A: Before the 1950s or so, as you’ve written about, this just wasn’t the case. Gangs didn’t exist. Prisoners weren’t “governed” — instead, they had a code which was mostly enforced by decentralized reputational mechanisms.
D: Absolutely. Before gangs took root, norms of good behavior were pretty well understood among inmates. It was clear when someone was following them and it was clear when someone wasn’t. People could individually choose to punish violations or not. That worked really well, but only in relatively small, relatively homogenous communities where there was a lot of agreement about what constituted a deviation. Reputation mechanisms were very effective only when prisons were small enough that you cared about your reputation because lots of people knew who you were and whether you were in good or bad standing.
But as prison populations in California grew through the ’50s and into the ’60s, it became too costly to know everyone’s standing. Reputation became a much less effective driver of behavior. And in the process of becoming less effective, we begin to see a lot of outbursts of serious acts of violence. During the ’50s and ’60s, there were large increases in homicides, stabbings, and riots. And it’s during this turbulent time when the first prison gangs emerge — initially simply for self-protection. But then once gangs are able to effectively protect themselves, it creates incentives for them to engage in a variety of different behaviors.
A: Can you say more about how these gangs actually function? If you’re an incarcerated person, how do they structure your life in prison?
D: I think most importantly is that when someone is new on a yard, the shot caller — either the gang leader himself or one of his colleagues — will go out and basically run through the rules: who holds the keys, who is in charge, how should you behave, and maybe the current state of politics. “This is the person who holds the keys to the yard. They’re the one who’s in charge.”
And then you’re going to program with them. That means, assuming you’re in good standing, you might work out with the gang, you might be tasked with doing certain jobs or activities for them, or maybe you just hang with them. That’s the primary way that gangs will guide you and put guardrails on what is and is not acceptable behavior.
A: An interesting implicit corollary here is that the thing that we think of as prison governance — prison guards and correctional officers — doesn’t sound like a major factor in maintaining order. Where are the correctional officers in all of this?
D: I would say it’s not that they don't matter, it’s that they don’t provide the totality of the governance in a way that many of us would assume. Guards of course regulate the perimeter and are responsible for preventing contraband and things like that from coming in. They will also quell large-scale disruptions — through tear gas, etc. And they will investigate assaults and homicides that happen inside prison. So they do govern in all of these ways.
But it’s clear that they don’t govern in their official capacity when it comes to daily social interactions. And then of course, when there are any disputes associated with the underground economy, those are also outside of the purview of officials. But those are two big, big parts of life in prison. That’s where gangs have the more dominant role in governance.
A: Riots, murders, and assaults in prisons have all been declining since the ’70s, which is really when gangs take hold. How much of that decline do you think is due to improvements in formal governance in prisons and how much is due to gangs?
D: We don’t have enough data to have any smart things to say about that. It’s clear that prison governance has gotten a lot better. For lack of a better term, the silver lining of mass incarceration is that corrections became professionalized and the technology to assist with it improved. Corrections today is a much more advanced field, largely because it’s deployed on a much wider scale.
But it’s also true that there are ways in which gangs stabilize communities. Large-scale disruptions are bad for gang business. Gangs have an incentive to regulate serious acts of violence, and even less-serious acts of violence. But the aggregate effect between the two is difficult to tease out.
A: Let’s get comparative for a moment here, because another thing that stands out in the literature is that the gang situation in California seems very contingent. Many of these groups formed along racial lines in the ’50s and ’60s, and they tend to have particular, often very complicated internal structures. You’ve pointed out how this is a natural consequence of rising prison populations. But then I want to know: Do other places with high prison populations have similar types of gangs? Do those gangs affiliate along similar lines? What about places that are more ethnically homogenous? How much can we say about whether this system arose as a product of particular conditions in California in the 1960s or as a result of more universal pressures?
D: Great question. If you look around the United States, places that have large prison populations — California, Texas, New York, Chicago — all have similar gang-like structures that are broken down along racial lines.
If you look at smaller prison systems — I was recently in the South Carolina Department of Corrections — they have a much smaller prison population, and they don’t have dominant gangs like those in California. Many of the conditions are the same. They’re both places of confinement. You’re forced to go there. You can’t leave. But there are no gangs emerging in the small ones. The social order that exists in the small ones actually looks a lot like the convict code that existed when California had a smaller prison system. They’re also not as racially segregated in those smaller prisons.
