Probable Causation

View Original

Bonus Episode 6: David Skarbek

David Skarbek

David Eil talks with David Skarbek about his book, The Puzzle of Prison Order: Why Life Behind Bars Varies Around the World.

Professor Skarbek is an Associate Professor of Political Science at Brown University.

Date: December 15, 2020

A transcript of this episode is available here.


See this content in the original post

TRANSCRIPT OF THIS EPISODE:

David E. [00:00:08] Hello and welcome to Probable Causation, a show about law, economics and crime. I am David Eil, your host for this very special episode. My guest today is Professor David Skarbek of Brown University. David holds a Ph.D. in economics from George Mason University and has published widely in economics, political science and crime journals. He's the author of the newly released book "The Puzzle of Prison Order: Why Life Behind Bars Varies Around the World," available wherever you buy books. David, welcome.

 

David S. [00:00:36] Thank you very much. I'm delighted to be in conversation with you.

 

David E. [00:00:39] Great. So before talking about the book specifically, how did you get involved in this kind of work? It's not you know what? It's what every person with a Ph.D. in economics ends up going into. What led you to this general subject?

 

David S. [00:00:54] Sure. I think that's a fair question. You know, since I was an undergraduate in California, I've been very concerned about the situation of mass incarceration in America and so when I got to grad school, I was sort of hoping to do some work on either, you know, prisons or criminal justice questions more generally. And in one of my classes that I was taking there was on constitutional economics and so we would use these sort of economic models of political constitutions to understand why they're structured the way they are, why they vary across countries, and why some are sort of more effective than others. And for that class, I had to write a paper and I sort of happened to know that one of the prison gangs in Northern California had actually a pretty extensive, fairly elaborate written constitution with sort of, you know, sort of a mission and sections and articles and a system of elections and checks and balances.

 

David S. [00:01:47] And so sort of in grad school for that class, I thought maybe I could see if these economic models of political constitutions would help explain the sort of economics of a criminal constitution. So I sort of just wrote that paper and it raised one question, you know, how to gangs organize internally, but it raised other questions. You know, why did the gangs exist in the first place? What are the consequences of them existing? And so I've just sort of been following that trail of questions that have been raised from paper to paper.

 

David E. [00:02:14] And then when you started this book project, was the idea just to kind of bring those paper ideas together. Did you have a kind of overarching idea in mind?

 

David S. [00:02:23] Well, I mean, I started thinking about this project actually about 12 years ago when I started doing a little bit of preliminary research on Latin American prisons. And I was trying to sort of think about how could you do a sort of comparative analysis of prison social systems and you know, for a variety of reasons, it was sort of too difficult to project. I sort of hadn't figured it out enough then. After my first book, which looked at California, however, I think I'd sort of come up with a better way of thinking through the comparative institutional analysis of prisons. And so this project was was basically to sort of start from the observation that there's a lot of distinctive characteristics of prisons in California, but the informal life of prisons in prisons globally varies tremendously and in a wide number of ways, in a lot of fascinating ways.

 

David S. [00:03:13] And so it was sort of trying to better understand initially the sort of American context and global perspective. And then I sort of realized that there's just there's a huge amount of really interesting variation to be explained when looking at prisons and different prison systems and sort of in a historical and contemporary perspective.

 

David E. [00:03:30] It seems to me that, you know, doing this kind of institutional analysis, especially integrating it into the kinds of institutions, also is typically done on institutions outside of prison you'd take a kind of rational actor view of incarcerated people, which may be at odds with the kind of perceptions that people have of incarcerated people generally. How do you think about kind of the decision making process of the typical person in a prison and how two audiences typically react to your model of them?

 

David S. [00:04:03] I think that actually the rational actor model might apply even more so in prison than in the sort of everyday context. And we learn this a lot from observing how prisoners act, what they spend their time doing and what they say or report back about their interactions with other people. And so prisoners consistently talk about prisons being a very strategic environment, they talk about how they have to think very closely about how their actions are going to be perceived. They think through very clearly, you know, what in economics we think of as signaling models they think about that in the context that they're operating in. And they say that there's significant consequences if they don't act rationally in prison the punishments have come very severely and much more severely than sort of lapses in sort of rational choice for most of us on the outside world.

