Through its publications, INSS aims to provide expert insights, cutting-edge research, and innovative solutions that contribute to shaping the national security discourse and preparing the next generation of leaders in the field.



News | July 5, 2017

Deconstructing the “Warrior Caste:” The Beliefs and Backgrounds of Senior Military Elites

By Susan Bryant and Brett Swaney Strategic Insights

The Warrior Caste

In May 2017, The Center for a New American Security (CNAS) published a study exploring the implications of the rise of a “Warrior Caste” in American society.1 The authors conclude the implications are mixed. On the positive side of the ledger, they report that the United States now possesses a more ready and professionalized military than ever before. On the negative side, this force consists of a dwindling percentage of the population, who are more isolated from mainstream American society than ever before. As a growing percentage of service members and officers are coming from military families themselves, the choice to join the military is becoming more akin to a decision to join the family business. The result is greater divisions between the “Warrior Caste” and the civilian population than the numbers would indicate if taken at face value.   


The diminishing numbers of those who serve should give pause to a country that prides itself on an ideal of “citizen soldiers.” Currently only .4% of the US population is serving in the military.2 The number of U.S. citizens who are veterans has also dropped significantly since 1980, when 18% of U.S. citizens were veterans. As of 2014, the number had dropped to 8% and is expected to continue its downward trend.3 Despite their starkness, the numbers alone represent only half of the equation. The other half is determining to what extent these warriors are representative of the population they serve. Do their values reflect or diverge from that of the population writ large? The question of who constitutes the members of this caste, how they lean politically and what they value is central for those who study American civil military relations. 

Getting substantive data on the political leanings and opinions of these “warriors” has been notoriously difficult, perhaps in part because the military would prefer to present itself as an “apolitical block,” removed from the partisan machinations of those whom they serve. Non-partisanship runs deep in the American military and is exemplified by a long tradition of officers who choose not to vote at all. Among the most famous of these adherents are General William Tecumseh Sherman and General George Marshall. More recently, this tradition has been embraced by General David Petraeus. Those who choose note to vote believe that doing so impedes their ability to render independent judgment removed from politics.4 Thus for multiple reasons, surveys on the political leanings of the military have been sporadic and as a result, longitudinal analysis has proven difficult. Despite these obstacles, periodic surveys have been undertaken that show military officers have become more conservative and increasingly identified with the Republican Party since the advent of the All-Volunteer Force in 1973.5  

In order to see if these characterizations remain accurate and to add another data point to the research, the authors undertook a survey of elite military opinion between October 2016 and February 2017. The survey group consisted of 278 military officers in the rank of 05 (Lieutenant Colonel/Commander) through 07 (Brigadier General/Rear Admiral). These military elites were surveyed through National Defense University (NDU) and were enrolled as current resident senior service college students or had completed the flag officer “Capstone” program in the last 24 months (2015-2017). Those surveyed included members of the Army, Navy Air Force and Marine Corps.6 Although the respondents were overwhelmingly full time, active duty personnel, 12% self-reported as members of the Reserves or the National Guard. The questions were broadly conceived and were focused on developing a picture of how these officers self-identified in terms of race, religion, and educational background, rather than on their views of contemporary political issues.7 That said, the timing of the survey did coincide with the general election and thus, the election no doubt influenced responses. Even so, the data collected is simultaneously illuminating and unsurprising.  

The responses demonstrate that the population surveyed was overwhelmingly homogenous in terms of gender, race, and religion. The respondents were 94% male, 89% Caucasian, and 95% Christian. Given that these students are attending resident programs at NDU, the numbers foreshadow the composition of those who will lead the military in the coming decade. The answer is overwhelmingly white, Christian men. The lifting of the combat exclusion ban on women in 2013 may eventually have an effect in terms of gender, but there is no reason to believe that there will be a significant change in the basic distribution of military senior leadership in the next ten to fifteen years. The findings are vulnerable to the critique that this population represents an extremely narrow and particular group of officers who came of age concurrent with the height of the GWOT, and who were molded by multiple tours in Afghanistan and Iraq. The authors recognize this and counter that these officers provide a unique bellwether into the mindset of those who will lead the Department of Defense in the next decade.  

In terms of voting preference, 55% of those surveyed reported they were Republicans, 10% Democrat, 3% Libertarian and 32% stated they were Independent. This compares to 26% Republican and 29% Democrat and 40% Independent in the general population.8 The high number of Republican respondents compared to the general population is unsurprising, as is the relatively low number of military elites identifying with the Democratic Party. However, the number of military elites who described themselves as “independent” is particularly interesting, as this number represents a large jump from previous surveys.9 If verifiable in future surveys, this trend mirrors an increasing percentage of the American population who self-identify as independent voters.10 However, given the timing of this survey, the results could also indicate a distancing from the politics of a particularly ugly election rather than a distancing from a particular party. It’s simply impossible to disaggregate without additional data. 

Voting Preferences

The results also hint at a split along gender lines. Women were more likely than men to self-identify as Democrat and liberal. This tendency has been recorded in other surveys as well.11 However, in this case, the overall low number of female respondents makes the data statistically insignificant and thus it is impossible to draw definitive conclusions. Nonetheless, both of these hints are interesting and should be followed closely in future in surveys.  

