Crime, safety, and policing

[00:00:27]

RICK WEISS: Hello, everyone, and welcome to SciLine’s Media Briefing on crime, safety, and policing. I’m SciLine’s director, Rick Weiss. And for those of you not familiar with SciLine, we are a philanthropically funded, editorially independent, and entirely free service for reporters based at the nonprofit American Association for the Advancement of Science. Our mission is simply to make it as easy as possible for you as reporters to include scientists’ sources and scientifically validated information in your news stories, whether those stories are about something sciencey, a real topic of science like COVID or climate change, or are topics about things going on in your community, as is very often the case. Issues like homelessness, like crime and safety. Where you might not realize it at first or think about it at first that there are a lot of scientists who actually study these topics and they can bring a lot of value to your stories.

Today’s briefing is part of a special series of briefings that we’re doing this month. It’s the fourth of six on topics that are in the news because of the lead-up to November’s election. Topics like abortion. Topics like the electoral system itself. And these are things that are covered, as I’ve mentioned, so often as simply matters of political opinion when, in fact, there’s some great research and evidence that you can add to your stories. Our hope is that when you do produce stories about these topics, you’ll include not just what people are saying about them, but what the evidence actually says about them. So please check out a link that we’ll put in the chat that will show you the rest of this month’s schedule. We’ve got two more briefings like this scheduled for next week. A couple of quick logistical points before we get started. We have two panelists today who are going to make short presentations of up to 10 minutes each. That will be followed by a Q&A period. If you have questions while they’re speaking or at the end of their presentations, please go down to the Q&A icon at the bottom of your Zoom screen, click on that, and tell us your name, your news outlet, and your question. And if you have someone you want to direct that question to, you can put that in as well. A full video of this briefing will be available within an hour or so after the briefing ends, and a transcript will be available a day or two after that—a couple of days after that.

Finally, I’m not going to take the time to give full introductions and bios for our speakers, those are on the SciLine website. I just want to tell you that you will hear first from  Charis Kubrin, professor of criminology, law & society at UC Irvine, who is going to give us sort of a big-picture snapshot of U.S. trends in various kinds of crime, particularly violent crime, and we’ll note several of the pitfalls that you need to watch out for as you report on this issue because it turns out that it really matters what kind of timescales or geographic scales or other scales you’re really talking about if you want to get it right. And second, we’ll hear from  Justin Nix, an associate professor in the School of Criminology and Criminal Justice at the University of Nebraska Omaha. And he’s going to focus on the relationship between crime and policing, another area of reporting where some real attention to detail, it turns out, matters. And on the relationship between policing and safety. And that includes actual safety and perceived safety, which turns out to be two very different things. So with that, let’s get started and I’ll turn it over to you, Kubrin.

[00:03:53]

CHARIS KUBRIN: All right. Thank you. All right. Can that—can everyone see that over there?

[00:04:04]

RICK WEISS: Perfect.

[00:04:05]

CHARIS KUBRIN: Wonderful. Well, first of all, thank you for having me. I’m really delighted to talk about crime in the U.S.: dynamics and social facts for reporting today. So we all know crime is always newsworthy, but never more so than during an election year. And if recent survey data have anything to tell us about what we’re in for this year, it’s that crime is going to be on the minds of most Americans as they head to the ballot box in November. Now, how do I know this? Well, recent data from Gallup tell us so. According to the data, 77% of Americans believe there is more crime in the U.S. than a year ago; 55% believe the same in terms of local crime. And a new high of 63% say the crime problem in the U.S. is very or extremely serious. Now, you may be wondering how these perceptions square with what we know about crime data. Well, it turns out in June the FBI released its 2024 quarterly crime report, and that compares crime trends from January to March of 2024 from the year prior, January to March of 2023. And what did we learn from this quarterly release? Well, violent crime decreased about 15%, homicide decreased a whopping 26%, property crime decreased as well. So lesson one, which I’m sure is not a surprise to any of you, is the large gap that exists between perception and reality when it comes to crime. Now, when this report was released, I, and many other criminologists, received calls from reporters asking us to make sense of the trends. And the biggest question that we always get asked is why. Why, in this case, did crime decrease so substantially?

But one of the things that I want to talk about today is that we cannot address the why about crime until we truly understand the what and we have a good handle on that. So today I want to talk about the what. I’ll share some fundamental social facts about crime in the U.S. Along the way, I’ll identify things to keep in mind as you’re reporting. And finally, I’ll close by reviewing some of the social correlates of crime that are of greatest interest today. And throughout the presentation, I’m going to focus on violent crime—homicide, in particular—for illustrative purposes, and because it’s the most serious crime and people are often the most concerned about it.

Now, one question I like to keep in mind as I am poring over crime statistics and trying to make sense of them, is the question “compared to what?” We cannot understand levels or trends in violence or crime without providing what I consider to be a comparative context. A key comparative context or perspective. This means assessing similarities and differences across social groupings such as time periods, places, and demographic groups. Once a comparative context is provided, the reality of violence comes into sharper focus, and we can gain a better understanding of what’s really going on when it comes to crime. And let me offer a few examples. I’ll start with time period, the time period over which crime statistics are compared. Now, the statistics I just shared with you cover a relatively short time period showing violence has increased, and substantially. But the story on understanding crime shifts quite differently if we choose a different time frame for comparison.

Here are two figures showing rates of violent crime offenses in the United States at the top and the rate of homicide offenses at the bottom from 1985 to 2021. And you can see on the left-hand side, on my left-hand side, we have the crime rate per 100,000 population. So on the one hand, if we focus on the recent past, say, the last few years of what’s been going on in terms of crime trends, we can see a pretty significant spike in violent crime, but especially homicide. Orient towards 2019 on the homicide graph and we see a huge jump up there in homicide. That is pretty troubling and quite different from this quarterly report that I just mentioned. On the other hand, though, if we take a longer snapshot, if we go all the way back to 1985, we see that even as homicide ticked up in this recent past in a troublesome way, it is nowhere near its all-time high in the 1980s. Now, this doesn’t mean we should minimize violence or accept the status quo. I always say one homicide is one homicide too many. My point is that time frames are crucial for contextualizing crime statistics. And I would be particularly wary of short crime trends or comparisons like year to year or even month to month. Time changes which are volatile and susceptible to noise and outlier cases really can mask overall trends, so it’s very important to consider different time frames.