 

David S. [00:04:51] Now, what's interesting about the prison context is we see a lot of people doing things that maybe at first glance seem irrational or seem sort of unusual. Right. So, you know, for example, you know, lots of prisoners get tattoos on their faces. Right very elaborate, very alarming tattoos on their faces. They join prison gangs that in theory, they're never allowed to leave, even after they join the gang and my sort of argument is that this is an indication of the irrationality of the actors, but that instead we have very rational actors who are facing very difficult constraints in the environment that they're acting. And so their rational response to these difficult and unusual constraints is what sometimes generates what appears to be unusual sort of outcomes and behaviors. So, I mean, the way that that's been perceived, I think varies from sort of intellectual community to community what I think an additional benefit in addition to explanatory power is that it gives prisoners agency. It doesn't sort of diminish them as sort of second rate decision makers. And it recognizes that they don't just passively respond to the context of incarceration, but they have an active role to play in sort of determining the context in which their daily life is going to unfold.

 

David E. [00:06:06] Yeah, I think in the book that was brought particularly to the readers attention by the prisons in South America, where government seems to be left kind of almost entirely to prisoners. And if you had a model where they're just ungovernable or can't make decisions at all, then I guess you would be very surprised by what happens in those prisons.

 

David S. [00:06:27] Yeah. And you know, in sociology, there's this big debate about sort of agency versus structure and, you know, where where do we need to focus our analytical lens is it on these sort of structural factors or is on individual decision making and sort of bringng them in the sort of new institutional institutional analysis stuff, I think has the right balance between recognizing that sort of deep structural factors, you know, prison officials decisions those are important constraints, but prisoners respond to them. And so there's a sort of dance between the influence of both these things, rather than simply saying that it's all all one, some deep structural factor or sort of all the result of atomistic action.

 

David E. [00:07:06] So one of the big factors that you talk about as far as driving differences in institutions and behavior between prisons, the size just the number of people in a prison. Why and how is that important?

 

David S. [00:07:20] Well, you know, it's important because I think that what fundamentally motivates the social organization of prisons, by my argument, is the need for governance. So governance institutions define and enforce property rights, they facilitate social and economic interactions and they aid in the production of collective or public goods. And in some prisons, officials provide the governance, but in many others, prisoners themselves have a demand to produce governance. And how that relates to the size of prisons is that when there's fairly small prison populations, prisoners can rely on very decentralized ways of providing governance, and they do so through sort of reputation based mechanisms. So in small populations, people know other people's social standing pretty well and they value their own reputation. So if prisoners decide to gossip about you, to ostracize you, to shame you, those things are hurtful. They sort of put you lower down on the social hierarchy and it's painful, and it means you're going to have less access to sort of resources, fewer people to sort of have support from.

 

David S. [00:08:27] And so in these small populations where people know reputations really well, prisoners can use these decentralized reputation based mechanisms at low cost they don't require a lot of resources, they don't require a lot of collective action. The sort of in a coordinated way, impose some structure. And so when those decentralized mechanisms work well, as I argue they do in small populations, then that's a sort of natural go to for the sort of informal governance through norms for prisoners to rely on. And I argue basically that in large populations they don't have access to that. And so they need to find alternative ways to provide governance.

 

David E. [00:09:05] You do describe in the smaller prisons that sometimes geography, where a prisoner comes from, becomes an organizing principle. Is that just kind of a part of the decentralized reputation that is incorporated informally? How does it differ from some of the kind of more formal and strict organizations in larger prisons?

 

David S. [00:09:26] Yeah. So I mean, I think in the in the new book, I talk about the prison system in England and Wales and compared to many places in the United States, including California, in England and Wales, they really strive to send prisoners to prisons that are sited close to where those people come from. And they have about five times as many prisons as they do in California. So they're actually able to spread out these prisons throughout the population centers in England and Wales pretty effectively. So as a result, when people go to prison, they go to prison and there are people from their own postcodes, the zip code equivalent in England, there's people from the same housing estates, the same schools that a new arrival to the prison has already gone to.

 

David S. [00:10:10] So in these communities, the importance of reputation is even more important. So not only are these small prisons, but they're small prisons with sort of dense pre and post prison social networks that are going to extend the life of the interaction and therefore increase the importance of maintaining good standing in that community. And to sort of, you know, sort of be a bit more clear what it means is when you show up to prison, you may already have a pretty well established reputation amongst the people who knew you prior to incarceration. While incarcerated rhey may gossip with friends and family who know your friends and family back home in town, and when you leave the prison at some point, it's very likely that you'll return to where you came from and other people who are incarcerated with you will likewise return to that community.

