Why Police Officers Rarely Change Jobs

Subscribe here: Apple Podcasts | Spotify | YouTube | Overcast | Pocket Casts

In 2020, the law professors John Rappaport and Ben Grunwald published an article called “The Wandering Officer,” where they found that police officers who had been fired earlier in their career were more likely to be fired again—and more likely to receive a “moral character violation” complaint. The research has informed the debate about police reform in the years since.

Four years later, Rappaport—along with Jonathan S. Masur and Aurélie Ouss—has published another article , revealing that, on average, policing is a stagnant job, meaning that police officers very rarely move between jobs and departments. The researchers got access to 20 years of data and found that roughly 70 percent of Illinois officers in their sample had held just one job, and another 21 percent had held just two. And it’s not necessarily that they’re choosing to stay in one job. A lot of policies have made it costly to switch agencies.

What the researchers found is that when police are stuck in the same law-enforcement agency, stuck with the same peers and commanding officer, they are also more likely to adhere to the “blue wall of silence”—an unwritten code to never implicate another officer of wrongdoing. As the authors write, “Unlike in other professions, it is terribly difficult for police officers to quit in protest or even to object to the orders they receive. If the policies and priorities of senior law enforcement management are misguided, the entire law enforcement organization is stuck with them.”

In today’s episode of Good on Paper , Rappaport joins me to discuss both of his papers and the factors keeping police locked in place.

The following is a transcript of the episode:

Jerusalem Demsas: On July 6, Sonya Massey was shot and killed in her Illinois home by a Sangamon County police officer, Sean Grayson. Grayson and another police officer had come to Massey’s home after she’d called 911 about a prowler . Once inside her house, Grayson asked Massey to remove a pot on the stove—an interaction which rapidly ended with him opening fire. He claimed to have feared for his life because of the pot of water. And, as a result, the 36-year-old mother of two is dead.

This is a horrible and all-too-common story. And in the weeks following Massey’s death, one additional fact came to light: Grayson had held five jobs in four years, all at different police departments in central Illinois. Now he’s been fired from a sixth. The AP reports that at one of those jobs, as a sheriff’s deputy in Logan County, “he was reprimanded for ignoring a command to end a high-speed chase and ended up hitting a deer.”

Why do police officers like Grayson keep getting hired?

Part of the answer comes from today’s guest, UChicago law professor John Rappaport, whose research on “wandering officers” revealed the extent to which previously fired officers find jobs in new departments and the structural incentives of small departments to keep hiring them.

But Rappaport’s newest paper complicates the thinking around police-labor mobility. While his earlier research shows that wandering officers are more likely to act improperly in their new roles, his new research reveals how other officers actually find it very difficult to move, exacerbating norms like the “blue wall of silence” that keeps police from reporting on each other.

My name is Jerusalem Demsas. I’m a staff writer here at The Atlantic , and this is Good on Paper , a policy show that questions what we really know about popular narratives. Today’s episode explores how we should think about the problem of police-labor mobility. Are there ways that police are too stagnant? Would there be fewer instances of police brutality and misconduct if it was easier to leave one police job for another?

All right. Well, hi, John. Welcome to the show.

John Rappaport: Hi to you. Thanks for having me.

Demsas: Yeah, of course. So I feel like this paper is kind of interesting. I love talking about—in my reporting, I report about the law, I report about economics, and I report about political science. I feel like we found all three in one place. And so I’m really excited to have this conversation.

But I want to start with the paper you wrote four years ago. And the context of that paper is that often police officers who are credibly accused of heinous acts of murder or corruption are revealed to have been previously fired or let go from other agencies. Not always, but this happens at a level where it started to cause a lot of questions: How often is this happening? Is this something that really explains a lot of police misconduct or just the police has bad employees in these instances?

So you wrote a paper called “The Wandering Officer,” and you and your co-author published this at The Yale Law Journal . And I’m hoping we can start there, and you can walk us through that paper. How many wandering officers are there, and how did you find that out?

Rappaport: Sure. So we were using a statewide dataset from the state of Florida, and we had read, like a lot of other people, all sorts of anecdotal accounts about these wandering officers, and we were really curious just how common they were. And when we ran the numbers, we found that about 2 to 3 percent of all police officers in the state of Florida at any given time are wandering officers. That is about 1,100 officers, give or take.