So each of these is consistent with a causal claim that big, diverse communities that can’t rely on official governance tend to form gangs. If you look at clan-based societies, they’re socially organized in a very similar way. And I hadn’t discovered this clan literature until I finished writing my book. But clans form when there are not strong and effective state based institutions, and when groups are large enough that they can’t rely on these informal mechanisms. So there’s a similar phenomena arising in certain times and places like we might expect.
A: To what extent do you see convergent evolution across the large prison systems? Are there similar gangs across the system owing to the movement of people between states?
D: It’s mostly not the same gangs. Sometimes it’s the same gang names. But it’s not like a franchise across state lines where the boss sends them the owner's manual.
This is an oversimplification, but a very common form of organization that emerges within gangs is a one-to-five-person group that is in charge on the yard, and they’ll have some collective decision-making. Everybody else follows the leader. But they typically have written rules that they make new prisoners read — and often memorize and recite –- before they can program with the gang. So there’s a pretty standard set of procedures. That’s partly because prisoners in general are subject to inconveniences surrounding sight, smell, and sound. So to the extent that gangs are trying to regulate these externalities, it’s not surprising that they would land on similar rules.
A: What kinds of rules are common across these gangs?
D: There’s a bedtime, after which you can’t be loud. There are taxes on drug sales. In the bigger prison systems there’s regulation of who lives in the community, with a particular eye to excluding specific people, usually sex offenders and former law enforcement officers. The collective responsibility of gangs in places like California means that they don’t want to be responsible for the actions of people who are deemed low status in the prison world. And then there are a lot of rules regulating drug sales, usually with respect to offering, managing, and collecting debt.
A: You argue very compellingly that gangs are fulfilling a kind of public good. I’m not sure this is how gangs conceptualize themselves — well, maybe it is. But these people aren’t exactly public-spirited. So how effective are they at actually providing governance?
D: I would actually say a lot of gangs would say that they’re providing a public good, and even go so far as to say that they’re part of an ideological or political project for their people. Whether that’s true entirely or at all is an empirical question. But they are motivated by calls to politics and to people.
Gangs are quite effective at getting contraband into prisons and regulating it. That’s not good from the prison’s perspective, but from the gang perspective it’s great. And they do regulate violence, but it’s difficult to assess how effective they are at doing so because we don’t have counterfactual cases to look at.
My case has never actually been that gangs reduce violence, in that violence is going down when gangs emerge. The appropriate counterfactual in my mind is not violence in prisons before and after gangs, because the confounding factor is that the population and diversity in prisons both skyrocket. The appropriate counterfactual is whether prisoners in large, diverse prisons today are more or less safe than they would be in prisons with the same population without gangs.
I think in that counterfactual, there would be a lot more violence and chaos until something like gangs emerge — or the prison hires dramatically more guards.
My thought experiment is: There’s something like 31 prisons in California. If we built another 100 prisons, I predict that gangs would become much, much less powerful because prisons would be a lot safer. They would be a lot easier to manage and control. There’d be a lot more effective official governance. And people wouldn’t need to turn to gangs for safety nearly as much.
A: How much do you think gang influence would be reduced by just dividing the same prison population into more, smaller prisons?
D: The prison size is sort of a proxy for information costs and effectiveness of reputations. There’s not one answer. There are multiple mechanisms through which it could affect things. Smaller prisons should make reputation mechanisms better, but of course moving between different facilities during a prison sentence will undermine their ability to do that.
But if you increase the recidivism rate and people are returning to the same prison facility over and over, that should make reputation mechanisms more effective. That's the case that we see in the gay and transgender housing unit in Los Angeles County Jail. Very small facility, incredibly high recidivism rate. People return to prison and they're treated like friends and family, at least according to one study, because there’s a bunch of familiar faces there.
In prisons in England, you’re incarcerated close to home as official policy. When someone arrives, even if it’s their first time in prison, lots of people already know them because they’re from the same neighborhood or housing estate. And because you’ll likely return back to the same neighborhood, it amplifies the importance of maintaining good standing. So we can think through predictions on all of these different prison organization changes.
A: Latin America also has very large prison populations, often with very high gang involvement. How similar is that situation to the U.S.?
D: It’s actually quite different, but that’s because prison officials provide extremely few resources. They don’t administer prisons. Often there aren’t any guards operating within a prison facility, and they provide essentially no social or economic governance within prison. For the most part, Latin American prisons are run as perimeter control lines.