 

David S. [00:10:58] So what you do in prison is sort of broadcast to a bigger audience and is relevant to people you're going to interact with for a longer time. And so in some pretty simple models of economics and repeated interactions, that provides a powerful shadow on a person's present interactions. You have a strong incentive to comply with the norms of the prisoner community because of this ongoing relationship.

 

David E. [00:11:23] Yeah, I think the example you describe of the I think the gay and transgender prison in L.A. seemed to reinforce that idea.

 

David S. [00:11:32] Yeah. And this is a very interesting housing unit in the Los Angeles County jail. It's there to house a community of people who are often subject to high levels of victimization while incarcerated gay and transgender prisoners. And there's a fairly controversial selection mechanism that determines who is housed in the gay and transgender unit and who is housed in the general population. We can talk about some of the sort of controversy surrounding that, but the consequence of this selection effect is that the people who are incarcerated in these particular dormitories, it's not only a small community and there's not only a lot of sort of pre prison relationships that are sort of refound or that exist also within the jail. But there's also a large degree of sort of a sort of homogeneity.

 

David S. [00:12:20] So in sociology, they use this term of social distance, which was more clear before the pandemic, you know, where it obviously has a very common public health meaning, but social distance in the sort of sociology and anthropology literature is looking at to what degree do the community that you interact with share your values, your worldview, your religion, your lived experiences. And to the extent that prisoners in this gay and transgender dorm are coming from relatively more similar past experiences and worldviews, it should be easier for them to agree on what are the acceptable norms, what constitutes a deviation from those norms, and what's the appropriate response to those norms.

 

David S. [00:13:02] And this is sort of most importantly, in contrast to the population in the general housing area, the Los Angeles County Jail, which houses people from an incredibly wide range of walks of life from all over Los Angeles County of all ages, different education backgrounds, different criminal commitments. In that community the sort of social distance, as they would say, is very high, there's a lot of diversity in that community compared to this sort of unique gay and transgender dorm where there's sort of more agreement about what the norms are and how we should live amongst each other.

 

David E. [00:13:37] Yeah I guess it occurs to me that also in in jails, there's a tremendous amount of turnover. You know, people are typically there awaiting trial or awaiting plea more often for a few months before either being released or going to prison and that might also contribute to a particularly unstable environment in jails. My understanding is that even though the typical person who's jailed is accused of less serious crimes than the typical person is in prison, my understanding is that jails are often particularly terrible places to be.

 

David S. [00:14:12] Yeah, that's exactly right. And so the sort of churn of jails works in actually two different ways, I think, or so I'm learning. I mean, so in the general population, yeah, most prisoners are relieved to move from their time in a county jail to their commitment out of state prison. And because state prisons are seen as places that are more stable, that are more orderly there's also more access to a variety of different programs, and amenties of these are probably aren't available at most county jails. The problem with county jails is the turnover is so rapid that there's not a lot of people there who are long enough to sort of entrench some norms and rules of behavior to sort of stabilize the volatility of sort of waves after waves daily and weekly of new people showing up, many of them who have never been incarcerated before.

 

David S. [00:15:00] And the incarceration environment is going to be, you know, a total mystery. They won't be aware of the sort of accepted norms in the jail and subsequently in the prison. So, yeah, county jails that churn, that turnover means that they're they're far more volatile than state prisons. On the other hand, what I found out at the gay and transgender housing unit is that there's an incredibly high level of recidivism. And what that means is that when many prisoners arrive at the gay and transgender unit, even for the first time, they often see sort of friendly faces of people that they knew previously. And of course, when people leave this particular unit, if they're incarcerated subsequently, which unfortunately something like 90% plus of them are within the next three years, they often return to sort of friendly and familiar faces.

 

David S. [00:15:49] So this is a combination of a far smaller community where there's a high degree of turnover, but instead of just creating volatility, it actually creates familiarity in the way that having pre prison social networks in the English context meant that you'd sort of know what you were getting into and people would know what you were about when you got there we're seeing similar dynamics in the gay and transgender unit. So in that instance, the sort of high degree of recidivism but of the same people from very similar communities, I think actually contributes to their ability to sort of maintain some norms of governance at a pretty effective level.

