Speaking Out: Why Retired Flag Officers Participate in Political Discourse 

Few other issues in U.S. civil-military relations are as controversial as the political activism of retired officers. Much of this modern-day controversy dates back to the 1988 presidential election, when former Marine Corps Commandant Gen. P.X. Kelley endorsed then-candidate George H.W. Bush in a primary, followed by former Chairman of the Joint Chiefs Adm. William J. Crowe, Jr. endorsing then-candidate Bill Clinton in 1992. Since that time, endorsements have been regular occurrences during election cycles. Events such as the “Revolt of the Generals” in 2006, in which six former generals criticized Secretary of Defense Donald Rumsfeld for his management of the Iraq War, further focused attention on retired officer speech.1 During the Trump administration, officers commented about everything from President Donald Trump’s policies with regard to the Middle East, to the need to keep the military out of domestic politics, to Trump’s leadership style.2

Yet, for all of the controversy that such events generate, we know relatively little about how retired officers, especially those at the senior level, actually think about the decision to comment publicly about political matters. Today, retired general and flag officers’ beliefs about norms are mostly inferred from survey questions about how officers view different political behaviors. Only rarely have the motivations for speaking out been studied beyond the occasional anecdotal response given in an interview.3 Even rarer are efforts to systematically compile their views in an anonymized survey setting. Do these former military leaders perceive informal rules — norms — about how they should engage in and relate to politics? What drives them to speak out or refrain from doing so?

This article seeks to address these questions through a unique survey of retired flag officers that asks respondents to report directly on their own views about the informal rules governing their engagement in politics. In carrying out our survey we examine evidence related to “the degree of ‘verbal’ acceptance of a norm’s claims by its addressees,” which is a key indicator of a norm’s robustness.4

The responses reveal a layered set of considerations that shape retired officers’ views of the norm governing whether to speak out in partisan contexts. We find evidence that retired flag officers do perceive such a norm and share views on the outermost contours of that norm, which designates some behaviors as off limits. While some respondents may question whether retired officers should be subject to constraints on making public commentary, the retired officers we surveyed clearly believe they should maintain limits on their partisan speech and actions. This reveals that the norm operates as a baseline that influences how military leaders think about the tradeoffs involved in speaking out.

Yet, we also find that retired military leaders have different interpretations about the boundaries of the norm, what counts as a violation, and, even more pointedly, whether it is at times justified to violate it. Moreover, while those in the survey report that they are aware of social pressures bearing on their decisions to speak out, they explain their actual choices as stemming from highly personal assessments about the wisdom of doing so. Our respondents indicate that external pressures from their social networks made up of other flag officers have had limited effects on their decisions to speak out or remain quiet. Nor do reflections on the instrumental value or efficacy of shaping debate seem high on the list: Whether their actions will actually change anyone’s mind or shift the debate does not seem to be a salient factor in weighing the costs and benefits of speaking out.

Instead, retired flag officers frame these as personal decisions in which they navigate their competing obligations to protect the institution and its members, act in the spirit of their oath to the Constitution, and stay true to their own moral compass.5 To be sure, such self-reporting on one’s motives may reflect some social desirability bias and may not reveal the full array of factors involved in the actual decision-making calculus of our respondents. Still, it is revealing that they expressly disavow social pressures as a main factor in their decision-making and instead emphasize their own volition and moral imperatives.

This article has important scholarly and practical implications for the study of civil-military relations and norms, more broadly. Empirically, the paper contributes by directly examining retired flag officers’ thinking about norms, rather than inferring them from surveys about their opinions on other issues. In examining their self-reported rationales, we thus learn how the actors who actually enact norms think about them and the factors that compel them to conform, or not, with their tenets. Specifically, we show that these officers perceive them and feel bound by them. Norms may be unspoken and unacknowledged, but findings from the survey suggest that the agents to whom they apply often likely perceive and reflect upon them. In the terminology of norms scholarship, there is a large degree of “concordance” about the salience of nonpartisan norms among retired flag officers,6 such that they exhibit high “validity.”7

At the same time, while there is widespread agreement among scholars that some actions by retired senior officers are off limits, flag officers reveal some variation in what behavior they feel is appropriate and what is not consistent with their professional obligations. There is diversity in how retired military leaders perceive their relationship to politics and political engagement and the boundaries of the norm of non-involvement in partisan politics. By documenting these viewpoints, we thus also add nuance to the debate about political and partisan behaviors in civil-military relations.

The analysis also has lessons for scholarship on norms more broadly. In particular, the findings below illustrate the elasticity of normative interpretations and the complexity of assessing norm violations. One lesson of our findings is that, while outsiders might identify an action as a violation of a norm, agents of those norms might see it differently. In other words, while much political behavior by retired flag officers is seen by scholars as violating a norm, that conclusion rests on a particular interpretation of the standards required of that norm.

We conclude that, both on behavioral grounds and in terms of attitudes expressed by military leaders, norms against partisan activism by retired military officers are very much contested.

In addition, the analysis illustrates the ambiguity inherent in many norms and how that complicates efforts to assess their robustness. Scholars of norms at times treat their tenets as singular and self-evident, with clear implications for the behavior and beliefs of the norms’ adherents. But our analysis suggests that norms can be vague and have poorly defined boundaries. This can create uncertainty about whether norms apply in a given situation and can justify varied interpretations of their boundaries. This insight about poorly specified norms, in turn, contributes to scholarly debate about whether contestation over how norms should be applied is norm violating, or norm enforcing.8 Our findings suggest that ambiguity promotes both tendencies. It does so by encouraging debate among the subjects of norms about what scholars refer to as their “applicability” in a given situation. Such debate supports normative constructs by reinforcing their salience as a basis for action. At the same time, debates over applicability also undermine norms by facilitating leniency in abiding by them among the norm’s subjects. This helps explain the seeming contradiction that retired flag officers both endorse norms that proscribe partisan political activity, but also at times think that actions with clear partisan implications fall within the boundaries of acceptable behavior.

This article proceeds as follows. First, we provide some background to normative arguments about retired officer speech. Scholars continue to debate whether retirement releases individuals from constraints on political activity.9 We review these debates as background and context for our analysis of the attitudes of retired officers toward partisan speech and activity. Second, we offer a conceptual framework that provides a template for assessing the degree to which a norm against political speech and activity by retired officers is robust, contested, or approaching “death.” This framework includes both observable behavioral indicators that this norm exists, as well as motivational elements related to officers’ attitudes about the appropriateness of that activism. We look first at what is known at present about the status of this norm, focusing on the period of the All-Volunteer Force. This section contends that, since the 1990s, there has been a steady movement away from a consolidated norm against retiree political activism — that is, whatever analysts might think about whether there should be a norm, in reality there is diminishing evidence that one exists.

Next, we turn to the empirical contribution of this article, which assesses key components of our framework related to how senior officers view the norm of non-involvement in partisan speech and activity. The approach, as noted above, is to evaluate the discursive support for this norm given by those subject to it — in this case, military leaders. We describe the methodology we used to survey nearly two dozen retired flag officers on the topic. Fourth, we report key findings from our survey, including the flag officers’ thoughts on the appropriateness of speaking out politically and the particular factors they weighed when considering whether or not to engage in public, political discourse. We then return to the framework and provide an overall assessment of the health of the norm against retired flag officers engaging in political speech today. We conclude that, both on behavioral grounds and in terms of attitudes expressed by military leaders, norms against partisan activism by retired military officers are very much contested. The article closes with a brief discussion of the implications of our findings for military professionalism and civil-military relations.

An Unsettled Norm

In civil-military relations, a trifecta of laws, regulations, and norms govern partisan activity by active-duty military officers. In retirement, officers remain bound by certain Department of Defense regulations regarding political activity, although these tend to be loosely enforced. Retired officers continue to be subject to the Uniform Code of Military Justice and theoretically could be charged with violating Article 88, which prohibits officers from using contemptuous words against the president and other elected officials, although many observers have pointed out that the likelihood of this occurring is very low. Retired officers who run for elected office also remain subject to provisions of Department of Defense Directive 1344.10, Political Activities by Members of the Armed Forces.10 These provisions pertain to the use of rank and former military status in campaign advertisements. However, this regulation goes to great lengths to encourage all servicemembers to “carry out the obligations of citizenship,” and there are few formal restrictions on the political speech of retired officers. In contrast, Department of Defense Instruction 1000.32, Prohibition of Lobbying by Former DoD Senior Officials,11 subjects retired flag officers to a one-to-two year cooling off period during which they are prohibited from lobbying their former military service. No such Defense Department regulations exist regarding political speech or partisan endorsements by retired flag officers.

