Armed servants

These men and women are the nation’s armed servants, empowered but subordinate, capable of wielding astonishing levels of coercive force, but expected to wield it only within narrow confines dictated by others.

The civil-military challenge is to reconcile a military strong enough to do anything the civilians ask with a military subordinate enough to do only what civilians authorize. How much autonomy can the military enjoy without violating the principle of civilian control? Alternatively, how much control can civilians wield without interfering disastrously in the conduct of the military mission? Civilian leaders decide these normative questions every day, for better or for worse.

The civil-military problematique is thus a simple paradox: the very institution created to protect the polity is given sufficient power to become a threat to the polity. This follows from the agency condition inherent in civilization. We form communities precisely because we cannot provide for all our own needs and therefore must depend on other people or institutions.

A direct seizure of political power by the military is the traditional worry of civil-military relations theory and has been a consistent pattern in human history. Less obvious but just as sinister is the possibility that a parasitic military could destroy society by draining it of resources in a quest for ever greater strength. Yet another concern is that a rouge military could involve the polity in wars and conflicts contrary to society’s interests or expressed will. And, finally, there is the simple matter of obedience: even if the military does not destroy society, will it obey its civilian masters, or will its latent strength allow it to resist civilian direction and pursue its own interests?

Democratic theory is summed up in the epigram: the governed should govern. People may choose political agents to act on their behalf, but that should in no way mean the people have forfeited their political privileges. Most of democratic theory is concerned with devising ways to ensure that the people remain in control even as professionals conduct the business of government.

That is, civilians are morally and politically competent to make the decisions even if they do not possess the relevant technical competence in the form of expertise. This is the core of the democratic alternative to Plato’s philosopher king. Although the expert may possibly understand the issue better, the expert is not in a position to determine the value the people will attach to different issue outcomes. In the civil-military context, this means that even if the military is best able to identify the threat and the appropriate responses to that threat for a given level of risk, only the civilian can set the level of acceptable risk for society.

The military can propose the level of armaments necessary to have a certain probability of being able to defend successfully against one’s enemies, but only the civilian can say for what probability of success society is willing to pay. The military can describe in some detail the nature of the threat posed by a particular enemy, but only the civilian can decide whether to feel threatened and, if so, how or even whether to respond. The military assesses the risk, the civilian judges it.

Huntington’s theory survives despite repeated attempts to repeal it. Consider just 3 core claims that Huntington makes:

That there is a meaningful difference between civilian and military roles
That the key to civilian control is professionalism
The key to professionalism is military autonomy

Fusionists claimed that the emergence of the grand coalitional warfare of WW2 meant that military and civilian political functions would be hopelessly entangled ever after.

The military officer is promising to risk his life, or to order his comrades to risk their lives, to execute any policy decisions. The civilian actor is promising to answer to the electorate for the consequences of any policy decisions. The military officer is expected to obey even stupid order, or to resign in favor of someone who will. The civilian is claiming the right to be wrong.

Coups are not a particularly realistic possibility in this country. There has never been a serious coup attempt, let alone even a temporarily successful coup in the US.

At the same time, the US has avoided all the environmental conditions that foster coups in other countries: catastrophic defeat in battle, collapse of the civilian political order, persistent underfunding of the military coupled with cronyism and corruption in the military personnel system, and so son.

The first step in bridging the divide is recognizing that politics pervade civil-military relations even if there is no coup. Politics is about deciding who gets what and how, and even more about deciding a group of actors is going to decide who gets what and how. A coup is one extreme form of this, but far less drastic examples include deciding whether to allow gays to serve openly in the military, or deciding whether to intervene to stop a civil war and, if so, whether to begin with a measured application of airpower or to use massive force. And in the politics surrounding these choices, the American case is replete with interesting conflict.

Agency relationships are ubiquitous: for example, stockholder-board, employer-employee, voter-politician, or Congress-bureaucracy.

Roosevelt, Huntington claimed, gave too much power and authority to the military during WW2, letting it make all strategic policy decisions.

For Huntington, this was not so much a relationship of cause and effect as it was a definition: “A highly professional officer corps stands ready to carry out the wishes of any civilian group which secures legitimate authority within the state.” A professional military obeyed civilian authority. A military that did not obey was not professional.

There are 2 prominent strands of analysis within this literature. One argues that agents work when monitored and shirk when not monitored, and so the solution to problems of agency lies in devising the optimal monitoring scheme, using intrusive and nonintrusive means.