 

David E. [00:16:25] By contrast, you describe in the larger prisons in California as they've become really quite big institutions, they've developed this racial structure where the different racial prison gangs kind of have formal understandings between each other and then is the governance within the racial gangs small enough that it can be done more informally? Or even intra racial relations are formalized to the same extent.

 

David S. [00:16:56] So in California, there's many levels of governance that are happening here within gangs, which, as you note, are sort of based along racial and ethnic lines. There is a combination of the same informal norms that I think would exist in just about any group of people who are sort of living or working together like I said, they also often have written rules and regulations. So some gangs require prisoners to sort of learn, you know, anywhere from 14 to 20 rules about how they're expected to behave on a daily basis. Some of these are simple things like don't throw trash on the tier unless it's being swept. Some of them are more about cultivating a sort of threat of violence in the sense that for some groups, prisoners have to work out one hour every day in order to sort of stay fit and protect their group and other of the rules are associated with illegality.

 

David S. [00:17:45] So if you're selling drugs, a certain percentage of the revenues have to go to a certain gang leader, for example. So within gangs, they have a lot of clear rules that are both sort of informal norms and others that are more written rules, but the racial aspect, what's interesting to me is that that strict racial and ethnic segregation that we see in places like the California prison system aren't replicated in many of these other ones. So in the English prison system, it's not nearly as as ethnically salient, in the gay and transgender dorm not true, in women's prisons in California ethnicity does not play in that crucial life defining characteristic and even in men's prisons before gangs formed. So going back to sort of the late 1950s and 1960s, what we know from people who either worked or lived in California prisons at that time is that racial characteristics were far less important and salient and influential during a period when society in general was full of more overt racists.

 

David S. [00:18:43] And so my argument essentially is that, you know, race is very important in some prison context, but not others. And I think that it becomes more important when prison populations become significantly larger. And the reason why is because in large prison populations, you know, as I've argued previously, you know, you can't know everyone's reputations then prisoners are looking for sort of shorthand ways to know who a prisoner affiliates with and who to hold responsible for a particular person's behavior. So in California today, you know, instead of an individual standing being most important, each person has to affiliate with the gang they operate in, what sort of political economists sometimes call the community responsibility system, which means that everyone in the prison has to affiliate with some gang and within a gang, each prisoner is responsible for each other members actions. So if one individual incurs a drug debt it's not only his responsibility to repay it, but the gang as a whole is held responsible.

 

David S. [00:19:44] So these mutual responsibility systems exert a tremendous amount of in-group pressure to ease or facilitate sort of social and economic activity across gangs. And so for the system to work well, because it's in a society of strangers, not one where we know everyone's reputations, it seems like a part of the racial segregation that exists is so that strangers can look at someone and based on one maybe the color of his skin and two maybe prominently place tattoos have a pretty good sense about which gang is sort of responsible for any particular person, any strangers actions or interactions with you. So that's why I think the race has become far more salient and much larger prison systems.

 

David E. [00:20:28] So in addition to these substantive rules that seem to be, you know, quite formal, written and widely disseminated, are there also kind of written procedures that are used to settle disputes both across gangs and within gangs?

 

David S. [00:20:45] Yeah, there's definitely procedure. So, you know, within a sort of gang or racial ethnic group, if you have a conflict with someone else in your group, you can't just get into a fight with them. You can't independently choose to attack them maybe when they're not looking depending on on the gang in the prison, there's a lot of variation, but it wouldn't be uncommon to require that a person go to the gang leader sometimes called a shot caller and basically say, look, I've got a beef with this guy. You know, we need to hash this out and the gang will find a cell that you can fight in that's low visibility and not likely to be observed by a correctional officer and you can go and sort of duke it out and resolve the problem.

 

David S. [00:21:25] Of course, you can also go to the, you know, the shot caller and ask them to just to sort of adjudicate some dispute, you know, without any sort of physical confrontation. When there's conflicts across gangs, it's typically you know, when it works well, I should say the shot caller from one gang will talk to the shot caller of some other gang and discuss the problem. There's a presentation of evidence to some degree to sort of substantiate what the sort of initiating complaint is and then there's a negotiation about how that's going to be resolved. Now, this negotiation is highly indeterminate and varies substantially depending on the relative threats of violence by both gangs, the negotiation and charisma of different gangs and the history of those two gangs, but again, in a sort of in a not untypical process, the gang that hasn't paid back a drug dealer is going to be given a few options. The gang can collect money from friends and family on the outside and pay it off.

