Introduction

The term “civil–military relations” refers broadly to the interaction between the armed force of a state as an institution, and the other sectors of the society in which the armed force is embedded. It is an intensely interdisciplinary area of research, reflecting the work of political scientists, military sociologists, and historians. Arguably, the field of civil–military relations really took off – at least in the United States – as social scientists became part of the war effort in World War II. Much of this early civil–military relations research focused on the individual service member and small unit cohesion (Stouffer et al. 1949–50; Gray 1959).

Subsequently, there have been several “waves” of civil–military research (Desch 1999:2). In the late 1950s and early 1960s, Samuel Huntington (1957), Morris Janowitz (1960), and Samuel Finer (1962) reoriented research away from individuals and toward the relationships among military institutions, societies, and governments in the post–World War II period. But given the belief that military intervention in the United States was highly unlikely, the focus of much postwar civil–military relations research was comparative, with special attention to emerging states in Africa, Latin America, and the Middle East (Shils 1962).

A second wave emerged in the 1970s in response to the belief that US–Soviet détente might create the conditions for international peace (Betts 1977; Perlmutter 1977; Nordlinger 1977). A third wave broke with the end of the Cold War and continues to this day. Much of this wave's research has focused on civilian control of the military in liberal democracies (Dunlap 1992; Weigley 1993; Kohn 1994; 2002; Luttwak 1994; Foster 1997; Holsti 2001) and the consequences of a possible “gap” between liberal society and a possibly distinct military culture (Ricks 1999; Guttman 2000; Feaver and Kohn 2001). It has not been unusual during this period to hear warnings about a “crisis” in civil–military relations (Foster 1997; Kohn 1994; 2002; 2008).

For the most part, those who study civil–military relations take for granted that there are significant differences between the leaders, institutions, values, prerogatives, attitudes, and practices of a society at large, on the one hand, and those of that society's military establishment, on the other (Welch 1993:507–11). Paradoxically, these differences apply even to military governments, since the distinction has to do with functional roles, not background. The functional role of the “political” leader, whether he wears a uniform or not, is to address the whole array of domestic and foreign affairs that the state may face. The functional role of the “military” leader is to run the military on a day-to-day basis (Brooks 2008:3).

The basis of civil–military relations is a dilemma: what Peter Feaver has called the civil–military problematique, which requires a given polity to balance two concerns. On the one hand, it must create a military establishment strong enough to protect the state. On the other, it must somehow ensure that this same military establishment does not turn on the state that established it (Feaver 1996).

The response of a polity to the civil–military problematique can be seen as a bargain negotiated among three parties: the citizens, the civilian governmental authorities, and the uniformed military. The purpose of this bargain is to allocate prerogatives and responsibilities among the parties.

Obviously, the terms of the bargain and the bargain itself will vary from state to state and, even within a single polity, may vary across time. For instance, in liberal democracies, the playing field is relatively balanced: all three parties have more or less of a say in negotiating the terms of the bargain. Although the citizenry may not be directly involved in drawing up the bargain, the bargain cannot be sustained without their acquiescence. For instance, during the period of the Early US Republic, important political and military leaders would have preferred a larger regular establishment, but strong opposition to the idea of a standing army rendered such a preference moot (Cress 1982).

This is not to say that in a liberal democracy, the parties to the bargain are equal. The liberal democratic civil–military bargain is the outcome of an “unequal dialogue.” It is “a dialogue, in that both [the civilian and military] sides expressed their views bluntly, indeed, sometimes offensively, and not once but repeatedly – and unequal, in that the final authority of the civilian leader was unambiguous and unquestioned” (E.A. Cohen 2002:247). In liberal democracies, the military, despite having a monopoly on coercive power, generally accepts its position relative to the other parties.

In authoritarian states, conversely, the role of the people at large is greatly curtailed. In the extreme case, the military may be coeval with the government, subject to the distinction made above. As has been said about Prussia, most states have an army but, in the case of Prussia, the army had a state.

From time to time throughout the history of a polity, certain circumstances – political, strategic, social, technological, etc. – change to such a degree that the terms of the existing civil–military bargain become obsolete. The resulting disequilibrium and tension lead the parties to renegotiate the bargain in order to restore equilibrium.

There are five sets of questions that lie at the heart of the civil–military bargain at a given time (Owens forthcoming). The first category concerns the issue of who controls the military, and how. In authoritarian or praetorian states, the question is largely moot. On the other hand, liberal societies often take civilian control for granted, but doing so begs several further questions: does civilian control refer simply to the dominance of civilians within the executive branch – the president/prime minister or the secretary/minister of defense? What is the role of the legislative branch in controlling the military instrument? Is the military establishment “unified,” that is, does it speak with anything like a single voice vis-à-vis the civil government? What is the nature of military advice? Should military leaders “insist” that their advice be heeded? What courses of action are available to military leaders who believe the civilian authorities are making bad decisions?

The second question is closely related to the first. What degree of military influence is appropriate for a given society? To what extent does or should the military intervene in domestic affairs? The extreme form of military influence is a coup d’état. Another form of military intervention in domestic politics is praetorianism. How does the government avoid or limit military intervention? For the most part, advanced liberal societies have avoided these forms of military intervention. But even in the case of a liberal society, it is appropriate to ascertain the proper scope of military affairs. To what extent should the military influence domestic or foreign policy? Should active duty officers lobby for programs and policies?