Another strand argues that monitoring is inherently inefficient and that optimal compliance comes from improving the quality of the agent and bringing the agent’s preferences more closely in line with those of the principal.

The contract is ritualized in the officer’s oath of allegiance and reinscribed through a myriad of cultural symbols, such as the privileged place assigned to the military in celebrations of national holidays.

As the external environment changes, for instance as threats to the state emerge or disappear, the civilians must revisit the problematique and make changes in the monitoring profile as needed. Of course, it is rather idealized to imagine civilians revisiting the civil-military problematique and deciding anew the question of delegation on every single security issue. In practice, the costs of reinventing the wheel probably make bureaucratic inertia (continuing the current pattern) attractive. Civilians still have the option, however, of changing the relationship on any issue; that they often choose not to simply underscores how they are sensitive to costs, as the agency model captures.

The problem in civil-military relations is not a lazy military — or at least this is not the only nor the most important problem. While individual military officers might be lazy or shiftless, the general problem is more complex than simply keeping the military industrious. In economic relationships, the principal and the agent have a different set of incentives regarding the basic work assigned to the agent: the principal wants lot of work for little pay, and the agent wants lot of pay for little work. In economic settings, its is not implausible to imagine that an agent would produce no work if he could get away with it. In the civil-military setting, this rarely makes sense. It is more reasonable to posit that both the civilian principals and the military agents want the same thing: security for the state. They can, however, disagree on how to provide that security, in general and especially in particular settings.

Working is doing things the way civilians want, and shirking is doing things the way those in the military want.

There is a difference between offering advice and “insisting that it is absolutely necessary to pursue a certain course of action if disaster is to be avoided” — where the latter might reach the level of shirking.

This is not to be confused with glory, or a desire for distinction, which connotes the militarism caricature. Rather, honor here captures the Aristotelian idea of getting credit for doing what is right, where right is acknowledged by others and defined according to some generalized conception of the good. The desire for honor derives partly from the basic human desire for legitimacy — the desire to gain peer approval. It is worth noting that even the most heinous scoundrels in history sought to justify their actions, however lamely and self-servingly.

In the civil-military relationship, the military preference for honor can be used to reinscribe the principle of civilian control; the military subordinates itself to civilians because, in a democracy, such subordination is recognized to be right and it would be dishonorable to do otherwise.

Senior army leaders believed that one of the lessons of Vietnam was that the army should never be deployed without the full support of the American public. Furthermore, the leadership believed that such support would come only if the civilian political leaders spent the political capital necessary to mobilize public opinion — and, finally, that civilian leaders would spend such political capital only if they were unable to deploy military forces any other way. In other words, the army believed that civilian leaders should not be given the opportunity to deploy combat troops on what appeared to be a “costless” mission; civilian leaders should commit to paying political costs, and those costs should be front-loaded.

Shirking by the US military usually takes 1 of 3 forms:

Efforts to determine the outcome of a policy calculus by giving inflated estimates of what a military operation would cost.
Efforts to determine the outcome of what a policy calculus with “end runs,” unauthorized public protest, leaks, or appeals to other political actors.
Efforts to undermine a policy through bureaucratic foot-dragging and “slow rolling” so that the undesired policy will never be implemented.

Likewise, some information is private to the principal. For instance, only the principal knows exactly know he judges various risks and how these judgments translate into preferences over outcomes.

The very act of hiring creates perverse incentives for the agent to misrepresent himself, which thereby increases the chances that the principal will hire a lout: it is hard to verify the true type, and the lout has a great incentive to appear even more attractive than a good worker.

Adverse selection could suggest that the civilians are liable to pick either people who are poor warriors (perhaps because they pick people like themselves) or good warriors who are likely to resent their authority. The decision to “hire” the military is revisited at regular intervals in decisions to promote or to fire certain individuals. Of course, as personnel proceed through the ranks the uncertainty should ease somewhat, since civilians gain more and more information on which to base such decisions.

Civilians could devise the strategy, deduce operations and battle plans therefrom, specify tactics to achieve those aims, outline logistics and equipment needed to accomplish all this, and direct the provisioning of the forces — in essence giving complete marching orders to the military. At the other extreme, civilians could simply tell the military, “Deliver us from our enemies,” and let the military decide all the rest.

The military can be monitored, therefore, by restricting the scope of delegation. At least insofar as military operations go, this type of monitoring takes the form of rules of engagement, standing orders, mission orders, and contingency plans.

Rules of engagement, then, are both a leash on the military and a information source for senior leaders, civilian and military.