 

David S. [00:22:21] It might be that the person who incurred a drug that to the gang has to work it off for the other gang, maybe assault a correctional officer or another prisoner that the other gang once harmed. And then the complaint gang might require or ask that the gang responsible for incurring this debt assault their own member to a severity that it satisfies the first gang, that the gang is taken it seriously. And, you know, this is done again for a lot of strategic and signaling reasons, which is that a gang is happy to assault their own member, especially if he incurred a debt, for example, that he shouldn't have, but they want a signal to the other gang, one, that we take this issue seriously and we're going to impose costs on ourselves as a consequence of it. But also signal number two is that nobody can use violence against our members except for us so it would create ambiguity if they sent their member over to be assaulted by the rival gang, for example. Right.

 

David S. [00:23:19] That would send a signal to other people in the prison or community that, oh, you know, especially if people don't know the details about what's going on, it would send send the idea that this gang is weak in some degree. So it's not much more formalized than a series of practices, but gang leaders vary in their ability to reduce conflicts to adjudicate them. It's their job to in the sense that other prisoners expect them to do it. If they don't do it well we have examples of prisoners writing to gang leaders at other prisons, asking that particular prison leaders be replaced. And many prisoners view these gang leaders as carrying out a legitimate role that although it causes problems at times on that, has a beneficial effect.

 

David E. [00:24:04] It sounds like generally, if you're just a rank and file person, it's going to be much lower expected costs for you to interact as much as possible within your own gang as opposed to across gangs where any disputes could be arbitrated in this more formal and possibly physically riskier way.

 

David S. [00:24:26] Yeah, I think that's true. And the gangs have a lot of rules that limit interactions across racial and ethnic lines. So there are rules know, again, they vary by prison and gang leader, but there are rules about, you know, whether you can, you know, eat lunch with someone from a different race, at a time where they could share a cigarette with them, whether you can play sports with them, whether you can work out on the same material. There's a lot of restrictions, limiting interactions with other groups that are driven from prisoners perspectives primarily to avoid the threat of sort of large scale disruptions between two large opposing groups.

 

David E. [00:25:00] Just so listeners get an idea. So when we talk about large and small, what's the size of a prison population for this kind of formalized racial governance starts to emerge?

 

David S. [00:25:10] I mean, today the average prison in California holds a little less than 3,500 people at the height of the prison population in California. It was an average of about 5,000 for a prison facility. There's a big range. So the largest prison still holds about 5,000, the smallest holds only about 2,000. These range from the lowest to highest security. And these sort of gang based governance or entrenchment is in operation in all of the prisons. If we look at the English prisons, their typical prison holds somewhere between 500 and 700 people. And so, you know, we're talking about sort of a fifth of the size of the typical California prison.

 

David S. [00:25:53] There are many people in prisons in England that hold fewer than 250 or 300 people. And the largest prison in England and Wales is still smaller than the smallest prison in California. So that's a little sense about how these things vary. I don't think there's a very clear threshold like 843 people and then it's a large, you know, rather than a small one. I think it's a continuum in that for a variety of factors, the size of the prison population is a decent proxy for the ability to rely on reputation based mechanisms, but when you're getting to communities of thousands of people, no one can know the reputation of a sufficient number of people in that community for it to have any teeth for to actually deter people from violating what would hopefully be accepted norms of behavior.

 

David E. [00:26:42] And is it just nimbyism and economies of scale that keeps the United States from building more and smaller prisons? Or is there some other cultural reason?

 

David S. [00:26:51] I think that's probably a big part of it. Some communities welcome prisons as sources for jobs. Obviously, some oppose them concerned about property values. The main argument that I hear is that it is economies of scale. And, you know, as your listeners I'm sure know, the idea is that, you know, if you have one large prison, then you only have to pay for one cafeteria to service 3,000 people rather than having four or five prisons and therefore four or five cafeterias. And you know, the basic rule of cost benefit analysis, you know, as I remember it, is that you have to count all of the costs and all of the benefits and this sort of cost benefit analysis that justifies very large prisons, I think ignores the costs that are associated with large prisons, namely that it encourages formation of gangs.

