A Deep Dive into Why Crime is Falling with John Roman

Crime is falling fast in America , and, as the many different opinions on the podcast can attest, almost nobody can agree on why.

My guest today has spent thirty years inside the question of America’s changing crime trends, first as a researcher at the Urban Institute, now as director of the Center on Public Safety at NORC.

John Roman has a theory about what actually drove the violence spike of 2020, and an even more specific theory about what’s driving the decline now, one that has less to do with new policing tactics and a lot more to do with $300 billion in pandemic relief money that quietly rebuilt the social infrastructure we lost.

In this episode, we talk about why the usual explanations for crime trends –  more cops, more task forces, more local initiatives, don’t hold up against the data. We get into the role police legitimacy potentially played after George Floyd, why the decline has lasted so long, and what the future might look like.

John K. Roman, Ph.D. is a senior fellow in the Economics, Justice, and Society Department and directs the Center for Public Safety and Justice at NORC at the University of Chicago.  His research focuses on the economics of innovative crime and justice policies and programs, cost-benefit methodology, and systems reforms, including justice system interactions with substance abuse, public health, adolescent development, housing, workforce development, and education. Dr. Roman has conducted research on behalf of numerous state and local governments, foundations, and federal agencies.

Transcript

SPEAKER_02 0:00

I'm Jeff Asher, and this is the Jeffalytics Podcast. Crime is falling fast in America, and as the many different opinions on this podcast can attest, almost nobody can agree as to why. My guest today has spent 30 years inside the question of America's changing crime trends. First as a researcher at the Urban Institute, now as director of the Center on Public Safety in New York. John Roman has a theory about what actually drove the violence spike of 2020, and an even more specific theory about what's driving the decline down now. One that has less to do with new policing tactics and a lot more to do with $300 billion in pandemic relief money that quietly rebuilt the social infrastructure we lost. In this episode, we talked about why the usual explanations for crime trends, more cops, more task forces, more local initiatives, don't hold up against the data. What do you know the role police legitimacy potentially played after George Floyd? Why the decline has lasted so long and what the future might look like. We also talk about what we still don't know about policing itself and why the research community has struggled to get this important information in front of the public in time for it to matter. Let's dive in. Thanks for having me on. John, what was that, like 10, 15 times I had to ask you before you finally agreed to come on the show? So I really appreciate it. Something like that. John, uh obviously we we go way back and we both sort of inhabit this space about writing about crime and trends and have gone through this sort of roller coaster over the last few years in terms of everything was up and explaining that and everything's down. What is your background? How did you get to be sort of in this world?

SPEAKER_00 1:41

I just sort of stumbled into this world. So I'm a failed journalist and a failed campaign person. I did both of those for a while out of college and then went to grad school and had a friend who was doing a study in the Michigan Department of Corrections, and this was the mid-1990s, looking at whether there were racial disparities in the way they handed out tickets to prisoners while they're incarcerated that caused them to change their security classification. And she asked me to do the statistical analysis for her, which I did and put it on my resume, and it caught the eye of a senior researcher at the Urban Institute who hired me. I was looking at the time for public health work, and she hired me into their justice center at the Urban Institute. And that's where I got my introduction to justice. But I've done nothing else for 30 years. So I feel like I've filled whatever gaps I brought with me to the field.

SPEAKER_02 2:40

We'll allow it. We'll call it a full resume. So what are you doing now?

SPEAKER_00 2:44

Well, right now, I mean, I think the key questions are: why is crime down? We we really have, you know, only a modest idea. We're at this incredible sort of reflection point where crime levels are down across the board, which is new. They were first, it was just shootings and homicides, and so it's gun, gun-related, but now it's down everywhere. Now we're seeing numbers approaching 1960s level, and we don't have a good answer for what is the causal mechanism, why is this happening, and much more importantly, what can we do next to perpetuate and sustain these really phenomenal low levels of crime and high levels of public safety? So that's really where I'm concentrating.

SPEAKER_02 3:30

And so we're gonna ask, you know, really easy softball questions here. But first off, is why did murder rise in 2020 and why is it falling now? I know you you've written extensively about this. You're one of the, I think, best supported in terms of just having a theory why all this happened. Can you walk us through it all?