The resources, to the extent that they exist, are brought by friends and family. And there’s either a delegation or a surrendering of governance to powerful inmate leaders. Sometimes that’s gangs, as in the case of Brazil, where gangs are very prominent. But in other places that are less criminally oriented, prisons are run more along civil society lines or as a sort of co-governance between inmate leaders — who are sometimes elected by the other incarcerated people — and prison staff.
A: That’s really interesting. Gangs in America also often have elected leadership. But what you’re talking about isn’t gang leadership, it’s just that the prison population more broadly will elect someone.
D: Well, somebody has to be in charge. Elections are one way of people converging on a candidate. Often it’s not an explicit race, it’s just, “Hey, this person has been around, they seem to have good sense, a steady hand. Let’s ask them to do it.”
It’s the same way we solve problems in a lot of settings. What’s unique is that it takes place in a context we often assume is tightly controlled by the state, when in fact many of these places are pockets of anarchy surrounded by the apparatus of the state.
A: But because criminal gangs in prisons often have control of a really lucrative drug trade, there are a lot of incentives to want to be in charge beyond being recognized as a broadly reasonable guy who would administer a prison well.
D: Without a doubt. It doesn’t select for people with the purest of intentions and incentives. When my first book came out, people would say, “Skarbek thinks gangs are good.” I don’t think that they’re “good,” certainly not in a general sense. When you think about rule of law concerns, we tend to think that powerful institutions should be bound by principles of the rule of law. Gangs of course aren’t. There are few systems of accountability for gangs and other inmate structures.
There’s no due process. There’s no equal treatment. And if a gang decides that you violated some rule and should be killed, you’re not involved in that deliberation. You won’t even know about it until they try to assault you. That’s the challenge in a setting where officials aren’t governing or governing well. There’s also incentive for people who want to profit while governing. But are there instances where those things produce a good outcome, or at least a more stable outcome than others?
A: The impression that I came away with from reading about internal prison governance is: It’s not that gangs are good, it’s that the guards are often willing to abdicate responsibility for the well-being of prisoners at best or are openly abusive at worst. With the gangs, there’s at least some accountability.
D: It’s not a very inspiring picture no matter who is in charge.
A: Gangs also have real informational barriers to governing their members, the people who affiliate with them. This is the problem that they in some sense exist to solve — that you can’t know everyone’s reputation. But at the higher levels, the 300-400 guys in gang leadership need to be able to track an enormous population of people across many different institutions. You can’t just email them. All their communication is monitored.
D: It’s pretty remarkable. But it doesn’t have to be perfect information — it just has to be better than the alternative. And now there are so many cell phones in prisons and that information barrier is falling rapidly, close to zero. Prison directors and wardens that I’ve spoken with say that cell phones are an unstoppable force that undermine all of our efforts to suppress what the gangs are doing.
A: It seems methodologically very difficult to do work on prison gangs and criminal organizations, which you’ve done for your entire career. There’s the challenge of getting raw data and doing ethnographic work in groups that have very strict codes of silence. And then, as an economist, you might want to make causal claims about gang governance structures or factors leading to these structures. Obviously you can't randomly assign prisoners to gangs and see what happens. So how do you handle this in your work?
D: Yeah, that’s a good question and it’s one I’ve been asking for the last 20 years. First, I don’t — and can’t — do causal inference in this space. But there are really rigorous ways of thinking about qualitative causal arguments from the field of comparative politics. That’s part of the reason I moved into political science. Since the ’70s, political scientists have been thinking hard about how we study institutions from a comparative perspective if we don’t have well identified causal identification — or even, for that matter, data. And there’s a lot of ways that we can approach these questions.
A second answer — and this is an argument I make to economists — is that even if we had numbers, that wouldn't capture a lot of the things that we care the most about. We can’t get data on the underground economy, obviously. But we also don’t typically look to numbers to describe social relationships: There’s no scale of racism from 1 to 10; you can’t quantitatively capture social relationships and how they change over time. A lot of what I’m interested in can’t be narrowed down to single dimensions or numbers.
I actually think that qualitative evidence does a better job of more fully and richly describing certain social relationships and institutions. And it’s that rich evidence, both quantitative and qualitative, that allows us to make much richer and detailed claims about causality. So instead of narrowing our measure of some concept to a single number, we have multifaceted, multidimensional, and much broader types of evidence to look to and to point to when making causal claims.