 

David S. [00:27:37] Gangs undermine rehabilitation. You know, they're not an ideal way that we want to run prisons. And, you know, they cause a lot of harm both in prison and on the outside of prison. And so I think that the sort of result of these large prisons in the U.S. has been from a either a systematic, you know, implicitly or explicitly ignoring of the costs that arise on the prison social order, the informal aspects, if you will, from really large prisons. And I think if you incorporated those costs, the sort of economies of scale, if not evaporate entirely, would indicate the optimum size of prison should be far smaller than 5,000 or 3,000 prisoners.

 

David E. [00:28:13] So you also discuss, you know, a lot of different kinds of extralegal governance among prisoners and then also a couple of organizations that you don't really credit as rising to the level of extralegal governing institutions, but are in some sense a collective of prisoners that have some group goals and one is this group called the Raiders at the Andersonville Prison and then also Muslims in a British jail. So what's kind of like the threshold for an extralegal governing institution? What are the characteristics we should look for in distinguishing that from these kinds like the Raiders?

 

David S. [00:28:54] Yeah, that's a good question. You know, I guess it really comes down to asking not only are these groups extralegal, but do they govern? Do they enforce property rights? Do they facilitate social and economic interactions, and do they provide collective and public goods? You know, there are a lot of groups in prisons that don't provide governance. There are some, like this group, the Raiders, that do the opposite of governance. They're violating property rights, they're hurting people, they're undermining social order. So they're not a governance institution simply in the fact that, you know, they're not doing anything to govern, quite the opposite, actually. In the context of the English system, there's sort of there is a there is a concern in the Daily Mail and other places that there is sort of rise in radical Islam in many English prisons. And so, you know, I sort of look at this particular case that's unrepresentative of the English prison systems where it seemed like sort of Muslim prisoners had banded together to some degree, had a lot of influence and if anywhere there was sort of forced conversions to Islam it was probably happening in this prison.

 

David S. [00:29:58] And, you know, I guess in my discussion there, as I try to be a bit unsure, I hope I communicate some lack of we don't have a lot of good evidence about that particular prison. So I hope I'm not sort of overly bold about how I describe them, but the structures that we see gangs investing in to produce governance, such as written rules and organizations, clear membership, initiation rituals, we don't see those really existing in the same way in the English context. Now it's clear that these prisoners are affiliating with each other it's based on, at least nominally on a religious community, and it's just unclear if they're then also having a second step where they're actively producing governance institutions to regulate the sort of everyday life of prisoners there. So I think that's sort of a case of we don't have quite enough information to know yet. It's not a very it's not along strong evidence in favor, but the absence of evidence is not evidence of absence. So at this point, I'd sort of say that both those groups are they're clearly organized. They're clearly having an influence, but it's just not clear that their influence is specifically in producing governance or governance institutions.

 

David E. [00:31:10] So the big failure of the Andersonville Prison is that extralegal governing institutions don't really emerge among the prisoners, even though they're also not really provided by the prison officials. And so you say that this is perhaps even more surprising given that at Andersonville, all the prisoners there are union soldiers and maybe you might expect soldiers to crave more order or be more obedient to order, but maybe it could be also that their leaders are not there the officers I understand were all taken to a different place, and so there's just no leadership to kind of organize, even if the people there would have been amenable to it. Is that a fair kind of thing to look for as not only obedience but also leadership and kind of more generally to the cultural norms of the world outside play themselves out within a prison itself in kind of establishing governing norms?

 

David S. [00:32:17] Yeah, I think that those are great questions. The Andersonville case, just to maybe give a little context, is is a fascinating what's one of the first ones that I got interested in. It's a prisoner of war camp built during the American Civil War during the last sort of 14 months of the war it was  in operation. It was essentially just a large field in Anderson, Georgia, there was a 15 foot walls and there were a few guards that prevented prisoners from escaping. The prison became notorious after the war because it had an incredibly high fatality rate, about a third of prisoners died who entered there. It was also tremendously overcrowded. At the height, there was more than 30,000 prisoners in what was essentially like a dirt field, and prison officials provided almost no resources.

 

David S. [00:33:01] They provided a very little bit of food on a daily or sort of every two or three day basis. There were no amenities or services. They didn't have a physical presence inside the stockade, so they weren't governing the everyday life of prisoners and while in Latin America, we often see prisons facing those same constraints, but Latin American prisoners can rely on visits from friends on the outside who have access to bring in economic resources in prison officials in Latin America don't stop prisoners from starting shops, restaurants, businesses within the prison. And so in a sort of strange way, these Latin American prisons are able to engage in economic activity because they have access to the outside world.