SPEAKER_00 3:49

Sure. I'm I'm happy to. So it is just a theory, but I think four things happened beginning in March of 2020 during the pandemic. So the first thing that happened, and probably the most consequential thing that happened, was that everybody stayed home. And businesses were closed and schools were closed, but also all the kinds of community supports that people who live in neighborhoods where there is recurring trauma, where there are a lot of shootings, where there is a lot of violent crime, those folks were stuck at home too. The kinds of supports that they could have gotten from neighborhood community organizations or schools or community centers or churches or whatever they leaned on, those were all closed. And so people were cut off from the good things in their lives in places with a lot of trauma. And that created this sort of toxic stew where you're home, the people you have long-standing disputes with are, you know, where they are because they're home too. And there are lots of guns around. And this created a recipe for a really intense moment of gun violence. And I think we saw that. And that explains most of what happened. There are a couple other things that happened at the same time that are really important. One is that there was a change in the way drug markets behave. People who don't buy a lot of street drugs, which hopefully is most people listening to this podcast, buy their street drugs at open air drug markets, but they also buy them at work and they buy them at school and they buy them at the restaurant or the bar that they go to. And since the everything was closed except the open air street markets, there was a lot of people going to those markets who didn't typically go. Plus, people were feeling a lot of trauma. You can see it in the National Household Survey on Drug Use and Health that there was a little spike in all kinds of drug use. So there's this pulse of demand in these drug markets that destabilized them. And because there's no rule of law in drug markets except for your gun, this created a lot of violence around those markets. Policing changed in ways in a different way than I think the public narrative has presented it. Police did pull back, but they pulled back by a purposeful decision, which is that policing agencies said to their officers, do less proactive policing, have less contact with civilians. We want to keep our officers healthy and safe and on the job. And so for a period of several months into 2021, there was less policing. And I think, and then there was just this huge amount of new gun purchasing. People were afraid during the pandemic. They bought guns to protect themselves. The data suggests that a lot of people bought guns who didn't use typically buy guns or hadn't bought guns in the past. Women bought guns at much higher rates. People of color bought guns at much higher rates. So a lot of people had guns who didn't typically have them. A lot of those guns leaked into the illegal gun market and they leaked very quickly. Studies from the Times suggest that the time from purchase to crime gun was much accelerated. So you have more guns, more volatile drug markets, less policing, people stuck at home, near people who they have long-standing disputes with, and it was just a recipe for violence, and that's exactly what happened.

SPEAKER_02 7:14

What role do you think something like trust? Obviously, it's it's such a complicated story because there's, yes, murder was up like 10 to 15% in the first quarter of 2020, and then it skyrocketed starting in the summer of 2020. So what role do you think something like? Obviously, we saw after the murder of George Floyd, this enormous increase in murder and all of these protests. What role do you think the this concept of trust or police legitimacy, uh, if people don't trust the police, the ghetto side uh book, the book ghetto side which talks about how that that role of police legitimacy is critical. Do you think that that's a major factor, or is that sort of maybe talked about too much?

SPEAKER_00 7:57

So I think it's a factor. I don't know if I think it's a major factor, and I and I know I disagree with a lot of people on this point. I think it was sort of the icing on top of the cake, is the way I would frame it. For about two months or after May 25th, 2020, which was when George Floyd was murdered, there's an elevated level of gun violence that does not fit the larger trend. And I think it's about this issue. It's about legitimacy and trust and an explosion of discontent with the way people were policed in specific neighborhoods. And those neighborhoods have a lot of violence. And so there was violence as a result. But I think the vast majority of the gun violence spike really didn't have anything to do with police legitimacy or perceptions of the legitimacy of the policing authority. And here's why. In places where there's a lot of historic violence, levels of trust in the police is very low to begin with. And so when we look at survey data from that period, we don't see a lot of change. Where we see the change is in neighborhoods that had typically had much higher levels of trust, more suburban places. So the people who lost trust after George Floyd were probably people living in places that didn't have that much violence, and there wasn't much new or additional violence in those places after May of 2020. So a little bit of an effect. They were very intense, but like a few days or a week in most places and not everywhere. What was everywhere all the time was the pandemic, the change in routine activities, the change in drug markets, the change in policing for officer health and safety. So I just think those were bigger mechanisms that had a bigger effect.

