Polycentric Defense.

• Author: Coyne, Christopher J.
• Position: Defense against internal and external threats - Report

Economic discussions of defense typically focus on the provision of security at the national level against external threats. Taking "the nation" as the relevant unit of analysis, economists model defense as a good that is jointly consumed, nonexcludable, and nonrival over a significant geographic space. This assumption leads to the conclusion that defense cannot be sufficiently provided privately due to free-riding issues and therefore must be supplied by a national government. What it, however, we shift our focus from "the nation" as the relevant unit of analysis of defense provision to more narrow and nuanced units? The purpose of this paper is to answer this question. The orthodox treatment of defense as a pure public good that must be provided by a centralized state suffers from three key issues.

First, it neglects the reality that many threats and therefore the defense against these threats do not affect everyone in a nation. For example, foreign invasions often target a specific area within a country'. When Russia invaded and annexed the Crimean Peninsula, for instance, the Russian government was not threatening to conquer all of Ukraine. As another example, consider that terrorist acts are typically highly local in nature. If, for instance, a major city within a nation is the target of a dirty-bomb threat, it is inappropriate from the standpoint of economic analysis to treat this threat as affecting "the nation" as a whole. This is not to deny that there could be threats that affect all members of a nation. It is instead to recognize that in practice most threats have externalities that are localized relative to the nation a whole. A full appreciation of defense, therefore, requires moving beyond "the nation" as the default unit of analysis.

Second, the orthodox treatment of defense overlooks the incompleteness of the state provision of defense, which is the result of several dynamics. One is that state efforts suffer from the well-known preference-revelation problem of determining the optimal quality and quantity of public goods. In addition, the state provision of defense suffers from a range of public-choice issues--for example, rent seeking, rent extraction, bureaucratic dysfunctions--that make the provision of security highly imperfect in practice (see Coyne and Hall 2019). Moreover, in many instances states are simply incapable of providing defense to citizens due to inadequate capacity or perverse incentives. In the face of incomplete provision of security by centralized nation-states, there is an important role for a variety of actors to play in supplying defense.

Finally, the orthodox treatment of defense assumes that the state providing defense is purely protective, meaning it does not violate the rights of its citizens. In reality, however, the state can and does use its powers to threaten the well-being of the citizens it is tasked with defending (see Coyne 2015). Under this scenario, the domestic state is an internal threat to citizens. The orthodox treatment of defense treats all threats as external and assumes away the paradox of government fundamental to constitutional political economy. As such, it has limited explanatory' power in scenarios where these conditions do not hold and where people face genuine internal threats from state power.

We address these issues by moving institutional context to the foreground. Rather than assuming that defense is in all instances a public good for the nation, we argue that the specific nature of the good depends on its institutional context. Goods that are considered part of defense are often associated with diverse scales of externalities. These goods can, therefore, be produced by diverse centers of decision making that operate at multiple scales rather than simply by a monocentric nation-state. Indeed, in practice, defense-related goods have historically been provided by a polycentric archipelago of public, private, and civil society organizations.

Our analysis bridges two literatures--the orthodox economic treatment of defense and the Bloomington School research program on polycentric systems of governance and social cooperation. As noted, economists typically model defense as a pure public good that must be provided by a monocentric authority. We leverage the insights of Elinor Ostrom and Vincent Ostrom and of the Bloomington School of institutional analysis based on their research (see Aligica and Boettke 2009; Tarko 2017; Boettke 2018; Aligica, Boettke, and Tarko 2019) to challenge the general applicability of the orthodox view of defense. Scholars in the Bloomington tradition have studied polycentricity in diverse arenas, including competitive market economies (V. Ostrom 1999, 2014, 49-58), scientific inquiry, federalism (V. Ostrom 1991, 1999, 2014, 49-58), common-pool resource conservation (E. Ostrom 1990), municipal policing (Ishak 1972; E. Ostrom, Parks, and Whitaker 1973; E. Ostrom et al. 1973; E. Ostrom and Whitaker 1974; Rogers and Lipsey 1974; Parks 1979; Boettke, Paiagashvili, and Lemke 2013; Boettke, Lemke, and Paiagashvili 2016; Goodman 2017), education (E. Ostrom 1996), social capital networks (Craig and Goodman 2019), and responses to climate change (E. Ostrom 2009). Vincent Ostrom briefly examines international order as a polycentric system and the implications this has for peace and defense (1991, 240-43). However, existing scholarship has largely neglected extending polycentricity to the provision of defense. (1)

We proceed as follows. The first section contrasts the orthodox public-goods theory' of defense with the polycentric alternative. The second section illustrates the prevalence of polycentric defense with a sample of real-world examples. And the third section concludes.

Theoretical Considerations

To understand the orthodox model of defense, one must start with Paul Samuelson (1954, 1955), who divided goods into two types: private goods and public goods. Private goods are excludable and rivalrous, whereas public goods are nonexcludable and nonrivalrous. Since Samuelson, national defense has been given as an archetypal example of a public good (see Coyne and Lucas 2016). Defense is asserted to be nonexcludable because it is very difficult to defend me from a missile strike or an invasion widtout defending my neighbor. Similarly, one person in a territory being protected does not tend to deplete security for other residents, and defense is therefore assumed to be nonrivalrous.

Since Samuelson wrote his seminal paper, however, various nuances have been introduced to the theory of public goods. New types of goods--for example, club goods and common-pool resources--have been introduced, and institutional context has become far more important to our understanding of the nature of goods (see, for instance, Buchanan 1965; E. Ostrom and V. Ostrom 1977; Cowen 1985; E. Ostrom 2009). These theoretical and empirical developments require rethinking national defense provision on two margins. First, defense should not always be considered a pure public good. Second, defense being a public good does not automatically' imply that it requires state provision, because under certain conditions private individuals can come together to provide defense just like priv ate individuals are able to come together to resolve the tragedy of the commons despite theoretical predictions to the contrary' (see E. Ostrom 2009).

This reconsideration of defense begins with a deep appreciation of how both the institutional context within which a good is provided and the scale at which it is consumed shape the nature of the good. As Tyler Cowen points out, "Nearly every good can be classified as either public or private depending upon the institutional framework surrounding the good and the conditions of the good's production" (1985, 53). This means that when the surrounding institutional context is varied, a good associated with defense can be more or less public in nature.

Consider the provision of missile defense (see Cowen 1985; Hummel 1990; Hummel and Lavoie 1994). The orthodox view is that missile defense is a public good. It is nonrival because defending someone from a missile attack does not make her neighbors more vulnerable to a missile attack. It is nonexcludable because defending a city from a missile attack does not permit singling out particular houses in the city limits and leaving them vulnerable. An appreciation of institutional context, however, suggests that these features of missile defense are not immutable.

If the marginal unit of national defense is the entire U.S. missile defense system, that unit may be nonexcludable and nonrival at the national level. However, a single antiballistic missile is rivalrous in that it cannot protect New York City and Los Angeles simultaneously. Once missile defense becomes rivalrous, it is no longer a public good. It is at that point either a common-pool resource or a private good, depending on the costs of exclusion, which also depend on...