 

David S. [00:33:42] In Andersonville, prisoners were again neglected like Latin American prisoners, but they didn't have access to people or economic resources on the outside world is just a dirt field. And so as a result, I argue there that there's not a lot of extralegal governance, there's not a lot of crime because there's very little steal from other people. They were so malnourished that there wasn't a lot of violent crime. So you don't need to invest in institutions to prevent crime and because of the lack of, you know, basically total lack of economic resources, you don't need to invest in economic governance institutions to facilitate conflicts that might arise. So sort of there's a lack of governance because there is simply nothing to govern, no activity to govern.

 

David S. [00:34:21] You raise the important question, which is maybe there's a confounding factor, which is that because these people were incarcerated as soldiers rather than for committing sort of a criminal offense, maybe they're accustomed to organization only when there's a sort of strong and clear leader there and I think that's a good alternative hypothesis for the outcome. What I do is when looking through the sort of record of prisoners at Andersonville, I tried to see if they were examples of governance that like proto leadership, like people who thought maybe it would be good if we organized and it turns out that there were examples of that in some limited realm. So there were people who organized religious activities, for example, and in the face of these sort of roving prisoners who started to prey on the prison more generally there were people who started to organize sort of groups of prisoners to sort of fight back against these Raiders.

 

David S. [00:35:15] And so, you know, leadership was, you know, maybe they didn't have the leaders that they were accustomed to in a sort of military setting, although it's unclear sort of how well-trained they were as soldiers, but, you know, there were people who were willing to take leadership roles who were sort of undermined or sort of unsuccessful for a variety of ways. So I think in general, like leadership is something that matters. I think we find it difficult, just like we do in thinking about entrepreneurship and economic models, to think about how to conceptualize leadership, but in each of the examples that I looked at, you know, there are clear examples of people who are stepping up to perform a leadership role to carry out the tasks of extralegal governance. And there's an audience that's sort of praising or undermining or attesting to the effectiveness or legitimacy or not in most of these, but I think those sorts of things are definitely very difficult to empirically document, I guess, in the sorts of source materials that I have available.

 

David E. [00:36:10] So I think as I hope listeners already appreciate, you've got so much rich ethnographic detail about each one of these settings, which makes for both a thrilling reading experience and also, you know, some really substantive answers to some of these kinds of questions like I raised with alternative explanations and so forth, but it's it's very different from the kinds of arguments that evidence that economists are usually used to reading, especially in this era of big and bigger data and a lot of focus on identification and so forth. So I think that definitely does your methods bring a lot of advantages to the problem, but is there a risk that you kind of tailor your theory around, you know, a few data points? And how do you kind of think about managing that risk?

 

David S. [00:37:00] Yeah, I mean, that's sort of the most important question in a lot of ways. Yeah. So in drawing on the sort of comparative institutional analysis, I mean, yeah, it's less causal inference, it's more focusing on the causal mechanisms about what gives rise to institutions, what sustains them, how and how well do they work and why do they change. And so the evidence that I use to sort of describe those mechanisms, it's not necessarily it's not usually quantitative, and it's less about sort of a clever identification strategy and more about understanding how people are sort of interacting strategically to generate some sort of stable equilibrium outcome. And what that means is that the qualitative evidence that I use tends to be much more multidimensional and multifaceted than the sort of thin data that's used in, say, studies of the relationship between police and crime.

 

David S. [00:37:52] We can get numbers on the number of police, the number of crimes we think they capture pretty well the underlying things that we're concerned about. But when you start thinking about institutions, they can be formal and informal, they can be decentralized, they can be centralized, they can be fragile, they can be robust, they can be antifragile, they can be personal based, they can be corporate based. So there's a lot of different characteristics that describe in any particular institution, I think really calls for more sort of rich qualitative evidence to sort of accurately describe what's going on and so that's why I actually think it's really valuable to engage sort of economic theories engaged with the rich qualitative ethnographic work of prison ethnographers. And so that's actually what I think is most fun about this project is it's bringing a set of theoretical hypotheses into a community that in general is pretty skeptical of sort of economic theory as an economic way of thinking.

