A Modest Proposal for Military Suicide and Military Sexual Assault


Editor’s Note: This is the fourth article in a series on defense reform in the next administration. Please read parts onetwo, and three here at War on the Rocks.


Military suicide and sexual assault remain intractable problems, despite extraordinary efforts by the Department of Defense and Congress. What more can be done in the next administration? A place to start is with better data. This suggestion might not seem that far-reaching, but it is actually nothing short of revolutionary. Better data will allow policymakers to tailor interventions, measure program effectiveness, and establish goals by which to measure success — tasks all but impossible today. While it seems hard to argue against collecting and publishing better data, the Department of Defense seems loath to do so. After all, opacity is not a bug, but a feature – indeed a central stratagem – in large bureaucracies.

Military Suicide

Suicide is one of the oldest and most troubling questions in the social sciences. It almost defies analysis, for, as Emile Durkheim wrote more than a century ago:

Each victim of suicide gives his act a personal stamp which expresses his temperament, the special conditions in which he is involved, and which, consequently, cannot be explained by the social and general causes of the phenomenon.

Every quarter, the Department of Defense announces the latest heartbreaking statistics about military suicide, breaking down the numbers by service and by component. We oversaw the Pentagon’s suicide prevention office when we led the Personnel and Readiness office, and every quarter the report frankly left us with the same nagging questions that animated Durkheim’s classic work: What were we to make of it all? While every suicide is a tragedy, was is right to say, as many do, that the military faces a “crisis” in suicide, what with that word’s implication of a growing problem? Were there trends hidden within the aggregate data that might suggest better prevention policies or uncover poorly designed ones? Which interventions were working and which were failing? How would we even know? It was all very hard to tell. In response, we commissioned new reports, hoping to learn more about the problem. Alas, this work has never been released to the public. But it is possible to at least partly recreate those reports using publicly available information. We do so here.

To start, we should offer a tedious but important note on methodology, as it is essential to place the military suicide rate in the context of the broader civilian rate. The young are more likely to engage in suicidal behavior than are the old, and men are four times more likely to commit suicide than are women. All other things being equal, the suicide rate in the military should therefore be expected to be higher than in the civilian sector, given that the military is mostly young and male. But all other things are not equal. There are two opposing tendencies at work in the military community that do not manifest in the civilian population. First, the military screens for depression and other behavioral health conditions that are often the drivers of suicidal ideation. This would suggest that the military suicide rate should be lower than the broader civilian society. Second, servicemembers gain familiarity with firearms, which could easily make that most lethal method of suicide more common in the military.

With that throat-clearing aside, we have taken the publicly reported data from the Department of Defense and compared it to the age-adjusted and gender-adjusted suicide statistics for the whole country (data compiled by the Centers for Disease Control and Prevention). The disaggregated data shows that the suicide rate in the U.S. military before the wars in Iraq and Afghanistan was about one-half of the national average. Today, the suicide rate in the active duty military (20 per 100,000 people) is still slightly less than the age and gender-adjusted national average (23 per 100,000 people), but the rate has more than doubled over the last 15 years, while the civilian rate has only increased slightly. Every military service experienced a similar increase in the suicide rate, with the Army and Marine Corps having had the most serious problems. The reserve component, on the other hand, has a rate nearly 50 percent higher than the national average. Here are more specifics:

