Stationary Bandits Research Articles

Monopoly of Taxation Without a Monopoly of Violence: The Weak State’s Trade-Offs From Taxation

Abstract This study presents a new economic perspective on state-building based on a case study in the Democratic Republic of the Congo’s hinterland. We explore the implications for the state of considering rebels as stationary bandits. When the state, through a military operation, made it impossible for rebels to levy taxes, it inadvertently encouraged them to plunder the assets of the very citizens they previously preferred to tax. When it negotiated with rebels instead, this effect was absent, but negotiating compromised the state’s legitimacy and prompted the emergence of new rebels. The findings suggest that attempting to increase taxation by a weak state in the hinterland could come at the expense of safety in the medium term and of the integrity of the state in the long term.

Fragile assets: Street gangs and the extortion business

This study examines how El Salvadoran street gangs make assets fragile and, hence, amenable to extortion in the territories that they control. Starting from prior Foucauldian inspired accounting research and the literature on stationary bandits, we propose that social actors and their assets are simultaneously enmeshed in relations of government and relations of sovereign power. Using longer-term participant observation data as well as interviews and archival data from El Salvador, the study shows how people and assets are placed into sovereign power relationships, thereby helping to make assets (and people) fragile and thus facilitate extortion. The study also suggests that it is the Salvadoran state, through its taxation practices and its failure to nurture marginal territories, that creates the conditions of possibility for the extortion business.

An empirical evaluation of active inference in multi-armed bandits

A key feature of sequential decision making under uncertainty is a need to balance between exploiting—choosing the best action according to the current knowledge, and exploring—obtaining information about values of other actions. The multi-armed bandit problem, a classical task that captures this trade-off, served as a vehicle in machine learning for developing bandit algorithms that proved to be useful in numerous industrial applications. The active inference framework, an approach to sequential decision making recently developed in neuroscience for understanding human and animal behaviour, is distinguished by its sophisticated strategy for resolving the exploration–exploitation trade-off. This makes active inference an exciting alternative to already established bandit algorithms. Here we derive an efficient and scalable approximate active inference algorithm and compare it to two state-of-the-art bandit algorithms: Bayesian upper confidence bound and optimistic Thompson sampling. This comparison is done on two types of bandit problems: a stationary and a dynamic switching bandit. Our empirical evaluation shows that the active inference algorithm does not produce efficient long-term behaviour in stationary bandits. However, in the more challenging switching bandit problem active inference performs substantially better than the two state-of-the-art bandit algorithms. The results open exciting venues for further research in theoretical and applied machine learning, as well as lend additional credibility to active inference as a general framework for studying human and animal behaviour.

On the Ends of the State: Stationary Bandits and the Time Horizon in Eastern Congo

We show that armed actors refrain from using their power to arbitrarily steal from an economy if, and only if, the armed actors' property rights over stealing from that economy are secure. By 2009, armed actors taxed, administered, and protected various villages in Democratic Republic of the Congo. We exploit the timing and targeting of an international military operation that permanently made taxing these villages impossible. Following the operation, these armed actors turned to violently expropriating the same villages. The findings suggest that the security of property rights over stealing, hence the stealing horizon, can sustain, or destroy, economic growth. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

On the Ends of the State: Stationary Bandits and the Time Horizon in Eastern Congo

On the Ends of the State: Stationary Bandits and the Time Horizon in Eastern Congo

On the Origins of the State: Stationary Bandits and Taxation in Eastern Congo

A positive demand shock for coltan, a mineral whose bulky output cannot be concealed, leads armed actors to create illicit customs and provide protection at coltan mines, where they settle as “stat...

Two bandits or more? The case of Viking Age England

The Olsonian distinction between roving and stationary bandits outlines the rationale behind the transition from anarchy to the emergence of the predatory state. This two-bandit model may, however, be expanded to include more bandit types. In the case of Viking Age England, local English kings were unable to monopolize violence and defend their realms against competing Viking raiders. As the Vikings’ time horizon grew, so did the accumulated value of more formal taxation, and bandit types evolved in four steps. The first step is the Olsonian roving bandit, who executed Viking hit-and-run attacks and plunders during the second half of the tenth century. The second step is the gafol bandit; gafol is payment for leaving, paid to, among others, Swein Forkbeard. The third step is the heregeld bandit; heregeld is a tax to support an army for hire; most notably Thorkell the Tall’s. The fourth step is the Olsonian stationary bandit, i.e. the strongest military leader among the Vikings, Cnut the Great, settled down as the new king. Overall, the Olsonian two-bandit model can be expanded to a four-bandit staircase model, in which the new gafol and heregeld bandit types explain the steps from anarchy and short-run raiding to long-run formal taxation in a predatory state.