 

David S. [00:38:46] And I think you're right that there's in dealing with a few cases instead of hundreds of cases, is there's a concern that the theory is being developed and tested by the same data and if that's the case, then if you're quote explaining those cases, you're not really explaining them. And so the way that I thought about this is actually that, you know, this theory of small and large community is decentralized, centralized community it's actually part of a much broader theoretical framework that is totally outside of the carceral context. So it's developed by people like Avner Greif, Elinor Ostrom, in studying either common folk resources, economic history and the medieval period and trade that took place, self-enforcing exchange that took place. And so in some ways the study of the prison is actually the out of sample test for me of theories of institutions and institutional change developed in a wildly different context and so in applying it to the prison context, then the next sort of concern for the qualitative researcher is am I cherry picking cases?

 

David S. [00:39:49] And if I do, I'm picking cases whose dependent variables align with my theoretical predictions. That's just a form of confirmation bias and bad social science. So what I do for the book is I try to select cases based on variation in the explanatory variables of this theory so not looking at first whether prisons have centralized gangs or not, but looking at variation in the size of the prison, the dense social networks, the social distance, how much officials govern or not and in selecting those cases, the goal at least, is that it's not confirmation bias because you're selecting cases first on what the theory predicts and then going to see sort of what the outcomes are. So that's sort of, I guess, you know, sort of in thinking about how to how to weigh the trade offs between a large number of cases within data versus a few number of cases with a very thick and rich data that latter that rich, thick qualitative data that the ethnographers provide, it allows us to substantiate and to describe that sort of multifaceted, multidimensional type of institution that I'm sort of at heart primarily focused on.

 

David S. [00:40:58] Maybe the last thing I'll say is also that, you know, this project is in part intended to try to convince prison ethnographers that comparative analysis is valuable. And the vast majority of prison ethnographies are single site or several sites in the same prison system. And I think that they're incredibly valuable in their own right. They provide a lot of important, valuable evidence about prison, social order, governance and a range of other things, but my argument in the book is also a methodological one, which is that we can use individual single site ethnographic studies as our own data in a broader comparative analysis. So in some ways I'm trying to push back against the prison ethnographer's who whose goal as in some ways is is to sort of overtly describe their particular cases. They focus on things like what is the personality of this particular guard and how does that affect the day to day? That may be an accurate finding, but if that guard is only in that prison and not in any others, then it may not generate any external validity or any generalizability.

 

David S. [00:42:04] So my goal is trying to sort of navigate between sort of rich qualitative evidence that I need to accurately describe institutions and institutional change, but also to move away from the level of detail here that prevents me from increasing my cross case explanatory variables. And so I'm trying to bring in the theory, get a fairly large number of cases and try to make that trade off about whether this is a theory that can explain present social order in a wide range of cases or whether it's limited to sort of, you know, like it's describing simply the ones that are in the book. And so I guess I'll leave it to readers to decide if I've been successful in doing that or not.

 

David E. [00:42:44] So last question, what's next for you and this part of research?

 

David S. [00:42:48] Well, I guess I have two different projects at the moment. One is that I'm working with a colleague, Danilo Freire, on survey experiments of vigilantism in Brazil. So there's a widespread practice of extrajudicial violence, and we're going to use sort of survey experiments to try to understand what sorts of situations do citizens think is a more legitimate reason to engage in extrajudicial violence, as well as trying to understand, again, through the survey experiments what the underlying mechanism is. Is it because they think the courts will take too slow to bring justice? Do they think that they won't be punished enough? Do they think that the police are corrupt? And then finally, we have a sort of information intervention to try to see if we can affect perceptions of legitimacy of lynching in Brazil. So we've got a sort of series of experiments on that and then I've been working on a project about trying to think about the problem of policing in America today. And so I've got a paper on de-bundling the police trying to think about how we can have sort of, you know, sort of more crime control, but also more justice at the same time.

 

David E. [00:43:58] Great. Thank you so much. My guest has been David Skarbek. His book is "The Puzzle of Prison Order: Why Life Behind Bars Varies Around the World." David, thank you so much. Thank you very much.

 

David E. [00:44:13] You can find links to the research we discussed today on our website probablecausation.com. You can also subscribe to the show there or wherever you get your podcasts to make sure you don't miss a single episode. Big thanks to Emergent Ventures for supporting the show and thanks to our Patreon subscribers the show's listener supported. So if you enjoy the podcast, please consider contributing via Patreon. You can find a link on our website. Our sound engineer is Caroline Hockenberry with production assistance from Elizabeth Pancotti. Our music is by Werner and our logo is designed by Carrie Throckmorton. Thanks for listening.