To be clear, we defined wandering officer in the paper as an officer who’s actually been fired before and then obtains another job. So that defined that way, we found about 2 to 3 percent.

Demsas: Are there ways that you’re worried Florida might be an outlier, or are there ways that you checked it against other states?

Rappaport: We’re just starting to do that, actually, and work here in 2024. I’m starting to try to look at other states. The data is not as good, but maybe we can find out a little bit just to see whether Florida seems way out of line. And the preliminary results suggest that it’s not way out of line. There’s, of course, a range from state to state, but I don’t think Florida is out of the mainstream in any respect.

Demsas: So your main finding, as you said, is that, in any given year, roughly 1,100 officers who had been previously fired were working for Florida agencies. It seems like a substantial number, but you also say that’s just 3 percent. How should we think about that? Is that a lot of people?

Rappaport: I think it’s in the eye of the beholder. When we wrote the paper, we genuinely didn’t know whether people would read it and think, Oh, they’ve just proven that this is not really a significant problem , or, They’ve just proven that this is a significant problem. And it seems to me, from the reception that the paper has received over the last few years, that people think it’s a serious problem and that even 2 to 3 percent is way too many.

Demsas: Well, what about you? What do you think when you see that number?

Rappaport: I mean, I think it’s high if you think about 1,100 officers, and then you think about all the different interactions that any given police officer is going to have over the course of, say, a year. I think we multiplied it out somewhere in the paper and said, This is probably hundreds of thousands of civilian interactions in the year .

And another important finding in the paper is that wandering officers go on to get fired or be subject to serious complaints of misconduct at about twice the rate of other officers. And so once you have that finding in mind, then you think, Eleven hundred sounds like a lot, and a couple hundred thousand interactions with people who are going to go on to get fired or commit serious misconduct at highly elevated rates is too much .

Demsas: I’ve been at this job for a little over two-and-a-half years now. That’s the longest job I’ve ever had in my life. And most of my friends who are either Millennials or younger, or even a little bit older than that, are very used to job switching. We’re used to hearing adults and employers say in the media that job switching means that you’re lazy or you’re bad at your job. But that’s not something we take, usually, very seriously. Why would that be different with police officers?

Rappaport: Well, I think it depends on what kind of labor mobility we’re talking about here. To shade into the newer paper a little bit, I think we may have too much of the bad kind of mobility and too little of the good kind of mobility. So I agree; I think that changing jobs is not inherently a bad thing, but when we’re looking at something like the wandering officer, it’s a little bit different from your just typical, average Millennial.

We’re looking at officers who’ve been fired. It’s hard to fire a police officer, so if they’ve been fired, that already says something significant about them. And if they’re then bouncing around in subsequent agencies, I think that looks a little bit different from you and your friends changing jobs.

Demsas: You also find that officers who were fired tend to move to smaller agencies with fewer resources in areas with larger communities of color. Why is that happening? Why is that an easier place for them to find work?

Rappaport: I think there’s just a greater demand and a smaller supply of officers in those areas. We’ve been hearing a lot in the news, especially over the last four or five years, about the hiring or the retention crisis in policing. I don’t know how real that is or how ubiquitous that is, but I know there are a lot of places that don’t have a ton of good applicants, and they’re the ones that tend to be reaching to hire these wandering officers.

Demsas: Should we expect to see in smaller agencies just higher rates of police misconduct? Is that what we see when we look at the literature?

Rappaport: That’s a good question. I’m not sure that we really know the answer very confidently to that question. But it’s a reasonable hypothesis based on what we do know.

Demsas: One thing that got me thinking, especially when you were talking about the need that smaller agencies really have, is the trade-off that they might have to make, right? Because if you’re a small, cash-strapped agency, and your choices are to hire a police officer that’s been fired before—but maybe he seems fine in the interview, and you want to give him a chance—what’s the risk of hiring that individual person versus the risk of not having an officer, which we know can often lead to also high rates of crime and victimization? How do we weigh those sorts of things?

