisons. We can’t say that this is everything because I’m certain that there are cases that for various reasons we just didn’t get to log. We didn’t know about. But you can see patterns and that is something that folks have commented on, saying, “That’s what would allow us to do an analysis through our expert eyes, and perhaps recommend changes in policies.”

I think that some [findings] are not surprising or are pretty straightforward. Almost all of the people who died were men. It happens everywhere across this country. This is not just an urban issue. It happens in the suburbs. It happens in rural areas.

There are other interesting patterns, for example, the use of sedatives by paramedics or medical personnel, sometimes at the request or direction of police. There were dozens and dozens of cases where someone was injected with — or sometimes it goes up their nose — a sedative, as part of their encounter with police. That was what was surprising to me, the extent to which that’s happening nationally.

Another thing I would say that I think is important is that if you look at the racial and ethnic breakdown of this, [the] non-Hispanic Black population in the U.S. is about 12% and non-Hispanic Black people were about 33% of our cases. So there’s a disproportionate impact. What’s also part of that, on the flip side of that is that approximately half of the people who died were non-Hispanic white people. And I think that in a lot of cases this is seen as an issue that affects only certain types of people. And for me it was quite interesting to see the extent to which conventional wisdom is not reflected fully in the data. Yes, there’s a disproportionate impact but if you look at it, the single largest group of people who have died in our database are white people, which I think is important because it shows to people who might not otherwise see this as an issue that affects them as something that they might need to be concerned about as well.

Weiss: Most of the people died in their homes or right near their homes. And so what that told me was that a lot of them were having some kind of crisis and either they called or a family member called. They were looking for an ambulance. They were looking for medical help. Instead, the police came. When that happened, all hell broke loose because most of the police officers did not know how to handle somebody in either a mental health crisis or in a drug crisis.

What do you hope the audience will take away from this documentary, the database, interactive story and the reporting?

Roosblad: This film is not necessarily a police film. [It’s] a film about human interactions that have gone wrong. And once we know how these interactions happen and what goes wrong, then we can find ways of preventing them in the future.

Pritchard: I see the publication of this database as a public service.

Everybody has an equal opportunity to go into it whether you are [for] defunding the police or whether you’re always on this side of the blue line. Whoever you are, you have facts now that you can look at to inform your opinions or your argument. Let the facts fall where they fall. At least we have them.

 

---