Taking a “Pro” Position on Principled Resignation

I come at the issue of principled resignation of senior military leaders from several perspectives: that of moral agency and consistency, of effectiveness, and of “using” lives well. Below I explain what I mean by each of these perspectives, then address some common criticisms of the position that I take.

Every one of us is a moral agent. Unless one is handicapped in a way that incapacitates one’s judgmental capacity or prevents one from acting based upon judgment, part of what it means to be a human being is to be free to make moral decisions about right and wrong and to be held accountable for these judgments and the actions which flow from them. Of course, no one is completely free; everyone’s freedom is properly limited in important ways. So the practical questions are usually what degree of freedom exists for a person in a specific case, and how should that person exercise his or her freedom responsibly.

American’s hold soldiers and their leaders responsible, for example, for the decisions and actions they take even in the midst of battle. Soldiers and leaders never become “mere instruments” or “killing machines.” In combat, the hardest decisions are often made by those on the spot, under the harshest conditions, and the highest risks. Certainly, some circumstances might mitigate judgment, explaining why we might accept behaviors in combat that would be unacceptable in other situations (Walzer, 2015, pp. 36, 40, 304–306, 311). These mitigations, however, prove the rule: Even in combat, we expect soldiers and leaders who fight on America’s behalf to remain moral agents. The reactions to the My Lai Massacre in Vietnam, the abuses in the Abu Grahib prison, the killing of civilians in the Iraqi city of Haditha, the “kill team” murders in Afghanistan, and the Marines urinating on Taliban corpses—all highlight the rule not the exception. We not only held responsible those who committed these acts but also their leaders who knew of these actions but did nothing to prevent them or said nothing afterward (Dubik, 1984, pp. 39–47). We may understand the difficulty, uncertainty, and urgency under which soldiers and leaders make difficult decisions and take actions, but this understanding has limits and does not erase the expectation of moral agency. Such is the tactical dimension of moral agency, but it has a strategic dimension as well.

If we hold that soldiers and leaders remain moral agents on the battlefield, we must also hold their seniors equally responsible for decisions and actions taken in higher headquarters or in the capital. Simply put, when the government issues stripes or bars to combat leaders, those accepting these promotions do not exchange their moral agency for their new rank. The same holds true for stars. Certainly, the conditions under which a general or flag officer exercises his or her moral agency differ from the sergeant and the captain. But equally certain is this: Moral agency is expected of the general just as it is for any other soldier or leader.

The exercise of moral agency is also important to any organization’s effectiveness—whether that organization is a corporation, a military unit, or a government bureaucracy. Albert Hirshman presents a very compelling case for such utility in his book, Exit, Voice, and Loyalty (Hirschman, 1979). “Voice” is the feedback that customers, partners, employees, and executives provide to help an organization understand in which areas it is doing well and in which it is not. Voice provides the grist for an organization’s self-analysis and subsequent adaptations which, in turn, assure improvements to the organization’s efficiency and effectiveness (Hirschman, 1979). Military units use after action reviews to elicit feedback. All high-performing organizations employ similar formal and informal feedback mechanisms to improve organizational behaviors (Hirschman, 1979, pp. 30–43).

Voice is especially important in the relationship between civil and military leaders who are coresponsible for making weighty war-waging decisions. Such decisions concern not only whether the nation should go to war or deploy its forces, but also how those forces are employed and for what reasons. Senior civil leaders rightly have the final decision authority as to political aims as well as military and nonmilitary strategies, policies, and campaigns necessary to achieve those aims. But those senior military leaders who are in dialogue about the efficacy of the final decisions are coresponsible for both the decisions-making process and its outcomes.

Coresponsibility was at play when General Harold K. Johnson, the Army Chief of Staff in the Johnson administration during the Vietnam War, considered resignation. In describing his situation, he said, “We had made our recommendations …. Our advice had been rejected and other courses of action were chosen, so we simply were good soldiers, and did what we were told to do” (Sorely, 1998, p. 268). Subsequently, General Johnson considered resignation but in the end did not resign. Assessing his own behavior after he retired, the General says, “I made the typical mistake of believing I could do more for the country and the Army if I stayed in …. I am now going to my grave with that lapse in moral courage on my back” (Sorely, 1998, p. 304).

General Johnson knew that the strategies and policies in Vietnam had deteriorated to the point of wasting lives in a war that could not be won at least not in the way it was fought. Further, he knew that he, his fellow service chiefs, other senior civilian officials, and the president were all partially responsible for that situation. The president made the final decision, no doubt about that. Part of General Johnson’s moral anguish, however, came from his realization that he was complicit, and he had not done enough.