Most economic applications stress controlling the agent through contracts that give agents an economic incentive to perform in ways that the principal wants; giving an agent a financial interest in the firm’s residual — that is, profit sharing — is a classic example. Political scientists have had difficulties applying this solution to political situations of agency because there is no obvious “profit” to be distributed. Profit sharing is straightforward when the organization’s basic output is easily measured and the main goal of the principal is greater efficiency. In political applications, the output (policy) itself can be in dispute or only imprecisely measured, and economic efficiency (more policy at less cost) is not the only desideratum.

Giving bureaus a fixed amount and allowing them to allocate it as desired is the most common use of slack in political organizations. The problem with using slack as a control mechanism is twofold. It requires that the principal consistently overpay for the service. And more important, slack does not solve the problem of ensuring that the policy output in fact accords with the principal’s desires and is not simply coming in at a lower cost.

Civilians promise autonomy to the military in matters of lesser import as an incentive for military acceptance of the ethic of subordination.

A slightly more intrusive form of monitoring involves using screening and selection mechanisms to ensure that only the right sort of agent enters into the contractual relationship.

By law in the US, civilians have a fair amount of influence over the makeup of the officer corps. Congress votes on all officer promotions, and the more senior and influential the promotion, the more carefully the application is screened by Congress.

The lower the rank of the military officers, the more frustrated they were with civilian viewpoints, and this suggests that promotion and selection procedures winnow out people who do not like civilian control, or at least that the views of those who get through the process are shaped by the organization.

In the civil-military context, an organizational norm that stresses obedience gives both civilians and the military a common expectation that the military will be subordinate. In this one respect, the principal-agent framework shares a finding of traditional civil-military relations theory: that a common culture of subordination — “professionalism,” “cult of obedience,” “norm of civilian control,” “ethic” — is a crucial component civilian control.

Interservice rivalry operates as another, slightly more intrusive fire alarm. Separateness of the American military services can bolster civilian control. Civilian intervention can be triggered when one service complains about another.

In the US case, the classic institutional check for the use of force is the Constitution’s division of military decisionmaking authority between the executive and the legislative branch; here, the framers of the Constitution treated the electorate as the principal and government as the agent. By extension, the presence of senior civilian officials in the DoD, over which Congress, through the Senate confirmation process, has some control, is also an institutional check. Likewise, the civilian staffs of congressional committees are another important institutional check, since they can block the activity of the executive branch.

The purpose of a separate militia (now the National Guard) was to be a last line of defense against the regular standing army, should it prove bent on a usurpation of control.

The role of the militia and the National Guard has declined more than the framers would have expected, however, and to a certain extent interservice rivalry can be thought of as a replacement check.

Pubic fire alarms like the news media also have another cost deriving from the fact that the civilian policymaking principal is himself an agent of a still more ultimate principal, the voter.

Police patrols include regularized audits and intrusive reporting requirements designed to turn up evidence of agent wrongdoing and, through regularized inspection, to deter moral hazard. Public investigative hearings and specific mandated reports and staples of congressional oversight and represent one of the more visible avenues of political control. Similar mechanisms operate within the executive branch to facilitate hierarchical control and some, like the Planning, Programming, and Budgeting System in the DoD, directly concern security issues.

These measures are costly in terms of civilian attention (not to mention dollars) but can mitigate somewhat the informational asymmetries in the civil-military relationship.

Contract incentives: Offer by civilians to use less intrusive monitoring in exchange for obedience.

Screening and selection:

Skill requirements for entrance into military.
Loyalty oaths.
Other accession instruments.
Professionalism.

Fire alarms:

The news media.
Defense-oriented think tanks.
Interservice rivalry.

Institutional checks:

Militia system and National Guard.
Interservice rivalry (sometimes).
Civilian staffs in Congress.
Atomic Energy Commission.
Confirmable civilian secretariat.

Police patrols:

Planning, Programming and Budgeting System and the budget process.
Civilian secretariat and Office of Secretary of Defense.
Restrictive rules of engagement.
Restrictive standing or mission orders.
Limits on delegated authority.
Audits and investigations.
Inspectors General.
Congressional Budget Office, General Accounting Office, Office of Technical Assessment.

Revising delegation decision: Intervening in a military operation to make a decision that was hitherto in the scope of delegated authority (e.g., picking bombing targets from the WH).

The military has been distinctive as an organization with a high reliance on coercion to enforce discipline — as in the famed Churchillian trinity of the British Navy: “rum, sodomy, and the lash.” Likewise, harsh punishments for battle cowardice was long an accepted means of forging an effective fighting force.