The third question concerns the appropriate role of the military in a given polity. Is it to fight and win the nation's wars or engage in constabulary actions? What kind of wars should the military be preparing to fight? Should the focus of the military be foreign or domestic? States have answered this question differently at different times and under different circumstances. For example, throughout most of its history, the United States Army was a constabulary force. It oriented itself toward large-scale conflicts against foreign enemies only in the 1930s. The end of the Cold War and the attacks of 9/11 have suggested new answers, e.g. a focus on “irregular warfare” (counterinsurgency and counterterrorism) as well as an openness to the use of the military in domestic affairs. What impact do such issues have on civil–military relations?

Fourth, who serves? Is military service an obligation of citizenship or something else? How are officers accessed and promoted? Is the accession and promotion of officers based on merit and achievement or political affiliation, social class, ethnicity, or religion? Obviously, such questions have been answered differently from state to state and even differently within a state at different times under different circumstances. Through most of its early history, the United States maintained a small regular peacetime establishment that mostly conducted limited constabulary operations. During wartime, the several states were responsible for raising soldiers for federal service, either as militia or volunteers. Conscription was the norm in the United States from World War II until the 1970s. Today the US military is a volunteer professional force. But even this force continues to evolve, as debates over such issues as women in combat and service by open homosexuals make clear (Moskos et al. 2000; Miller and Williams 2001). Other states pursue different approaches.

Finally, how effective is the military instrument that a given pattern of civil–military relations produces? All of the other questions mean little if the military instrument is unable to ensure the survival of the state. If there is no constitution, the question of constitutional balance doesn't matter. Does effectiveness require a military culture distinct in some ways from the society it serves? What impact does societal structure have on military effectiveness? What impact does political structure exert? Is the effectiveness of militaries in some developing states degraded as a result of their primary role in ensuring domestic security and regime survival? What impact does a given pattern of civil–military relations have on the effectiveness of strategic decision making processes (Brooks 2008; Desch 2008)?

In general, there are two lenses through which to examine these questions. The first is the institutional lens, which focuses on how the actors in a polity, including the military as an organization, interact within the institutional framework of a given polity's government. The most influential institutional theory of civil–military relations was advanced fifty years ago by Samuel Huntington in his seminal work, The Soldier and the State (1957). The primary concerns of institutional theorists are control of the military, the proper sphere of the military, and the ability of the military to maintain its effectiveness in protecting the interests of the state – in the face of a “social imperative” that may be hostile to the military “way.”

The second lens is sociological or cultural. This lens focuses on the broad question of military culture vs. liberal society; the role of individuals and groups, e.g. women, minorities, enlisted servicemen and women within the military and the relationships among them; the effectiveness of individual service members in combat; small unit cohesion; the relationship between military service and citizenship (to include the civic republican tradition); the nature of military service (occupation, profession, etc.); and the relationship of militaries and the societies from which they stem. The origins of the sociological perspective on military affairs can be traced to Morris Janowitz’s 1960 book, The Professional Soldier (Burke 1993; 1998).

A variation of this perspective is “concordance theory,” which rejects the idea that “healthy” civil–military relations necessarily require a distinct separation between the civilian and military realm. Israel, for instance, has little separation between the two and yet civil–military relations seem stable (Schiff 1995; 2009).

Questions of civil–military relations are complex. It is unlikely that one analytical approach will provide anything close to the whole picture. A “central task of the political sociology of the military is to look at both the military institution and the political system and to determine how the special institutional characteristics of a particular military establishment shape its response to influences coming from the political system” (Stepan 1971:55).

Of course, an important reason for studying civil–military relations is to determine what constitutes “good” and “bad” relations. Such a determination is of more than merely academic interest. It has implications for the very survival of a polity. As the civil–military problematique would suggest, the worst-case consequences of dysfunctional civil–military relations would include catastrophic failure on the battlefield leading to the defeat of the state in a war or the seizure of the government by the military itself.

But dysfunctional civil–military relations may generate other adverse outcomes short of the catastrophic ones. For example, poor civil–military relations may lead to failures in strategic assessment (Brooks 2008). This is true during both war and peacetime. In the case of war, poor strategic assessment may contribute directly or indirectly to defeat on the battlefield because strategic leaders, both political and military, fail to share information or cooperate in other ways (Snyder 1984; Brooks 2008). A case in point is the US war in Iraq. Many observers have contended that most of the problems the United States faced in this conflict were the result of Secretary of Defense Rumsfeld's management style (Herspring 2008), the insulated nature of the Bush administration, or Rumsfeld's penchant for simply overruling advice that did not support his preferences. However, Brooks argues persuasively that these problems, especially with regard to post-conflict planning, resulted from civil–military pathologies created by earlier debates over “transformation” in the Pentagon between Rumsfeld and the uniformed military. These pathologies resulted in oversight mechanisms that weakened strategic coordination (Brooks 2008:226–55).

In the case of peacetime, poor strategic assessment may lead to an overestimation of an adversary's capabilities, resulting in the wasting of resources on defense. On the other hand, underestimating adversaries’ capabilities leads to the allocation of too few resources to defense.

Criteria for judging the health of civil–military relations might include: (1) relative harmony between civilians and the military; (2) the effectiveness of the armed forces in executing their missions; and (3) constitutional balance. “Good” civil–military relations would seem to exhibit some combination of the following: (1) comity and a low number of disagreements between civilian and military decision makers; (2) success in war and peace and the absence of policy–strategy “mismatches”; and (3) a lack of encroachment by either party on to civil–military decisions on the “turf” of the other.

Some authors dispute the notion that harmony and comity necessarily make for “good” civil–military relations, arguing that tension between civilian authorities and the military is healthy. For instance, a low number of disagreements between civilian and military decision makers may simply mean that the civilians have appointed “yes men” who can always be expected not to “rock the boat” (E.A. Cohen 2002). This charge was frequently leveled against former US Secretary of Defense Donald Rumsfeld.