Here’s the thing, though. National trends, no matter what time frame examined, belie critical variation and violence that exists across place, whether states, cities, or neighborhoods throughout the United States. If we think about the recent quarterly crime statistics I reported, it is the case that not all places equally experienced such substantial declines in violence and homicide. Some of these places even saw substantial increases in violence and homicide. Let’s start with variations state by state. So this is a figure that shows rates of murder in the United States in 2022 listed by state. And it’s not important that you look at every single state and figure out what its rate is. What I would rather you do is look at how each state compares to the national average. You can see that the national average for the United States as a whole was 6.3 homicides per 100,000 population. It’s marked there. You can see many states were well above that average, Louisiana, New Mexico, South Carolina. Check out the District of Columbia with a rate of 29.3 per 100,000. On the other end of the spectrum, many states had much lower-than-average homicide rates for that year. But we can go even more local by looking at variation across cities, both within a state and across states. Here is a figure that shows homicide rates in 2023 by a select number of cities. I got these data from a report that was published by the Council on Criminal Justice, a nonpartisan think tank of which I am a member. They do a lot of reporting on crime trends. They are an excellent resource, by the way. But you can take a look at a list of cities that they examined homicide rates for. Again, taking into account the national average, you will see cities like Baltimore and Memphis had much higher than average. Whereas, Austin and Arlington, for example, are much lower than average. But we can go more local yet by examining variation across neighborhoods within a given city.

Now, it’s a well-established fact that crime and violence do not occur randomly in a city, but rather, are concentrated in certain neighborhoods within a city. And the FBI does not publish neighborhood crime data. And I should say, for a lot of the past slides I’ve shown you, that’s FBI data. I have a slide with lots of information about the different types of crime data, including from the FBI, that you could report on. But in terms of neighborhood-level data, that is not made publicly available by the FBI, so criminologists such as myself have to go and collect it from police departments, which is a very arduous task. I have a lab that I codirect with another professor at UCI where one of our main tasks is going around and collecting neighborhood-level crime data for as many cities in the United States as possible. And when we plot that neighborhood-level data, we will see the extreme variation that exists by neighborhood. So this is a figure that shows us violent crime rates in Los Angeles neighborhoods in 2020. You can see that the neighborhoods are represented here as census tracts, a proxy for neighborhood. You can see the L.A. city boundary. And the lighter shade, the yellow, some of the early greens, indicate lower violent crime rate communities, compared to the darker blues which reflect communities that have higher violent crime rates.

Now, I want to push this point about variability even further if I can by emphasizing a really important finding in criminology. And that is that most neighborhoods, even high-crime neighborhoods, have streets where no violence has occurred at all. And this relates to a saying that criminologists have that I’m particularly fond of, and that is that crime is hyperlocal. And you can see that when we examine a map of street segments in the city of Los Angeles. So moving from neighborhoods to street segments. So you can see on my left here we’ve got Los Angeles street segments for the year 2020 with shading. Very difficult to see. When we zoom into a portion which shows you just one neighborhood there, you will see, literally, street by street the extreme variability that exists in terms of violent crime events. So the key point here is that a very large portion of violence is often found at a relatively small number of streets in any given neighborhood. And this social fact reveals something important we often lose sight of, and that is that a neighborhood might be extremely disadvantaged, it might have very high crime rates, but the majority of the streets in that neighborhood have little or no crime. In other words, violence – and homicide, especially – is a truly rare event.

But it is also the case that homicide is rarer for some groups than it is for others. If we look at variation just by two demographic characteristics, sex and race, we will see this in action. If we look at homicide offenders for the year 2022, FBI reports that 80% are male. We can compare that to the proportion of men in the U.S. population, which is 49%. If we look at race and, in particular, African Americans, we see that while they comprise 52% of homicide offenders, that’s much higher than their representation in the population. So males and African Americans, it turns out, and we can talk about other groups, as well, are overrepresented as homicide offenders in the data. And it turns out this is true for victimization data, as well. So I don’t have time to talk a lot about this, but one thing I want to emphasize is you can imagine if, instead of examining variation by sex or race or age, for example, we combined all of that to look at intersectionality, you could see that we would get very sharp contrasts. In other words, if we compared information on homicides, say, among young African American males versus older white females, that variability might come in even sharper focus. My takeaway point, though, from all of this, is that violence, and especially homicide, is rare, but it’s more common in some groups, in some places, and in some time periods than others. And what this really means is that violence is not idiosyncratic or random, as I so often hear people talking about it or see it reported on the news. You’ll hear stories, you know, this one—how did this happen? It seems so random. Rather, it is socially patterned. Of course, the million-dollar question is what explains this social patterning.

Now, in trying to understand violence, we often look for something in the individual. Something inside that person, whether it’s mental illness, or their inability to control their emotions, or some other individual deficiency. And while individual risk factors are indeed important, focusing solely on individual-level factors cannot explain the temporal, geographic, and demographic patterns of crime that I just described. So researchers often look for patterns of social organization and social arrangements that promote violence in our society. And in a slide that is woefully incomplete and insufficiently detailed, you can get a sense of the range of theorized social correlates of crime and violence. And these are important sort of above and beyond the individual-level risk factors. And they range from things like exogenous shocks. And what do I mean by an exogenous shock? Well, when something happens in our society that fundamentally alters social organization in society and the way we go about doing things; the COVID-19 pandemic, for instance. We can look at factors like people’s perceptions of institutional legitimacy, economic conditions, demographics, criminal justice policy and practice, guns and gun markets, drugs and drug markets, policing, police-community relations, as we’re going to hear from Nix here in a minute. It’s really important to remember that these social correlates of crime and violence vary by levels of analysis. So if you’re focused on street segments or neighborhoods, you would want to identify the causes or the correlates that make the most sense at that level, compared to examining national trends, where you would look for forces that are much bigger and broader and apply to all, for example, neighborhoods in the U.S. You also—we also know that these social correlates vary by crime levels versus trends. So we can look at what’s happening at one point in time and try and understand what caused that, but it’s also—it could be a different set of correlates if we examine trends.