The purpose of this civil–military relationship, especially in the dialogue associated with the conduct of war or use of force, is to identify the most prudent solutions to the problems that the seniors face not merely to demonstrate who is in control of whom (Dubik, 2010, 2015). Voice is a very important part of this dialogue, for it ensures that the one who makes the final decision hears what he or she needs to hear not just what he or she wants to hear. Voice is sometimes very disturbing and difficult to deliver, but every participant in the dialogue leading to the final decision must exhaust all voice options and avenues to ensure all aspects of a problem and all dimensions of risk are understood before a decision is made (Cohen, 2001, pp. 208–204).

George C. Marshall scholar Mark Stoler recalls, for example, serious disagreements, even what he called “blowouts,” among senior political and military leaders over multiple issues (Conversation with the author, June 14, 2016). The debates over whether victory through airpower was possible, whether to invade North Africa or France in 1942, balancing operations and supplies between Europe and the Pacific as well as balancing equipment for U.S. forces and that sent to allies, priorities for industrial output and raw material usage, and manpower allocations to the armed forces and industry—all involved multiple, and sometimes very heated, civil–military exchanges. These are examples of voice at its best.

Some organizations can be deaf to criticism, however, some avoid important-but-uncomfortable realities, and some are blind to facts that don’t fit biases. Other organizations simply deny they have any deficiencies or ineffective strategies. Still others co-opt those who dissent or present alternative views. Anyone who has worked in a bureaucracy quickly becomes familiar with all of these behaviors. In these kinds of circumstances, “exit,” or resignation, may be the only way to get the organization’s attention. Exit may be the last resort to demonstrate that something is seriously awry or that ground reality doesn’t match senior leader reality. And exit may be the only alternative to salvage one’s integrity and not “go along” with some serious wrong.

This was the kind of situation that General Johnson was reflecting upon when he concluded that he made a mistake in not resigning. Voice, in any real sense, was mostly absent in the administration of President Lyndon B. Johnson. President Johnson openly distrusted his military advisors, sought to keep the chiefs from opposing his Vietnam policy, concealed the finality of his decisions from senior military leaders, and remained ignorant of the joint chiefs’ actual opinions (McMaster, 1997, pp. 50, 84, 89). Further, the chiefs themselves served more as technicians for planning that was actually conducted in the Office of the Secretary of Defense “than strategic thinkers and adviser in their own right” (McMaster, 1997, p. 147). In sum, exit may be the only way to help an organization to focus on what it would prefer to ignore even to its own detriment (Hirschman, 1979, pp. 21–29, 44–54, 106–128; Felice, 2009, pp. 1–124, 179–198).

An organization that limits voice or exit, whether overtly through a matter of formal policy or more subtly through culture or innuendo, may make it easier on itself in the short run but not better in the long run. Certainly there are institutional risks associated with both voice and exit, especially exit. But there are also institutional risks associated with remaining silent or remaining in position in the face of serious wrongs or serious, sustained, and avoidable ineffectiveness. Judging what decision to make amid conflicting risks is what moral agency is all about.

Finally, institutional risks are not the only things at stake; the highest stake is life itself, for life is the coin of war’s realm. War risks, damages, changes, and often ends lives: the lives of the innocent who are always caught up in every war and the lives of the citizens who become the warriors who fight on behalf of their nation. War can also substantially alter the life of the political community itself, even if the war does not involve an existential threat. The set of national security organizations involved with waging war, as well as the set of civil and military leaders who head these organizations, form the very heart and brain of wartime decision-making. These leaders should set a very high bar to ensure the lives they use are “used well” not used poorly or wasted.

In all democracies, political leaders retain final decision authority on all significant war-waging issues—war aims, strategies, policies, and campaigns. Military leaders, however, are coresponsible to make sure those final decisions are as good as is possible given the vagaries and realities of war. This is the failure that General Johnson saw in himself.

In war, even when things go well, people die and lives are destroyed—true on the battlefield and true in the capital. When the stakes are this high, all Americans should raise the bar for proper use of moral agency for all those involved. Having the right to final decision authority does not entail a right to be wrong, no one has the right to be wrong in using another’s life. Rather, the right to final decision entails an obligation to use that authority carefully and properly. Mistakes made when this obligation is met can be understood, and mistakes made when this obligation is avoided, disregarded, or obstructed fall into a different moral category.