SPEAKER_02 9:54

And why do you think it lasted so long? I mean, I guess in the grand scheme, it was 2020 to 2022. So it wasn't that long. But you know, the pandemic conditions had eased by end of 2021, and yet violence persisted at being high. Why do you think that was the case?

SPEAKER_00 10:10

Yeah, I just, you know, it's we talk about violence as an epidemic, and I don't know if I love that metaphor because that suggests it spreads to anybody who's unvaccinated, right? And really, violence spreads to other people who are at high risk of being victims or being a party to violence. And so I don't know if I love the epidemic metaphor, but but it is certainly true that there is a cycle of retaliation when there's gun violence. Somebody who's shot, people retaliate, that creates new victims, that creates new animosity and tension. Animosity isn't even a big enough word, and that creates and that perpetuates the cycle of retaliation. It's important to note that through much of 2020 and 2021, the community-based organizations that we now look to to break this cycle of retaliation, the community violence, intervention, and interrupter, that most of that funding wasn't online yet. These programs were very small and in very few places. And a lot of those folks were being encouraged not to have a lot of citizen contact for the same reason the police were being encouraged not to have a lot of citizen contact. Because, you know, they didn't their the leadership of those groups also didn't want them to get sick. So I think that allowed all this to last longer than it might have in a different situation. But in a different situation, it probably wouldn't happen anyway.

SPEAKER_02 11:35

All right. So now we're through 2022 and we start to see violence fall. And at first it was like large for the time, and then it accelerated in 2024 and accelerated again in 2025, and it's still down a lot in 2026. Why you've written a lot about this, and I think that you've got a very uh comprehensive and sensible theory of this. So why do you think it's falling so significantly now?

SPEAKER_00 11:59

Well, first I have to say kudos to you because we wouldn't have known any of this without Jeff Asher. That's how we got Jeff Asher is our early warning system. So we knew that that's the same thing.

SPEAKER_01 12:10

Payment I give to every guest to say something like this. So thank you.

SPEAKER_00 12:14

The tide was changing, and we and we could start thinking about these things way in advance where of where we would have if we'd had to wait for FBI national crime statistics. And we had a year of extra time to get ahead of this stuff. So that was really important. Um I think a lot. So my theory, and again, this is my theory, and I I don't, you know, I don't know how widely accepted it is, not terribly, would be my disclaimer. But beginning in 2021 and 2022, the federal government at the time, and and this included Congress, passed a couple of pieces of legislation that were really important. There was the Bipartisan Safer Communities Act, which pumped $15 billion across the country into violence intervention programs and policing and all kinds of things to stop this spike. The bigger investment, though, was the American Recovery Plan Act, which we all lovingly call ARPA. ARPA pumped $300 billion into state and local governments. And it's hard to express how big a number $300 billion is. But it's it's something like if you took all of the spending during that same time period on all police, all jails, all prisons, all community supervision, it's approximately it's a little smaller than that, but it's of that same magnitude. It's like almost like we doubled the size of the justice system. That's the scam. So, what did that money go to? A little bit of that money went to policing, and it helped to replace police officers who got sick or left during the pandemic. Most of that $300 billion went to state and local governments to do the things that they do, which are to fund teachers, to fund community centers, to fund social workers, to fund behavioral health scientists. And the thing about it is those are the people who work directly with young people who are at the greatest risk of violence and victimization. So these are people who are working with adolescents and young adults to help them create opportunity and to remove risk conditions that are in front of these young people. And so it stands to reason that if you work directly with the people at the greatest risk, work really hard to remove the obstacles and the conditions that create gun violence and other kinds of offending, uh, that you'll see pretty rapid reductions in crime. Now, the other thing I would just say is, and it got no attention at the time, between March 2020 and May of 2020, we lost 1.25 million local government jobs. And these weren't furloughs, these weren't people left and then came back. These were jobs that were just gone. That's essentially the size of the, you know, the US military, just gone in three months. And it took until the end of 2024 for those jobs to come back. And again, these are the teachers and counselors and people working with adolescents and young adults. But as soon as the ARPA money started going to local governments to get those people, get those jobs filled, get people back in those roles. That's the same moment that crime tipped and stopped going up and started its decline, which is now ongoing.