  • There were 265 suicides in the active component of all military services in 2015. This number has been roughly stable since 2011, when it spiked at 321. By comparison, in 2001 and 2002, before the current conflicts started, the active duty military saw only 142 suicides each year.
  • The rate of suicide in the active component in 2015 was 20 deaths per 100,000 people. In 2011, the peak year for active component suicides, the rate climbed to 23 per 100,000 people. In 2001 and 2002, the active component suicide rate was just slightly above 10 per 100,000 people.
  • Suicides in the civilian population are increasing. Indeed, the national suicide rate is at a 30-year high. Adjusted for age (to make comparable to the youth of the military) and adjusted for gender (to make comparable to the male-female mix of the military), the suicide rate in broader society has climbed from 19.4 per 100,000 people in 2001 to 23 per 100,000 in 2015. There are many explanations, all controversial, to explain this phenomenon.
  • In the active component Army, the rate of suicide has increased from about nine deaths per 100,000 people in 2001 to more than 23 per 100,000 in 2015. As late as 2004, the suicide rate in the active component Army was 10 deaths per 100,000 people. At its maximum, which occurred in 2012, the rate rose to 29.9 per 100,000, almost three times the pre-war rate.
  • In the Army National Guard, the rate of suicide peaked in 2013 at more than 33 deaths per 100,000 people. While falling from that sad height in more recent years, the rate of suicide has hovered at about 30 deaths per 100,000. The Army Reserve has only a slightly lower suicide rate. Among world militaries, this rate is matched only by that of the post-Cold War Russian army.
  • In the active duty Navy, the suicide rate in 2001 was 9.6 per 100,000 people, about the same as in the active component Army in that year. This rate was relatively stable through 2008, after which it climbed rapidly. In 2012, the suicide rate in the active duty Navy spiked at 18 per 100,000 people, before falling again to about 13 per 100,000 people in 2015.
  • In the active duty Air Force, the suicide rate in 2001 was a little less than 10 per 100,000 people, the same as in the active duty Army and Navy. The suicide rate was stable through 2007, when it began an increase which remains unabated to this day. The worst year for suicide in the Air Force was 2015, the most recent for which data is available. In that year, the rate was 19.3 per 100,000 people.
  • In the active duty Marine Corps, the suicide rate has generally been higher than in the other services. In 2001, for example, while the Army, Air Force, and Navy had about nine suicides per 100,000 people, the Marine Corps had 16 per 100,000 people (a total of 28 suicides). This number is volatile, though, due to the small size of the Marine Corps; a few additional suicides each year can dramatically affect the prevalence. The Marines are also the youngest service and have the fewest women, factors that would lead one to expect a higher suicide rate than in the other military services. The rate of suicide in the Marine Corps reached an apex in 2009, with 25 per 100,000 people (51 suicides). As in the Army and Navy, the second-worst year was 2012, with a rate of 23 per 100,000 people.
  • In 2014, the suicide rate for women in the active duty military was 11.4 per 100,000 people; among civilian women, the suicide rate was about eight per 100,000 people. Historically, the female suicide rate in the military has been less than six per 100,000 people. In other words, over the last 15 years, the female suicide rate in the military has about doubled its historical average and is now substantially higher than the civilian rate for females.








So, is there a “crisis” in military suicides? We believe that the answer is yes. It is true that the suicide rate in the military is less than the national average, and that point should not be overlooked. It should also not be forgotten that the suicide rate is lamentably rising in our society as well as around the world, for that matter. But it is startling to have seen the overall military suicide rate double over the last 15 years. The fact that the military does not accept (and eventually separates) servicemembers with behavioral health problems strongly persuades us that the suicide rate in the military should always be expected to be far less than in the broader civilian population. It isn’t, not by a lot. There is particularly a problem with female suicide in the military, which exceeds the national civilian female rate, even when it should be much lower. And the reserve component rate is frightening. Taken together, we believe this constitutes is a crisis.

Our favorite military theorist (of a sort) always insisted on asking “What can be done?” When it comes to overdetermined problems like suicide, that’s not an easy question to answer. Lots of programs are in place to prevent suicide, and their efficacy can be debated. It is impossible to know what the suicide rate would be in their absence, and some no doubt work well. But we believe that an obvious move is for the Department of Defense to make clear its desired endstate: a suicide rate a fraction of the national average (when adjusted for age and gender). More specifically, the next secretary of defense should announce his or her ambition to have a suicide rate of no more than 10 per 100,000 people — that’s a reduction of more than 50 percent. Each service and each component should then measure their programs against this metric. While this suggestion might seem like a low-calorie offering, we can only say that success will be had only when we clearly state our desired goal, something we don’t do today. It is hard to have a strategy if you don’t have ends to which to link your means.

The second thing we must do is publish more extensive and detailed data about suicides and suicide attempts. This data should be broken down not only by service and component, but also by age, gender, unit, geography, deployment history, and means of death. This information — as well as all other data collected on suicides by the Department of Defense — should also be made freely available to academic researchers. To give one example of how this might help, we learned at P&R that women in the military are four times as likely to use a firearm in committing suicide as are their civilian counterparts. Indeed, in 2014, two-thirds of the women in the military who committed suicide used a gun; in the civilian world, only one-third of female suicides are by firearm. This finding is not that surprising, given that women in the military are likely to have more familiarity with, and access to, guns than civilian women. But this suggests that certain interventions might be particularly successful in preventing suicides among military women. We ourselves don’t know what these interventions might be, but offering data that explains in greater detail the tragic phenomenon of military suicide is essential if others are to find innovative solutions.

