Self-governance in San Pedro prison.

• Author: Skarbek, David B.
• Position: Report

History provides many examples of self-governing communities that maintained extensive economic exchange and order despite the absence of a government that enforced the law or protected property rights. In such cases, individuals engaging in economic activity with different groups or with others in a single group must develop self-enforcing mechanisms that will induce cooperation, such as abiding by contracts. Individuals facilitate cooperation by various means, including reputation, signaling, and commitment mechanisms, which allow self-enforcing exchange in a diversity of situations. For example, Edward Stringham (2003) identifies the self-enforcing mechanisms that facilitated the rise of seventeenth-century stock exchanges in Amsterdam. Exchange occurred among medieval traders with the assistance of institutions such as the law merchant (Greif 1989; Milgrom, North, and Weingast 1990). Peter Leeson (2007b, 2008) identifies the signaling and commitment mechanisms that facilitated exchange in precolonial African trade. Emily Schaeffer (2008) examines a modern-day manifestation of these types of self-enforcing reputation mechanisms in the case of Hawala traders, who act as intermediaries in international financial transactions. For a contemporary stateless society, Benjamin Powell, Ryan Ford, and Alex Nowrasteh (2008) find that Somalia's relative economic performance has improved during its period of statelessness. (See Stringham 2007 for an excellent collection of related essays on anarchy and the law.) These studies demonstrate the remarkable variety of situations in which people benefit from mutually beneficial exchange and establish order without the assistance of the state.

In these studies, no particular bias exists that causes the people involved to be relatively uncooperative, so it is unclear how robust the outcomes are compared to those for less-cooperative individuals. (1) To understand better the role of the agents' type in self-enforcing exchanges, I examine here how inmates of the San Pedro Prison in La Paz, Bolivia, where inmates must govern themselves, overcome particularly serious impediments to the establishment of order. Cooperation among prison inmates is especially unlikely because criminals have high discount rates, lack the ability to exclude noncooperators from their community, and cannot migrate away from predatory groups. Nonetheless, San Pedro Prison has not deteriorated into a predatory environment in which a single group abuses others. On the contrary, inmates have secure, long-term property rights in their housing and other valuable resources and engage in extensive economic exchange; and outsiders voluntarily associate with and even live among the prisoners.

Past research on self-governance in a prison environment examined the Andersonville prison camp that operated in Georgia during the War Between the States and concluded that such an institutional setting was likely to produce a dominant predatory group (Hogarty [1972] 2006). Opening in 1864, Andersonville quickly filled with captured Union soldiers. The prison received more than forty thousand inmates, 40 percent of whom died (Costa and Kahn 2007). The survivors suffered at starvation levels of subsistence and were plagued by sickness and disease, and many endured physical abuse by other inmates. Bands of "Raiders" brutalized, pilfered, and abused newly arriving prisoners. Thomas Hogarty ([1972] 2006) cites MacKinlay Kantor's (1955) depiction of the Raiders' robbery of these "fresh fish." (2) The leader of the Raiders, John Sarsfield,

shouldered forward and wrenched the [blanket] roll away from the nearest prisoner. The man hallooed, Sarsfield knocked him flat, the balance of the fresh fish leaped toward Sarsfield, Sarsfield's Raiders swatted, stabbed, kicked. This fight was over in less than a minute. Six of the [new prisoners] lay on the ground and the rest had fallen back into the watching throng--several others shy of their blanket rolls, as was the first man. All of the new-come prisoners were bleeding, two were unconscious. Sarsfield's Raiders were the richer by eleven blanket rolls filled with combs, socks, extra shoes, Bibles (these could be bartered), gilt melaineotypes, housewives, knives, eating utensils, and name-it-if-you-like. (Hogarty [1972] 2006, 107)

Only after another group of prisoners, known as the Regulators, grouped together with the assistance of the prison guards were the Raiders overthrown and order restored.