SPEAKER_02 15:42

What um, I don't know if you've seen this, Ian Adams and co-author who paper open. Uh Scott Mortgows, M-O-U-R-T-G-O-S. They have a very interesting paper and it's it's a preprint. Have you seen this? I have not. So this is great. It's a very uh preprint, it's heavily academic. I'm gonna ask you to comment on it. It's perfect. The theory that they have is what they basically call criminogenic entropy. That, like you take all of this uh they have, I think, 15 different factors and you track them over time, over the last 40 years. It actually matches really well with the rise in crime in the 80s and the 90s, and then the drop over the last 20 or 30 years. With the sort of COVID being the outlier, you had this like once-in-a-lifetime thing that caused this bump. Do you have thoughts? And the way it's very technical, the paper, but the concept is basically like things are getting better because society is improving. And I guess that's kind of the bigger picture view. Do you have any thoughts on that as an idea of why this is continuing year after year, well below where we've ever been?

SPEAKER_00 16:53

I I love this question. So I will happily pontificate about a paper I've never read.

SPEAKER_01 16:59

That's a pre-print.

SPEAKER_00 17:00

So it's not even No, but but I think the larger point, the larger point for me would be the paper side, right? That if you look historically at the number of police in America and the crime rates, what you see is that crime rates match pretty well. So from 1998 until 2018, 2019, the number of police per capita in the United States is absolutely flat. And and crime is sort of finishing its great decline of the 1990s, but then it's pretty flat in the 2000s and it goes down a little bit in the 2010s and then back up again, but it's pretty flat. It's in a narrow range. And what that says to me, I am going to land this plane. What that says to me is that what really matters with crime and justice that gets it to break out of a narrow equilibrium are things outside of the criminal justice system, right? That if you look at the explanations for the great American crime decline of the 1990s, where all crime drops by half or more, that the things that most people would say reasonably caused that crime decline, yes, there was mass incarceration, we talked about that. Yes, there was the end of the crack epidemic, and we could talk about that. There was also, that was when SSRIs appeared in the mass market. And we finally had mass market drugs for anxiety and depression and psychopathy. And they were, these prescriptions were handed out in the tens of millions. And we had a generation coming of age into their most crime prone years that had not grown up inhaling fumes from leaded gas. And we had people getting their, you know, aid for family-independent children, what we used to call welfare. Those payments weren't coming in a check that they cashed and then walked around with cash in their pocket in the as they in the 1980s. By the 1990s, they got a debit card. So they were less of a target for crime. And so there were all these things going on in society that helped to pull the crime rate down that had nothing to do with how many cops were on the street or how many people were locked up or how long they were locked up. So I think that principle that as society is healthier, crime goes down, I think is exactly right.

SPEAKER_02 19:25

And what do you think of the concept that it's falling because the kids these days are all on their phones? Essentially, we've lost this social socialization because everyone's staring at their phones, which is bad, but also it means that they're not going out and robbing and killing each other.

SPEAKER_00 19:44

I mean, I think there's some merit to that. I mean, again, I'm a big believer in routine activities. And so if your routine activities is that you're inside looking at a screen and not outside, where you're both somebody who's a target, you know, that you're susceptible to being victimized or you're, you know, able to do something you shouldn't do, I think that makes all the sense in the world. There's not a lot of great evidence on this yet. There is some evidence from the moral panics around video games that video games had a suppressing factor to crime, whether that was just a routine activity story where young men were playing a lot of video games instead of committing crimes, or whether they got some of the aggression out through the video game instead of out on the street. We don't really know what the causal mechanism is. But overall, it makes a lot of sense that how people spend their time affects their criminal risk both directions, perpetrator and victim. So I think there's probably some merit to it.

SPEAKER_02 20:47

What do you think is the biggest misperception today related to US crime or policing trends? I think it's the policing trends.