And then finally crime type. I’ve focused on homicide here but, of course, there’s all different kinds of violent crimes, as well as many property crimes that have their own set of correlates. So I will stop there, but the last thing I want to say before I conclude is that in addition to research on crime trends, more generally, I do research specifically on two key issues. That is immigration and crime and criminal justice reform and crime. And given these are likely to be kind of key issues as we lead up to the election, during the Q&A, I’m happy to answer questions on those specific topics, as well. So, thank you.

[00:18:53]

RICK WEISS: Fantastic. Thank you for that wonderful introduction, Charis. And a reminder to reporters, all of these slides will be available immediately after the briefing so you can really dig into them. Tons of information there, a great intro. And let’s jump over to  Justin Nix.

[00:19:16]

JUSTIN NIX: OK. Thank you. And I really appreciate the opportunity to speak with you all today.

[00:19:20]

RICK WEISS: Justin, I think that’s not the screen you want.

[00:19:22]

JUSTIN NIX: Is that the notes? Of course, it wouldn’t be.

[00:19:25]

RICK WEISS: It’s your—it’s your desktop, I think.

[00:19:28]

JUSTIN NIX: All right. Apologies.

[00:19:29]

RICK WEISS: It’s okay.

[00:19:30]

 JUSTIN NIX: We practiced this.

[00:19:40]

RICK WEISS: There you go.

[00:19:41]

JUSTIN NIX: OK. Apologies for that. So for this talk, I’m going to try to stick to high-level trends, meta-analysis. Meta-analyses that are synthesizing large bodies of literature on what we know about the various topics that I’m going to cover. And so I would say that we can say that we have a high degree of confidence in the things that I’m going to talk about. Starting right off the bat with the relationship between police presence and crime. And so one of my favorite studies here was a retrospective study of 240 large cities over a 40-year period. And what the researchers found was that for every additional 10 or 17 officers hired, about one life was saved. And in per capita terms, these effects were twice as large for black versus white victims. But beyond just having the police, in criminology, it’s well-accepted by now that it really depends on what police are doing in terms of the beneficial crime outcomes that you’ll see.

So when police are focused on problematic places, particular problems, and people who are most at risk, right—people, problems, and places—they tend to have more of an effect than simply sitting back and waiting on 911 calls to come in or driving around randomly and waiting to happen upon a crime. And so I list several meta-analyses here that have synthesized dozens of research studies across a variety of contexts—big, medium, and small jurisdictions, some in the U.S., some in other jurisdictions—and collectively, these things tell us—these studies tell us that when police focus on making pedestrian stops, they can have about a 13% reduction in crime on average compared to other areas of the jurisdiction that don’t see that focused effort. If police focus on disorderly behaviors, right, things that aren’t necessarily criminal, things like graffiti, panhandling, loitering, disruptive behaviors on the subway, for example, they can see as much as a 26% reduction in crime on average compared to places where they just do what I’ll call “business as usual” policing. And like Charis talked about, we know about hotspots in communities, right? Street segments, corners, 2- or 3-block radiuses in neighborhoods that generate upwards of 50, 60% of all calls for service in a city, right? So we’re talking about less than 2% of places in a jurisdiction generating 50 to 60% of calls for police service. And so not surprisingly, when police focus on those places, they can see a significant reduction in all types of crime.

And lastly, here I list problem-oriented policing. This work goes back to the late ’70s, but the idea is that police can focus on things that aren’t even necessarily criminal. Problems are things that repeatedly end up generating 911 calls or calls for police service, right? And it could be just about anything you can think of. Police are uniquely positioned to get stakeholders involved and come up with a solution that might involve a criminal justice response, or it might involve something less traditional than that. But the end goal is to solve problems rather than slapping a band-aid on it by arresting someone, or the like. With all of these studies, I would mention that we always have to be cognizant of harmful effects, right? Anytime we increase police presence or increase police contact with citizens, we do run the risk of bad outcomes, right? A use of force or a viral video can undo all of the progress that might be made in the aggregate. So we always have to be mindful about the tradeoff here, essentially.

If we move on to trends in police violence, which is something I spend a little bit more of my time as a researcher on, I would hope that by now we’re all familiar with datasets assembled by folks like The Washington Post’s Fatal Encounters, Mapping Police Violence. But here, I wanted to highlight a new dataset put together by researchers at the Cline Center for Democracy at the University of Illinois. They call it the SPOTLITE Dataset. And it’s more comprehensive than all of those other datasets because it also tracks when officers discharge their firearms with the intent to strike a person. So not only do they have people that police have shot and killed, they have people in this dataset who they’ve shot and injured, as well as shot and missed. And so it’s more comprehensive than those other datasets. And from it, we know that the average day in the United States involves 6 to 8 people being victimized by use of lethal force by police officers, with 3 or 4 of them actually dying as a result. From that graph on the left where we just have incidents, on the Y-axis, and it’s a bar graph there from 2014 to the most recent year that they’ve released, you can see a slow upward trend over time. And I think you can also see a spike there in 2020 that corresponds with the spike that we saw on Charis’s slides around 2019, 2020 in homicides and in violent crime more generally. Over on the right, you can see some significant geographic variation. Where you see more red, these are higher rates of fatal police violence. So I should mention that this study that I’m referencing did not use SPOTLITE, it was published before SPOTLITE was released. And where you see blue is lower rates of fatal police violence.

So some clear geographic disparities there, where the Northeast are using—or we see fatal police violence rates that are significantly lower than in the Southwest and in the Midwest. From these datasets, we also know that 95% of those who are killed are men. About 1 in 4 displayed signs of or were known to have a mental illness. And there are known racial disparities here, as well, right? So black people are more likely to be killed by police, and this is primarily by way of they’re being more likely to be stopped and arrested by police, and I’ll get into some of that a little bit later. But that’s a fact that’s not really up for dispute anymore, these disparities are real. The extent to which individual-level bias, or organizational-level bias generates them is a lot trickier scientifically to tease out, but the disparities are real, so we really shouldn’t be debating that anymore.