Rappaport: Right. It’s not an enviable position to be in, I think. Especially before the research came out, you might even think that some managers and small-town agencies would think that a guy who has been fired before is going to behave better because even though it’s, in general, very hard to fire an officer, Here’s one that got fired, so he knows he’s not invincible, right? And he’s going to fall in line when I hire him this time. And that’s what he’s saying to me in the interview , as you suggested. But now, after seeing the research and realizing that, at least on average, wandering officers get fired and receive serious misconduct complaints at twice the rate of other officers, it’s hard to follow that line of thinking any longer.

But there are other trade-offs, as you say. You have to think about hiring this guy or hiring a rookie. Now, he may be more likely to commit misconduct than a rookie, but he might also be cheaper than a rookie because we have to train a rookie. It’s going to be a long time before a rookie can be out on the streets by himself doing the same kinds of things that this veteran officer might be able to do.

You posited that we’re a small, cash-strapped agency, so that’s another trade-off I have to make. And then, as you say, there’s even a third trade-off, which is: Should I hire this guy or hire nobody? And that’s going to be, obviously, very context dependent, but there are situations in which you could imagine a reasonable manager concluding it’s better to have somebody than nobody.

Demsas: The main reason you’re here is because you also wrote a new paper very recently. And when I first heard about “The Wandering Officer” paper from a few years ago, and it got a lot of attention—I don’t really focus on criminal-justice reform as an issue I report on—but what I generally took away from it is that we should kind of be suspect, very suspect of any officer trying to get a job in a new area, any officer that’s been fired before.

And I don’t know if I formed the strongest policy beliefs off of it, but it did feel like the main takeaway was that police mobility was bad. Is that how you saw it? Did you feel like others read your paper as sort of an indictment of police-labor mobility?

Rappaport: I think some people might have, but that’s, maybe, over-reading the paper to some extent because we are talking about mobility just among a very small segment of the policing profession, rather than mobility of your average officer.

Demsas: I’m hoping you can talk to us about this new paper, then. You have a new paper out. It’s called, “Labor Mobility and the Problems of Modern Policing.” What’s the top line of that? What are you trying to do? What did you find?

Rappaport: Well, first of all, we do a couple of things in the paper. There’s an empirical part of the paper, and then there’s a conceptual or theoretical contribution. In the empirical part of the paper, we wanted to just try to figure out: How much lateral mobility is there for police officers? I had heard, anecdotally, from someone in the policing world that, Hey, it’s really hard to change jobs if you don’t like your job in policing.

And this was news to me, maybe partly because of the wandering-officer research. But this was news to me that there are these police officers who are at agencies, and they really don’t like their job, but they feel like it’s really hard for them to move.

And so I wanted to figure out whether this was, first of all, accurate and whether it was a general phenomenon and, second of all, try to think a little bit about why that was the case and think about: What are the implications of that phenomenon if it turns out to be real? And so that’s what we try to do in this paper.

Demsas: How do you do it? I know you’re looking at Illinois data now. Does Illinois also have a good dataset available? Or why’d you pick them?

Rappaport: Illinois has pretty good data. Actually, in an appendix of the paper, we replicate most of the results using the Florida data to try to stave off any concern that this is something unique to Illinois. There was one good thing about the Illinois data that’s not true of the Florida data, which is that we can see officers’ ranks—police officers, detective, sergeant, and so on. There’s not really one specific way to measure labor mobility, so we look at it from a few different perspectives.

The first thing we do is just look to see, Well, how many different jobs do police officers in our dataset have? And it turns out that 70 percent of the officers in our dataset have only one policing job in Illinois, and another 21 percent have only two jobs.

Demsas: Wow.

Rappaport: And this looks very similar in the Florida data. And so right away that tells you that most people are finding a job and they’re staying put, or they’re making one move in their career and that these officers who are bouncing around from agency to agency to agency are something of an anomaly.

Demsas: Why aren’t they moving? Why aren’t they moving more?

Rappaport: This is a big question, right? This is a big part of the heart of the paper, is to try to figure out: What are the causes of this friction that keeps officers from changing jobs? And I think there’s a lot of different reasons here. We sort them into what we call nonlegal and legal reasons. In terms of the nonlegal reasons, we’re talking about all sorts of different things—like, for example, the fact that most police departments are natural monopsonies. They are the only game in town when it comes to policing.