Having the final decision authority and using it properly are two dimensions of the same moral phenomenon. “Proper use” involves setting a climate in which the voice of senior political and military leaders is the norm for achieving sufficiently prudent and effective decisions. Proper use must also allow exit under the right conditions, even if as a last resort. Neither voice nor exit can be allowed to threaten a democracy’s civil control over its military. And neither can be allowed for petty nor petulant reasons. When a senior military leader is involved, neither voice nor exit are appropriate when merely cover for matters of service or other parochial advantage. As the General Johnson case shows, both options must be reserved for substantive and important issues that correspond to the weightiness of the decisions at hand and the lives at stake when those decisions are executed.

So my position is that principled resignation of a senior military leader, a form of exit, should be permitted, always as a last resort, only under narrow circumstances and never in a way that threatens civil control.

Of course not everyone agrees (Feaver, 2017; Kohn, 2017). I’ll address six of the more common counter arguments. The first two of which involve issue of role differentiation. One declares that the electorate, not a subordinate appointed official, holds the final decision authority—the president in the case of the United States—responsible for his or her decisions. This counter, however, assumes that the quality and substance of any decision is separate from the decision itself. To hold a person responsible for a decision requires assessing both the process of making that decision and the quality of the decision. Those involved in the process of making important war-waging decisions are coresponsible to ensure the process is as thorough and objective as possible, for the process often determines the quality of the decision.

To allow this counter-argument is to allow a senior military leader involved in an important decision to say, “The quality or substance of a decision is not my responsibility. I merely provide input and advice then obey regardless of the outcome.” In describing General George C. Marshall’s relationship to President Roosevelt, David Kaiser says, “Marshall’s personality did not allow him simply to accept FDR’s demands and let the President take the consequences” (2014, p. 245). Marshall, in testimony before the Senate Committee on Armed Services and the Committee on Foreign Relations on May 7, 1951, says that the joint chiefs, the secretary of defense, and the president “are responsible for the total security of the United States.” Marshall never doubted that the president held final decision authority, but he also never doubted that “it becomes the hard duty of those responsible to consider what is the wisest course to follow in such matters” (Stoler & Hold, 2016, p. 504). When speaking of wartime responsibility, Marshall uses the plural, not the singular—revealing his sense of personal responsibility with respect to the quality of the decision at hand and the actions that flowed from that decision.

The president’s role is strongly differentiated because his or her position includes the authority to make final decisions. This differentiation does not, however, extend to the decision-making process itself or the quality of its product. Here, the roles of those who participate are much less differentiated. In fact because neither the civilian nor the military leader has a corner on relevant knowledge with respect to complex war-waging issues, both perspectives are necessary. Thus, they are functionally equal in this regard. And they are morally equally in that both are coresponsible for the outcomes of the decisions they reach, for both civil and military leaders must execute those decisions.

The other counter in the role differentiation category states that the scope of the president’s responsibility is wider than any military leader regardless of rank or position. Of course, this description is accurate. The president may chose, rightly, an alternative being considered for reasons other than military. History is replete with such choices. Franklin Delano Roosevelt’s (FDR) overrode Marshall, for example, in the decision to invade North Africa in 1942, and key in the president’s calculations were domestic political considerations. By the time a senior military leaders reaches a position that is involved with strategic war-waging decisions, he or she understands fully this dimension of strategic decisions. One of the essential purposes of the civil–military dialogue at this level is to allow the most prudent courses of action—all things considered, military and nonmilitary—to emerge or at the very least the most imprudent to fall from consideration. At the strategic level, any one of the nonmilitary considerations—whether diplomatic, fiscal, economic, or political—is as legitimate a factor as are military considerations. Marshall speaks directly to these factors. He says,

we probably devoted more time in our discussions, our intimate discussions of the American chiefs of staff, to such matters [political and other nonmilitary aspects] as any other one subject, because we were very fearful that we might find our campaign upset by some political gesture.

Marshal claims that the balance of military and political issues was “a delicate balance back and forth” (Bland, 1996, pp. 414–415). Further, everyone involved in decision-making at this level realizes, or should, that he or she could be wrong—that’s why the back and forth. So, saying the scope of the president’s responsibility is larger than that of any military leaders is not to present an objection to principled resignation but to describe the conditions within which a strategic decision is made.