Thinking about people’s perceptions of the police, one of my favorite things to show my undergraduate classes is the annual Gallup Confidence in Institutions polls. And they actually just released their 2024 data about a week ago. So over on the left is the percent of respondents. This is a survey of about a thousand-some-odd Americans each year, every June, going back to 1993 for the police. Even further for other institutions. And the green thick line there is people who express a great deal or quite a lot of confidence in the police. And then I’ve got it plotted from ’93 up until 2024, which was a couple of weeks ago. And what I think you should see here is that compared to the early 2000s, the police have lost confidence, or perceived confidence, from the American public pretty slowly over time. There’s a downward trend there. Although, year over year, from last year to this year, there was an 8-percentage point spike, which was the largest on record since Gallup started asking about the police. It does bounce around from year to year, so I would caution against overinterpreting that. Again, thinking about Charis’s point to take into context more time when you can, you see the downward trend going back to the early 2000s.

I think another takeaway here is that police tend to stack up pretty well against other institutions, they rarely fall behind more than just small businesses and the military in terms of inspiring confidence. They fare much better than the criminal justice system as a whole. They fare much better than, sorry to share, the TV news and the newspaper media, and Congress is down there at an abysmal 9%. So for what that’s worth, that can kind of tell you where the police stack up to other institutions. And I’ve not plotted them all here or it would just be a jumbled mess, but a select few here. And Americans’ confidence in pretty much all institutions tends to be kind of going down since the early 2000s, except for maybe small business. When we think about subgroup differences, over on the right, I’ve got the green dots showing where that group was last year in the poll and the blue or purple dots for where they were a couple weeks ago when Gallup ran their poll. And so you can see some pretty significant age disparities where older people tend to have more confidence in the police. Younger people had a lot more movement year over year in their confidence in the police, it went from 27% to 43%. The racial disparities play out here, as well, where white people have more confidence than people of color. Although, again, a large jump among people of color. And then with party identification, we see Republicans have a lot more confidence in the police than independents or Democrats. We don’t tend to see gender disparities pop quite so much in this poll, women and men tend to express similar levels of confidence in the police, and there are a myriad reasons why here that, hopefully, will be evident throughout the talk here. Shifting gears now to the factors that influence police responses to crimes. Over on the left, this is a series of graphs put together by the Pew Research Center.

Using data from the Bureau of Justice Statistics, they document that over 30-some-odd years, roughly 40% of violent victimizations even get reported to police in the first place, and even fewer property victimizations get reported to the police. And so then on the right side of that graph, you can see that among those that are reported, about 40% of violent crimes get cleared versus 12 to 17% of property crimes. And so one thing that influences police response to crimes is citizen cooperation in generating public safety, right? Citizens have to be willing to pick up the phone and call 911 when they see something or when they’re victimized. And this could also take the form of providing information to aid with investigations, testifying at trial. But if people are reporting less than half of all victimizations to begin with, the police are—it’s a tall order, right, to do something about crime when half of it goes undetected. And then property crimes like burglary, even when they get reported are going to be pretty hard to solve, right? Victims often don’t realize until hours or days later, like that time I left my garage open, and somebody stole my bike. I called the police and told them, but I never expected that they would actually be able to do anything about that.

So we also know that offense severity, right—you see if you were to—if I were to plot homicide clearance rates here, there would be a lot of variation. But the national level would be higher than what you see down there on property crimes. So offense severity tends to matter, right, as well as workload for officers in terms of how they respond to crimes when they become aware of them. From a meta-analysis that Tammy Kochel and her team did back in 2011, we know that non-white suspects are more likely to be arrested than white suspects. And this is true after we control for things like severity of the crime, presence of witnesses, evidence, victim preferences, prior records. Controlling for all of those things does not wash the race effect away, right? So again, I think the evidence tells us that race of the suspect matters in terms of whether the police solve the problem by making an arrest. And lastly, here I list legislation and prosecutorial discretion. We’ve learned that when states take efforts to decriminalize certain behaviors, or when prosecutors go on the record saying they won’t prosecute certain offenses, that affects the way police do their jobs and they often pull back from enforcing the behavior altogether. This is by no means an exhaustive list, by the way. This is just what I came up with for today.

Lastly, when we think about feeling safe, I think  Kubrin’s talk gave us a good glimpse into how divorced from reality people’s perceptions of crime are from the actual objective data. And I’ll start with the police. One thing that we’ve learned over the last couple of years with a series of experimental studies done by Justin Pickett and his colleagues, is that black Americans are uniquely fearful of the police. In his first study, he showed that they were more fearful of being stopped by cops without good reason than they were of being robbed. And in a more recent study that I actually coauthored with him, we found that black Americans were less fearful of being mistreated by non-white cops and female cops, as well as cops who were wearing body-worn cameras. So one thing the police can do to help alleviate feelings of fear would be to take serious calls for diversification. We’ve been talking about it since the ’70s, and progress was made for a while, but it’s halted for various reasons. And so to the extent that we can make progress on that, we may make progress on making black Americans feel less fearful of the police. Lastly, I’ll mention here that signs of physical disorder—again, I mentioned this earlier in my talk. Things that aren’t even necessarily criminal, right? Graffiti, litter, panhandling, disruptive behaviors on the subway. These things make people feel less safe. They make people less—or, excuse me, more afraid to take the subway to get around, right? And so in a recent analysis of nearly 100 studies in the U.S. and elsewhere, the authors found that strictly speaking in the U.S. when they measured mental distress and fear in the individual, they found that disorder was significantly related to those feelings.