SPEAKER_00 20:57

So when so crime is way down, we know from Jeff Azher that crime went down in 23, it went down more in 24, and then it accelerated in 25. And Jeff's early numbers, you're welcome to, I'll talk about you like you're not here. But if you want to correct me, go ahead. But it suggests that 26 so far is another year where crime rates continue to fall pretty substantially. And when I read the newspaper accounts of what's going on, what I hear is this town created this task force, and that town created this new squad, and this town hired some officers. And I say, look, this crime decline is everywhere. It's happening in rural areas and suburban and urban. It's happening in red states and blue states and big cities and in the countryside. Um, it's happening all at once. The cause must be national. And so what your local law enforcement agency does matters, but it matters in the same sense as when We talked about legitimacy and the threats of legitimacy around George Floyd's homicide as being a thing that pushed something that was up a little higher. In this case, if you're doing really good evidence-based policing, I can absolutely see that taking this downward trend and making it even more downward. But it doesn't explain the trend. And the great thing that's happening in one place doesn't affect another police intervention in another place. And so the idea that suddenly we've had 90,000 local law enforcement agencies innovate simultaneously in ways that are bringing crime down everywhere, there's no data to support that. So I don't think that's what's going on.

SPEAKER_02 22:45

So how do you communicate that when you constantly like, you know, City X is seeing an increase? Well, we got to do more for the police. City X is seeing a decrease. It must be because the police did X that was also happening at the same time the decline started. How do you how do you communicate? How do you get people out of that pattern of thinking about the solutions specifically in that lens?

SPEAKER_00 23:09

I mean, if I was in marketing, I would love to market police, right? Because when crime is up, we need more police to fight this scourge. And when crime is down, it's look how well the police are doing. Let's give them even more resources. I think somebody other than the research community has to be the one, be the messenger here. So I've seen, I think it's Mayor Scott in Baltimore talk a lot about how ARCA funds helped Baltimore. And he talks about how it helped them fund more police officers and more police tech, but he talks even more expansively about how it allowed them to open all the pools and all the community centers and hire more teachers and hire more behavioral health specialists and to fund their community violence intervention programs and their local neighborhood nonprofits, and that those were really the things driving down crime. So, you know, you have to be careful because the I say this all the time. I'm not saying that policing doesn't work. What I'm saying is America is not willing to fund more policing than we have today. It's not willing to accept less, but it's not willing to fund more. So we are in an equilibrium. And the only way to break out of that equilibrium and have more public safety is to do other things outside the system. And we seem to, because we had this amazing natural experiment around COVID in the way we fund state and local governments, we seem to have found some things that work and we should really focus more on them. But nobody cares when I say that.

SPEAKER_02 24:37

Because you're describing COVID as an amazing natural experience.

SPEAKER_00 24:41

Right. Well, that's probably not going to win me uh yeah, you guys.

SPEAKER_02 24:45

But just try try potty training triplets and then we'll see how I it's a thankless. Anyhow, that's neither here nor there. So why do you think I mean this is the key challenge that I try to address as frequently as I can, but like we know these trends. You and I talk about these trends, we like intuit these trends. People don't always believe them. And it's taken a long time just to get people to believe the basic, which thing that's been happening for three years of murders falling. So one is my my two questions are why do you think this is the case? And what can we do about it?