Unfortunately, “except for the most obvious cases, there is no consensus in the recent literature as to what constitutes ‘good’ civil–military relations or ‘effective’ civilian control of the military” (Snider and Carlton-Carew 1995:16). In the case of liberal democracies, the absence of coups would seem to set the bar too low. Most scholars focus on the extent to which civilian preferences prevail when there are differences between the civilians and the military (Huntington 1957:83–5; Feaver 1996; Desch 1999:4–5; Kohn 2002). Others equate healthy civil–military relations with the maintenance of civilian values and the lack of military domination of society (Millett 1979). Of course, a critical measure of good civil–military relations is success in war. Paradoxically, this may require civilian intervention in military affairs, generating significant civilian–military friction (E.A. Cohen 2001).

Determining “bad” civil–military relations is less difficult. The two polar extremes of bad relations are militarism and de-bellicization. The former is the dominance of military institutions, values, prerogatives, attitudes and practices, etc. within society (Ekirch 1956; Vagts 1959). The latter is the denigration or even complete extirpation of military virtues from a society (Huntington 1957; Guttman 2000; Sheehan 2008), the most dangerous consequence of which is defeat in war. While some have suggested that the United States is moving toward one extreme or the other (Lasswell 1941; Ekirch 1956; Bacevich 2005), the evidence to support such claims is weak. The real civil–military relations issues for most polities, including developing states, concern the mutual influence of the civilian and military sectors of society. “The problem of the modern state is not armed revolt but the relation of the [military] expert and the politician” (Huntington 1957:20; Welch 1992).

Influences on Civil–Military Relations

A number of factors influence the civil–military relations of a state. The first of these are its history and culture. All too often, students of civil–military relations treat history and culture as peripheral issues, “intervening variables” that lie between the “real” factor to be studied – the independent variable, whatever it may be – and the dependent variable, the state or pattern of civil–military relations (Desch 1999:11). But the substantial differences in patterns of civil–military relations between Prussia and Great Britain in the nineteenth century and between the United States and Israel today are directly attributable to differences in culture and history (Craig 1955; Strachan 1997; Schiff 2009).

The political institutions of a state also exert a strong influence on its civil–military relations by allocating relative power to civilian and military leaders. Clearly, different regime types will exhibit different patterns of civil–military relations (Janowitz 1964: Perlmutter 1970; S.P. Cohen 1984; 2002; Avant 1994; Millett and Maslowski 1994; Bland 2000; Peri 2003; Lewis 2006; Schiff 2009). The military may be dominant, subordinate to civilian control, or share power (Brooks 2008:33–4). Even in highly militarized regimes, the military may only be one constituent part. For example, in the Soviet Union, the military had to compete against the Communist Party apparatus and the state security system, the KGB, for influence (Nichols 1993). The People's Liberation Army (PLA) faces similar challenges in China.

In liberal democracies, civil–military relations are affected by the constitution of the state and the statutes and practices arising therefrom. In such polities, civil–military relations are complicated by the vast array of players in both the civilian and military realms.

The former consists of the executive and legislative branches of government, both of which are further divided. The executive branch includes the president or prime minister, the appropriate cabinet officers, especially the secretary/minister of defense, advisory committees, e.g. the National Security Council in the United States, and non-cabinet civilian appointees such as the service secretaries. The fact that the interests of political appointees and career civil servants are not always the same and that the interests of both may differ from those of the uniformed military has an important impact on civil–military relations. Nor are legislatures monolithic, consisting as they do of members from a number of political parties. Structure matters as well. The national legislature of the United States is bicameral. In the United States, as in most liberal regimes, the legislative branch does most of its business in committees.

The same goes for the military realm, which usually includes a number of uniformed services. For many years in the United States, the services were the main players on the military side. The result was often a high degree of interservice rivalry, which reached its peak in the United States during the “defense unification” debates after World War II. This extreme manifestation of interservice rivalry played out not only within the newly formed Department of Defense, but also in Congress and the press (Caraley 1965; Keiser 1982; Boettcher 1992). While competition among separate services for mission and resources may contribute to civil–military pathologies, such competition may also have some beneficial effects, e.g. division of labor, a prudent focus on planning future forces, and innovation (Sapolsky 1997; Owens 2006). But since the passage of the Goldwater-Nichols Defense Act in 1987, power has flowed from the services to the commanders of the “unified commands,” the joint regional and functional organizations that are tasked with the actual conduct of operations and deployments.

Nonetheless, the individual services still exert a great deal of influence on US policy. As Huntington observed, each military service is built around a particular “strategic concept […] which defines the role of the service in national policy, public support which furnishes it with the resources to perform this role, and organizational structure which groups the resources so as to implement most effectively the strategic concept” (1954:483). These paradigms, which the late Carl Builder called “masks of war” (Builder 1989), have shaped the services’ institutional approaches to influencing policy, especially in Congress.

The power of the Chairman of the Joint Chiefs of Staff (CJCS) was also enhanced by Goldwater-Nichols (Locher 2002). Previously, CJCS was merely the spokesman for the Joint Chiefs of Staff, a corporate body consisting of the four service chiefs, who in their collective capacity were the source of military advice to the president. But Goldwater-Nichols made the Chairman, not the Joint Chiefs of Staff as a whole, the primary military adviser to the president and the secretary of defense. Along with the large Joint Staff, the Chairman per se has become a major player in civil–military affairs, despite the fact that he is not in the chain of command. Precedent is also important. In the case of the United States, the principle of military subordination to civilian authority seems to have been internalized by each generation of officers based on the precedent set by George Washington at the end of the American Revolution. Most analysts agree that the likelihood of a coup d’état in the United States is low, but others wonder if the power of this precedent has been weakened over the past few years. But the low likelihood of a coup in the United States does not mean that military actors cannot still find other ways to undermine balanced civil–military relations.