Despite the regulations cited above, once individuals retire, their behavior regarding political activity is governed largely by norms — social conventions about what behaviors befit a former officer.12 Among the most important of these norms is the ethic of nonpartisanship. Few in or out of uniform question the necessity of the norm of nonpartisanship for active-duty servicemembers. Subordination to civilian authority requires the military to avoid institutional involvement in partisan politics or even the appearance of it. However, the appropriateness of retired flag officers participating in public, political discourse is a topic that sparks debate among both scholars and practitioners.

Normative arguments that proscribe retired flag officers from speaking out on partisan issues are generally made on three grounds. First, many observers note the unique position that they hold as the elite, senior-most leaders in the military — especially retired four-stars, whom Richard H. Kohn refers to as “princes of the church.”13 The public political pronouncements of a retired four-star are different from those of a retired lieutenant colonel. Although some note that retired flag officers who draw a military pension and remain subject to the Uniform Code of Military Justice maintain formal ties and obligations to the military, even in retirement,14 most who argue against the propriety of retired flag officers engaging in partisan speech and making endorsements do so from a normative standpoint, not a legal one. That is, they do not hinge their normative argument on the technical detail that retired flag officers are still bound by military regulations. Instead, they argue that retired flag officers have a responsibility to avoid using their rank to interfere in partisan politics because of the special role they play in guarding the reputation of the profession they served — a role that is not severed in retirement.15

Some argue this exchange of military prestige for political power tarnishes the reputation of the military profession and threatens to erode the public’s high confidence in the institution — the very factor that compelled politicians to seek out such endorsements in the first place.

The second component of the normative argument against retired flag officers speaking out on partisan political issues is the perception that they still speak for the institution, and therefore, their involvement in partisan political speech could be perceived as reflecting the current sentiments of an ostensibly nonpartisan institution. Recent survey research has corroborated some of these concerns, showing that most Americans fail to draw clear distinctions between active-duty and retired flag officers, and that most assume that retired flag officers reflect the views of those serving on active duty.16 Scholars and practitioners who advise against retired flag officers publicly airing their partisan political beliefs voice concern that doing so will cause the American public to view the military as a partisan actor and will cause the legislative and executive branches to view the military as suspect, thereby further weakening civilian control.17

The third component of the normative argument against retired flag officers speaking out pertains to the exploitation of military service for political influence or gain. Jason Dempsey has termed this “the paradox of prestige,” whereby politicians seek endorsements from retired officers in order to leverage the esteem associated with the military’s high public confidence levels.18 Some argue this exchange of military prestige for political power tarnishes the reputation of the military profession and threatens to erode the public’s high confidence in the institution — the very factor that compelled politicians to seek out such endorsements in the first place. It is the specific act of partisan campaign endorsements — in which retired flag officers put their rank, name, and military service behind a candidate for elected office but face no electoral accountability themselves — that critics find most troubling.19 Those who find fault with partisan campaign endorsements often argue that if retired flag officers want to fully enter the political process, they could run for office themselves and face the full scrutiny of the American electorate.20

These proscriptions against retired officers engaging in domestic politics have been commonly cited by many scholars of civil-military relations, as well as publicly referenced in senior officers’ public remarks. Still, although the position is less commonly argued, some analysts and retired flag officers themselves contest whether such a norm is necessary or appropriate. There are four main arguments furthered by these proponents of retired flag officers speaking out on partisan political issues. First, they contend that, as retired servicemembers, these individuals are no longer bound by the restrictions they faced while on active duty.21 Therefore, retired flag officers are free to speak their minds publicly on any issue as an expression of their first amendment rights.22 Constraining such speech would be anti-democratic.

The second argument is that their years of service to the nation and military experience affords them unique insights and wisdom that the American public would benefit from hearing.23 As this argument goes, it would be imprudent to deprive the voting public of such insights at election time. Some go further and suggest that it is imperative to hear the political opinions of those who were willing to sacrifice their lives in service to the country.24

Third, those who support retired flag officers speaking out on political matters point to the long tradition of former military officers being involved in partisan politics throughout American history.25 George Washington, Ulysses S. Grant, and Dwight D. Eisenhower served as both generals and presidents, and the republic did not collapse, nor did civil-military norms fray.

A fourth argument — one that has been increasingly deployed since Trump ran for president in 2016 but is not unique to recent times — is that the current stakes are so high that retired flag officers have a special obligation to speak out.26 This particular argument acknowledges the validity of the norm of nonpartisanship for retired flag officers but insists that during extraordinary times the norm must be sidestepped.

In short, there are competing views about the norm governing retired officers’ political speech, although those that argue against a norm of their abstaining from political speech have traditionally been less influential than those who see the need for sustaining it. Below, we discuss what is known empirically about the state of the norm, beginning with a framework for analyzing it.

A Framework for the Norm of Retired Flag Officers’ Political Speech

While scholars and practitioners often note that the norm proscribing retired flag officers from speaking out politically is, at times, debated both within and outside of the military, minimal research has been done to conceptualize what the norm itself looks like and to assess its health. Absent a framework to consider what such a civil-military norm might look like in practice, attempts to assess its robustness are incomplete at best. Thus, below we outline what a norm against retired officers getting involved in domestic politics might involve in principle. Specifically, we present a framework that defines norm robustness along a continuum and includes indicators of a variety of observable behaviors, as well as motivational elements.

Our research builds upon past scholarly work on norm development and robustness that suggests that a norm “life-cycle” exists, whereby norms emerge, are broadly accepted, and then are internalized.27 We also integrate the potential for norms to erode or “die” into our framework.28 The framework assesses norm robustness not just with respect to incidences of noncompliance by individual flag officers, but with respect to how those violations are received by other audiences, including the public or other members of the profession.29 When describing the dynamics of norm death, we thus follow existing approaches that emphasize behavioral non-compliance and external reactions as core indicators of erosion. Yet, as noted above, we also extend the discussion to include whether the norm’s subjects believe in the norm, as expressed in their discursive endorsement, or verbal acceptance, of its tenets.30

Table 1. Defining Features of Norm Robustness

Death of a Norm	Contested Norm	Robust Norm
Officers superficially internalize norms	Officers inconsistently internalize norms	Officers deeply internalize norms
Officers no longer perceive offending behavior as violations; few, if any, costs for violating norms	Officers are conflicted on propriety standards and costs for violating norms	Norms are viewed as costly to violate
Violations are frequent and rarely addressed within the profession	Violations are occasional and inconsistently addressed within the profession	Violations are rare and swiftly addressed within the profession

At the far-right end of the spectrum in Table 1 is the idealized version of a vibrant, robust norm against retired officers getting involved in domestic politics, where officers have deeply internalized the norm and perceive that violating it will carry significant costs. Internalization is a far more deliberate proposition than being aware of a norm’s existence and understanding it. General comprehension of a norm is required for a baseline level of awareness, but internalization suggests a deeper level of commitment. When norms are deeply internalized in a profession, any debates that emerge tend to be one sided, since its members do not see violating those norms as a legitimate option. In fact, when norms are most firmly entrenched, there may be relatively little debate at all about their appropriateness.

A norm has, as Martha Finnemore and Kathryn Sikkink describe, “‘a taken-for-granted’ quality that makes conformance with the norm almost automatic.”31 For example, in wartime, U.S. military leaders do not deliberately weigh whether to violate the law of armed conflict, even though such violations do occur occasionally. This lack of consideration is not solely due to the clear rules that prohibit violating this law, such as the Geneva conventions and the Uniform Code of Military Justice. A code of conduct has been instilled in these leaders from the very beginning of their service. A deep internalization of norms is also buttressed by formal rules and swift sanctioning for violations. An inconsistent internalization of norms, on the other hand, may be accompanied by ambiguous rules and an uneven enforcement of them. Hence, when a norm is deeply internalized, violations that do occur are rare and quickly addressed within the profession and violators sanctioned.