And so we know that there are things police can do to police disorder. I should mention that there are other things we can do. Easier things that don’t necessarily involve the police such as fixing street lighting, clearing out abandoned lots, doing things—you know, picking up litter, scrubbing graffiti, right? Things to make people feel more comfortable walking around in their communities can reduce—reduce fear and mental distress. But I should also mention that these outcomes are much harder to measure, right, than whether a crime occurred. Now you’re tapping into how a person feels on any given day, and there’s a lot of stochastic noise, I think, that goes into measuring that and measuring it accurately and reliably. So with that, I’ll stop there. Here are a couple of ways to find out more about some of my other research and those of my colleagues here at UNO. And then those slides that’ll be shared—I’ve got everything linked so that you can go and see for yourself what I was talking about. So, thank you.

[00:32:40]

RICK WEISS: Fantastic. Another great presentation with so much data and so much information. And interestingly, beyond that, an appreciation recognition that this is not all about, you know, hard policing and crime, it’s about psychology, it’s about sociology, it’s about people’s feelings. Such a big part of the puzzle. So making it harder, I guess, to study and to get answers to, but important to include that. We’re going to start the Q&A section of this briefing now. A reminder to reporters, if you have a question, please go down to that Q&A icon at the bottom of your Zoom screen and let us know what you’d like to ask. I do like to start these briefings with one question from me which asks our speakers to address something not solely in their role as experts, but also in their role as news consumers themselves. Based on the news stories that you see coming out every day on this topic, one way or another, is there something you’d like to say to reporters about something you see in these stories that you appreciate, and you’d like to see more of, or something you see that just keeps bugging you and you’d like to sort of get the word out to maybe get a corrective going. And Charis, I’ll start with you.

[00:33:50]

CHARIS KUBRIN: Sure. I mean, that’s a great question. So I guess, one thing that I’m not particularly fond of is the fact that often, reporters will report on what end up being what I call kind of n of 1, or more idiosyncratic, or atypical crime events. And what concerns me is they don’t necessarily reflect the true crime problem that we would see if we were looking at the data as a whole. And this limits our understanding of real threats to society and, in some cases, reinforces problematic stereotypes. Let me just give you a concrete example. We’ll talk about homicide just because I gave my presentation on that. If you look at the data, I’ve worked with large datasets on homicide, you will see that a significant portion of homicides result from kind of mundane arguments that escalate, often in the presence of precipitating factors, say, drugs, alcohol, firearms, and that occur between people that know each other, acquaintances, family members, significant others, so on and so forth. But if you look at sort of an array of news stories on homicide, often what’s reported are random killings between strangers or types of homicide that are relatively rare if you look at the data. I mean, there are so many, for example, reports of homicides among serial killers, for example. And I teach a homicide and suicide course at UCI, and one of the things I tell the students is serial killers, and that kind of horrific event, which are important and not to minimize it, but they represent less than one half of one percent of all homicides. So we get an outsized concern around a particular type of homicide, whereas we minimize kind of the routine violence that occurs every day that is much more threatening to society. One last point on that in terms of stereotypes.

I do work in immigration and crime. One of the things I’ve noticed in the reporting on that specific topic, is that often when crime is reported and it’s committed by an immigrant, particularly an undocumented immigrant, you know, the headline will read something like undocumented immigrant committed such and such a crime. And it’s headline after headline sort of connecting the words immigrant or immigration and crime, reinforcing in the minds of Americans that the two go hand in hand. I’ve done research on this, published a book last year, that reviews all the studies on what we know about immigration and crime, and it turns out that that concern, that connection, is greatly overstated. And I point to an example of you would never see a headline in our news that says native-born American commits such and such a crime. Right? So just being careful and not reporting things that both don’t reflect what the data have to say and that also can reinforce existing stereotypes.

[00:36:52]

RICK WEISS: Great advice. Thank you.  Nix?

[00:36:56]

JUSTIN NIX: Yeah. I think I would actually echo what Charis said during her presentation about not—or about being mindful of the comparison, right? When we think about what crime did in the first couple of quarters of this year. What are we comparing that to? When we think about police violence going up over time, is a year-over-year jump of 10 additional instances of police violence significant? Well, it’s 10 more people who were killed, which is 10 too many. But when we can bring into the fold more data and try not to overreact to what might prove to be a blip, right? So if we go back to that Gallup poll I showed, an 8-percentage point rise in confidence in the police. Well, it was also rising from the all-time low since Gallup had started, right? And so there might be a little bit of just regression to the mean there, and there’s just noise from year to year to begin with. So I always see around January or February these stories get written about what the crime data shows from last year or what the police violence data shows from last year, and sometimes it can be a bit—it’s over-reactionary. Obviously, sometimes it proves to not be a blip, right? The homicide spike in 2020 persisted for a year or two, but now it’s back—it’s going back in the right direction. So just being careful not to write stories that are overreacting in the moment.

[00:38:10]

RICK WEISS: And if those blips are real, like ’20 to ’22, to what extent is that an exogenous force-related thing like the pandemic? Interesting. Great. OK. Well, we have the first question here that actually plays off of something I was pretty curious about when you, Justin, talked about the surprisingly high percentage of crimes that don’t get reported. And this question asks, to what extent are unreported violent victimizations victims of domestic violence? Or is there anything else you want to add about why so few crimes are reported?

[00:38:43]

JUSTIN NIX: Yeah, we know that victims of domestic violence, in particular, are hesitant to come forward, whether they’re afraid of their abuser retaliating, whether they’re afraid that they won’t be believed, or that they might be revictimized by law enforcement and the criminal justice system, or whether they just don’t want to see their abuser go to jail, right? So there are a myriad reasons, but we do know that reporting is especially low among that group of victims, in particular. So I know that answers the first part. Was there a second part that I missed?

[00:39:14]

RICK WEISS: Just whether there’s a number you can put on the percentage of these things that are domestic violence?

[00:39:20]

JUSTIN NIX: I believe the National Crime Victimization Survey does break it out by—or at least, there is a report that has been done, specifically, dealing with domestic violence victims and their reporting. And I know that it’s lower. I can’t put a hard number on it off the top of my head, though.