If you work in a small city in Illinois, there’s only one police employer in your city, which is very different from if you were a mechanic or you were a plumber or you worked in health care. There’s probably lots of different jobs that you could get even in a small city. But with policing, there’s very few, often only one option within your commuting zone.

It makes it really hard to change jobs, especially when you couple with that residency requirements that a lot of police departments have, that you have to live in the city where you work. Those might be a good thing for a lot of reasons, but they do make it harder to change jobs, because it means that changing jobs means you must move your home.

Demsas: To me, there’s obviously a lot of research and push to get police to live in the places they work, right? There’s a lot of people who argue that it makes people more grounded in the communities that they’re in, that it’s really good if people have been there for a long time. There may even be domain-specific knowledge that they’re getting, whether it’s the knowledge that they’re getting of walking that community, knowing the people, knowing what the hotspots are on their own, feeling comfortable in it, not feeling on edge.

But then also, even on just a more practical level, regulations might shift from jurisdiction to jurisdiction. Is it the case that this sort of stability might outweigh the costs? Like, there are a ton of benefits that might be accruing that are, maybe, difficult to measure?

Rappaport: It is possible. We raise this as a potential concern. But we say in the paper that there are both costs and benefits to limited labor mobility for police officers, and we’re not exactly sure which one to weigh out. We are concerned about the potential costs of low mobility. But there are benefits, as you say. You might attract a certain kind of candidate who wants to make a commitment to that neighborhood or that town.

There’s actually research—Elizabeth Linos at Harvard has done some research about what kind of job announcements attract people into the policing profession. And she finds that people are more attracted by job announcements that talk about job stability than they are by announcements that talk about public service, and that this is especially true for women and people of color. So we don’t dispute at all the fact that there are lots of potential upsides to having limited mobility in the profession. But there are some downsides, we think, too.

Demsas: And what are those downsides? If you’re staying in place, what is the big cost that’s happening there? For some people, that feels like the dream, that you get to have a job that you get to stick in for a long time.

Rappaport: If you like it.

Demsas: ( Laughs .) Yeah.

Rappaport: If you like your coworkers. If you like what you do. If you like your superiors.

I think, again, there are multiple costs we could talk about. The first thing that comes to mind is just this question of stagnation—the idea that you may well have a lot of mismatch in your agency. You have a lot of officers who aren’t necessarily well matched in terms of the skills that they’re being asked to use, but they can’t make that lateral move that would allow them to get a better match.

This kind of thing could lead to burnout. You could have officers who are unhappy on the job. They don’t like their job. Maybe they don’t like their coworkers. They don’t like their partner. But it’s very hard for them to move, and so they may burn out. They may decrease their effort. They might even start breaking the rules more often.

I also think that—although it’s true that job stability might attract certain candidates into the profession—limited lateral mobility might also deter other kinds of candidates from entering the profession. And maybe the type of people who want a career that’s going to allow them to bounce around and be upwardly mobile—maybe those people are not entering the profession to begin with.

Demsas: It feels like some of the things that you’re talking about that are restraining police mobility seem almost just insurmountable. Like, the fact that there are going to be large gaps in distance between different police jurisdictions just seems like it obviously has to be that way. You can’t really have a thick labor market for policing. You’re not going to have the Chicago [PD], the Atlanta [PD], the LAPD all in the same spot for someone to be able to bargain between for jobs. So is this something that’s just insurmountable? Or how much of it are factors that are just a part of policing’s function and form in America?

Rappaport: Yeah. I think there are some factors, some sources of friction that are insurmountable, to use your word. But there are others that we didn’t get to yet that I think are much more malleable. There’s a whole set of laws, I think, that make it more costly for police officers to change jobs.

The first thing that comes to mind is the pension system. Unlike a lot of private-sector workers in this country, police officers don’t have 401(k)s. They don’t have defined-contribution retirement plans. They have defined-benefit plans, which are also called pensions, right? And pensions typically have a very long vesting period—could be something on the order of 10 years.