Norm degradation can be a function of slow and steady erosion, punctuated by dramatic or egregious lapses, but a single instance or case may not be sufficient to indicate widespread norm slippage.

At the opposite, far-left end of the spectrum, a norm is likely defunct and no longer effectively regulates the behavior of members of a profession. In the case of retired officers being politically active, officers superficially internalize the norm — if they internalize it at all — and there are few, if any, perceived costs associated with violating it. Norm violations are frequent and are rarely or poorly addressed within the profession or by the public.

In the middle of the continuum, the norm is contested, and officers inconsistently internalize it. Violations are occasional and are inconsistently addressed or sanctioned publicly or by other retired officers.

Table 2. Indicators of Norm Robustness

Death of a Norm	Contested Norm	Robust Norm
Retired flag officers frequently issue partisan campaign endorsements	Retired flag officers occasionally issue partisan campaign endorsements	Retired flag officers rarely issue partisan campaign endorsements
Retired flag officers frequently criticize elected leaders publicly	Retired flag officers occasionally criticize elected leaders publicly	Retired flag officers rarely or never criticize elected leaders publicly
Majority of retired flag officers are comfortable speaking out on political matters	Some retired flag officers are uncomfortable speaking out on political matters	Majority of retired flag officers are uncomfortable speaking out on political matters
Majority of active-duty officers approve of retired flag officer political speech	Active-duty officers are split on their approval of retired flag officer political speech	Majority of active-duty officers disapprove of retired flag officer political speech

In Table 2, we have identified specific indicators of the health of a given civil-military norm — in this case, the norm that retired flag officers should refrain from political speech. Observable indicators include the extent to which they issue partisan campaign endorsements, publicly criticize elected leaders, and are comfortable speaking out on political matters, along with how active-duty officers view retired flag officer political speech. Other indicators, taken up in later sections of this article, relate to attitudes professed by flag officers about the norm (i.e., how they view the norm itself, its boundaries, and its applicability to their own behavior).

Assessing the frequency of various behaviors is a function of both the prevalence of political activities — how often and how routine they are — and the proportion of individuals engaging in them at any given time. For example, retired flag officers rarely issuing partisan campaign endorsements or publicly criticizing elected leaders are indicators that the norm is robust, whereas a majority of retired flag officers being comfortable speaking out on political matters and a majority of active-duty officers supporting them in doing so are indicators that the norm is defunct. Norm degradation can be a function of slow and steady erosion, punctuated by dramatic or egregious lapses, but a single instance or case may not be sufficient to indicate widespread norm slippage. Conversely, a large portion of retired flag officers — if not the majority, certainly a sizeable minority — routinely engaging in partisan campaign endorsements and public criticism of elected leaders over time would indicate the death of the norm.

The Status of the Norm

Before turning to our analysis of retired flag officer attitudes toward the norm of retired officer speech, we briefly review what is known about the robustness of the norm against their engaging in partisan activity. Specifically, the discussion focuses on the period following the establishment of the All-Volunteer Force, which fundamentally transformed the military’s relationship to society. Norms governing the political activity of retired officers have arguably been unsettled throughout U.S. history. The switch to an all-volunteer force in 1973 marks the beginning of the contemporary era of civil-military relations in the United States and has special relevance to our analysis of the current status of the norm.

We contend here that the most visible and pronounced degradation of these norms since the start of the All-Volunteer Force era began in the early 1990s. Although it is difficult to identify the high point of norm robustness during the All-Volunteer Force era, we can pinpoint the first real signs of its relative deterioration.

That the normative decline began in the early 1990s is especially noteworthy because this predates both the current era of partisan polarization and the advent of social media — both of which could plausibly explain greater political activism among retired flag officers. While numerous factors are likely responsible for the degradation of norms among retired flag officers beginning in the early 1990s, we identify three possible, albeit interrelated, explanations: increasing partisanship among officers; the Republican Party’s ownership of national defense issues along with lingering effects from the Vietnam War within the cohort of flag officers who served in Vietnam as junior officers; and the emerging prestige associated with the U.S. military as confidence in the institution began to grow.

In short, retired officer political activism and the degradation of a norm against non-involvement coincide with an increasing self-identification of officers as partisans themselves, and in particular with identification with one of the two dominant political parties.

As chronicled in his Foreign Policy Leadership Project surveys of military and civilian leaders, Ole Holsti found that partisanship among the U.S. military’s officer corps increased between 1976 and 1996.32 Fewer than half of the officers whom Holsti surveyed in 1976 self-identified as partisans, with independents constituting the largest block at 46 percent.33 By 1992, however, only 26 percent of senior officers self-identified as independent and 67 percent as partisan.34 Of note, the growing proportion of military officers who identified as Republican outpaced the proportion of civilian leaders who did the same, creating a civil-military gap along partisan lines that has not abated. In 1976, 33 percent of military leaders and 25 percent of civilian leaders self-identified as Republican, but by 1996, 67 percent of military leaders self-identified as Republican, compared to just 34 percent of civilian leaders.35

Holsti’s findings of rising partisanship within the officer corps was echoed in the late-1990s by Peter Feaver and Richard Kohn’s landmark Triangle Institute for Security Studies survey on the civil-military gap and journalist Tom Ricks’ in-depth report on Marine Corps basic training.36 The officer corps’ predominant affiliation with the Republican Party and a higher rate of partisanship among junior service members compared to their civilian peers37 have since been confirmed in more recent survey research conducted in the post-9/11 era.38 In short, retired officer political activism and the degradation of a norm against non-involvement coincide with an increasing self-identification of officers as partisans themselves, and in particular with identification with one of the two dominant political parties.

Coinciding with the increase in the proportion of senior officers identifying with the Republican Party in the early 1990s was the solidification of the Republican Party’s ownership of national defense issues and lingering effects from the Vietnam War within the cohort of senior flag officers. John Petrocik coined the term “issue ownership” to describe how parties tend to own and emphasize certain issues in campaigns and elections and how voters tend to trust one party over the other to handle those issues.39 The Republican Party’s issue ownership emerged during the Nixon administration when the public began associating the Democratic Party with the failures in Vietnam and a growing anti-militarism.40 One study found that, in 1968, the Republican Party began placing significantly more emphasis on the issue of national defense in its party platform than the Democratic Party. In fact, by the 1980s, the Republican Party referenced national defense in its party platform seven times more than the Democratic Party did.41 By the end of the Reagan administration, with its massive defense build-up, the Republican Party solidified its issue ownership of national defense. This continued in the early 1990s with the decisive military victory in the Gulf War under the presidency of George H.W. Bush. National defense and military issues had become increasingly and more systematically partisan affairs in U.S. domestic politics.

Not only were military officers more likely to self-identify as Republican in the 1990s, but they also viewed the Republican Party as better equipped to oversee the military and handle defense issues.42 Relatedly, the flag officers who were leading the U.S. military in the early 1990s had been junior officers who came of age during the Vietnam War. No one better represented this generation than Gen. Colin Powell, who served as chairman of the Joint Chiefs of Staff from 1989 to 1993. Powell, like many of his contemporaries, was profoundly impacted by his experience in the Vietnam War, which later influenced his view of civil-military relations as chairman. In his memoir, Powell reflected on the hard lessons his generation of officers learned in Vietnam:

Many of my generation, the career captains, majors, and lieutenant colonels seasoned in that war, vowed that when our turn came to call the shots, we would not quietly acquiesce in halfhearted warfare for half-baked reasons that the American people could not understand.43

Powell’s reflection is evocative of the “stabbed-in-the-back” narrative that emerged among a generation of Vietnam veterans. While not unique to the Vietnam War, this narrative posits that the uniformed military did all it could do to win the war, but civilian political leaders failed by not resourcing the military adequately to fight the war or by having unclear political objectives.44 This narrative often allows military officers to pin defeats in war on their civilian overseers while avoiding accountability themselves.