[00:39:35]

RICK WEISS: Great. Here’s a question from Kylie Cameron from KMUW Public Radio in Wichita. In Kansas and Wichita, we recently had crime data discrepancy issues with low city—with how city police and then the state agency reported on crime. The city used per victim, the state did it per incident. This led to a lot of inaccurate reporting in our community. Is this something you often see?

[00:40:01]

CHARIS KUBRIN: So I actually thought that was going to—that question was going to go in a different direction because we’ve changed how we are collecting and reporting crime data by the FBI, which I’m happy to talk about at another time. We’ve gone from the summary reporting system that was used for decades, to NIBRS, which is the National Incident-Based Reporting System, and there’s been a lot of hiccups with that. To be discussed another time. But the point is an excellent—or the question raises an excellent point that I want to make, which is the denominator is really important, right?

So, first of all, one of the key issues I often see about reporting that’s problematic is that people will report offenses, whether it’s homicides or robberies or property crimes, without any denominator. And they’ll make comparisons across states, forgetting that states have very different population sizes, or cities have population sizes, or neighborhoods have different population sizes. So my first point about this is you have to have a denominator there. Then the question becomes, well, what is the denominator? And that depends, really, on the kind of—the answer to the kind of question you’re asking. So I think it’s very important to pay attention to that. And I would not ever, ever look at—report crime data or look at the reporting of crime data that does not standardize and then create a rate of some sort. You’ll notice all the slides that I showed in my presentation were for, at the national level, rates per 100,000 population. At lower levels, it could be per 10,000 population.

[00:41:36]

RICK WEISS: Mm-hmm. Great. Anything you want to add to that, Justin? Great. Okay. In terms of social correlates of crime, is there enough data yet to correlate the emptying of many downtowns due to work-from-home trends with increased crime in central cities, or decreased?

[00:41:55]

CHARIS KUBRIN: Boy, this is a really investigated area right now by criminologists is trying to understand how an exogenous shock like the pandemic influenced our movement during the pandemic, post-pandemic, and what implications that had for crime. And it’s just a very complicated puzzle because, in many respects, during the pandemic, we were all hunkered down, and we saw huge decreases in certain types of crime. A lot of the instrumental types of crime that involve a burglary or a robbery, for example. But then we’ve seen rises in others. So I think we’re still figuring out the role that any of these one factors plays, but it’s important, in general, to look at how broad societal changes impact the movement of populations, including in and out of city centers, with implications for crime.

[00:42:53]

JUSTIN NIX: Yeah, I would echo that. And going back to the point about the denominator, I’m aware of at least one study by Chalfin and Massenkoff where they tried to factor in the fact that a lot of us were hunkering down and not going outside and, presumably, putting ourselves at risk to be a victim. And they came up with a way to try to measure time spent away from the home as a denominator. And I think that they found that violent victimization was actually—or, you were at a greater risk of being violently victimized in 2020 and 2021, and that these—that the measures that they come up—or the results that they come up with don’t necessarily square neatly with what the violent crime data might tell you, in part, because our denominators were so off for those first two years where we were fundamentally changing our behaviors and most people putting themselves less at risk. But if you were having to go out in public, you were at a greater risk, according to that one-off paper.

[00:43:44]

RICK WEISS: Interesting. Here’s a question. Are there any signals of state gun laws when looking at state-level violent crime rates? Basically, do states with more lenient gun laws show higher rates of violent crime?

[00:43:59]

CHARIS KUBRIN: This is just a very controversial issue and the research on it is expansive. I don’t feel qualified to simply answer that question in a short nutshell. What I can say is—I mean, part of it involves the challenges of measuring gun ownership. We only have measures of gun availability; they’re imperfect. But I would say the public health literature shows, you know, there’s no doubt that guns play a role. The question is the size of the role, what impact, how it matters in—like over the course of the pandemic where a lot more people purchased guns. So there’s a lot of—I guess—I guess it would be much easier to talk about this in greater detail with very specific questions and pointing reporters to studies than to kind of give a global answer on the role of guns and gun markets.

[00:44:56]

JUSTIN NIX: Yeah, I feel the same way, and I would—I’m glad you mentioned the public health literature. I would throw out names like Cassandra Crifasi and Dan Webster. I think they’ve done a lot of good work on—when Missouri repealed their purchaser permit laws and things you can see that correlate with community violence. I think that part of the challenge here is estimating how many guns are out there and then teasing out handguns versus long guns. We know handguns are far more likely to be used in crimes. So, yeah, it is—it’s pretty tricky literature, but I do think that—and I can say there is literature on police violence. I know that wasn’t where the question was directed, but where there are more guns, there tend to be more instances of police violence in both directions, right? Officer-involved shootings as well as officers being victimized by gunfire, right? So we know it increases opportunity, but those questions as to how much and what are the circumstances therein, it’s a lot murkier, I think.

[00:45:53]

RICK WEISS: And the references you made to Crifasi and Webster, are they at an academic institution we can share with reporters? Do you know where they’re based?

[00:46:01]

JUSTIN NIX: Is it Johns Hopkins that they’re at, Charis?

[00:46:04]

CHARIS KUBRIN: I don’t—I don’t remember.

[00:46:05]

RICK WEISS: We’ll see if we can follow up.

[00:46:06]

CHARIS KUBRIN: And I was also thinking of the work of Dan Semenza and others.

[00:46:09]

JUSTIN NIX: Semenza, yeah.

[00:46:09]

CHARIS KUBRIN: Yeah.

[00:46:10]

JUSTIN NIX: He’s at Rutgers, I believe.

[00:46:12]

RICK WEISS: OK, good. That’s helpful. Thank you. Here’s a question directed for you,  Nix. You mentioned that it takes more than 10 added police officers to save one additional life from homicide. Is that considered a good return on investment? It seems unsustainable to keep growing police forces by that magnitude.