There are numerous examples of coups triggered by a civilian decision to punish some part of the military for an earlier disobedience. Such a coup effectively neutralizes any ability to punish.

The 3rd set of punishments, and by far the most prevalent, involves variations on forced detachment from the military — the military equivalent of firing. The military is distinctive as a profession with both an up-or-out career path and a very generous prize for those who stay in a “full” term, historically 20 years. The rewards for “making it to retirement” are substantial — a guaranteed pension equivalent to roughly 50% of one’s last base pay, payable immediately — and they are won after just 20 years of service. Moreover, the rewards climb exponentially with seniority.

There is more than a little irony in the fact that the military that prevailed against world communism is also the American institution that most closely approximates the idealized benefits of a socialist society. But these rewards accrue only to someone who makes it to retirement, that is, someone who successfully earns each successive promotion and so is not discharged before 20 years are served.

A slight variation on this involves failing to promote an officer. Officers serve with a time-limited commission. They become eligible for promotion by accession-year cohorts, according to a strict and linear time-in-service calculation. A promotion board of more senior officers decides which officers in a cohort will be promoted and which will not.

Not promoting an officer can thus be equivalent to firing that officer. Competition for promotion intensifies the more senior the officers get.

There are 3 administrative discharges: the honor discharge (under which all privileges are maintained), the general discharge (an intermediate discharge reflecting some minor disciplinary infraction, under which certain benefits may be maintained), and the general discharge under other han honorable conditions (a more severe dismissal, usually given in lieu of a court-martial, under which most benefits are lost).

And there are 2 punitive discharges that carry the stigma of a felony conviction, that can be issued only by a court-martial, and that deny virtually all benefits: a bad conduct discharge, and, the most severe of all, a dishonorable discharge.

As a point of departure I assume only 2 players, a civilian principal and a military agent. This assumption is off-putting to traditional civil-military scholars because, as we know, there are in fact multiple principals (the president, Congress, the SecDef, and so on) and multiple agents (4 services, the quasi-autonomous Joint Staff, more or less independent combatant commands, the National Guard, and so on).

How can we tell whether the costs of intrusive monitoring are less than the value of working minus the value of shirking, and so on?

Like all social science, agency theory can do no more than make a probabilistic claim — that given certain values for the equilibrium conditions, we expect that one set of outcomes is more likely than another set of outcomes.

It treats civilian control not as a once-and-done choice between Weberian ideal types but rather as an ongoing decision about how to monitor the delegation of responsibility to the military. It reflects the fact that civilians and military leaders confront the problems of agency on a day-to-day basis.

The navy is always the least intrusively monitored of military forces because connectivity is much harder to maintain.

MacArthur defended his behavior as justified by the officer’s oath, which was to defend the Constitution, not to obey the commander in chief.

It involved the full gambit of shirking, including foot-dragging obstruction within the administration, repeated attempts to revisit decisions that were made by the president, calculated leaks to the media, speeches and articles to public audiences, and even a carefully crafted subversive rewrite of the army doctrine so as to make it incompatible with the undesired national military strategy.

Even in the cases where the military could be said to have shirked, there is no documented case of the military refusing to use force when expressly ordered to do so. Rather, military shirking took the form of vigorous opposition before civilian principals rendered a final decision.

In the Central American case during the early Reagan administration, it appears that the army deliberately refused to develop contingency plans for military intervention so as to reduce civilian options and compel the president to decide against military action.

The decision whether to use force is binary, while the decision how to use force is a continuum. If civilians and the military disagree on whether to use force, it is relatively easy to see which side prevailed. The nature of a “how” decision is much harder to characterize and so it is much harder to ascertain whose preferences prevailed.

In peacetime, civilians are seldom exposed to the intricacies of military planning, and, in wartime, when civilian intervention in the details of military policy is much more likely, soldiers often interpret policymakers’s injunctions in ways that allow them maximum operational discretion.

For peacetime military activities, the Planning, Programming, and Budgeting System established by SecDef McNamara in the 1960s institutionalized a form of intrusive monitoring that gave civilians far greater access to information than had been available previously.

A nonintrusive monitoring regime involves the extensive use of indirect mechanisms of civilian oversight: contract designs, screening and selection mechanisms, and fire alarms that monitor the military and then alert the civilian principals when they detect shirking. These categories include such measures as offers by civilians to use less intrusive monitoring in exchange for obedience, skill requirements for entrance into military service, loyalty oaths and other accession instruments, the ethics of professionalism, the news media, defense-oriented think tanks, and even inter-service rivalry.