Powell’s formative experience in Vietnam and the lessons he drew from that war help explain the activist role he played as chairman, where he felt unconstrained in publicly telling civilian leaders how and under what conditions the military should be employed. The Powell Doctrine, with its criteria for the use of force — have clear objectives, use decisive military force, use as a last resort — were conditions he gave to civilian leaders, not the military. Perhaps even more provocative than the Powell Doctrine was the New York Times op-ed he published on the eve of the 1992 election called, “Why Generals Get Nervous,” in which he outlined his opposition to then-candidate Clinton’s proposal to employ the military in Bosnia.45 Aided by his widespread popularity, Powell wielded an extraordinary amount of political skill that ultimately served to increase the stature and power of the U.S. military while constraining civilian leaders.46

What is noteworthy is that Powell made all of these public statements on how the military should be used — along with thinly veiled criticism of civilian leaders’ proposals on the use of force — while on active duty as chairman. In this light, it is hard not to view his public stance as greenlighting a new brand of political activism among his retired flag officer peers — fellow Vietnam veterans — in the early 1990s. If the sitting chairman could criticize elected leaders’ policies and publicly weigh in just weeks before a presidential election, surely retired flag officers could do the same. Powell’s actions played a role in undermining a norm of retired flag officers avoiding partisan involvement by legitimating political speech by virtue of his stature and actions.

The third and final factor that may help explain why the normative decline among retired flag officers began in the early 1990s has to do with increasing public confidence in the military. According to Gallup’s annual confidence in institutions poll, only 50 percent of the public had “quite a lot” or “a great deal” of confidence in the military in 1981, and on average, 58 percent expressed “quite a lot” or “a great deal” of confidence in the military throughout the 1980s. In the first half of the 1990s, however, that figure jumped to 66 percent, likely propped up somewhat by the Gulf War victory.47 The early 1990s did not witness the high levels of public confidence in the military that we saw during the post-9/11 wars, but confidence in the military after the Gulf War and following the 1980s defense build-up was nonetheless on an upward swing compared to the late 1970s and early 1980s. This matters, because when public confidence in the military is high, politicians and elected leaders are more incentivized to use the military for partisan gain, and retired flag officers are more inclined to oblige by leveraging the prestige associated with their rank and service for partisan causes.

Taken together, these factors — increasing levels of partisan identification among officers, the solidification of the Republican Party’s issue ownership of national defense and military matters, a cohort of retired flag officers emboldened by their experiences in Vietnam to publicly criticize politicians, and increasing prestige associated with military service — help explain why norm erosion among retired flag officers began in the early 1990s.

One of the two most significant indications of this decline is the advent of and then dramatic rise in partisan campaign endorsements. The origin of partisan endorsements can be dated to Crowe’s endorsement of the Democratic presidential nominee, Clinton, in 1992. Crowe’s endorsement was notable because it set off a cycle of endorsements, especially for Republican candidates, in every subsequent presidential election.48 In the 2000 presidential election, 85 retired flag officers, including five former service chiefs, endorsed George W. Bush for president.49 By 2012, Mitt Romney assembled the largest list (500) of retired flag officers to date, which included former Chairman of the Joint Chiefs Gen. (ret.) Hugh Shelton and several former service chiefs such as the recently retired commandant of the Marine Corps, Gen. James Conway.50

What’s more, retired officers whose notoriety comes from regular appearances on cable news or other partisan media outlets can, in many cases, draw even more partisan audiences than elected politicians.

In addition to endorsement lists published by campaigns, retired flag officers began making notable appearances at presidential nominating conventions. The presidential conventions in 2016 stand out. These featured Lt. Gen. (ret.) Michael Flynn leading Republican National Convention attendees in a “lock her up” chant, and Gen. (ret.) John Allen marching a coterie of former military servicemembers on stage to militaristic music during the Democratic National Convention.51 Yet, these were hardly the first conventions to feature endorsement speeches by retired flag officers. Just a year after retiring as commander of U.S. Central Command, Gen. (ret.) Tommy Franks appeared at the 2004 Republican National Convention to publicly endorse Bush — just as Gen. (ret.) Wesley Clark did for John Kerry that same year at the Democratic National Convention.52

Recent examples aside, the first notable retired flag officer endorsement was Crowe’s endorsement of Clinton. In his memoir, Crowe partly attributed his endorsement to wanting to dispel the myth that all military officers were Republican.53 In fact, Clinton’s election may have helped spur greater political activism among retired flag officers who self-identified as Republican in the 1990s because they viewed Clinton as a threat to the Republican Party’s almost-assured issue ownership of defense and military matters after 12 years of consecutive Republican presidential rule. Of note, Zachary Griffiths and Olivia Simon found that, since Crowe’s endorsement, retired flag officers have endorsed Republican presidential candidates eight times more than Democratic candidates.54 This does not suggest that all retired flag officer political activism since then has been by and for Republicans — it has not. Rather, it helps explain the spark that set off the normative decay in the 1990s.

In addition to endorsements during elections, a second indication of growing involvement in partisan politics is a surge of public statements by retired officers that have partisan bearing, if not overt partisan content. While public commentary about foreign policy and strategic issues by retired officers has long been common, there has been an uptick in the incidence of expressly partisan commentary and action in the form of commentary and op-eds written by retired officers in which they criticize elected leaders and violate the norm of nonpartisanship.55 There has also been an increase in partisan commentary on social media, as retired flag officers have sought to “grow” their followings by appealing to partisan constituencies.56 Analysis of their social media accounts suggests that these individuals can attract ideologically coherent (if narrow) follower networks. What’s more, retired officers whose notoriety comes from regular appearances on cable news or other partisan media outlets can, in many cases, draw even more partisan audiences than elected politicians.57

To be sure, there has been push-back to this norm erosion among retired flag officers. Adm. (ret.) Michael Mullen and Gen. (ret.) Martin Dempsey, both former chairmen of the Joint Chiefs of Staff, forcefully called for an extension of the nonpartisan ethic into retirement, especially for flag officers. And both have spoken out on numerous occasions about how partisan endorsements threaten to damage the military’s credibility as a nonpartisan institution.58 While Gen. Joseph Dunford, Mullen’s successor, may not have been as vocal in calling for officers to adhere to the norm of nonpartisanship in retirement while he was chairman, he acknowledged that he personally will continue to avoid partisan politics in retirement, just as he did on active duty, and has since written of the dangers of politicizing the military.59 The very fact that these men felt it was urgent to make such public statements about the importance of adhering to the norm may be an indicator of its decline.

Moreover, the three most recent chairmen seem to be outliers by at least some measures in explicitly proscribing retired flag officer engagement in expressly partisan actions. Recent years have witnessed several high-profile instances of retired flag officers publicly criticizing the sitting president. The think tank New America identified 230 retired flag officers who spoke out against Trump while he was in office.60 Even Mullen violated his own proscriptions against speaking out in a commentary piece in The Atlantic, entitled, “I Cannot Remain Silent,” in which he questioned Trump’s leadership amid his crackdown on George Floyd protesters in Lafayette Square.61 In addition, recent survey research has found that many active-duty military officers are supportive of retired flag officers speaking out publicly on political issues, further calling into question the durability of the norm for retired officers.62

All of this discussion of behavioral violations, in turn, provides the larger context for the inquiry into retired officer views about norms. The survey aims to explore flag officers’ discursive endorsement of or verbal support for norms. In other words, despite the mounting evidence that the norm against partisan speech and activity by retired flag officers is contested, how they actually think about the appropriateness of speaking out has not been well studied. Moreover, while there has been debate about the appropriateness of retired officers engaging in political speech, there has been surprisingly little study of their motives for speaking out when they choose to do so. One exception is the study by Griffiths and Simon, which found that the majority of retired flag officers who endorsed candidates likely did so not out of ideological motivations or material incentives, but because of personal connections within their peer network of retired flag officers.63 In addition, how they evaluate the potential effect of their comments on public debate has not been studied. Some research suggests that military officers speaking out can shift public opinion at times.64 Other research, however, shows that it often does little to shift public sentiments. Rather, such political signals can cost the retiree and the military institution considerable credibility with the public, even if this is highly conditional on partisan identity.65 Nonetheless, less is known about whether the efficacy of speaking out is a major factor shaping the decision by retired flag officers to do so.