[00:46:32]

JUSTIN NIX: Yeah, I would agree. And if you dig into that study, right, they speak about some of the harmful effects that that had, right? The larger agencies tend to make more low-level arrests. And these low-level arrests often deteriorate perceptions of legitimacy, and there can be a whole list of issues that that creates, as well as opening the door for instances of uses of force, excessive force, and the like. I should mention that that was a retrospective study of cities with 50,000 or more population, so we shouldn’t be thinking about this with respect to the modal police department, which is a small jurisdiction that probably has 10 cops total. But in larger cities, I think it’s born out over that time that each 10 to 17 officers abated approximately one homicide. So the tradeoff becomes if you don’t think that that’s a good return on investment, we need to see evidence for what the alternative would be. And the really tricky part there is that most of the evidence we have on alternatives to policing as far as a way to disrupt violence or prevent violence don’t happen in a vacuum. They happen while police are still operating in the background doing the things that they’re doing. And so it just becomes a lot trickier, I think, to estimate the marginal effect of whatever program you might have in mind as a means to reduce violence.

[00:47:46]

RICK WEISS: Just while we’re on that topic if I can inject a side question. Do you have an answer to what explains the geographic difference you described between police violence in, say, the Northeast versus some of those other areas of the country? Is that a matter of policy? Is it a matter of more bias in some parts of the country than others?

[00:48:03]

JUSTIN NIX: I don’t think we have super conclusive evidence. I’m working on a couple of papers now with some colleagues that I’ll hopefully be able to share soon. But I think it’s a function of a few things. I mean, up in the Northeast we tend to see stricter gun laws than we do out west of the Mississippi. But I think also, just more fundamentally, Western cities tend to be a little more spread out geographically. So if you think of a city like Phoenix that covers 500-some-odd miles and a city like Boston or Philadelphia, these old school East Coast cities that are really densely populated and compact, it’s just a further drive, on average, to trauma care that might save a person’s life, right? And so when I look at fatality rates for police shootings, I see that same sort of divide along the Mississippi where in Philadelphia, New York, Atlanta, Boston, fatality rates tend to be quite low, 20 to 30% of police shootings result in death. Versus when we go out to Phoenix and L.A. and these other western cities, fatalities tend to be 50% or more, right? And so that could explain some of the difference. But if I’m honest, in the end, it’s probably more complex than that and we still just don’t really know.

[00:49:10]

RICK WEISS: Here’s a question directed to you, Kubrin. Can you talk a little more about options for obtaining city and neighborhood-level crime data and any weaknesses of those data that reporters should keep in mind?

[00:49:22]

CHARIS KUBRIN: Sure. So a lot of the data I was reporting on come from the FBI that collect information from police departments and then publishes it annually and quarterly, and that sort of thing. And for the longest time, FBI crime data—by the way, this is reported crime data. I like to keep that in mind. It’s not crime data, it’s reported crime data. And it’s actually response to reported crime data, to be clear, because we don’t really know the true number of crime. We call it the dark figure of crime because so much crime goes unreported, as we’ve been speaking about throughout this webinar. So it’s really a subset of everything that’s out there.

That aside, they—in 2021, the FBI switched from its summary reporting system, which provided an aggregate monthly tally of all crimes, to NIBRS, which is the National Incident-Based Reporting System. And the good news is that NIBRS counts many more offences and provides much greater detail about them like age, sex, race of victims, and the circumstances of the crime. So it goes real in-depth. The bad news is the conversion to NIBRS has been challenging. We’ve got lots of police departments throughout the United States, included in major cities, that have not gotten on board and reported. It’s getting better. But that first year was quite challenging, so—and there’s been a lot written about this transformation from SRS to NIBRS. I encourage you to check out Jeff Asher’s website Datalytics where he writes a lot about the missingness of the data and the challenges with reporting. And also a great new book by a very famous criminologist Richard Rosenfeld. He just published a book called Crime Dynamics that talks a lot about crime data and their trends. But you can go right onto the FBI’s webpage and access their Crime Data Explorer tool. I have a slide in my presentation that I didn’t share but that provides information on how to access that data. And you can go right in there and you can select out states, select out cities, do things over time. It’s a real fun tool and provides lots and lots of information for reporters.

[00:51:44]

RICK WEISS: Fantastic. Let’s see. I’ve got a question here for —I think it’d be for you, Nix. Does increased mistrust in the police affect the ability of police to actually do their jobs?

[00:51:59]

JUSTIN NIX: Yeah, I think so. The work of Tom Tyler, a psychologist at Princeton comes to mind here where he talks about the process-based model of policing and what police can do to earn the trust of citizens by way of treating them respectfully and giving them a voice, right? And the opposite becomes true when people don’t feel very trusting of the police. When they trust the police, they don’t call 911 to report crimes. They might take matters into their own hands and initiate retaliatory violence. When they don’t trust the police, they might not cooperate with ongoing investigations. They might not provide information that could help prosecute a perpetrator of a crime, right? So yeah, I would say that when people don’t trust the police the work of the police becomes that much more challenging.

[00:52:47]

RICK WEISS: And I’m just looking—I’m having a technical glitch on some of the questions coming in. Here’s one. Do you have any insights into how Americans factor in or prioritize their feelings about crime when they’re deciding on who to vote for in an election? We also had a question that asked whether a 4-year time scale is an appropriate time scale to consider comparative crime data. In other words, to say during this administration, you know, crime rates were X, and during another administration, they were Y. So the first part of this question about how Americans prioritize it might be relevant to that Gallup data that some of you showed.

[00:53:29]

CHARIS KUBRIN: Yeah, I’ll start off and then turn it over to Justin. So crime is always on the minds of Americans as they go to the ballot box, and it’s been, for decades, a key decision point for voters. But what’s interesting, and I don’t have the data with me right here, so I can’t give you exact numbers, but in the latest Gallup poll, I noticed that it’s not a top priority. The economy, immigration, other things are kind of, at least, for this particular year, trumping that, if you will. So it’s there and it’s important, but I think its relative importance shifts over time. This is where we’d like—I’d really like to get longitudinal data.

[00:54:15]

RICK WEISS: Justin, anything to add there about comparing?