Common Objections 3 and 4 concern the nature of morality. According to the third objection, principled resignation should not be permitted because morality has no substantive content, it is simply “gut feeling.” Such an extreme position motivated Michael Walzer to write Just and Unjust Wars in 1977. In talking about the Vietnam era antiwar movement, he says,

We suffered from an education which taught us that [morality] had no proper descriptive use and no objective meaning. Moral discourse was excluded from the world … of social science. It expressed feelings, not perceptions, and there was no reason for the expression of feeling to be precise. Or rather, any precision it achieved had an entirely subjective reference. (Walzer, 2015, pp. xi–xii)

Walzer’s book ultimately demonstrates that we do argue about the moral dimension of war and that such arguments are not merely descriptions of the feelings or state of temper of those arguing. Those arguing—whether common citizens, political elite, military leaders, or representatives of the media—adduce evidence and present reasons in a mutually comprehensible way. They attempt to argue in defense of alternatives during a decision-making process as well as explain and justify decisions made. Further, Walzer shows that these arguments can be studied and evaluated as to their cogency (2015, pp. xiii and xv). One of Walzer’s conclusions applies to the public discussion concerning America’s post 9/11 wars: “The moral world of war is shared not because we arrive at the same conclusions … but because we acknowledge the same difficulties on the way to our conclusions, face the same problems, talk the same language” (2015, pp. xiv–xv). Simple put, morality involves far more than “gut feelings.”

The fourth counter position acknowledges morality’s legitimacy but claims war requires two moralities. One applies to decisions and actions on the battlefield, and the other to decisions and actions in the capital. Soldiers and leaders on the battlefield have a choice, according to this argument, whether to follow an order or not. Moral agency at the tactical level is not in question. At the strategic level, however, this position holds that senior military leaders do not have the choice to resign. In its starkest form, the claim is that principled resignation is unethical: The choices at the strategic level are only advise, protest, and obey. Voice, yes; exit, never.

Exit for matters of principle, according to this view, challenges civilian control of the military because exit de facto opposes the legitimate authority of civil leadership. This position again conflates having the right of final decision authority with the quality of the decisions that result from use of that authority. One may object to a specific decision without questioning the underlying legitimacy of having the right of final decision authority. When considering resignation, for example, General Johnson was not questioning the authority of the president. He was questing the wisdom of the decisions.

Certainly, senior leaders incur additional institutional responsibilities upon assuming their positions, responsibilities that do not exist for their juniors at the tactical level. And in this sense, the fourth objection has some validity. Stuart Hampshire captures this difference in his essay, “Public and Private Morality,”

… the assumption of a political role, and of powers to change men’s lives on a large scare, carry with them not only new responsibilities, but a new kind of responsibility which entails, first, an accountability to one’s followers, secondly, policies that are to be justified principally by their eventual consequences, and thirdly, a withholding of some of the scruples that in private life would prohibit one from using people as a means to an end and also from using force and deceit …. In addition to moral seriousness in justice and fairness … in private life, there is political seriousness in the user of power [in public life.] … In particular, it requires priority for duties of careful and responsible calculation [i.e. decision-making] [and] for the virtues of prudence and ‘cleverness,’ cleverness being that Aristotelian component in moral virtue which political responsibility demands. (1978, p. 52)

Senior military leaders do not assume a political role in the same ways elected and appointed civil officials do, but military leaders participate in the exercise of the political role of others. In accepting their position, therefore, the senior military leader acknowledges that he or she is accepting new kinds of responsibilities. Political seriousness, as Hampshire uses the term, as well as prudence and cleverness, leave room for individual conscience. One does not become a “mere instrument” at the strategic level any more than one does at the tactical. The “I was just following orders” justification has its limits whether one is a private or a general.

Marshall acknowledged this limit during testimony on May 8, 1951, then Senator Styles Bridges (Republican, New Hampshire) asked Marshall what he would have done if a difference with President Roosevelt “had become an absolute difference of opinion, where you thought the best interests of the country were not being served by the President or the administration?” Marshall responded, “I probably would have resigned as Chief of Staff” (Stoler & Holt, 2016, p. 510). Speaking of the FDR administration during the war, David Kaiser observes that the “freedom with which senior officials spoke their mind even in opposition to their own administration’s current policies was a remarkable tribute to the mutual confidence between Roosevelt and his subordinates” (2014, p. 238). In this case, a robust voice precluded the necessity of exit. But Marshall’s response to query shows that he kept exit as a possible alternative.

The public life of a senior political or military leader is a life that includes moral hazards. Public life sometimes involves, even requires, taking morally disagreeable decisions and actions. War itself, whether on the battlefield or in the capital, is a realm where the morally justified and the morally repugnant can coexist in the same act. A “certain level of roughness is to be expected by anyone who understand the nature of the activity [whether politics or war]” says Bernard Williams in “Politics and Moral Character” (1978, p. 62). But conscience is not absent or forbidden in public life for (a) even obligations to the state have limitations, (b) the exercise of power “is a form of individual expression, not diminished but enhanced by the institutions and offices upon which it depends” (Nagel, 1978, p. 77), and (c) conscience is an essential element of character.