Methodology

From December 2020 to January 2021, we surveyed a select number of retired flag officers who served in the U.S. military on their views regarding participating in public political discourse.66 The survey queried respondents on the types of political activities they have engaged in and their views on the propriety of such activities through a mix of multiple-choice and open-ended, free-form text questions. Of the 39 retired flag officers to whom we sent the online survey, 23 completed the entire survey, yielding a 59 percent response rate. In order to protect the anonymity of members of such a small, elite population, we collected limited demographic information from survey respondents.

Tables 3, 4, and 5 display the branch in which respondents served, their rank when they retired, and their self-reported partisan identification. 

Table 3. Select Demographics of Sample: Branch/Service 

Branch/Service	n
U.S. Air Force	2
U.S. Army	18
U.S. Marine Corps	1
U.S. Navy	2
Total	23

Table 4. Select Demographics of Sample: Rank/Grade

Rank/Grade	n
O-7	3
O-8	6
O-9	5
O-10	9
Total	23

Table 5. Select Demographics of Sample: Partisan Identity

Partisan Identity	n
Democrat	1
Independent	13
Republican	6
Other/Unaffiliated	3
Total	23

Several points regarding the sample are worth noting. First, with only 23 respondents, the sample is by no means representative of the broader population of retired flag officers, which is typically estimated at more than 7,500 individuals at any time. Our small sample thus has limitations, and we stop short of generalizing our findings across all flag officers, active or retired. Rather, the aim of this project is to obtain deeper insights and opinion data from a small, elite group of individuals. Our insights are therefore illustrative, not definitive. Many of the questions we asked respondents were open-ended questions and were designed to elicit deeper perspectives from a select group of individuals. Moreover, existing scholarship on the motivations of retired flag officers with regard to their public, political discourse is limited — a reflection of how difficult it is to survey this particular population. For all of these reasons, we sought to obtain personal, in-depth perspectives from a select group of retired flag officers. But we do not attempt to characterize these views as conclusive or reflective of all retired flag officers today.

Second, the sample is heavily skewed toward retired Army generals. One of us is an active-duty Army officer, and another is a retired Army officer, and the sample is somewhat of a reflection of our professional networks and the snowball sampling techniques we employed in identifying potential survey respondents. Although most respondents in this survey served in the Army, there is little to suggest from past research that branch of service has a significant impact on officers’ political views or their adherence to civil-military relations norms.67

Therefore, the respondents were evaluating their options under high-stakes, real-world conditions, versus in a simulated setting in which the exigencies of deciding whether to speak out were less intense.

Third, the inclusion of nine retired four-star officers in our sample drawn from each of the services is unique and provides novel insights into how the senior-most flag officers think about issues related to political speech. Four-stars, including service chiefs and combatant commanders, have arguably given more thought to the intersection of military service and politics than any of their flag officer peers at the one-, two-, or three-star level, and their insights are of particular value to our study.

Fourth, while the sample is small, it is an ideologically mixed group. We sent the survey to retired flag officers who we knew had been politically vocal in the past, as well as to those who have refrained from such activity, and we asked them to forward the survey to their own networks of diverse, retired flag officers. Although our sample is not representative of the entire population of retired flag officers, it does reflect the diversity of opinions on the propriety of their speaking out.

Lastly, it is worth noting that we fielded our survey during a volatile period in American politics (following the 2020 presidential election). As such, the survey takes place in the shadow of Trump’s presidency. We should expect that retired flag officers faced strong external pressures to speak out during the Trump administration, given the significant partisan polarization in American society at the time coupled with the prestige associated with the U.S. military.68 Examining retired flag officer attitudes in this context has some advantages in that the pressure to consider speaking out was more than hypothetical. Therefore, the respondents were evaluating their options under high-stakes, real-world conditions, versus in a simulated setting in which the exigencies of deciding whether to speak out were less intense. As the responses demonstrate, the discussion about speaking out was not abstract — this was a decision that these individuals were actively making and, in some cases, acting upon. At the same time, the particular nature of the context meant that Trump’s actions were in the background, framing their assessments of the costs and benefits of speaking, which in many cases these surveyed flag officers made explicit. Whether and how they would have responded five years prior is difficult to ascertain.

Findings

The responses to the survey yield three important insights. First, retired flag officers were aware of and influenced by civil-military norms, yet they varied in how they thought such norms should constrain their behavior in retirement. Second, their social networks matter, but the respondents did not identify peer pressure as a decisive factor in whether they decided to speak out. Third and of most significance, responses from the retired flag officers we surveyed reflected conflicting loyalties and obligations that, in turn, shape the degree to which they are willing to speak out on political issues.

Norms as a Baseline

First, our findings reveal that retired officers are very much conscious of and influenced by their perception of the norm that military personnel should abstain from engaging in public discourse. They are also aware that this norm has implications for retired officers. Norms thus provide a baseline against which those in our survey evaluated engaging in public discourse. They drew a distinction between active-duty personnel engaging in political commentary and retired officers doing so, suggesting that retired officers should not be held to the same strict standards that those on active duty are (see Figure 1). Nonetheless, they also agreed that they and their peers adhere to certain unspoken codes of conduct when it comes to speaking out on domestic political matters.

Figure 1. Views on Appropriateness of Speaking Out on Political MattersThat they actively grappled with these norms is also apparent in their open-ended responses. As one respondent put it, “The default setting must be to stay of out of political discourse (not policy discourse), however, each individual must apply his or her best judgement on communicating, as an exception.” Other comments implicitly affirmed that retired officers were bound by norms, and that weighing in publicly while invoking military service might violate them. As another respondent wrote,

I do not object to retired/former military officers running for office or serving in administration cabinet positions and obviously in that role they can and must speak out on political issues and do so as members of their political affiliation, but when they cross that line as political figures, they become politicians and must be careful not to use that military title for solely political means.

Still others framed the norms in terms of their importance to U.S. civil-military relations: “[P]artisanship among retired [general officers] is the quickest path toward having a litmus test for four-star nominations by presidents. … [T]hat could be highly detrimental to American civil-military relations.” In sum, the respondents’ answers reflect a general awareness — and concern for — standards of normative propriety with respect to political speech by retired officers. Whether in the form of “unspoken codes of conduct” or informal fears about military politicization, these responses indicate, at a minimum, the salience of norms when rendering a judgment about these forms of political activity.

Interpretations Varied

While the surveyed officers assuredly incorporated normative considerations into their responses, their understanding of the “red lines” implied by those same norms varied significantly. Few of the retired flag officers we surveyed interpreted these norms to prohibit speech under any circumstances, but those who did had strong opinions on the matter. Of the few who took that view, one noted that “three and four stars should agree before their promotions to NEVER comment … unless they are running for office.” Those on the other side of this issue cited that the first amendment applies to all citizens, including retired flag officers.

The vast majority of our respondents had more nuanced views. Several acknowledged the qualitatively different sets of circumstances that exist when a retired officer declares political candidacy. Many saw that as the one case in which it was definitely permissible to speak about domestic politics, because a retired flag officer would have crossed the line to being a full politician at that point. Others acknowledged that, although there should be no absolutes preventing such political speech, retired flag officers should largely refrain from speaking on political issues or, at the very least, carefully weigh the implications of doing so. One retired officer reflected, “I agree that retired flag officers should generally be reticent to speak out on political issues … but they should not be prohibited from doing so. They have earned the right, but like all rights, they should exercise it with care and circumspection.”

Respondents were closely divided on other political activities such as publicly criticizing a politician for politicizing the military and donating to a political campaign or political action committee.

In order to refine this analysis with an eye toward specific forms of political engagement, we asked respondents which kinds of political activities they had engaged in, and, regardless of whether or not they had engaged in those activities, which ones made them feel the most uncomfortable (see Figure 2). The most frequently cited political activities that a majority of retired flag officers indicated they had participated in were writing op-eds, speaking out publicly about respecting the oath to the Constitution, being quoted or interviewed in print media about politics or the military, and being interviewed on television about foreign policy. Relatedly, few respondents indicated that these activities made them uncomfortable.