Political Institutions, Resources, and War: Theory and Evidence from Ancient Rome

How does the size of the governing coalition affect the amount and type of violence in society? This paper combines models of the selectorate with stationary bandits to predict that larger coalitions substitute away from fighting for private goods (such as plunder) towards fighting for public goods (such as defense), but does not necessarily reduce the overall scale of fighting. To test this idea, I empirically investigate how Rome's transition from Republic to Empire affected military patterns. I find that the Republic engaged in more battles overall and that Republican battles had more of a public good component. Overall, this paper furthers our empirical knowledge about the ancient world and brings the data to bear on contemporary debates about the causes of peace and war.

Calculating Bandits: Quasi-Corporate Governance and Institutional Selection in Autocracies

This paper applies the logic of economic calculation to the actions of autocrats. We model autocrats as stationary bandits who use profit-and-loss calculations to select institutions that maximize their extraction rents. We find in many cases autocrats achieve rent maximization through creating and protecting private property rights. This in turn yields high levels of production, with expropriation kept low enough to incentivize continued high production. Importantly, while this leads to increasing quantities of available goods and services over time, it does not lead to true development; that is, the coordination of consumer demand with producer supply through directing resources to their highest-valued uses. We apply our model to the authoritarian governments of Singapore and the United Arab Emirates, showing how they function as quasi-corporate governance organizations in the business of maximizing appropriable rents.

On the Origin of States: Stationary Bandits and Taxation in Eastern Congo

When do states arise? When do they fail to arise? This question has generated scholarship across all social sciences. A dominant view is that states first arise when violent actors impose a “monopoly of violence” in order to extract taxes (Carneiro, 1970, Tilly, 1985). A fundamental fact affects all existing studies: systematic data are the creation of states. There is therefore no statistical evidence on the causes state formation. As a foundation for this study, I organized the collection of village-level panel data on violent actors, managing teams of surveyors, village elders, and households in 380 war-torn areas of DRC. I introduce optimal taxation theory to the decision of violent actors to establish local “monopolies of violence”. The value of such decision hinges on their ability to tax the local population. A sharp rise in the global demand for coltan, a bulky commodity used in the electronics industry, leads violent actors to impose “monopolies of violence” and taxation systems in coltan sites, which persist years after demand collapses. A similar rise in the demand for gold, easier to conceal and more difficult to tax, does not. My findings support the view that the expected revenue from taxation, determined in particular by tax base elasticity, can explain the first stages of state formation, and that temporary changes in the economic environment have persistent effects on local state capacity.

Field experiments on irrigation dilemmas

It is often assumed that irrigation systems require a central authority to solve coordination problems due to the asymmetry in position and influence between those located at the head-end of a system and those located at the tail-end. However, many examples of complex irrigation systems exist that are self-organized without central coordination. Field experiments on asymmetric commons dilemmas are performed with villagers in rural Colombia and Thailand. Our experiments show that there is a dynamic interaction between equality in the use of the common resource, and the level of the contributions to the creation of a common resource. Inequality in the distribution of benefits in one round triggers lower levels of group contributions, reducing efficiency and triggering even more inequality in contributions and distribution of the resource among players.The upstream players act as “stationary bandits”. They take more than an equal share of the common resource, but leave sufficient resources for the downstream players to stimulate them to continue their contributions to the public infrastructure.After 10 rounds, players can vote on one of three allocation rules: equal quota, random and rotating access to appropriation of the resource. The rotating access is most often elected. The resource dynamics in the second part of the experiment depend on the rule elected. With the quota rule, the stationary bandit metaphor is less relevant since taking equal shares of the resource is enforced. With the rotation access rule, the players act strategically on the rotating position. They invest more when having the first access to the resource compared to less favorable access. And when they have first access they extract the main part of the common resource. The rotation rule led to a reduction of the performance of the groups. With the random access rule there is no such strategic investment behavior and participants remain investing equal and similar levels as in the first 10 rounds.The experiments show that a necessary condition of irrigation systems to self-organize is the development of norms to allocate fair shares of the water in order to recruit sufficient labor to construct and maintain the physical infrastructure. The different allocation rules do not increase efficiency, but they did increase equality of the earnings.

Head-enders as stationary bandits in asymmetric commons: Comparing irrigation experiments in the laboratory and the field

The emergence of large-scale irrigation systems has puzzled generations of social scientists, since they are particularly vulnerable to selfish rational actors who might exploit inherent asymmetries in the system (e.g. simply being the head-ender) or who might free ride on the provision of public infrastructure. As part of two related research projects that focus on how subtle social and environmental contextual variables affect the evolution and performance of institutional rules, several sets of experiments have been performed in laboratory settings at Arizona State University and in field settings in rural villages in Thailand and Colombia. In these experiments, participants make both a decision about how much to invest in public infrastructure and how much to extract from the resources generated by that public infrastructure. With both studies we find that head-enders act as stationary bandits. They do take unequal shares of the common-pool resource but if their share is very large relative to downstream participants' shares, the latter will revolt. Therefore for groups to be successful, head-enders must restrain themselves in their use of their privileged access to the common-pool resource. The comparative approach shows that this result is robust across different social and ecological contexts.