Military Sexual Assault

As with suicide, sexual assault in the military is a scourge from which relief does not seem anywhere in sight. Dozens of legislative provisions and countless training programs have made little headway in eliminating the problem. Despite this, we do not counsel pessimism. While sexual assault might seem to be a much-studied problem, the Department of Defense has much more to do before it can fully understand the issue. It needs more data to reach a granular understanding of exactly who is committing sexual assault, who are the victims of sexual assault, the relationship between various forms of sexual misconduct, and the outcomes of cases in the military justice system. None of this is accounted for in any great detail today.

To do what we urge, there are three critically needed policy reforms. The first is something that is entirely missing: a uniform system of evaluating the efficacy and impact of any of the Department of Defense’s programs, initiatives, laws, regulations, and directives that have gone into effect since the establishment of its first sexual assault policy in 2005. Every task force and panel that looked at the issue of military sexual assault called for such measures, yet the will to execute this recommendation has not yet materialized. A new program or major statutory change should never be launched without designating metrics for measuring its success, as well as preparations for the collection of key information to assess outcomes. The secretary of defense has total authority to ensure that this critical work is done — something that is next to impossible in civilian jurisdictions across the country. He should insist upon it in the next administration.

The second reform is a more comprehensive database of sex-related cases. Amazingly, the Department of Defense is currently unable to provide aggregate data for all sexual assault, domestic violence, and child neglect, abuse, and sex crime allegations and their outcomes. The Sexual Assault Prevention and Response Office (SAPRO) collects sexual assault reports by adults against adults. It does not collect information about the sexual assaults of children, spouses, or intimate partners. The Family Advocacy Program (FAP), on the other hand, which possesses information about cases involving children and families, does not have the ability to collect data on command actions or the legal dispositions on those cases falling under its jurisdiction. Making matters worse, the approach to determining whether a case is categorized as belonging to SAPRO or FAP is completely subjective for reporting purposes.

For military justice statistics specifically, there is currently no reliable source of data available anywhere within the Department of Defense. While the SAPRO annual report offers a considerable amount of data about sexual assault incidents, there is neither consistency nor quality control on the information collected and relied upon to make sweeping assessments about the efficacy of the responses to such matters. Much of the data is erroneous or incomplete, as has been identified through the work of the Judicial Proceedings Panel, established by an exasperated Congress in 2012. The JPP is the first initiative to collect uniform, military-wide sexual assault case data directly from case documents, but the lack of uniform standards across the military services for recording and maintaining case information renders the data available incomplete and difficult to analyze, even when taken directly from case records. Additionally, the unitary sentencing requirement in the military — wherein one sentence is adjudged regardless of the number of offenses the accused is convicted of — makes any meaningful analysis of military sentencing data completely impracticable. The military sentencing scheme stands in contrast to civilian courts, in which a sentence is imposed for each offense convicted, and a judge determines whether the sentences run consecutively or concurrently. Critics of the system have long advocated removing the unitary sentencing structure and supplanting it with offense-specific sentencing. The Pentagon finally recommended such reform to Congress in the proposed Military Justice Act of 2016.

A complicating factor that should be reviewed is the use of “locally filed” reprimands. We learned at P&R that an astonishing number of sexual assault perpetrators had previous run-ins with the military justice system. In many cases, the perpetrators had a rather frightening number of such scrapes – sometimes six or seven. Why were these people allowed to remain in the military? One reason is that many forms of administrative punishment do not follow the service member over the course of a career. Instead, the reprimand is filed only at the installation where the incident occurred, and the record is destroyed upon a change of station. So a service member could have DUIs, other alcohol-related offenses, or a host of other sometimes rather serious (even violent) missteps and these will not follow him or her to later commands. (Whether an administrative sanction is filed locally or goes into the official personnel record of the service member is up to the discretion of the commander.) The military services pride themselves on giving young people a fresh start in life and resist the idea of their troops having a “dirt file.” Nonetheless, the fact that many administrative actions are filed only locally – especially in the Army, with which we have the most experience – means that repeat offenders can sometimes evade serious consequences, since they establish a new, clean record every time they PCS. This is, in our mind, a loophole whose time has long since passed. In fact, it would be smart to look more broadly at the whole administrative reprimand process, which is fraught with problems for both the accused and the military service.