A: A lot of the secondary research you use comes from anthropologists, sociologists, criminologists, etc. When you’re reading qualitative research, what makes something useful for you? There's a very wide range of approaches and ideologies within those fields.
D: For my second book, I read hundreds of studies of prison social order. There’s a winnowing process where you say, “Some of this doesn’t seem very credible, some of it’s off topic,” so you get rid of it. You want to engage with those people that are looking at it in as objective a social science way as possible. There’s no shortage of more activist writing on prisons — much of which I’m a fan of — but you have to find the pieces that are most valuable.
I’m not an ethnographer, and some ethnographers are opposed to comparing across ethnographies. But as a comparative political scientist, I’m all for it. It just requires a detailed reading and trying to figure out whether a piece is a serious piece that’s representative of a broader phenomenon.
A: We’ve talked a bit about public goods provision. A lot of your work is about applying a lens of rational choice theory to criminal activity. There is a niche that these organizations fill. People within them are behaving rationally in response to incentives. I’m curious why this is a lens you find productive in your work, and whether you can also speak to its limitations.
D: I’ve actually tended to emphasize it less lately, and for the same reason that I emphasized it before, which is that criminologists are hesitant about the approach. I wanted to advocate for rational choice in my first book, but then it ended up being more of a distraction than anything, so in my second book, I don’t think I said anything about it. So it really depends on the audience I’m speaking to. I think rational choice and historical analysis based on institutions is the winning combo for these sorts of questions. It grants a lot of autonomy and agency to the people I’m studying. It presumes that if they’re engaged in some ongoing and sustained practice, they’re not idiots for doing so. It takes me as the one who is ignorant of the relevant constraints and costs and assumes they know these things better than I do.
So the goal is that when you see something that seems puzzling, you learn enough about the context in which it’s operating to make sense of it. What are the perceptions, beliefs, and constraints facing the individual actors? What’s the story from their perspective that seems to make sense? They’re not calculators of pleasure and pain. They’re not adding up probabilities or expected values when they make decisions. But in a lower-powered, rational choice way, I think they’re engaged in purpose and strategic behavior within a potentially high-stakes environment.
A: You’re an economist who moved into criminology. What were the dominant theories of gang formation that you were responding to in that field?
D: There were two common explanations. One was an argument that as legal rights changed, prison officials became less oppressive, and that opened up the possibility for gangs to emerge and take back control. John Dilulio is a political scientist who wrote a great book on governing crime. He puts the social order of prisons as caused by the quality of prison leadership and, very often, the specific personalities involved.
And two, maybe most directly, is this sort of “importation theory”: that we have gangs in prison because there are gangs on the outside, and when their members are incarcerated, they import their cultures and values. But neither of those, I felt, explained the empirical record as far as I could see it.
But there’s a larger issue, which is that the sorts of theories that I like aren’t very common in criminology. Criminology doesn’t have the sort of parsimonious style and comparative static prediction models that I like and find intellectually compelling. Some of the criminological theories, in my opinion, don’t have the clear empirical implications that allow you to test across different hypotheses. So I was pushing back against a messier sort of theorizing in the field, rather than a specific theory or author.
A: If you could conduct your magic wand trial — where you’re not constrained by methods, data, or ethics — what’s the experiment you would want to run?
D: The most straightforward — and obviously fantastical — study would be randomizing population size and ethnic fragmentation in prisons. Women’s prisons don’t have gangs. So for example if we locked up 20 times more women, would they then form gangs? That’s basically what my argument would predict. If that was ethical and feasible to do, I’d be curious about. Since it’s not, we did do an agent-based model.
A: There was one sort of natural experiment in California, which was the series of bills known as realignment that aimed to reduce the prison population. Are there any indications that had an impact on gangs in the prison system?
D: I haven’t focused on California in quite a few years now, but my casual observation is that gangs have moved to what they call a good-neighbor policy, where there’s less tension between racial and gang groups. Gang warfare is less common today than it was in the ’90s and early 2000s, and that correlates pretty well with realignment.
The other side of realignment that’s interesting is that those policies only affected prisons. There was a ballooning of the jail populations in California. And again, the sort of anecdotal reports I got from people working and living in prisons is that the gang stuff was becoming a lot more important in the jails. Both of those observations are what I would guess, though with a larger error term.
