Central Authority and Order*


  • *

    We are particularly grateful to Peter Bearman for his comments. Additional assistance came from Joshua Baron, Colin Deval, Robert Faulkner, Daniel Geffen, Robert Jervis, Jonathan Monten, Tammy Smith, Charles Tilly, Quentin van Dooselaare, and Kenneth Waltz. The three anonymous referees at Sociological Theory gave comments that were both incisive and extremely useful, for which we are also grateful. We divide responsibility for the remaining errors. Address correspondence to: Emily Erikson, University of Massachusetts at Amherst, Department of Sociology, Thompson Hall, Amherst, MA 01003–7525. Tel.: +413-545-4969; Fax: +413-545-0746; E-mail: erikson@soc.umass.edu, jmp84@columbia.edu.


Strong central authorities are able to effectively manage costly defection, but are unable to adequately address lesser conflicts because of limits to their ability to monitor and enforce. We argue, counterintuitively, that these limitations build cooperation and trust among subordinates: the limitations contribute to the production of order. First, limits to authority leave space for locally informed decentralized enforcement. Second, central authorities act as powerful but incompetent third parties whose threatened interventions increase incentives to cooperate and, therefore, to trust. We outline the mechanisms by which a strong central authority enforces order and test their utility by considering the secondary literature on rates of conflict in strong, weak, and capricious states. We supplement this evidence, based on association, with a close examination of diverse case studies: baseball umpires, commercial contracts, and domestic disputes. By analyzing these case studies, we isolate and describe the mechanisms by which central authorities produce order in varied settings. We find that central authority may be effective, but the majority of this effectiveness derives from an indirect influence on dyadic relations rather than direct intervention. The state interacts with local communities, but each operates according to distinct logics. The particular character of their interaction produces four mechanisms useful in the production of order. We briefly explore implications for the operation of law as well as the production of generalized trust.

For it turned out that great acts of authority were so clumsy that experience itself has made known that only goodness of government brings prosperityMontesquieu (1995)