However, precedent is not always determinate. The military establishments of India and Pakistan were both shaped by the British military tradition, but the civil–military relations of the two countries are vastly different (Schiff 2009).

Changes in the international security environment also influence civil–military relations, although writers have disagreed about the direction of that influence. Harold Lasswell contended that a greater level of external threat would move civil–military relations in the direction of a “garrison state” (Lasswell 1941). Michael Desch, on the other hand, has argued that states facing high external threats and low internal threats have the most stable civil–military relations (Desch 1999).

Technology has an impact on patterns of civil–military relations. For instance, the destructive power of nuclear weapons not only increased the role of civilians in the development of strategy, but also reduced the leeway of the military in operational and even tactical matters (Feaver 1992). The proliferation of information technology increases the potential for civilian involvement in operational details. Social forces play an important role as well in shaping civil–military relations, as illustrated by debates over racial integration, women in combat, and open homosexuals serving in the military (Moskos et al. 2000; Miller and Williams 2001; Williams 2008).

The character of conflict affects civil–military relations. Changes in the kinds of war the military is expected to fight have potential implications for the process of strategic decision-making, the composition and operations of military organizations, and the interagency cooperation process during the course of a conflict. Patterns of civil–military relations during traditional interstate war may well differ from those during the conduct of a counterinsurgency. Heavy reliance on special operations forces has implications for congressional oversight and the increasing reliance on contractors will continue to complicate civil–military relations.

Finally, there is the changing concept of “soldier-hood.” In the case of the United States, the idealized model for military service throughout most of American history was the “citizen-soldier.” A civilian most of the time, he answered his country's call in times of emergency, returning to civilian pursuits once the emergency had passed. Since the end of the draft in 1973, the citizen-soldier has given way to the long-term professional, a soldier akin to the Roman legionnaire (Abrams and Bacevich 2001; E.A. Cohen 2001). Other countries face similar issues (S.P. Cohen 1984; 2002; Gal 1986; Haqqani 2005).

Theories of Civil–Military Relations

As the Prussian “philosopher of war,” Carl von Clausewitz, observed: “theory exists so that one need not start afresh each time sorting out the material and plowing through it, but will find it readily to hand and in good order” (Clausewitz 1976:141). A theory seeks to illuminate comprehensively and systematically the link between cause and effect.

At a minimum, a theory must be able to describe usefully the phenomenon or phenomena under investigation and explain it. This is the empirical function of theory. But theories often are employed to do two other things. First, based on its description and explanation of the phenomenon under examination, a theory may be used to predict, at least in a general way, what might happen under similar conditions in the future. A theory may also serve as the basis for prescribing policy, for translating “is” to “ought.” This normative function links the descriptive and predictive qualities of the theory to the policy of the state.

Applied to civil–military relations, a workable theory would meet the minimum requirement to describe and explain the nature and characteristics of different patterns of civil–military relations. But most theories of civil–military relations are also used to specify general conditions either conducive or detrimental to healthy relations (Desch 1999:11). Finally, a theory of civil–military relations may prescribe what steps a state must take in order to achieve or maintain healthy relations.

Clearly this meaning of theory is less formal than that found in the physical sciences. It is more based on intuition, experience, and an understanding of the rules arising out of practice. What Michael Handel observed with regard to theory in war applies with equal force to theories of civil–military relations: “the development of the study and theory of war is (and probably will remain) in a pre-Newtonian, pre-scientific, or non-formal stage” (Handel 2001:xvii).

The need for a theory of civil–military relations is driven by the aforementioned civil–military problematique (Feaver 1996). In order to ensure its security, society delegates the authority for the use of force to a subgroup within society. How does society ensure that this subgroup does what it is supposed to – protect society from its enemies, both foreign and domestic – without turning on society itself?

The problematique implies that there are two polar dangers for a society when it comes to civil–military relations. If the military is weakened in order to ensure that it will not turn on society itself, it may face defeat on the battlefield. But if the military is given everything it needs to ensure that it will prevail on the battlefield, it may be in a position of political dominance, able to dictate policy to the civilians. These correspond to de-bellicization and militarism.

The extreme case of military dominance is a coup. But even short of a coup, there is always the possibility that the military will not do what civilian authorities want it to do. For instance, in the United States the uniformed military has often employed such techniques as leaks to the press, lobbying the public and Congress, “foot-dragging,” and “slow-rolling” to thwart the policy goals of civilian authorities (Feaver 2003; Owens forthcoming).

Samuel Huntington: An Institutional Theory of Civil–Military Relations

The theory of civil–military affairs prevalent in most Western liberal polities is based on a distinction between the civil and military realms. This approach can be traced to the practice of eighteenth and nineteenth century European states and to the theory of war advanced by Clausewitz. Rebecca Schiff calls this approach “separation” theory (Schiff 2009).

Although the Prussian state that Clausewitz served is often seen as the exemplar of militarism, his formulation of war as a continuation of politics or policy (politik) by other means (Clausewitz 1976:87) implies a distinction between political decision makers and the military. “No one starts a war – or rather no one in his senses ought to do so – without first being clear in his mind what he intends to achieve by that war and how he intends to conduct it. The former is its political purpose; the latter its operational objective” (1976:579).