[00:54:18]

JUSTIN NIX: I couldn’t say it any better myself. I saw the same Gallup poll, that it slipped a little bit in terms of where it normally comes in, but it’s always salient in the minds of voters. It’s one of the key issues that they’re going to think about, even if it’s not as high on the list this cycle as it was maybe the last cycle. And looking at the 4-year trends, I think I’m a fan of smoothing out the data and trying to look for trends when you take a 4-year snapshot, right? But keeping in mind that these national-level estimates, again, are going to mask a lot of state, city, and neighborhood-level variation that Charis talked about in her presentation. And so just keeping that in the back of our minds. I think that at the federal level, there are things we can do. Providing funding for jurisdictions to help fight crime. But more often than not, what matters is the hyper-localized response that’s going on in jurisdictions and how they work together with the community to solve problems.

[00:55:12]

CHARIS KUBRIN: Yeah, that’s exactly right. And can I just add one thing, which is that—is 4 years enough? Is it too little? Is it too much? It’s like when a student asks me how many interviews do I need to do to make this study legit, right? And, I mean, I always go back to what is the question you’re trying to answer, right? So if you’re interested in a question that speaks to a political administration, then 4 years makes complete sense. As long as that is identified at the outset, right? So there are cases where we do want to know about very short time trends, and there’s other questions that we have about longer. They’re all important questions, but I think the key point is to have that be part of the investigation. Often, I’ll get reporters who will say, “What’s going on with crime?” And, OK, I don’t even know where to start with that question. Are you talking about over time? Across place? This crime? That crime? I mean, so the more that the specific question or interest is identified, and delineated, the better we can answer those questions. And if 4 years makes sense, given the question, sounds good to me.

[00:56:25]

RICK WEISS: Right. OK. One more question. And I want to remind reporters as we get ready to top this off at the top of the hour, you will be confronted with a very short survey as you log off today. It’s just 3 or 4 questions. It’ll take you half a minute. It’s so helpful to us for you to take those few seconds to fill that out so we can keep producing briefings like this that serve your needs. So please take those extra moments to fill out the survey as you leave. But I have a couple more questions I want to squeeze in as we just about wrap up. One of them has to do with the data you mentioned,  Nix, about comfort levels of people as diversity increased on police forces. One thing you didn’t address was whether the disparities in race-based arrests go down as police forces become more diverse. And I wonder if there’s any data on that?

[00:57:18]

JUSTIN NIX: That’s a great question, and I don’t think the matter is settled. I don’t know that increasing diversity is going to change anything operationally about things that police departments currently prioritize on any scale larger than maybe in one city, right? If there was sort of a revamping of not just who the police are but what the police do. But I can tell you that those studies that I cited were experimental designs where the one thing was manipulated was the picture of the police officer that people were looking at and asked to evaluate. And we saw some pretty—pretty significant differences across the race of the officer. But to the extent that that might translate to less aggressive policing or fewer discretionary arrests, I think that it would have to be coupled with other reforms before we’d see a serious return on investment.

[00:58:05]

RICK WEISS: OK. Great. Well, we are just about at the end of the hour, and I do want to give each of our speakers an opportunity to give one take-home message for reporters. If there’s one thing for each of you that you just really want reporters to walk away with today and keep at top-of-mind for them, this is your chance to do it. And, Charis, I’ll start with you.

[00:58:25]

CHARIS KUBRIN: Okay. Well, I think maybe I’ll say something about the causes or the why because I think that’s what’s of interest to most people. They want to get right now to what’s going on. And I guess, even as we don’t necessarily know the answer, something to keep in mind in reporting, which is that often, we’re searching or seeking out sort of the singular magic bullet that explains that crime trend, right? And I hear this all the time. It’s the pandemic. It’s criminal justice reform. It’s the economy. And when we do this, I think we fail to recognize a real fundamental fact about crime, which is that it is multidimensional and caused by many, many factors. It’s complex. So rather than treat the causes of crime as mutually exclusive or pit them against each other, which often happens—it’s this. No, it’s this—I think it’s really beneficial to remember that they are overlapping and interrelated. So I would be wary of someone—any seasoned criminologist espousing a single-factor explanation for something. That is always something that I think it’s always more complex than this police policy went into place and did this, or the pandemic did this. There’s many different forces going on at one time. So there is no magic bullet explanation, which, of course, means there’s no magic bullet solution when it comes to combatting crime or making crime rates go lower. We have to think about all of the factors that matter at different levels of analysis, from street segments, to neighborhoods, cities, and states, at different periods of time, and then, of course, across the demographic groups that I mentioned earlier.

[01:00:10]

RICK WEISS: Fantastic. Thank you. And  Nix.

[01:00:13]

JUSTIN NIX: Yeah, I agree. No magic bullets. And if your expert is giving you a magic bullet, find yourself a new expert. The truth is it’s a very complex world that we’re trying to understand.

[01:00:23]

JUSTIN NIX: For me, it’s that the evidence is pretty clear that policing matters to some extent. Policing can reduce crime. But I think it’s also true that police in many places can do a lot better. And for anybody comes and says, well, police don’t reduce crime. Well, we know that that’s not true, there’s dozens and dozens of studies we can point to that show things that they can do to have an effect. The key question becomes are the police in this place or in that place operating at maximum efficiency, right? What are the tools that they need to operate where that tradeoff is something that we’re willing to accept? Where they’re reducing crime and they’re minimizing harm. And maybe there are a lot of other things we can do that don’t really involve the police and their respective goals and missions, but the truth is it’s a pretty tangled web and it’s going to take a little of—all the stakeholders. But increasingly, I see that argument that, well, we know the police can’t do anything. And that’s not true. That’s a pretty nihilistic view and I don’t think it’s—I think it’s quite dangerous, honestly. It doesn’t really do us any good. So pushback on that would be my takeaway.

[01:01:30]

RICK WEISS: Some great advice from two really very clear and generous speakers today. Tons of information there for you to work with, reporters. I so appreciate your contributions,  Nix, and Kubrin. Thank you, reporters, for your attention to these kinds of details and for sticking with the data and the facts as closely as you can as you cover this complicated story. And we look forward to seeing you at future SciLine media Briefings, including two more on election-related issues next week, Tuesday and Thursday. Check out the SciLine website. Thank you all for attending, and we’ll see you at the next SciLine media briefing. So long.