I have mentioned the North-Wallis-Weingast “Natural States” analysis in several posts. It is sufficiently interesting and important, however, to deserve a stand-alone analysis. In a nutshell, I find many things about their analysis quite persuasive, but in the end, I believe that it exaggerates the stability of natural states, and fails to analyze adequately the factors that determine the stability (or instability) of natural states.
Overall, I am quite sympathetic to the NWW view that those with a comparative advantage in violence will utilize their power to extract rents. I also agree that rival violence specialists have incentives to cooperate, agreeing to divide rents among themselves and avoid conflict that would dissipate these rents. That is, there is some truth to their contention that the natural state “uses limited economic entry to generate economic rents which are used to create credible commitments among elites to support the current regime and provide order.” However, economic logic and history both demonstrate that these agreements among violence specialists–either within a natural state or between natural states–are often not self-enforcing. Put differently, cooperative arrangements among violence specialists often break down, leading to violence within states or between states. Moreover, NWW pay insufficient attention to the costs of “negotiating” the cooperative arrangements among violence specialists. Due to these costs, they may fail to arrive at stable agreements. Furthermore, the costs incurred to arrive at cooperative arrangements can waste substantial economic resources.
In some ways, NWW are applying efficiency wage theory, or the Stigler-Becker theory of honest enforcers, to the study of political organization. Rents are costly (there are distortions at the margin relative to the first best), but they can support second best arrangements. The question arises, however, whether there is an “invisible hand” operating in the “market among violence specialists” that will result in such a second best outcome. Although NWW acknowledge that natural states are prone to violence and civil wars, overall they convey the impression that the natural outcome in natural states is for the “efficiency wage” for violence specialists to induce them to eschew violence, and to encourage stability and enough economic specialization and trade to enhance their rents. I think this is overly optimistic.
In essence, NWW posit that violence specialists will form a cartel. Indeed, firms in an industry always have an incentive to collude to restrict output and raise prices, and share in market power rents. The “folk theorem” of game theory states that if cartel members discount the future sufficiently lightly, the collusive outcome is self-enforcing; the threat of a rent-dissipating price war in response to a defection from the cartel agreement induces the cartel members to adhere to the agreement. NWW essentially argue that the threat of a real civil war that would dissipate rents induces the members of a violence cartel from seizing the rents of others.
But the experience with cartels–even when they were legal, and even when they were legally enforceable (as in pre-WWI Germany)–is that they are seldom stable, folk theorem or no folk theorem. Cooperation tends to come undone, leading to episodes of “price wars,” and often the complete collapse of the collusive agreements.
Similar considerations raise doubts about the stability of violence specialist cartels. Violence specialist cartels are not stable under all parameterizations. In particular, if violence cartel members discount the future too heavily, collusion cannot be supported. A breakdown of collusion within a state leads to civil war. A breakdown of collusion between natural states leads to a war between them.
The dreary litany of civil and international wars in Europe, for instance, suggests that stability is the exception, rather than the rule. This is not surprising, given: the short lifespans of monarchs and the aristocracy (which means that they likely discounted the future quite heavily), the incompleteness and imperfection of information about past actions and violence capacity of cartel members, and the possible entry of new violence specialists not part of the original agreement.
The interaction between violence specialists/elites within a state and inter-state rivalries also plays an important role. Members of a state largely secure from the threat of foreign conquest are likely to have a lower discount rate than those of a state under threat of invasion. A stable cooperative arrangement among the violence specialists is more likely than the secure state than the insecure one. This conjecture has empirical content. Specifically, the creation of an international order that raises the costs of inter-state aggression should lead to greater internal stability in natural states.
Thus, if you believe as I do that self-enforcing cartels are the exception, rather than the rule, you will recognize that NWW’s characterization of natural states is unduly optimistic. I also have some difficulty with their suggestion that the development of rent-creating institutions is intended to achieve stability. They say that the “natural state solve[s] these problems [of political instability] by creating and maintaining privileged rights to valuable resources.” This implies that the conscious purpose of the privileges is to create stability. I rather think that the violence specialists use force to seize the privileges, and that under nearly ideal conditions a balance of terror among the violence specialists results in a stable outcome, whereas under less ideal conditions the outcome is violence and strife. That is, the natural state and its elaborate array of privileges is not the (second best) solution to a mechanism design problem intended to maximize some objective function. Instead, the stable natural state is an almost accidental equilibrium that only occurs under an unusual confluence of circumstances.
NWW also pay little attention to the process of negotiating an agreement among violence specialists. When there is heterogeneity among bargainers, and/or information is imperfect and incomplete, it is often very costly to reach agreements. In the context of violence specialists, uncertainty about the capacities of competing specialists and heterogeneity in the distribution of rents among them can impede the consummation of agreements. The process of “negotiation”–if it can be called that, even as a metaphor–in this context is likely to entail considerable exercise of violence, perhaps lasting over long periods. A considerable fraction of the rents can be dissipated during the process of “negotiations” among violence specialists, as the chronic financial difficulties of monarchs and aristocrats throughout European history attests.
Libecap’s interesting work about the unitization of oil fields illustrates the difficulties of negotiating arrangements to split rents. Libecap shows that although there are clear gains to owners of property with access to a common pool of oil to operate that pool cooperatively, such agreements are typically reached only after a substantial fraction of the rents accruing to the pool are wasted through excessively rapid, non-cooperative extraction. Analogously, although as NWW suggest it would be jointly rational for violence specialists to reach an agreement to secure their rents, it is often individually rational for them to expend considerable resources in an attempt to gain a larger share of the total rents. Again, a reading of European history suggests that this is practically important.
NWW raise the interesting question of the transition from natural states to “open access” states that create far fewer rent-generating privileges for elites. As they note, there are relatively few examples of this transition, and there may be different mechanisms leading to the transition. The destruction of the natural state by an external invader who does not impose a new natural state in its place–as occurred in German and Japan post-WWII–is one mechanism. The history of England suggests another. Specifically, the financial distress of a monarch involved in rent seeking foreign conflicts with other natural states weakens the monarch’s bargaining position relative to non-elites, who agree to finance the monarch in return for a reduction in elite privileges.
In sum, NWW’s concept of a “natural state” is a useful way of understanding the operation of most states throughout history. Its emphasis on the use of violence, or the threat of violence, to secure privileges that produce economic rents for the violence specialists is appropriate and illuminating. Where they go astray, in my opinion, are their suggestions that (a) the elaborate structures of preferences in natural states are designed to secure stability, and (b) that stable, self-enforcing “cartels” among violence specialists are the rule, rather than the exception. The violence specialist cartel analogy, which is implicit in the NWW natural state concept, suggests that instability may indeed be the rule. That is, although sometimes rents can support self-enforcing agreements, they often cannot. In my view, exploration in a political economy context of the conditions that support cooperation, and a study of these conditions historically, will advance our understanding of political and economic evolution, and particularly of the transition from natural states to open access orders.