Actual war as opposed to war in theory (absolute war) “is only a branch of political activity” that “is in no sense autonomous” (Clausewitz 1976:605). “The character and general shape of any war should mainly be assessed in light of political factors and conditions” (1976:607). Since “absolute” war is theoretically unlimited, driven by its own logic to the extreme of violence and exhaustion, it is the political purpose of war that makes war “rational,” providing war with a purpose beyond its own logic.

The most influential theory of civil–military relations in the West in general and the United States in particular is the institutional theory advanced half a century ago by the eminent political scientist Samuel Huntington in his book, The Soldier and the State (1957). The Soldier and the State has had a great and lasting effect on American thinking about the way the military interacts with civilian society, especially within the uniformed military. Indeed, the US military has come to endorse many of its general conclusions and has made it central to its civil–military relations education.

Huntington's main descriptive or empirical claim was that American civil–military relations were shaped by three variables: first, the external threat, which he called the functional imperative; and two components of what he called the societal imperative, i.e. “the social forces, ideologies and institutions dominant within the society” (Huntington 1957:2–3).

The first component of the societal imperative is the constitutional structure of the United States, the legal-institutional framework that guides political affairs generally and civil–military affairs specifically. The second is ideology, the prevailing worldview of a state. Huntington identified four ideologies: conservative pro-military, fascist pro-military, Marxist antimilitary, and liberal antimilitary (1957:89–94). He argued that the fourth was the dominant ideology of the United States (1957:143).

Huntington contended that both components of the societal imperative – the constitutional structure and the American ideology of antimilitary liberalism – had remained constant throughout American history. Accordingly, the entire burden of explaining any change in civilian control or level of military armament would have to rest with the functional imperative, i.e. the external threat (1957:2, 156).

Huntington further contended that liberalism was “the gravest domestic threat to American military security” (1957:457). “The tension between the demands of military security and the values of American liberalism,” Huntington continued, “can, in the long run, be relieved only by the weakening of the security threat or the weakening of liberalism” (1957:456). Thus the requisite for military security is a shift in basic American values from liberalism to conservatism. Only an environment that is sympathetically conservative will permit American military leaders to combine the political power which society thrusts upon them with the military professionalism without which society cannot endure (1957:464).

According to Huntington, America's antimilitary liberal ideology produces “extirpation” – the virtual elimination of military forces – when the external threat is low, and “transmutation” – the refashioning of the military in accordance with liberalism, which leads to the loss of “peculiarly military characteristics” – when the external threat is high. In Huntington's view, the problem for the United States in a protracted contest such as the Cold War was that while transmutation may work for short periods of time during which concentrated military effort is required, e.g. a world war, it would not assure adequate military capability over the long term.

In the context of the Cold War, Huntington argued that the ideological component of America's societal imperative – liberal antimilitary ideology – would make it impossible to build the forces necessary to confront the functional imperative in the form of the Soviet threat to the United States and to permit US military leaders to take the steps necessary to provide national security. Thus the predictive element of Huntington's theory held that without a change in the societal imperative, the United States would never be able to build the military forces necessary to confront the USSR.

The prescriptive or normative element of Huntington's theory was to suggest a way for the United States to deal with the dilemma raised by Feaver's civil–military problematique: how to minimize the power of the military and thus make civilian control more certain without sacrificing protection against external enemies. His prescription, which he called “objective civilian control,” has the virtue of simultaneously maximizing military subordination and military fighting power. Objective control guarantees the protection of civilian society from external enemies and from the military themselves.

In Huntington's prescriptive or normative theory, the key to objective control is “the recognition of autonomous military professionalism,” i.e., respect for an independent military sphere of action. Interference or meddling in military affairs undermines military professionalism and so undermines objective control (1957:83).

This constitutes a bargain between civilians and soldiers. On the one hand, civilian authorities grant a professional officer corps autonomy in the realm of military affairs. On the other, “a highly professional officer corps stands ready to carry out the wishes of any civilian group which secures legitimate authority within the state” (Huntington 1957:84). In other words, if the military is granted autonomy in its sphere, the result is a professional military that is politically neutral and voluntarily subordinate to civilian control. Of course, autonomy is not absolute. Huntington argues that while the military has responsibility for operational and tactical decisions, civilians must decide matters of policy and grand strategy.

While objective control weakens the military politically, rendering it politically sterile or neutral, it actually strengthens the military's ability to defend society. A professional military obeys civilian authority. A military that does not obey is not professional.

At the opposite pole from objective control lay Huntington's worst-case situation, “subjective control,” which constituted a systematic violation of the autonomy necessary for a professional military and produced transmutation. Huntington argued that subjective control was detrimental to military effectiveness and would lead to failure on the battlefield by forcing the military to defer to civilians in the military realm (1957:80–3).

Morris Janowitz: The Sociological Response to Huntington

Morris Janowitz offered an early critique of Huntington from the standpoint of sociology (Janowitz 1960). The sociological perspective does not ignore the central civil–military relations question of institutional theory: civilian control. However, it focuses most of its attention on the relationship between individuals in the military and civilian society. With regard to Huntington's civilian–military divide, Janowitz argued that the distinction between the civilian and military roles that lay at the heart of Huntington's theory had been blurred by the emergence of nuclear weapons and limited war. For Janowitz, this state of affairs was only the latest manifestation of the way in which emerging technologies and the political interaction between civilian and military elites were causing the two spheres to converge.