Stationary Bandits. Lessons from the Practice of Research from Sicily

Stationary Bandits. Lessons from the Practice of Research from Sicily

Cooperation, Games, and Ecological Feedback: Some Insights from Bali

Cooperation, Games, and Ecological Feedback: Some Insights from Bali

Nietzschean Development Failures

Government policies, and substantial external resources and technical assistanceprovided over the course of decades, have neither eliminated broad poverty nor resulted inequitable income distributions in the poorer countries of the world. This paper explains thedevelopment failures with reference to Nietzschean behavior where the strong act withoutethical restraint and the rule of law does not protect the weak. While Nietzschean behaviorviolates principles of social justice, there are also inefficiencies. The Nietzschean strong whorule have no incentive to adopt efficiency-enhancing policies. Efficiency in a Nietzscheansociety is also greater, the higher the leisure preference of the weak and the less the weakare capable of producing. Labor productivity is low because the weak do not consistentlywork. These are the outcomes when the strong behave as roving bandits. When the strongbehave as stationary bandits, efficiency is enhanced but income distribution can be expectedto remain unequal. The Nietzschean perspective on development failure is compared withalternative explanations for the sustained plight of the poor in poor countries and the unequaldistributions of income and wealth.

Theft by Bandits and Taxation by Kings: A Critique of Mancur Olson on State-Formation

In Power and Prosperity (2000), Mancur Olson argued that bandits who permanently monopolise theft in a given location would limit their own requisitions, protect their victims from theft carried out by others and possibly even provide public goods for their victims. Such stationary and monopolist bandits would effectively transform themselves into autocratic rulers of pristine states. I argue that Olson's ideas do not constitute a satisfactory theory of stateformation for two main reasons. First, kleptocracy, which is a condition of competitive theft, represents an equilibrium according to Olson. So there exists no reason to believe that theft will become monopolised. Second, in Olson's theory, if a monopolist bandit exists, then a monopolist on violence presumably also exists. So having a ‘monopolist on violence’ seems to constitute both ‘cause’ and ‘effect’ in his account. However, if properly revised, his basic insights might nevertheless furnish an important hypothesis from which a more satisfactory theory of state-formation could be constructed.

THE TWO SIDES OF THE FORCE: ECONOMIC ANALYSIS OF PRODUCTION AND APPROPRIATION

There are two ways for homo economicus to make a living. He can produce useful goods and services for exchange with other producers –the way of production and exchange– and/or he can try to appropriate a larger slice of whatever others produce –the way of appropriation and conflict. Until the 1960s, economics was almost exclusively centered on analyzing production and exchange for which it was enough to focus only on the bright side of homo economicus, that is, on his propensity to exchange for what he wants. Since then, interest in analyzing appropriation and conflict has been increasing, exposing his dark side –that is, his inclination to grab what he wants. More recently, a synthesis of both analyses has been driving the research on the governance of private and public transactions as well as the growing consensus that almost all transactions need governance (in the sense of a restraining influence over the parties to transactions). Although much remains to be done, a meaningful analysis of governance assumes the dark side of self-interest and deals explicitly with the costs of appropriation and conflict. This section contains three papers that illustrate progress in the economic analysis of appropriation and conflict. First, Herschel Grossman looks at recurring distributional disputes between social groups to determine the conditions for their settlement through a self-enforcing constitution or for them to become neverending civil conflicts. Second, Andres Solimano discusses the threat of huge losses due to civil conflict and terrorism with special reference to Latin America. Given the limits to preventing the losses, Solimano looks at the complementarities between private and public insurance to share them. Finally, Edgardo Barandiaran high-lights the relevance of Aoki’s comparative analysis to explain the institutional arrangements to protect property from autocrats (Olson’s stationary bandits) and their implications for the emergence of the democratic state. Cuadernos de Economia, Ano 40, No 121, pp. 607 (diciembre 2003)

The investment theory of wars: Belligerent dictators in the McGuire/North-model of autocracy *

In the early 90s Martin C.McGuire and Mancur Olson Jr. developed aneconomic explanation of autocracy. Thedifferentiation between roving andstationary bandits is the core of theirtheory. The authors claim that an``invisible hand'' leads to a conversion ofroving bandits into stationary bandits. Inthis respect, stationary bandits are``public goods providing kings'' who reignpeacefully. Thus, war is hardly a rationalpolitical option. However, this resultstands in contradiction to the overwhelmingevidence for wars in human history.Therefore, war as a political option isintroduced. It is shown that theMcGuire/Olson argument challenging the ideathat the welfare of people ruled bystationary bandits is always higher thanwelfare of people ruled by roving banditsproves to be valid only under certaincircumstances.