If you change your job—you start working at one agency. You don’t like it. You want to change. If you change your job before the 10-year mark, then everything that your employer contributed to your pension is going to go up in smoke. It’s just gone. You can’t take your service credit with you, meaning if you start over at agency No. 2, you don’t get credit for the hours that you put in at agency No. 1 toward your pension. You have to start over at zero.

You also are going to lose your rank. So if you’re a sergeant at agency No. 1, and then you move to agency No. 2, you start back as a police officer. You’re also going to lose your seniority. So seniority is a little bit different from rank. Seniority is usually governed by a collective-bargaining agreement, and it gives you access to benefits like vacation days and equipment and good shift assignments, right? Who takes the graveyard shift? That’s determined by seniority. These kinds of things are legally created, and I think they are potentially changeable.

Demsas: Why do these things exist, right? Why is it not possible to port your rank or to port your seniority or to bring your pension along? And you would imagine there might be a lot of variation across a country of 90,000 different localities. Why is there so much stagnation here?

Rappaport: Yeah. So here, too, I think there’s a host of different reasons at play. One way to think about the question is: Why do police officers tolerate this limited mobility, right? They’re thought to be politically powerful—they’re a politically powerful lobby, politically powerful unions. And limited mobility is typically, all else equal, thought to be bad for employees. It’s good for employers because it suppresses wages and it reduces turnover costs, but it’s bad for employees because it suppresses wages. And so why do police officers put up with this?

I think the first thing you have to see is that the very same rules that make it hard for an officer to lateral out of an agency also make it hard for other officers to lateral into the agency and compete with that officer for promotions, shifts, resources, and so on. And so these rules that restrict lateral mobility are fundamentally protectionist rules.

Now, why would police officers want protectionist rules? I don’t know exactly, but we can think of a couple of hypotheses. One is that maybe police officers think that they’re unlikely to want to move. They think of themselves, like you say—they’ve found the dream. They’ve got a good public-sector job. They’re serving the public. They have a pension coming to them if they stay put, and they think they’re unlikely to want to move. If you think you’re unlikely to want to move, then you don’t really care about rules that make it hard for you to lateral out. You do still care about rules that make it easy for other people to lateral in and compete with you. And so you want to clamp down on that.

But even if it’s not the case that cops believe they’re unlikely to want to move, there’s a sort of jurisdictional collective-action problem going on where I think police officers, in the aggregate, on the whole, would collectively be better off with more mobility because it helps to increase wages, among other things.

But for any given officer, the scope of her influence is really limited, meaning she can have sway, at most, over the rules that are in place at her agency or in her state. She could potentially put pressure on the relevant decision makers to change those rules. But she can’t do anything about the rules in the neighboring agency or the neighboring state. There’s no way that she can put pressure on them to try to make those rules more welcoming to potential transfers like her.

And so when you’re in this kind of scenario, it’s in each individual police officer’s interest to seek those protectionist rules that decrease external competition. And the officers in the other jurisdictions are all going to make the same decision. And so we end up with this sort of protectionist equilibrium that is understandable for any given officer but might be making officers, on the whole, worse off.

Demsas: I want to throw another potential hypothesis at you, too, which is that: Is it possible that the police unions are not, maybe—there’s a principal-agent problem here?

There may be some people—if you’re already in the police union or the police union staff themselves are maybe more responsive to the people who have been there a long time, of course, they themselves feel seniority. The people who staff police unions or who are elected are going to be long-term players there too. And so they have an interest in maintaining their status as a player in local politics and also a player in maintaining power and bringing resources to people who have been there for a long time, and they’re less interested in police officers as individuals or the larger labor sector or labor market of police officers?

Rappaport: Yeah. I think that’s completely plausible. Maybe to restate it slightly would be to just say that the people who have the power in terms of lobbying and influence, that subset of people think they’re unlikely to want to move, because they’ve got it made, right? And so maybe they’re not thinking about what your typical officer in year three wants; they’re thinking about what they want.

[ Music ]

Demsas: All right, time for a quick break. More with John when we get back.