Figure 2. Self-Reported Political Activities and Associated Level of DiscomfortSimilarly, the types of political activities that the fewest number of respondents reported having engaged in were also the ones that made the most respondents uncomfortable. For example, only 9 percent of respondents had spoken at a rally for a political candidate, but 76 percent indicated that this was an activity with which they were uncomfortable. In addition, while 32 percent of respondents had endorsed a political candidate running for office, 71 percent were uncomfortable with this. Respondents were closely divided on other political activities such as publicly criticizing a politician for politicizing the military and donating to a political campaign or political action committee.

In addition, many respondents distinguished between commentary focused on foreign policy or related matters and commentary having to do with domestic politics. In response to a question that asked what event or action made speaking out necessary, or brought them the closest to speaking out publicly, one retired flag officer noted:

Speaking publicly on issues of history or personal philosophy or in leadership discussions in educational environments, or discussing lesson[s] learned from personal experience are all valid and important means for retired senior officers to use their voice and I routinely engage in those opportunities. Educating and informing politicians on specific national defense policy or historical issues, not related to advocacy for a specific party or candidate, or conducti[ng] similar discussion with the media in a non-partisan manner, are also ways to use your voice.

This was a common theme, although one warned that it was difficult to make hard-and-fast rules related to these distinctions, observing that the term “‘politics’ needs a more granular explanation as almost everything has a political dimension these days.”

Still others emphasized that designating particular actions as permissible or off-limits in the abstract was difficult. The context in which such statements might take place mattered. One officer commented:

Speaking out against the regime in Nazi Germany is an example most retired officers would agree would have been right to do. Yet not having any specific guidelines beyond this extreme example is profoundly unhelpful, and lends itself to wildly varying interpretations. ‘Lock her up’ chanted by a retired 3-star at a national political convention is not the same level of speaking out as [James] Mattis speaking out in the aftermath of the Jan 6 attacks at the Capitol, yet some would facilely throw both sets of comments in the same box.

Social Networks Matter, Peer Pressure Does Not

Our survey also provides insights into the social networks of retired flag officers and the extent to which these retired officers view these networks as playing a role in either encouraging or discouraging their participation in public political discourse. As depicted in Figure 3, the majority of respondents indicated that they routinely keep in close contact with other senior retired officers and discuss issues pertaining to the U.S. military with them. However, respondents were split on how often they spoke about domestic politics with their retired flag officer peers, with about half indicating that they often or always did, and the other half indicating that they did so sometimes or rarely.

Figure 3. Frequency and Nature of Retired Flag Officer Engagement with Other Retired Flag Officers

The majority of retired flag officers we surveyed also indicated that their peers encouraged them to speak out on foreign policy and military matters as shown in Figure 4. Of note, 69 percent of retired flag officers indicated that their peers had urged them to endorse a candidate running for office while only 46 percent indicated that their peers had urged them to refrain from such endorsements.

Figure 4: Retired Flag Officers’ Self-Reported Peer Influences

Figure 5. Factors Influencing Retired Flag Officers’ Thoughts About Speaking Out

Nevertheless, while the majority of respondents indicated that they keep in close contact with their retired peers, and while more than two-thirds of respondents indicated that their peers had urged them to endorse a candidate or speak out on political issues, respondents cited peer pressure as the least important factor influencing their decisions about speaking out (see Figure 5). In other words, although retired flag officers acknowledged that they have received some peer pressure from their retired colleagues, they claim that such pressure did not affect their decision-making regarding whether to speak out on political matters.

We asked respondents to elaborate on the pressures that they had contended with to either engage or not engage publicly about politics more broadly. Intriguingly, while citing incidents in which they were urged to speak on defense or foreign policy, or even to endorse a candidate, roughly one-third of respondents indicated that they faced no pressures one way or the other. Several cited friends, family, and peers as sources of pressure. Many of these respondents simultaneously reported that they had both been encouraged to speak out and discouraged from doing so by members of their social networks. Despite acknowledging these pressures, once again few retired flag officers indicated that they shaped their decisions of whether to speak out.

Conflicting Obligations

As discussed previously, the decision among retired flag officers to engage in any form of political activity reflects a complicated landscape of often contradictory pressures or perceived obligations. This network of conflicting loyalties mirrors the same opposing forces documented among their active-duty counterparts. For example, serving officers are instructed on the necessity of honoring authoritative command channels such as the executive and Congress. However, they are also bound by a variety of other allegiances, ranging from the operational requirements to the mission, professional norms of conduct, supervisory responsibility for subordinates, and support to democratic governance, to name a few.69 When circumstances place these obligations in tension, servicemembers may be forced to discern — often without clear guidelines — which take precedence. In the case of retired officers, similarly conflicting loyalties may be at play, but the respective weight of each on individual decision-making is likely to change once the officer is no longer bound by formal regulation. For example, deep-seated professional norms are likely to carry over into post-service life but may clash with newly unmuted partisan beliefs or personal ethical convictions that could not take prominence while still in uniform.

Several cited concerns about the impact that their political speech might have on those currently serving on active duty, weighing both the pressures it might put on senior military leaders and an imperative to speak out because their active-duty peers were prohibited from doing so.

To investigate the motivations underlying the decision of whether to speak out politically, we asked respondents a number of open-ended questions about the considerations guiding them. The specific lines they drew about when and why speaking out might be appropriate varied, as noted above. But they shared in common a reflection that there were tradeoffs to consider in making such decisions. As such, the responses reflect the competing obligations that officers in a democratic state face, including efforts to safeguard the well-being of the institution, to protect servicemembers and currently serving military leaders, as well as to be true to their oath to support and defend the Constitution.70 The retired officers we surveyed seemed attuned to the potentially competing considerations in weighing whether to speak out.71

These tradeoffs emerge as prominent themes in many comments. One significant tradeoff of speaking out was the risk it posed to the reputation of the institution and its status as a non-partisan body:

The average American does not differentiate between retired and active flag officers. So, when retired flags make public political declarations, the American public tends to believe that active flag officers think similarly, and hold those positions. The military then gets dragged into the political fray.

Several cited concerns about the impact that their political speech might have on those currently serving on active duty, weighing both the pressures it might put on senior military leaders and an imperative to speak out because their active-duty peers were prohibited from doing so. One retired officer noted:

The dominant pressure I have felt is the internal pressure of not making my successor’s job any harder than it already is. That said, during the Trump administration I have also felt the pressure of knowing that I can speak when others inside the administration — including military officers — may not be able to speak. In that way, the last four years have been truly unprecedented leading me on occasion to unprecedented actions.

Yet, another observed:

I remind myself that I am not accountable for our national security policy any longer and that those who are accountable — those still serving and in key positions — deserve the opportunity to do their jobs without worrying about how those of us who have retired will “grade” their work. Therefore, the circumstances in which retired Generals and Admirals should engage in public discourse should be few, based on unique expertise, and constructive.

Still others made a finer distinction between the interests of the currently serving military leadership and the institution itself. As one respondent put it, “I try to balance what is best for the institution in the long-term vice [versus] what is convenient or supportive of current senior military personnel. I will ALWAYS err on what is best for the long term of the institution.”

Another prominent theme related to having an obligation to the public and to potentially enriching public knowledge and debate about issues within their area of expertise. As one put it, “I believe part of my responsibility is to use the experiences I had in the military, and what I know about the actions of the military, to better inform the American people (the vast majority of whom have no military experience and do not know the reasons for various actions by the government).” Another added, “I also believe that reinforcement of the military role to ‘support and defend the Constitution’ can require education in the public domain but requires a factual, historic, non-partisan approach to that dialogue, both on and off the record.”

One respondent explicitly framed the decision to speak out in terms of tensions between the public good and the consequences for currently serving military leadership. As he or she put it:

Quite often retired officer commentary is not helpful to the discourse. And I personally observed what a specific Chief of Service and a former CJCS thought of the public input retirees were providing. However, I believe retired Sr. military expertise should be part of the public discourse. They’ve earned the right to voice a civil and informed opinion.