Hogarty argues that the events at Andersonville show that situations of self-governance will necessarily be chaotic. In particular, he finds evidence confirming the hypothesis that self-governing communities will give rise to domination by a group of people--the most "criminal" ones--over other, productive members of the community, and the dominators will attack the others to benefit themselves. The dominant group will consist of the most criminal individuals because, according to Hogarty, they will "'feel more at home' than they did in the society of rules" ([1972] 2006, 108, 100). (3)

In general, prisons differ from a situation approximating anarchy in several important ways that make order more difficult to achieve. First, as Virgil Storr notes, prisoners have few opportunities for productive enterprise (2006, 120). (4) Some trading occurred in Andersonville, but the extent of the market was greatly restricted. With few options for productive enterprise, it is more likely that people will come into conflict, biasing prisons toward more violence and less order. Second, prisoners have few resources with which to protect themselves from roving bandits in the facility. Inmates at Andersonville, for example, lived in tents in an open field. Inmates in typical prisons in the United States lack the materials and authority to construct secure buildings to protect their property and lives. Third, exit from prison is impossible. As Storr (2006) argues, freedom of exit is an important mechanism for checking the power of potential predators in a community (see, for example, Nozick 1974), and the inability to exit makes violence more likely than in an ideal situation of anarchy. Fourth, the Andersonville prison was necessarily temporary in nature: the imprisoned soldiers expected to be released at the war's conclusion. They believed that the Union would participate in a prisoner-exchange program and free them from Andersonville very soon (Marvel 1994, 92, 157). Their short time horizon discouraged capital accumulation and reduced the potential gains from repeated dealings. Most prisons prevent inmates from investing resources productively and thereby prohibit an important method of keeping order in anarchy. (5) Thus, prisons represent a particularly difficult situation for self-governing mechanisms, relative to an ideal state of anarchy, where production and exchange are possible, exit may occur, and group selection processes can sort members. (6)

In this article, I examine successful self-governance in a Bolivian prison, which, because the facility differs from the typical prison, provides insights into the mechanisms that facilitate order. Inmates effectively govern themselves because of two institutional features of this unique prison that move it closer to a situation of anarchy than most prisons. First, inmates are allowed to engage in economic exchange within the prison--for example, by opening restaurants, offering carpentry services, and operating commissaries. In addition, prison guards allow nonprisoners to enter the facility, so economic exchange with people outside the prison walls creates a greater division of labor and improved economic opportunities. The freedom to exchange raises the cost of engaging in conflict and provides resources to protect an inmate's person and property.

Second, secure property rights and well-established markets (including for inmate-owned cells) exist in the prison. These markets, which inmates expect to persist into the future, provide incentives to invest resources in productive ways. Because the inmates are the residual claimants, owners of prison real estate have incentives to create and enforce rules to protect their property's value. These rules include homeowners' associations and committees to adjudicate disputes and resolve conflict. The primary factors that facilitate establishment of a self-governing community are an increasing division of labor, economic exchange, residual claimants with secure property rights, and well-established markets that people expect to persist. I find in this case that even under the unfavorable conditions that a prison presents, such as a biased agent type and immobility, self-governing communities can accomplish order.

Obtaining accurate information about the activities in an inmate-governed foreign prison presents a challenge because no official documents report on the facility. Three types of firsthand accounts of San Pedro Prison, from diverse perspectives, provide evidence. First, an inmate incarcerated for almost five years for smuggling cocaine recounts in detail his experience in San Pedro Prison (Young and McFadden 2003). He explains the prison's organization, the inmate economy, formal rules created by the inmates, and the social norms. (7) Second, official reports prepared by governments and nongovernmental organizations--including reports by the U.S. Department of State, the National Lawyers Guild, and the Inter-American Commission on Human Rights--provide evidence.

Finally, a unique source of evidence consists of the reports of nonincarcerated people who have taken inmate-led tours of the prison. Prison administrators do not officially approve of these tours, but for a small bribe to the guards, tourists can enter the prison. According to one journalist, the tour costs $35 and lasts about two hours, and the inmates give tours to approximately fifty people a day (Baker 2009). For an extra fee...