SPEAKER_00 25:19

So I think I give more, I give people more credit than maybe others do, right? Like, so I think we were in a situation for a long time where the American public just didn't have very good information. So the violence spike happens in 2020. In 2021, the FBI changes how police departments report data. So absent third-party intervention like Jeff Asher, Council on Criminal Justice, major cities chiefs, and dork, my own uh group did this for a while. We wouldn't have known until the fall of probably 2023, what had happened in 2020, because we didn't have 2021 data. They didn't release the 2021 data until the fall of 2022. So the next year we had good data after 2020 was 2022 that they didn't release until 2023. So, what how is the public supposed to make anything of data like that, right? When it's so delayed and incomplete that you just can't help but but sort of wave your hands at it. I think the public is sharper than we give them credit for. I think there are people whose politics are their identity and they're never going to be moved off of a position, and that happens on both the left and the right. But I think for the vast majority of the public, if you give them timely data that is well explained, I think you can change people's minds. And but you have to reach people where they are. And I think the old model of let's wait for the data and then let's write a journal article and then let's get it peer-reviewed. And that all happens like five years after the whatever the event was that generated the data. Um, and then it gets cited by six other people over the next five years, that that model isn't changing minds, right? And the this conversation is part of a new way of communicating. And I hope a lot of people who listen to you listen, not just because they think you're interesting and your guests are interesting, but because they're thinking about better ways for better communication strategies for themselves, because the research community has an enormous amount to contribute. Um, and it is sort of muzzled itself. And hopefully, hopefully that will happen less in the future. And I think you have to be optimistic that if you say more and you say it clearly and your logic is accessible to other people, that you can change people's perceptions in ways that are really positive.

SPEAKER_02 27:57

And if there was a data source, you were the the czar of crime data or policing data, and there was one source that we aren't collecting now that you'd like to see collected, what would it be?

SPEAKER_00 28:09

Yeah, I think we we need to know more about what police do. We have almost no visibility into what happens to an officer on the street, right? We don't know you report for your shift. We don't know what time you go on duty, we don't know what you do while you're on duty, we don't know, we don't know anything. I remember I was working on a project in New York City, and they had this sort of comp stat for public housing. And one of the things the officers were charged with doing was doing wellness visits up in these high-rise buildings, and they were supposed to call in to dispatch to say, hey, I'm leaving my vehicle, I'm going into this high-rise, you know, I'll be about 10 minutes just so that they could keep tabs on the officers. And they wouldn't even do that. They just do whatever they were going to do. So we don't know what they do. We don't know how many officers are on the street. We don't know how they spend their time. We don't know why we spend so much time on overtime. And because we don't know any of this stuff, we don't know, we can't figure out what the most productive use of police are, right? We can't figure out do we have the right number of police? Are they in the right parts of the city? Are they doing the right things with their time? Do we need different training? Do they need different technology? So we don't, we don't know anything about any of it. And we're left trusting them. And there are times of high trust and low trust, as we've discussed. But it doesn't, you know, if you were running a big business that way, and there are a lot of billion-dollar police departments, and you didn't know where any of your consumer reps were any at any time of the day on any day, you you would want to practice your business a different way going forward, I suspect.

SPEAKER_02 29:47

Yeah. I mean, I love that idea. I think that the way forward with that is something like collecting a national repository of calls for service. Yeah. And then, I mean, the key is that we don't know what is a good response time? You know, what is a city like Los Angeles or Detroit or New Orleans relative to the national benchmark? And we have none of that. And so I don't know how you do it without collecting, you know, tens or hundreds of millions of thousands, millions of rows of calls for service data and just analyzing it.

SPEAKER_00 30:16

And I mean, and you know, when you do look at these calls for service data in some places, what you get is the time the officer calls in and says, I'm responding to when they show up at the scene, but you don't know the call came in at 10. You don't know all three of those things when the call came in, when the officer said, yes, I'm going to the scene and when they got there. So we don't know how much Americans are willing to pay for faster service. We don't know if they care. We don't know if it's important or not. Uh, in many major cities, everything in seriousness from a burglary on down to less serious things. A lot of places you just call in, you get a police cadet or you leave a voicemail and they call you back and say, here's your report number for your insurance company, and nobody ever comes out. And so it does a little bit of a disservice. It says these crimes aren't solvable. And so people don't necessarily want more policing because they don't see the things that police do as solving anything. And so we're stuck. We're sort of trapped in this equilibrium. We can't make the policing better. We can't help officers do their job better. We leave the pol the citizen completely uninformed. It's it's just a poor place to be. And it's really no different from where we were in the 1950s and 60s. It's sort of astonishing how far everything else has moved and how little this is.

SPEAKER_02 31:33

What do you see as sort of the future of, I guess I'm constantly asked and constantly punt on the question of like what's next for with the murder rate and violent crime? Is this going to continue or is this something that you think is sort of we're going to get back to that sort of leveling off that you talked about in the 2000s?