[ Break ]

Demsas: One implication of your paper also plays into these questions about police misconduct, right? Because if you feel like you can’t move, or you feel like you’re probably going to be stuck at this agency for the rest of your life—and the people around you are going to be responsible for your promotions; how your day goes; what shifts you get; how great your everyday, day-to-day career feels—you’re much less likely to report on your fellow officers. Did you find evidence of this? Or how do you think about that and the potential impact of keeping officers stuck in place?

Rappaport: We don’t find empirical evidence of this. This is not something we were able to test with our data. But as a theoretical matter, I think it’s very plausible that, when officers have limited mobility, they’re basically, as you say, tied to their employer and their colleagues. And they take on a very strong, self-interested incentive not to rock the boat and to comply with whatever their superiors want and to conform with the norms that are created by their peers or the peers who came before them, right?

And that’s true even when those norms are pernicious and even when those norms look like the blue wall of silence, right? The blue wall of silence is a norm pursuant to which officers refuse to inculpate each other and potentially even lie to protect fellow officers from allegations of misconduct. Even if you join an agency and you think the blue wall of silence is a terrible idea, it is incredibly risky for you to push back against that norm. If you do, you might be ostracized. You might be put on the graveyard shift. You might have people not respond when you call for backup. And then you’re stuck. There’s nothing you can do about it, because it’s so hard to change jobs and start over.

Demsas: Yeah. And it seems like here is a place where there’s an interplay between your first paper and this one, too, in that you also know that if you leave, there’s a high likelihood that you’re seen as a bad apple to other agencies. And so it actually kind of implicates you, too. Even though it is true that the average person who gets fired from a police station or is likely to have police-misconduct problems, people also might feel like, Oh, just leaving might give me that stench of being a bad apple , as well.

Rappaport: Yeah. That’s exactly right, especially in a world in which it’s difficult for potential future employers to really verify these things.

Demsas: So what should the policy response be for someone reading both these papers? If you’re a state legislator or maybe you’re even the head of a police agency—you’re chief of police—and you’re hiring other officers, right? When you look at both of these papers, what sorts of things would you want people to be doing?

Rappaport: I think one potential reform that implicates both lines of research is improved information flow. As I said a few minutes ago, sometimes I think agencies hire a wandering officer because they don’t even know that they’re dealing with a wandering officer. Either they didn’t have the resources to do a good background check, or the records that they were looking for weren’t available, or maybe the officer is not technically a wandering officer, because he was allowed to leave voluntarily in the face of misconduct accusations rather than being fired. And it’s really hard to know what’s going on, because there’s no centralized database that keeps track of police misconduct.

I think if you had a centralized database at the state level or even at the federal level, this kind of information flow would do a couple of things. I think it would make it easier to identify wandering officers or officers who raise some of the same concerns as wandering officers, but it would also bolster confidence in agencies when they’re looking to hire lateral transfers. Because right now, as you said a few minutes ago, we might assume that, Well, this person’s on the lateral market. They must have done something wrong. Why else would they be leaving? But the more information we have—the more transparency there is—the more comfort we can take that we really know about this person’s background.

Demsas: I wonder, too—because I think you touched on this in one of your papers—about the efficacy of using a national database here. How much would it solve if you actually had a lot more information about why people were leaving? And is that something that would cut against the interests of a lot of these local players?

Rappaport: I think it depends on what’s in your database. There is already one kind of national database, which is called the National Decertification Index. So something we haven’t really talked about yet is the fact that police officers are a licensed profession, just like a lawyer is licensed or a barber is licensed. When you want to be a police officer, you have to be licensed in the state where you want to work. This is something that creates some friction on interstate moves because if you want to move to a new state, you have to get a new license.

There is a national database that keeps track of—or tries to keep track of—officers who have been decertified, meaning they’ve committed misconduct that’s serious enough that they actually not only were fired but had their license pulled. So they were told, You can never be a police officer in the state of Florida again , right? That’s put into a national database—sometimes—with the idea being that before Georgia, let’s say, certifies that officer, they’ll look in the database and see that the officer was decertified in a prior state.

That’s not a lot of information, because not that many officers are decertified. That’s really only the worst of the worst of the worst. And I think to achieve the kind of benefits that we’re talking about here, you would need to have a more inclusive database that covers all terminations and maybe even misconduct that doesn’t lead to terminations.