Like Huntington, Janowitz focused on the meaning of a professional officer corps. But while Huntington saw military professionalism as a fixed standard, an “ideal-type” based on a strict division of labor between the uniformed military and civilians, Janowitz conceived professionalism as dynamic, changing in response to new sociological conditions. Janowitz argued that, given the central place of the US–Soviet rivalry in both international and domestic politics, even a professional military could not avoid some degree of politicization.

With regard to Huntington's functional imperative, Janowitz contended that in the nuclear age the military needed to adopt a new military role and military selfconception – that of a constabulary force. According to Janowitz, “the military establishment becomes a constabulary force when it is continuously prepared to act, committed to the minimum use of force, and seeks viable international relations, rather than victory” (1960:418). Obviously, the constabulary concept blurs the distinctions between peace and war. Accordingly, the soldier comes to resemble a police officer instead of a warrior. This leads to politicization of the military and raises a challenge to civilian supremacy, as the military attempts to use the political system to resist unwelcome policy direction.

Janowitz's solution to the new problems created by the constabulary concept was to reject Huntington's concept of objective control of the military. Instead of autonomy, he prescribed greater civilian oversight of the military at all levels. He pointed out that civilians possess three main mechanisms for controlling the uniformed military: the budget process; the allocation of roles and missions; and advice to the president concerning the use of the military to advance US interests in the international realm. But the military, Janowitz argued, had found ways to undermine civilian control (1960:363–7).

Although Janowitz proposed a number of external mechanisms for strengthening civilian control, he, like Huntington, ultimately fell back on professionalism. But unlike Huntington's professional officer who eschewed politics altogether, Janowitz's officer corps would be politically aware and possess functions and expertise that overlapped with those of its civilian counterpart.

Janowitz's heirs have taken military sociology far beyond his own conclusions, generating a vast and rich literature that examines how the military and civil society have shaped each other. But when it comes to the critical issue of ensuring civilian control of the military, Janowitz did not really go much beyond Huntington. He merely concluded that to ensure civilian control, Huntington's “self-imposed professional standards,” the basis of objective control, needed to be supplemented by a “meaningful integration” of military and civilian values (Janowitz 1960:420).

Despite Janowitz's sociological challenge to Huntington's institutional theory of civil–military relations, the latter still dominates the field. The reason for this continued dominance is its elegance as an ambitious treatment of civil–military relations and the fact that his prescriptions for how best to structure civil–military relations continue to find a very receptive ear within the American officer corps (Feaver 2003:7).

Most recent attempts to reconstruct the theoretical edifice of civil–military relations constitute refinements of Huntington and Janowitz rather than providing a new theoretical alternative (Avant 1994; Desch 1999). Exceptions include Peter Feaver, who has offered a civil–military relations theory based on a “principal–agent” framework; Rebecca Schiff, who argues on behalf of a theory of concordance that takes issue with the “separation” theory that she attributes to both Huntington and Feaver (Schiff 2009); and Risa Brooks, who, while maintaining the civil–military distinction, has usefully modified aspects of existing theories to create a “distributional approach” that aids in understanding how various patterns of civil–military relations influence strategic assessment (Brooks 2008).

Peter Feaver: Agency Theory and Civil–Military Relations

Feaver argues that although Huntington's theory is “elegant,” it doesn't fit the evidence of the Cold War. For instance, one of Huntington's testable hypotheses was that a liberal society (such as the United States) would not produce sufficient military might to survive the Cold War. But in fact, the United States did prevail during the Cold War despite the fact that the country did not abandon liberalism (Feaver 2003:27). The continued divergence between civilian and military preferences during the Cold War casts doubt on the predictive power of Huntington's empirical theory.

The same problems affect Huntington's prescriptive theory. During the Cold War, the military became more “civilianized,” the officer corps more politicized, and civilians habitually intruded into the military realm (Feaver 2003:37). Feaver concludes that the disjunction between Huntington's theory and the available evidence requires another theory. To provide such an alternative, Feaver turns to “agency theory.”

The problem that agency theory seeks to analyze is this: given different incentives, how does a principal ensure that the agent is doing what the principal wants him to do? Is the agent “working” or “shirking”? The major question for the principal is the extent to which he will monitor the agent. Will monitoring be intrusive or non-intrusive? This decision is affected by the cost of monitoring. The higher the cost of monitoring, the less intrusive the monitoring is likely to be.

The agent's incentives for working or shirking are affected by the likelihood that shirking will be detected by the principal and that the agent will then be punished for it. The less intrusive the principal's monitoring, the less likely that the agent's shirking will be detected. Feaver argues that shirking by the military takes many forms: the most obvious is disobedience, but it also includes “foot-dragging” and leaks to the press designed to undercut policy or individual policy makers.

Feaver posits four general patterns of civil–military relations: (1) civilians monitor intrusively, the military works; (2) civilians monitor intrusively, the military shirks; (3) civilians monitor unintrusively, the military works; and (4) civilians monitor unintrusively, the military shirks. He then shows that Huntington's postulated outcomes are in fact special cases of his own more general agency theory of civil–military relations: Huntington's “objective control” corresponds to pattern (3); his “subjective control” corresponds to pattern (1) (Feaver 2003:119).