SPEAKER_00 31:50

So I think that there's two things to say. So one thing is maybe we'd come back to is where's policing headed? Because I do think we're about to see a major shift in how policing happens in America. But with respect to the short-term trends, I you know, I'm not, I wish I could be more optimistic, but I'm not. And I see things ticking up probably as soon as this summer. And the reason is very simple, right? All that pandemic support for states and local governments is either over or has to be spent down by the end of this year. And most of it's already been spent. So local governments are in the position of trying to replace that money so they can continue their services. And over time, they, you know, the gaps appear and those services start to fall by the wayside. And we're just seeing less and less of this of all these successful enterprises. So that's not good. And I think that, you know, there were particularly around shootings, or in the pandemic, there was a period where an astonishing number of people got shot all at once. And I think that I hear from police all the time and prosecutors all the time that there are a lot of guys who are willing to shoot back who are involved in this lifestyle, but there aren't that many who are willing to shoot first. And a lot of the shoot first guys got shot, got arrested, got incarcerated, and there was a gap that takes time to fill. And I think it's from what I hear from people on the streets starting to get filled. So I'm I'm a little pessimistic, I have to admit.

SPEAKER_02 33:22

That's really pessimistic. I was assuming you'd go down to, you know, we're gonna hit zero murders by 2032 or something like that.

SPEAKER_00 33:28

Asymptotics are real, right? As the rate gets lower and lower, it is harder to make gains. That's just the way math works.

SPEAKER_02 33:38

So my last question is much more general. What role does AI play in your work? And do you have any thoughts on sort of how AI fits into this field and this area of study going forward?

SPEAKER_00 33:52

Yeah, I don't know. I mean, I use AI, I'm a super user, so I have accounts with both, you know, Anthropic and OpenAI. I I use it as a giant search engine and a thesaurus for the most part. I definitely use it to help me code. Um, it's gotten very good for that. And I think it's gonna allow us to do more analysis more quickly. Now, whether we do more good analysis or not, it is entirely dependent on the human beings behind it. Because, you know, when it goes back to the joke I made about marketing police, right? When violence goes up, we hire more police. And so there are more police patrolling places with more violence. So now you have more police and more violence. Does that mean more police cause the violence? No. Well, no, but the data will show you that it does. Um and so human beings, really good researchers, have to be behind all this stuff, really thinking hard about what the data are telling them. And it's not just enough to have great data architecture and have this brilliant coding assistant to help you run your models faster. You still have to have the right model and you still have to deal with the fact that nature and human behavior are super messy and complicated. And there are reasons why the hard questions are hard and will remain hard. So I think we'll do a better job overall. I think we'll see more good papers, but I don't know that we're getting any closer to solving the biggest problems, which is, you know, why do people commit crime and what kind of solutions are the best solutions, and how all of that is all mixed together and unpacking it is very hard. So I am optimistic. It's all helpful, but um, you know, training young people how to do this really well is was the solution 50 years ago. It's the solution today, and it'll be the solution 50 years from now. There we go. Well, John, what's next for you and where can people find you? Uh I what's next is is more thinking. Um I have a sub stack called external processing that a lot for people to sign up for. I'm direct the Center on Public Safety at NORC at the University of Chicago, and we have great research at NORC.org that I would invite people to take a look at and see if it's helpful.

SPEAKER_02 36:06

And what does Nork stand for?

SPEAKER_00 36:09

So Nork used to stand for the National Opinion Research Center. We were founded at about the same time as Gallup in the 1930s, but somewhere along the line we got acronymized, acronized. I don't know. Now we're just NORC.

SPEAKER_02 36:23

All right. John, thank you so much. This has been fantastic. I really appreciate it. Great conversation, and everyone should check out External Processing to his newsletter.

SPEAKER_00 36:31

Thank you. Thank you for having me on.

SPEAKER_02 36:33

Thanks for listening to the Jeffalytics Podcast. Be sure to subscribe and to learn more, head on over to ahdatalytics.com for more information and previous episodes. If you like what you heard, please leave a glowing review, which will help others to discover the show. Until next time, I'm Jeff Asher.