Most scholars work within established paradigms. Few have the temerity to challenge them—and fewer still succeed in causing fundamental reconsideration. Peter Feaver has accomplished the latter task. He has taken on the dominant long-standing theoretical framework for American civil–military relations and, in the process, provided a more accurate view of them in the early twenty-first century.

Samuel Huntington offered a paradigm in his 1957 book The Soldier and the State. He asserted that only “autonomous military professionalism” provided the best fit between national security needs and “objective” (as contrasted with “subjective”) civilian control over the armed forces. Other frameworks for analyzing civil–military relations have been advanced since then: bureaucratic politics, convergence theory, the “postmodern” military, shared responsibility between civilians and military, and structural theory. None has as detailed an empirical foundation or (in some cases) as ambitious paradigmatic aims as either Huntington or Feaver.

The Soldier and the State was written at the height of the Cold War. The challenge in the mid-1950s was clear, unambiguous, and based on both an internationally recognized state and its allies. It assumed combat against a uniformed enemy, fighting conventional warfare, with a threat of nuclear bombs in the background. This represented (in Feaver's summary) a fundamentally ideological viewpoint. Huntington argued that America required a shift from liberalism to conservatism to ensure maximum security, since the constitutional framework could not be changed. Feaver developed his arguments, by contrast, after the collapse of the Soviet Union and disappearance of its overt threat. (Writing was completed before the launching of the “war on terror” and the invasion and occupation of Iraq; nonetheless, the challenges posed by them provide additional support for his arguments.)

The theoretical heart of Armed Servants lies in agency theory, set forth in Chapters 2 and 3. Developed by economists and management specialists, but infrequently used by political scientists, agency theory is usually presented as the “work or shirk” trade-off. Those responsible for supervision—the “principals”—wish to extract the maximum effort for the minimum cost. Those who carry out the tasks—the “agents”—hope to get away with the greatest reward for the least work. Translating this model from economic to political realms leads to major problems, however. The issues are compounded in applying them to armed forces (at least those willing to accept governmental control, rather than substitute their own preference, as through a coup d'état or other irresistible pressure). Feaver suggests three reasons. First, military agents prefer to be told what policy to pursue. They wish to deal with threats from a “position of advantage, controlling the tempo and the scope of the conflict” (p. 63), resulting in a preference for offensive operations. Second, military agents wish to have their accomplishments given appropriate recognition by others—or, in other terms, they manifest a “general military preference for honor, or a desire for respect” (p. 65). Thirdly, they wish to carry out their tasks with minimal civilian interference and oversight. Feaver adds an important caveat, however: Both civilians and the military are “imperfect judges of what is necessary for national security” (p. 65).

Application of agency theory, like any rational actor approach, presupposes that those involved act in accordance with their interests. Two types of costs are involved: electoral costs of time and effort, and policy costs of divergence between preferred and actual policy outcomes. Feaver discusses them both as a general issue of costs and as a game in formal terms. His conclusion drawn from his model merits citation: “[T]he costs of monitoring, the expectation of punishment, the strategic calculus of the actors … are nonetheless essentially absent in traditional civil–military relations theory” (p. 113).

Shirking exists within the American military, Feaver asserts. Although insubordination is rare and a coup practically unthinkable, three forms exist: 1) inflating costs of military operations (thus predetermining the outcome of policy calculus); 2) carrying out “end runs,” leaks, and various appeals to political actors; and 3) bureaucratic foot dragging so that an undesired policy or policies will never be implemented. However, civilians retain the right to punish; more important, the American armed forces see themselves as agents of civilians, and hence accept sanctions.

The book's empirical foundation is rich and detailed: 10 tables summarizing points made in the text; 33 tables presenting data. For those more comfortable with case studies, there are summaries of American involvement in the 1990–91 Gulf War, Somalia, Haiti, Bosnia, and Kosovo. Force was employed in all, requiring both a policy decision by the principal and its implementation by the agent. Looking at 29 different cases, from Berlin in 1948 to the invasion of Panama in 1989, Feaver finds that the American military worked in 23 and shirked in the remainder, a clear indication of civilian dominance: “[T]he empirical record does not support a picture of a renegade military resolutely thumbing its nose at civilian leaders” (p. 151).

This book appears at a propitious time for a reconsideration of American civil–military relations. Secretary of Defense Donald Rumsfeld has embarked on fundamental transformations (building on changes proposed earlier, but bringing a strong sense of urgency and unwillingness to accept orthodox Pentagon lines of argument). The “war on terrorism” and the ongoing insurgencies in Afghanistan and Iraq for which the armed forces seemed ill prepared add further impetus to the need for reconsideration. Does agency theory provide a better guide than Huntington's “objective civilian control” or the plethora of approaches noted earlier?

In Civilian Control of the Military (1999), Michael Desch focused on four “clusters” of cases. However, only part of one involves the United States directly, examining the much-studied Cold War period. Breadth and depth, yes; but those interested in the theory and practice of American civil–military relations will find Armed Servants more satisfying. Also broad ranging in global terms is Charles Moskos, David Segal, and John Allen Williams, eds., The Post-Modern Military (2000). The authors set forth 11 variables (for example, perceived threat, force structure, or major mission), then apply them to a dozen states. Moskos's chapter on American civil–military relations is interesting, albeit just over 16 pages long.

Summing up: Agency theory now counts as a significant theory in civil–military relations. Feaver's book applies it to the United States; others should utilize it in other settings.