Perhaps the most salient tension in the responses, however, related to their oath to support and defend the Constitution. Those we surveyed seemed acutely attuned to this obligation and it appears to have weighed heavily in the calculations of many. This is reflected in their responses to questions about motivating causes for political activity in which they might have engaged. Overall, retired flag officers listed the importance of democratic institutions, supporting and defending the Constitution, and personal moral or ethical convictions as the most important factors in considering their public engagement (see Figure 5).

Indeed, many felt that their oath to the Constitution required them to engage in public discourse under exceptional circumstances. Specifically, when we asked what event made speaking out necessary or brought them the closest to speaking out, roughly two out of five explicitly referenced events that occurred in recent years, such as the forceful disbursement of peaceful protesters in Lafayette Square in June 2020.72 In all, nearly half of respondents reported having made explicit or implied references to political events within the past couple years.

One respondent described his or her decision to speak out in the following way:

It really was a culmination of events linked to the President’s [Trump’s] disrespect for the Constitution that caused me to speak out, write an Op-ed, and for the first time publicly           endorse a Candidate. The President’s recent statements and activities in the Summer and Fall that questioned whether or not he would accept the results of the Election, his reference to “my Judges and Obama Judges”, “my Generals’, etc. all demonstrated he did not respect the Constitution.

Another respondent described the pressures that might compel him or her to speak out: “When the actions are harmful to protecting the Constitution or Constitutional norms, when the military is being asked to follow illegal, immoral, or unethical orders, or when the military is being used for over partisan purposes.” And yet another wrote: “Lafayette Square rhetoric and actions by government officials, including POTUS, using the military as props for domestic political power. And, more recently, actions by POTUS to overturn [the] electoral process, calling right wing extremists out against the Capitol, and abusing power.”

Often, although couching their considerations in terms of their oath, respondents explained that they saw their decision as a personal choice based on moral and ethical concerns. As one respondent explained, “I felt it was time that I make sure I spoke out on Constitutional issues … and I wanted to be on the ‘right side of History.’” As another framed the decision: “Although I remain skeptical of retired officers speaking publicly, I have concluded that when the protracted actions, behaviors and policies of political leaders threaten the foundations upon which the nation was established, it would be wrong — a dereliction of duty — to remain silent.” Still another explained:

I believe there are circumstances when it is acceptable — and required — for retired military to engage on public discourse, as stated in the first answer I provided. If there is partisan activity that goes against our Constitution or constitutional norms, our laws, or our national values, I believe it is imperative for retired military to engage.

Lastly, we asked respondents what has been the response of their fellow retired flag officers regarding their choice to engage or not engage in public political discourse. Nearly half of respondents indicated that the feedback they have received has been mostly positive or supportive, while approximately one-fifth reported that the feedback they received was mixed. While some respondents were discouraged from public commentary by peers “advocat[ing] strict non-negotiable abstinence,” others found their retired colleagues more accommodating, “recognizing it as an exception that was needed,” or found that peers “actively persuaded [them] to speak out more.” No respondent indicated that the majority of the feedback they received was negative. This is likely a function of the social circles that retired flag officers maintain following active service, where an opportunity exists to curate peer networks to others of like mind.  Hence, while the respondents did not cite peer pressure as a factor in their decision to speak out, they nonetheless reported that members of their social networks often validated their decision to do so.

Discussion

Our findings reveal a nuanced understanding of norms among the retired flag officers we surveyed. Despite being sensitive to norms that govern officer involvement in politics, many of our respondents indicated they were not rigidly bound by them. Many were willing to suspend or qualify their adherence to the norm that discourages retired flag officers from participating in partisan politics. The variety of explanations confounds a simplistic assumption that all retired flag officer speech must be the result of political opportunism. Some cited moral and ethical concerns, while others invoked the Constitution and respect for democratic processes — a nod to the many crises and extraordinary circumstances during the Trump administration. Several indicated that there were certain circumstances that required norms to be sidestepped. This self-expressed permission structure is worthy of note in that it cannot be ascribed to a lack of awareness. Many norm violations among active-duty servicemembers can be explained by incomplete or shallow socialization. However, we cannot attribute the same lack of recognition to these retired flag officers, the most elite and well-socialized stratum of the officer corps. Most of our respondents acknowledged that an unspoken code of conduct exists that guides how retired flag officers engage on political matters in public, and many voiced discomfort with certain partisan activities. However, strong minorities in our sample reported engaging in those activities and thus jettisoning the norm of partisan neutrality.

Of note, while the majority of our respondents indicated that they maintain robust peer networks of their fellow retired flag officers, many of whom pressured them to either speak out or refrain from speaking out publicly, they also cited peer pressure as the least influential factor in their decision-making. This stands in contrast to Griffiths and Simon’s findings, which suggested that personal connections with peers were the motivating factor behind campaign endorsements.73 It is striking that the majority of our respondents acknowledged the existence of peer pressure but then dismissed such pressure as not influential in their decision-making. Given the military’s hierarchical nature, it seems improbable that respondents were as immune to peer pressure as they indicated. Their responses in our survey could be a product of social desirability bias: Respondents may have felt compelled to downplay the impact of peer pressure in favor of loftier motivations such as moral and ethical considerations, support and defense of the Constitution, and the importance of democratic processes.

Yet, despite that, we find a great deal of contention over the boundaries of the norm and the scope and depth of their commitment to it.

At the same time, the fact that they did not report normative pressures as an explicit factor in their decision-making may itself indicate a lack of robustness of the norm. When a norm is vibrant, an individual who is thinking of violating that norm might be expected to express concerns that they would be going against social expectations. The fact that the retired flag officers whom we interviewed were comfortable stating that they felt free to make decisions based on their own calculus and volition may be revealing in this regard. This finding from our analysis may also have larger lessons for understanding norm internalization. Yet another way of interpreting these findings is that, although peer networks may have played a critical role in enabling or facilitating campaign endorsements, letter signing, and other forms of political activism, retired flag officers had other pre-existing motivations for engaging politically. Regardless, the role that peer networks play in retired flag officer political activism merits further research. 

Assessing the State of the Norm Today

In this section, we analyze insights from our survey respondents in the context of the framework introduced earlier in this article. Contrary to those that argue that there should be few limits on the political activism of retired flag officers, we find strong support among our respondents that a norm against engaging in partisan speech and actions does and should exist. Yet, despite that, we find a great deal of contention over the boundaries of the norm and the scope and depth of their commitment to it.

Overall, given our findings that most retired flag officers are uncomfortable engaging in partisan political activity — even if some nonetheless indicated that they had engaged in those activities — coupled with trends in retired flag officer political speech and campaign endorsements over the past 35 years, we assess the norm against this population engaging in political speech to be contested. Despite the increase in partisan campaign endorsements by retired flag officers since the late 1980s, the percentage of those who endorse candidates remains fairly low at about 5 percent of all living retired flag officers.74 Nonetheless, the increase in the number of endorsements in the past several decades coupled with the high-profile nature of recent endorsements, including senior retired flag officers speaking at both party nominating conventions in 2016, raises the salience of campaign endorsements in our estimation. Because of this, we assess these to be indicators of a contested norm, rather than a robust norm.

As we indicated at the outset of the paper, and as reflected in the open-ended responses of our respondents, Trump served as a motivating factor not just in terms of making campaign endorsements but also in retired flag officers’ decision to publicly criticize the president. It may turn out that public criticism of the president reached its peak in the Trump administration, and in future years such criticism may recede, causing us to re-assess this particular factor as indicative of a robust norm. However, the “Revolt of the Generals” in 2006 that well preceded Trump and open letters by “Flag Officers 4 America” that were critical of President Joe Biden cause us to evaluate this as an ongoing indicator of a contested norm.75

Lastly, recent survey research provides insights into how active-duty officers view retired flag officer political speech — another indicator that speaks to the health of the norm. In 2009, a survey of active-duty Army officers found that 68 percent of respondents agreed or strongly agreed with the statement that it is proper for retired generals to publicly express their political views. However, in a similar survey of Army officers conducted from 2017 to 2020, only 49 percent of Army officers agreed with that statement.76 The decline in support among active-duty officers for retired flag officers speaking out on political issues could be in response to the increase in high-profile campaign endorsements, the spectacle associated with retired flag officers at the 2016 nominating conventions, or the various controversies that retired flag officers were involved in during the Trump administration. In 2009, these survey results might have indicated a defunct norm, but the more recent results are further evidence of a contested norm.