Demsas: I’m curious here, again, about the demographics because, in your paper, you find that there’s greater mobility for white officers and also for less-educated officers. And you say that those are principally attributable to voluntary rather than involuntary moves. Why would there be greater mobility for those two groups, and how does that fit into the story you’re telling here?

Rappaport: We don’t really know. And we suggest that those findings be interpreted with some caution because they don’t exactly replicate in the Florida data. They’re sort of directionally similar, but the magnitude of the results is different. You could tell a lot of different stories about why white officers or white-male officers have more mobility. They might just have more labor-market power in general. I don’t know why less-educated officers would have more mobility. It’s a mystery to me.

Demsas: You’ve referenced this a little bit in your last couple answers here, but I’ve been thinking about how much of this is just a classic occupational-licensing story leading to declining mobility. We know that there are tons of jobs that require licensing—public-sector jobs but also private-sector jobs—and that’s led to a serious decline in interstate mobility.

For me, I’m trying to think through because you’re offering a lot of different reasons why you would find these frictions in the labor market for police. And I’m trying to think about how you would order your concerns, like which ones you would find to be most binding for police officers. Is this a classic occupational-licensing thing—if you were to just make it a lot easier for people to switch between jobs, and maybe you would have standardization across states about what the job was and the laws that exist across jurisdictions—to be standardized? Or is that a very subconcern that you have when it comes to the frictions that exist in this market?

Rappaport: My sense is that certification licensing is one of the smaller concerns. Partly, I get this from message boards. We spent a lot of time, when writing this paper, on police-officer message boards to try to figure out: What are police officers talking about? Either current police officers, prospective police officers, and especially police officers who are talking about changing jobs or thinking about changing jobs—what are they concerned with? What do they say is holding them back from changing jobs?

And the thing is that certification really only comes up when you’re talking about interstate moves. And I think the number of potential interstate moves is probably dwarfed by the number of intrastate moves. And things like pension portability and rank-and-seniority portability, those are implicated even for people who want to make intrastate moves, and I think those are probably more important on the whole.

Demsas: Did you find anything else interesting from perusing those? Did you learn anything you didn’t expect?

Rappaport: When we started reading them, it was all new to us, and so I think we didn’t know what to expect.

Demsas: What sorts of things are they talking about on there, mostly?

Rappaport: They’re talking a lot about pensions and about rank and seniority, and using all sorts of metaphors like, Getting a job in policing is like being married; you should only do it once and expect it to last a long time , this kind of thing. Or, you know, My wife wants to move to Georgia, but I can’t take my pension with me, so I’m going to veto that , right? This kind of thing. And so it was actually really satisfying confirmation of what I think the theory would predict.

As lawyers, we sat down, and we identified these sources of law that seem to be creating these frictions, and then to go onto the message boards and see that, Hey, this is the same stuff that the cops are talking about on these message boards , was really a nice confirmation of the real-world applicability of the theorizing we were doing.

Demsas: So I write about local government a lot, and the main critique I’ve had about it is how local government’s lack of accountability to the mass public has allowed interest groups to fill the void and basically set the agenda for local politics. There’s a political scientist at Berkeley—her name is Sarah Anzia—and she finds that these local groups are very present and influential in local politics. For example, she finds that police and firefighters’ unions’ influence correlates with greater spending on their salaries, and cities with more politically active police unions are less likely than cities with less-active ones to have adopted body cameras.

And, as you’ve laid out clearly, a lot of what’s happening in policy governance here is happening at the local level, and police unions seem to play a large role here. Is a big part of the reform you would want to see to do with police unions? Is there something to do here with, maybe, moving a lot of the governance of policing up to the state level? Do you have thoughts on that?

Rappaport: That’s a really good question. I have some other work about police unions, where we find that collective-bargaining rights increase the incidence of violent misconduct. That’s a reason separate from all the others that I have concerns about police unionism.

It’s a really good question whether it would be a good idea to move more of the governance up from the local to the state or even, you could say, federal levels. I think some things—it would make a lot of sense. For certain reasons, it would make a lot of sense to do that. I do think that greater standardization could help a lot.