The data also indicated that respondents were highly educated. Of those surveyed, 88% have at least one Masters’ degree. In addition, 15% of the flag officers also had a terminal degree (PhD, JD or MD). This statistic, although impressive, requires unpacking. There has been an increasing trend towards “in house” advanced degrees for military officers over the past two decades. More and more military officers are earning accredited masters degrees at military colleges and universities, and fewer have been receiving them at civilian schools. Although the survey instrument did not ask for the source of the degrees, it is reasonable to assume this is true in this case. 

As an example, only 22% of Army officers selected for Brigadier General between 2008 and 2016 have graduate degrees from civilian institutions.12 This phenomenon of “in house” educational credentialing is problematic on two levels. First, it further isolates future military leaders from the population they serve and also from the future civilian leadership they will serve with and for should they reach the pinnacle of military service. Second, schooling by and for military populations increases homogeneity of thought among senior military personnel. It is one more factor that contributes to the creation of a “Warrior Caste” and could also contribute to a group think mentality among its members.  

The CNAS report also considers the increasing importance of familial connections in the decision to join the service. The NDU survey also reinforces this trend. Of the respondents, 41% indicated that they come from military families. However, the NDU survey diverged significantly from previous surveys in terms of where the respondents call “home.” Although very few (4%) came from New England or the Pacific Northwest, there was no overwhelming majority of Southerners. Approximately 10% of the respondents came from the West Coast, 7% hailed from the Southwest and the remainder were split relatively evenly between the South, the East Coast and the Midwest. 

This divergence can be largely explained by looking at the source of commission. Fully 30% of the respondents graduated from one of the military academies. These institutions work to balance admissions evenly from across the United States. This fact likely accounts for the discrepancy between this survey and the geographic concentration of the force writ large. 

The consolidation of the warrior caste serves to highlight the growing differences between the military and the society they serve. Although far from truly alarming, there are differences between the backgrounds, values and political leanings of the elite members of the “Warrior Caste” and those of the society they serve. This disconnect inevitably diminishes a sense of shared sacrifice which will have consequences on the ability of society to serve as a check on the use of force, the political sustainment of force, and for the consideration of veteran affairs. The homogeneity of the warrior caste and the increasing prevalence of “in-house” education should also raise serious concerns about the threat of group-think, and remind us that a diversity of backgrounds, perspectives, and beliefs are vital to fostering creative and flexible military leaders with the capability to meet the many hydra headed challenges we face. That said, this survey represents the thoughts of a small, and very select segment of the American military at a single moment in time. The true value is the ability to continue to sample this population annually and develop a trend line in furtherance of research on civil military relations. National Defense University’s Institute for National Strategic Studies is committed to developing this database, updating it on an annual basis and allowing it to be used for academic study. 


  1. Amy Schafer, Generations of War: The Rise of the Warrior Caste and the All-Volunteer Force (Washington, DC: Center for a New American Security, May 8, 2017), available at <>.
  2. "Welcome to DMDC," Defense Manpower Data Center, last modified May 25, 2017, available at <>.
  3. Gretchen Livingston, “Profile of U.S. Veterans Is Changing Dramatically As Their Ranks Decline,” Fact Tank (blog), Pew Research Center, November 11, 2016, available at <>.
  4. Charles G. Kels, “The Non-Partisan Military” Armed Forces Journal (August 5, 2008), available at <>.
  5. For previous survey data refer to Peter D. Feaver and Richard H. Kohn, eds., Soldiers and Civilians: The Civil Military Gap and American National Security (Cambridge: MIT Press, 2001); Jason Dempsey, Our Army: Soldiers, Politics and American Civil Military Relations (Princeton: Princeton University Press, 2009); and Heidi Urben, “Civil Military Relations in a Time of War: Party, Politics and the Profession of Arms,” (PhD diss., Georgetown University, 2010), available at <>.
  6. Respondents in the grades of 05-06 were enrolled as full time senior service college students at the Eisenhower School for National Security Studies, National War College and the Joint Advanced Warfighting School. The respondents in the rank of 07 had been enrolled as Capstone students at National Defense University between 2015 and 2017. 
  7. “Capstone Questionnaire,” Institute for National Strategic Studies, February 2017, available at <>.
  8. Jeffrey M. Jones, “Democratic, Republican Identification Near Historical Lows,” Gallup, January 11, 2016, available at <>.
  9. For a comprehensive summary of past major surveys of military partisanship, see Heidi Urben, “Civil-Military Relations in Time of War,” (PhD diss., Georgetown University, 2010), 7-50, available at <>.
  10.  Jeffrey M. Jones, “Democratic, Republican Identification Near Historical Lows,” Gallup, January 11, 2016, available at <>.
  11. Heidi Urben, “Civil-Military Relations in Time of War,” (PhD diss., Georgetown University, 2010), available at <>.
  12. For a discussion of Senior Leader Talent Management see Michael J. Colarusso and David S. Lyle, Senior Officer Talent Management: Fostering Institutional Adaptability, (Carlisle Barracks, PA: Strategic Studies Institute, 2014). The numbers of General Officers with Civilian Graduate Degrees was derived from the publicly available biographies on the Army General Officer’s Management Website



Colonel Susan Bryant, DLS, USA is a Senior Military Fellow in the Institute for National Strategic Studies at the National Defense University. Brett Swaney is a Research Analyst in the Institute for National Strategic Studies at the National Defense University. The views expressed in this op-ed are the authors’ own and do not reflect the official policy or position of NDU, the Department of Defense, or the U.S. government.