Feaver uses the Cold War to test Huntington's prescriptive theory. Huntington had argued that the best way to ensure both military effectiveness and subordination to civilian control was through pattern (3) – objective control (Feaver 1996:33). However, it turns out that the civil–military relations pattern during the Cold War that most corresponds to the evidence is pattern (1) – Huntington's nightmare civil–military scenario – subjective control (Feaver 2003:178). Indeed, agency theory predicts that pattern (1) will prevail when there is a wide gap between the preferences of the civilians and the military, when the costs of intrusive monitoring are relatively low, and when the military thinks the likelihood of punishment for shirking are fairly high. Feaver argues persuasively that the evidence from this period supports these hypotheses. Yet according to Huntington's own criteria for professionalism – expertise, responsibility, and corporateness (Huntington 1957:8–10) – the US military remained highly professional despite extensive civilian intervention.

Of critical importance in establishing Cold War civil–military relations was the firing of a popular military hero (MacArthur) by an unpopular president (Truman). This dramatic action shaped the expectations of the military concerning the likelihood of punishment for shirking during the Cold War period (Feaver 2003:129).

Feaver explains the post–Cold War “crisis” in civil–military relations in a way that integrates a number of features that arose in the 1990s – the end of the Cold War, a growing gap between civilian and military elites, the personal history of President Clinton, the creation of a powerful chairman of the Joint Chiefs of Staff by the Goldwater-Nichols Defense Act of 1986, and the occupation of this office by a popular, politically savvy general – Colin Powell (Feaver 2003:180–233).

Feaver argues that civil–military relations pattern (2) prevailed during the 1990s: civilians monitored intrusively, the military “shirked.” The cost of intrusive monitoring went down. The preferences of civilian and military elites diverged in many important ways, increasing incentives for the military to pursue its own preferences. Finally, the expectation of punishment for shirking decreased as a result of the election of Bill Clinton, whose equivocal relationship with the military made punishment unlikely. Combined with a powerful and popular military leader and an absence of consensus regarding security affairs across the executive and legislative branches, the civilian principals were in a relatively weakened position vis-à-vis the military agents (Feaver 2003:190–210).

Feaver observes that civil–military relations are obviously better when there are good civilian leaders and worse when civilian leadership is bad. One issue is how to hold civilians accountable to the same or greater degree than the military is held accountable. Bad policy, after all, presumably comes from civilian principals.

But Feaver's focus is on the military institution in a democratic polity.

[E]ven when the military is right, democratic theory intervenes and insists that it submit to the civilian leadership that the polity has chosen. Let civilian voters punish civilian leaders for wrong decisions. Let the military advise against foolish adventures, even advising strenuously when circumstances demand. But let the military execute those orders faithfully. The republic would be better served even by foolish working than by enlightened shirking.

(Feaver 2003:302)

After all, the claim that the military should not do what civilians want because what they want is bad for the country shapes the rhetoric of every coup leader who justifies his seizure of power as the rescue of a state from the consequences of an inept government.

Concordance Theory

Rebecca Schiff has argued on behalf of a theory that questions the assumptions that underlie both Huntington and Feaver: that a theory of civil–military relations should be based on the physical and ideological separation of the military from the political institutions of a state (Schiff 2009:32). Concordance theory is concerned with predicting and preventing military intervention in the domestic affairs of a state. Schiff contends that the best way to avoid such an occurrence is to achieve concordance among three “partners” within a polity – the military, the political elites, and the citizenry – on four issues: the social composition of the officer corps; the political decision-making process; the method of recruiting soldiers; and “military style.” Such a cooperative relationship may involve separation but does not require it.

According to Schiff, concordance theory resolves two problems associated with separation theory, as explicated by Huntington and Feaver. The first is the tendency of separation theorists to treat the particular institutional arrangements arising from the experience of the post–World War II United States as universal, applicable to all states regardless of their particular historical conditions and culture. This is especially relevant to the cases of developing states because it means that they “need not adopt the traditional Western model of civil–military relations in order to achieve greater political maturity (Schiff 2009:33).

The second problem is methodological. Separation theory's institutional analysis alone “fails to take into account the cultural and historical conditions that may encourage or discourage civil–military separation. We can see this deficit, for example, in the post-revolutionary US example – a stark contrast to the post–World War II United States from which separation theory is derived” (Schiff 2009:33).

Criticism of concordance theory includes the charge that it is merely a variation of Huntington's concept of “fusion,” the demand “that military leaders incorporate political, economic, and social factors into their thinking” (Huntington 1957:351) and that “military leaders assume non-military responsibilities” (1957:353). Critics also claim that the predictive aspect of concordance theory falls short with regard to the correlation between agreement among the three partners and military intervention on the one hand and to the likelihood of coup d’état on the other (Wells 1996).

But, according to Schiff, these objections ask concordance theory to be something it is not: a theory capable of analyzing all facets and problems in the realm of civil–military relations. Its causal objective is more limited: to predict the likelihood of military intervention in domestic affairs, which, Schiff claims, it does successfully (Schiff 1996:42–3).

Civil–Military Relations and Strategic Assessment

Risa Brooks argues that patterns of civil–military relations affect national security because of their impact on strategic assessment. Brooks identifies two variables that determine the pattern of civil–military relations: (1) the intensity of preference divergence between political and military leaders with regard to corporate, professional, and security issues; and (2) the balance of power between political and military leaders (political dominance, shared power, military dominance). These two variables interact, generating “logics” that affect the institutional features of strategic assessment (Brooks 2008:2–34).

Next she identifies four sets of institutional processes that constitute the element of strategic assessment. The first is the routine for information sharing. The second is strategic coordination regarding the assessment of strategic alternatives, risk and cost, and the integration of political and military policies and strategies. The third is the military's structural competence in conducting sound net assessment. The fourth is the authorization process for approving or vetoing political-military actions (Brooks 2008:34–42).