Right now, we have about 18,000 law-enforcement agencies in the country, and there’s very little standardization in most states. So the type of paperwork that a police officer has to fill out when he engages in a certain kind of policing activity is going to differ from agency to agency to agency.

This also creates friction. It creates a drag because now if I want to change jobs, I go from Agency 1 to Agency 2, I have to relearn all of Agency 2’s paperwork and all of their policies, which might look very different from Agency 1’s, right? And so I do think that increasing standardization—it would reduce the value of this agency-specific knowledge and make it easier, less costly for officers to change jobs, which might have some of the benefits that we talked about earlier.

So I also think that we might think about siting responsibility for training at the state level or even federal level because one downside of mobility is that when there’s more mobility, the employer is going to be less likely to want to invest a lot of resources into training, because they know they would be training an employee who might pick up and leave, right?

And similarly, actually, from the employee’s perspective, they’re going to be less likely to want to invest a lot of resources into learning agency-specific knowledge if they know that they might pick up and leave in a few years.

But if training is orchestrated more at the state level or even the federal level, well then, the benefits of that training can be internalized within the entire state or within the entire country. And so you don’t have to worry as much about employers and employees failing to invest in training.

Demsas: Well, our last question always on the show is asking: What is an idea that you had that you originally thought was a great idea, but it turned out to only be good on paper?

Rappaport: Okay, so I do have something in mind, which is: floating.

Demsas: Okay.

Rappaport: Have you ever done this?

Demsas: No.

Rappaport: Do you know what I’m talking about?

Demsas: You mean down a river?

Rappaport: ( Laughs .) No. It’s a very specific kind of floating. This is a relaxation type of activity. Actually, what’s especially bad is that my wife got it for me for my birthday. But you go to this place that looks kind of like a spa, and they have these pools of water that are highly salinic, so they’re full of salt. You’re very buoyant in the water, and they’re body temperature.

Demsas: Okay.

Rappaport: And then they turn out all the lights. And you get down in this pool of water, and you float on your back, and it’s kind of like a sensory-deprivation thing, right? You’re not seeing anything. You’re floating. You feel weightless.

Demsas: Yeah.

Rappaport: Turns out this sounded good on paper to me, but I thought the whole thing was terrifying.

Demsas: Yeah.

Rappaport: I learned for the first time that I could be claustrophobic. I never thought of myself as someone who was claustrophobic. But when you lock me in a little room with no lights, floating in a tiny, little pool of water, I felt very claustrophobic. I ended up sitting up in the pool of water, just waiting for the session to be over.

Demsas: ( Laughs .) How long were you in there?

Rappaport: It was an hour.

Demsas: Oh my God. Are you able to just leave, or does someone have to let you out?

Rappaport: You could have, but I was playing mind games with myself and trying to—again, this was a birthday gift, so I was trying to find a way to appreciate it.

Demsas: No. I do understand. I can totally see how someone gets me—I’m like, Oh, this is gonna be very fun. I’ll lie in a pool . But I just feel like it would be better if the lights were on. Wouldn’t it be nicer if there was some, like, lovely little light? Like, why is that better?

Rappaport: It’s supposed to be meditative, you know. You turn out all the lights. You close your eyes. And a lot of people fall asleep, actually, while doing this. You’re that buoyant that you could actually fall asleep. But yeah, it didn’t work for me at all.

Demsas: Well, hopefully John’s wife is not one of our many listeners on Good on Paper .

Rappaport: Maybe for this episode she is. I don’t know.

Demsas: Well, thank you so much for coming on the show. We really appreciate it.

Rappaport: Thanks for having me. I appreciate it.

[ Music ]

Demsas: Good On Paper is produced by Jinae West. It was edited by Dave Shaw, fact-checked by Ena Alvarado, and engineered by Erica Huang. Our theme music is composed by Rob Smierciak. Claudine Ebeid is the executive producer of Atlantic audio, and Andrea Valdez is our managing editor.

And hey, if you like what you’re hearing, please leave us a rating and review on Apple Podcasts.

I’m Jerusalem Demsas, and we’ll see you next week.