The fifth and sixth arguments against principled resignation concern potential costs. The fifth objection says that the cost of a senior leader who resigns is the abandonment of the soldiers for whom he or she is responsible. Soldiers cannot leave, the argument goes, so their senior leaders cannot either.

If exit is exercised as the last resort, however, the cost is not about abandoning one’s soldiers. Rather, such a choice is better be understood as a sacrifice for them or for the nation. The cost incurred by the senior leader may be loss of position, perhaps forced retirement or formal judicial or nonjudicial action that results in “fading away” once the media surge ran its course. Personal cost, however, is outweighed by other considerations.

A Civil War example addresses this objection. After President Lincoln was assassinated and Andrew Johnson became president, Johnson “told Grant the he intended to hang those who had served as general officers in the Confederacy, starting with Robert E. Lee.” The president and Grant had a sharp exchange with the president wanting to know exactly when these men could be tired. Grant said “never so long as they do not violate their paroles.” The president “refused to accept that, thinking that because Grant was under his orders he was under his thumb.” Grant then “threatened to resign. Johnson blinked” (Perret, 1997, p. 386).

Of course Grant knew that his popularity, much greater than President Johnson’s, which mitigated the personal risk associated with his threat to resign, but foremost in his mind were other costs: those to the nation, to his soldiers, and to the innocent who would suffer should President Johnson act on his desire. To put it mildly, the country was in significant turmoil with the real possibility that the South might resort to guerrilla warfare. Grant simply could not claim “I gave my advice, and the president decided,” for he knew that he would bear some of the responsibility for letting that decision get made and for the consequences that would surely follow. Far from abandoning his soldiers with his threat to resign, Grant showed his loyalty to them and to the nation.

The last objection involves a more complicated argument about the cost of exit. Here, the cost is to the military institution itself and to the civil–military relationship so necessary to prosecute war. The argument is that even thoughts of resignation, let alone threats or actual resignation, could increase the tendency of civil leader to vet candidates for senior military positions to ensure their loyalty to whatever decision a senior political leader takes. This fear of resignation could force civilians to distance themselves from their military advisors, thus decreasing the likelihood of proper military participation in important wartime decisions. The result, according to advocates of this counter argument, could be an overall reduction of the quality of national defense decisions.

These potential costs are significant and real. Anyone considering resignation must take them seriously. They may, under certain circumstances, be reasons not to resign. But a deeper look at this objection reveals something quite disturbing: The foundation upon which this objection rests is a belief that civilian leaders neither understand nor respect their military subordinates and must rely, therefore, on control and on defining loyalty as absolute obedience. Such a foundation suggests a troubling pathology concerning the U.S. civil–military relationship.

The causes for that pathology may have its roots on both the civil and military sides of the equation. Too many generals and admirals, for example, believe that in war, their only job is to execute the policy maker’s war aims, requiring of them only an “end state” and the time and resources to accomplish the job. This belief is mistaken. And too many political leaders believe, falsely, that in war, their job is to make policy decisions, getting only episodic input or options from military leaders who are excluded from policy discussions. The military false belief over emphasizes autonomy, and the political over emphasizes power, control, and compliance. While convenient because each can “blame” the other when things go wrong, such attitudes are based upon a false understanding of the purpose of the civil–military dialogue: discovering and implementing the wisest and most prudent course of action of those actually available. This is the kind of output that should come from a proper civil–military dialogue if war is to be waged successfully—and it takes the right leadership environment and both civil and military leaders to have this kind of dialogue.

The pathology that this last objection to principled resignation identifies is an important problem, but its solution—to reject even the thought of resignation—is misaimed. The real solution lies in insisting on more common understanding, more mutuality, and more collaboration. The question of exit would rarely, if ever, come up if the participants in the civil–military dialogue better understood how important and necessary each other’s perspectives were and better understood the real purpose of the dialogue. The FDR–Marshall relationship is a positive example of voice precluding the necessity of exit and the Lyndon Baines Johnson–military relationship, a negative one. It does not seem too much to expect that both senior political and military leaders act toward each other in ways that recognize that both are sworn to protect and defend the constitution and that the American people hold both responsible for the lives and resources used in war. America has always been better served when such a relationship exists, and it has been ill served when it does not. The more logical and effective way to end the discussion about principled resignation is not to deny moral agency but to create an environment from which a proper civil–military relationship can arise.

Principled resignation of a senior military leaders is a complex, thorny, and risk-ridden affair. It should be. But its difficulty should not become a reason to deny its necessity. Moral agency, consistency, effectiveness, and the significance of “using lives well” all argue that, while it may be a rarely used final resort, resignation plays—or might play—a critically important role in waging war well.