By the mid-1970s, the list of organizations focusing on military and foreign affairs had ballooned to include dozens of journals and a wide spectrum of organizations. By the end of the Cold War, the number of interest groups lobbying the federal government on defense and foreign policy issues was estimated at more than 900, and this does not even include the sizable military-expert community in academia. Compared with virtually any other country of interest, the US has enjoyed the most robust of “fire alarm” communities, with a vigorous, attentive, and relatively free press regularly reporting on the doings of the military and forcing the pace of civilian oversight.

The effect of security classification was to move a large portion of defense affairs out of the public eye. To some extent, the reliability of these monitoring systems was also undercut by the simultaneous rise of an aggressive public relations operation by the official defense department. In a sense, the efforts of the fire-alarm community to detect and report on shirking may have been countered somewhat by the ability of the military establishment to withhold and shape information.

The institutional model is the traditional paradigm of military service, in which effort is valued on normative grounds and members are motivated by intangible incentives like honor and duty to country.

The occupational model resembles employment in the civilian marketplace, where prestige is based on levels of compensation and members are motivated by tangible incentives like pay and benefits.

Huntington compares MacArthur with the generals who resisted Hitler during the 1930s, chastising both sets: “Both the German officers who joined the resistance to Hitler and General MacArthur forgot that it is not the function of military officers to decide questions of war and peace.”

A recurring theme in his discussion of Korea is the indecisiveness and vacillation of the civilian leadership at that time. Perhaps MacArthur’s low estimation of President Truman’s resolve vis-a-vis the Chinese translated into an equally low estimation of Truman’s resolve vis-a-vis MacArthur.

The constitutional framers’ desire to prevent abuse of power by an overstrong civilian executive led them deliberately to divide authority for military affairs between the executive and legislative branches.

The threat of being thrown out of the military — and in most cases of military shirking, forced departure from the service is the dominant form of punishment — may have been a sufficiently severe replacement for the more traditional physical forms of punishment.

The agency perspective begins with the premise that civilians have the right to be wrong. Civilian principals have the right to ask military agents to do something that ultimately proves costly, foolhardy, and even disastrous. Military agents have an obligation to advise honestly about the consequences of proposed course of action, but in the final analysis they must obey even dumb orders.

General Westmoreland complained that civilian efforts were made to tailor the forces sent to Vietnam in order to send nuanced signals to the North Vietnamese, without regard to the military mission the new reinforcements were meant to meet.

He suggests that the appropriate military response in the early days of the Vietnam war would have been to resign in protest; he accuses the military of being “over-professionalized” — more prepared to follow orders than to exercise independent professional skill and judgment. According to this view, the goals pursued by Kennedy and Johnson were fundamentally flawed. The chiefs were foolishly and unjustifiably loyal to President Johnson in carrying out his orders and forgot their higher oath of allegiance to the Constitution which, according to this view, obligated them to seek to work with Congress to thwart the administration’s policies.

The conventional wisdom is that intrusive monitoring proved very costly. A failure to delegate, which flowed from the civilian failure to trust the military for fear of another MacArthur episode, was directly responsible for the military’s failure to innovate tactically and thereby come up with a way to win the war in Vietnam. “In Vietnam, the military has been handicapped by civilians who won’t let them go all out.” “In wartime, civilian government leaders should let the military take over running the war.”

Our problem was not so much political interference as it was a lack of a coherent military strategy.

The dominant military lesson from Vietnam is that there was too much civilian control during the war, not too little. Even the military’s insistence on the need for public support before embarking on a risky operation, or the need for clear goals (exit criteria, in modern parlance), are in fact veiled criticisms of the ways civilians ran the Vietnam war. The military learned that civilians cannot “stick it out” over the long haul, that civilian interference produces disasters and ties the military’s hands, and that civilians do not understand the proper use of force. It is striking the extent to which the Vietnam experience “taught” the military to doubt the wisdom of submitting to civilian control.

The goal civilians pursued, the substance of “working,” was to preserve South Vietnam without conquering North Vietnam — in other words, an avoidance of defeat rather than a quest for victory. This was quite clearly what civilian leaders wanted, and they refused to pay for anything more. It was the policy the civilians asked for and, by and large, it was the policy the military delivered. So far as civil-military relations go, civilians have a right to be wrong. This time they were.