Conclusion

Our unique survey sample — retired general and flag officers, over half of whom were three and four stars — yields important new insights into how normative considerations impact retired flag officers’ decision-making process of whether to speak out publicly on political matters. These officers, many of whom served in the military for 35 to 40 years, are the most well versed in professional norms, including the norm of nonpartisanship. Past research has shown that the socialization process for officers takes time, and lengthy time in service and higher rank are often correlated with a greater adherence to professional norms.77 Moreover, many of these officers served at the highest levels of government, working closely with and advising civilian political leaders. Senior retired flag officers, by virtue of their selection process, professional military education, time in service, and professional experience, are the most socialized and sensitive to civil-military norms and political considerations. Nevertheless, we show that even among that sample, norms against retired officer involvement in partisan speech are contested, and increasingly so.

Specifically, while those in our sample did perceive the existence of a norm against partisan speech by retirees, they also raised questions about its boundaries and, in some cases, supported violating it openly. Importantly, it is not so much that they disagreed about whether partisan activity should generally be off limits, but whether retired officers should always be bound by that rule. In line with other survey research that shows that regard for norms of nonpartisanship have steadily eroded more broadly within the active-duty military, the findings from our anonymous survey of retired officers may indicate a decline in the robustness of those norms more universally.

This point is worth expanding upon, as it highlights a larger observation about norm robustness and what is required practically to sustain a norm. Just because a norm is vibrant and robust today does not mean it cannot become a contested norm, or even a dead norm, at some future point. Absent deep internalization, constant teaching, clarity, and well-specified standards that are agreed upon and enforced, civil-military norms can and will deteriorate. We present our model about the norm of retired flag officer political speech along a continuum, indicating a degree of movement along norm adherence. Some respondents in our survey indicated that recent events required them to speak out publicly on political matters, suggesting a one-time deviation from norm adherence for them, while other respondents reflected a greater fluidity in their normative interpretations. Regardless of their individual motivations, however, repeated norm violations in the aggregate combined with minimal social disapprobation in response to those violations move a norm from robust to contested and raise the possibility of further deterioration toward the norm becoming defunct.

What is unclear from this study alone is how quickly a norm can deteriorate or, conversely, how long it might take and what actions might be required for a norm to rehabilitate once it has deteriorated. Given our assessment that the norm pertaining to retired flag officer political speech is contested, the issue of rehabilitating a norm merits increased attention by scholars and practitioners concerned about this development.

Ultimately, time will tell whether the attitudes professed and actions supported by the retired flag officers in our sample were merely an artifact of a particularly fractious moment in American politics, or whether they presaged a broader sea-change in retired officer attitudes and behavior. Regardless, for scholars and practitioners who are concerned about the health of U.S. civil-military relations today — and who believe a norm against partisan non-involvement by retired officers is a key pillar of the military’s nonpartisan ethic more broadly — our findings are not reassuring.

 

Risa Brooks is Allis-Chalmers Professor of Political Science at Marquette University.

Michael A. Robinson is an active-duty Army strategist and an adjunct associate professor in the Security Studies Program in Georgetown University’s Edmund A. Walsh School of Foreign Service. He is the author of Dangerous Instrument: Political Polarization and U.S. Civil-Military Relations (Oxford University Press, 2022).

Heidi A. Urben is a professor of the practice and director of external education and outreach in the Security Studies Program in Georgetown University’s Edmund A. Walsh School of Foreign Service. She is the author of Party, Politics, and the Post-9/11 Army (Cambria Press, November 2021).

The views expressed in this article are those of the authors and do not reflect the position of the U.S. Army, the Department of State, or the Department of Defense.

Acknowledgements: We are grateful to Peter D. Feaver, Zachary Griffiths, Carrie Lee, participants at the 2022 annual conference of the American Political Science Association, participants at the 2021 biennial conference of the Inter-University Seminar on Armed Forces and Society, and the anonymous reviewers for their helpful feedback and suggestions.

 

Image: Gage Skidmore (CC BY-SA 2.0 DEED)

 

Appendix: Survey Instrument

1. Have you ever engaged in the following?
   1. 1. Seriously considered running for office?
      2. Been urged by another retired officer to speak out on foreign policy?
      3. Been urged by another retired officer to speak out on matters dealing with military discipline or organizational values?
      4. Been urged by another officer to endorse a candidate for political office?
      5. Been urged by another officer to refrain from endorsing a candidate for political office?

1-Yes
2-No

2. How often do you do the following?
   1. Keep in touch with other senior retired officers?
   2. Speak with other retired officers about issues pertaining to the US military?
   3. Speak with other retired officers about issues pertaining to domestic politics?

1-Always
2-Often
3-Sometimes
4-Rarely
5-Never

3. Do you feel there are unspoken codes of conduct that you and other senior officers follow with respect to speaking out about domestic political matters?

1-Yes
2-No
3-Other (please specify)__________

4. Do you think that the standards governing political commentary for retired officers should be different from those governing active-duty personnel?

1-Yes
2-No
3-Other (please specify)__________

5. What event or action would you say made speaking out necessary, or brought you the closest to speaking out publicly?

          (free-form text response)

6. What pressures did you contend with in your decision to engage with/remain outside of public discourse on politics?

          (free-form text response)

7. What are your thoughts on the belief that there are no circumstances in which retired military engagement in public discourse on politics is acceptable?

          (free-form text response)

8. On a scale of 0-10 (0 = not at all important, 10 = very important), how did the following factors influence your thoughts about speaking out?
   1. Influence or pressure from peers or colleagues?
   2. Organizational or professional norms or standards?
   3. Personal moral/ethical convictions?
   4. Support and defense of the Constitution?
   5. Importance of democratic institutions or processes?

9. How many of the following activities have you engaged in since retiring? (select all that apply)
   1. Wrote an op-ed regarding politics or the military?
   2. Interviewed on TV about foreign policy?
   3. Quoted or interviewed in print about politics or the military?
   4. Spoke publicly about respecting the oath to the Constitution?
   5. Co-signed a letter about national security matters?
   6. Co-signed a letter on another subject?
   7. Made a financial contribution to a political campaign or PAC?
   8. Publicly criticized a politician for politicizing the military?
   9. Endorsed a political candidate?
   10. Interviewed on TV about domestic politics?
   11. Spoke at a rally for a political candidate?

10. The previous question asked you to select which political activities you have engaged in since retirement. Whether or not you have engaged in them, which, if any, make you the most uncomfortable? (select all that apply)
    1. Wrote an op-ed regarding politics or the military?
    2. Interviewed on TV about foreign policy?
    3. Quoted or interviewed in print about politics or the military?
    4. Spoke publicly about respecting the oath to the Constitution?
    5. Co-signed a letter about national security matters?
    6. Co-signed a letter on another subject?
    7. Made a financial contribution to a political campaign or PAC?
    8. Publicly criticized a politician for politicizing the military?
    9. Endorsed a political candidate?
    10. Interviewed on TV about domestic politics?
    11. Spoke at a rally for a political candidate?

11. What has been the response of your fellow retired general and flag officers regarding your choice to engage/not engage in public political discussions?

          (free-form text response)

12. What was your branch of service?
    1. Air Force
    2. Army
    3. Marine Corps
    4. Navy

13. At what rank/grade did you retire?
    1. O7
    2. O8
    3. O9
    4. O10

14. Generally speaking, how do you think of yourself politically?
    1. Strong Republican
    2. Not very strong Republican
    3. Independent
    4. Not very strong Democrat
    5. Strong Democrat
    6. Unaffiliated
    7. Other__________

15. If you consider yourself an Independent, toward which party do you generally lean?
    1. Lean toward Republican
    2. Don’t lean toward either party
    3. Lean toward Democrat