Brooks then hypothesizes how the various configurations of power and preference divergence affect the quality of strategic assessment, using case studies to illustrate the relation between various patterns and strategic assessment. She hypothesizes that the combination of political dominance and low preference divergence leads to the “best” strategic assessment. The combination of shared power and high preference divergence leads to the “worst” strategic assessment. Other combinations lead to “poor” or “fair” strategic assessment (Brooks 2008:42–54).

Of course, the quality of a state's strategic assessment is not the only determinant of a state's success or failure in the international arena. The competing strategies of other states and other exogenous factors may well trump even the best strategic assessment. Michael Desch has employed a similar methodology to show that the alleged military advantage of democratic states in international relations is overstated (Desch 2008).

Problems with the Contending Theories and Suggestions for Further Research

There is no more important question facing a state than the place of its military relative to civil society and the roles that the military exercises. The coercive power that a military institution possesses always makes it, at least theoretically, a threat to the regime. Clearly, there are many possible patterns of civil–military relations that provide different answers to the five questions posed at the beginning of this essay.

As the survey of contending theories of civil–military relations suggests, there is no “general” or “unified field” theory that successfully explains all of these patterns (Bland 1999). Nor, given the variety and complexity of civil–military patterns is one likely or desirable.

Institutional theory and agency theory focus on control of the military and the military's role. Sociology usefully investigates the question of who serves. For a variety of reasons, Huntington's institutional theory remains the dominant paradigm for examining civil–military relations. First, it deals with the central problem of such relations: the relation of the military as an institution to civilian society. Huntington was the first to attempt a systematic analysis of the civil–military problematique. Second, despite the claims of many of those who look at US civil–military relations through the lens of sociology, analytically distinct military and civilian spheres do appear to exist.

But as the discussion above of Feaver's critique of Huntington makes clear, there are many problems with Huntington's argument. In 1962, S.E. Finer argued that Huntington had severely understated the problem of civilian control (Finer 1962:7–10). He contended that a professional military does not necessarily keep officers out of politics, but indeed might incline them to engage in politics (1962:207 ff.). He also observed that differences in national experience limit the applicability of Huntington's theory.

Schiff agrees, arguing that Huntington's theory is particular to the American experience and is therefore not applicable to other countries (Schiff 2009: chs. 5–7). Indeed, she argues that it does not even apply to the United States during all historical periods (2009: ch. 4).

In addition, empirical studies have not confirmed some of Huntington's key assertions or predictions. It is also the case that some of Huntington's historical arguments are questionable. For instance, the US Army in the late nineteenth century was not nearly as isolated as Huntington contended it was (Gates 1980; Owens 2007). This particular problem illustrates the importance of keeping historical context in mind when examining civil–military relations. Historians such as Russell Weigley, Richard Kohn, and Lawrence Cress have made significant contributions to the study of civil–military relations (Kohn 1975; Cress 1982; Weigley 1993). Finally as Eliot Cohen, perhaps Huntington's most accomplished student, has pointed out, some of the most successful democratic war leaders have paid very little attention to the divide that Huntington's objective control demands (E.A. Cohen 2002).

When it comes to the question of civilian control of the military, Feaver's agency theory corrects some of the flaws in Huntington's theory. Agency theory seems to do a better job of describing the problem of civilian control than Huntington's theory. It is also does better with regard to the predictive aspect of the theory. One reason for this is that agency theory does not depend on the non-rigorous and therefore problematic concepts of professionalism and autonomy to predict how and under what circumstances civilians will best be able to control the military instrument.

Finally, it follows that if agency theory fulfills both the descriptive and predictive functions of a theory better than Huntington's institutional theory, its prescriptive element will also be more useful than what Huntington laid down. Nonetheless, critics argue that as applied to civil–military relations, agency theory achieves analytical rigor by severely limiting its scope. The theory is too parsimonious; it fails to explain enough in the world.

Janowitz and the military sociologists also provide useful insights, especially regarding the question of “who serves?” and related issues. The writers who take their bearings from Janowitz have indeed moved the question of demographics, ethnicity, and recruitment to center stage in a way that transcends the American experience (Moskos et al. 2000; Williams 2008; Schiff 2009). But even as they argue that the concept of separation between the two spheres is theoretically and empirically flawed, these writers still maintain the analytical distinction between the military and civilians.

In the case of concordance theory, critics charge that the definition of military intervention sets the bar too low to be meaningful. The cooperative relationships that are necessary to avoid military intervention themselves look like intervention unless the standard for civilian control is merely the absence of a military coup. The utility of the idea of “concordance” is somewhat in question as it is not clear that the degree of harmony among civilian and military leaders is a “good” or “bad” thing in terms of either political control or military effectiveness.

In many respects, the current state of theorizing about civil–military relations brings to mind the story of the three blind men examining an elephant. Since each can only sense what he is touching (the trunk, a leg, and the tail) and has no concept of the elephant as a whole, each concludes that the beast is something different from what it really is. Despite the lack of an overarching framework for analyzing civil–military relations, the various areas of the field offer many rich “pastures” in which researchers may graze.

Research agendas might well include: additional examination of the emerging civil–military patterns of such emerging powers as China, Russia, and Iran; ascertaining a theory of civil–military relations of Muslim states; follow-up work to Risa Brooks’s excellent study of the impact of civil–military relations on strategic assessment; the civil–military implications of the expanded roles of contractors on the battlefield and increased reliance on special operations forces; the civil–military implications of the increased utilization of airstrikes by unmanned aircraft; the impact of popular will on effectiveness in various sorts of warfare, e.g. counterinsurgency; and further research into the impact of an increasing “civilianization” of the military on military effectiveness.