The third and critical if controversial reform to improving the Department of Defense’s response to and prevention of sexual assault is a very simple one. It involves definitions. The current definition of sexual assault in Department of Defense policy is very broad: “intentional sexual contact characterized by use of force, threats, intimidation, or abuse of authority or when the victim does not or cannot consent.” The term encompasses “a broad category of sexual offenses,” including “rape, sexual assault, aggravated sexual contact, abusive sexual contact, forcible sodomy (under the Uniform Code of Military Justice, this includes both forced oral or anal sex), or attempts to commit these acts.” The amount of confusion caused by this self-referential definition, upon which all sexual assault prevention and response training is based, cannot be overemphasized.

To better understand differences in outcomes, sexual offenses should be classified by the Department of Defense in three basic categories: (a) sexual assault, which constitutes an unwanted penetrative sex act; (b) unwanted sexual contact, which constitutes an unwanted, non-penetrative sexual touching; and (c) sexual harassment, which constitutes repetitive and unwanted sexual words and actions. While these are far more simplistic than the prevailing definition, for policy and training purposes these distinctions are easy to understand and yet remain consistent with military and civilian legal definitions. These distinctions are not intended to minimize the very real anguish of victims of any type of misconduct, but they do allow a better understanding of the problem. For example, we learned at P&R (from internal data) that “unwanted sexual touching” in the military occurs most often in the workplace, and is usually a follow-on from sexual harassment. Penetrative sexual offenses, in contrast, were not foreshadowed by sexual harassment in the workplace and, according to the data we saw, very rarely were related to the “unwanted sexual touching.” This hints at different remedial approaches, solutions obscured by today’s totalizing definition. Moreover, sexual harassment is not even considered a part of the Department of Defense’s sexual assault policy at all and is handled not by SAPRO but by the Equal Opportunity Office. The potential confusion between sexual harassment and some forms of unwanted sexual contact necessitates that these forms are viewed together on the same continuum of harm. Perhaps most importantly for assessing outcomes, reporting processes about sexual misconduct should be aligned so that reports in all three categories are uniformly and consistently tracked and monitored. The simplified definitions would make it much easier for victims to consistently articulate which category of misconduct they experienced. This approach would also facilitate an important, if sensitive, policy discussion about the difference between sexual misconduct that constitutes a crime punishable by confinement and potential lifetime sex offender registry versus that which constitutes workplace misconduct for which one should more appropriately be fired or demoted, but not necessarily criminally prosecuted.


Military suicide and military sexual assault are among the most sensitive issues facing a leader at the Department of Defense. Whatever else can be said about the Pentagon’s approach to these issues, it cannot be gainsaid that extraordinary attention and resources have been devoted to combating both problems. We can also say that the people who oversee prevention efforts in the Department of Defense SAPRO, Suicide Prevention Office and all of their military service counterparts have a dedication that is unsurpassed in the Pentagon. But more must be done.

The reforms we recommend here are an attempt to move the Department of Defense from an awareness of the vexing problems of suicide and sexual assault to an understanding of those problems. Understanding precedes action, after all. And, if these reforms we discuss appear modest in scope, they prepare the battlefield for more intense action. Better data more clearly presented combined with better definitions more clearly delineated are the scaffolding on which more effective preventions efforts will be built. As General Gordon Sullivan, the former Chief of Staff of the Army famously said, the intellectual comes before the physical. He wasn’t talking about military suicide or sexual assault when he offered that advice, but it holds true here nonetheless.


Brad Carson served as acting Undersecretary of Defense for Personnel and Readiness from 2015-2016. Previously, he was Under Secretary of the Army and General Counsel of the Army. He also was a U.S. Congressman. He is a veteran of Operation Iraqi Freedom. He can be contacted at bradrogerscarson@gmail.com.

Morgan Plummer departed the Obama administration after serving on the immediate staffs of the Under Secretary of Defense for Personnel and Readiness, Under Secretary of the Army, and Deputy Secretary of Defense. Prior to that, he served more than ten years on active duty in the U.S. Army. He can be contacted at mcplummer1789@gmail.com.

Note on data: The date used in this article were compiled by the authors using Annual Reports of the Department of Defense Suicide Prevention Office, Centers for Disease Control Web-Based Injury Statistics & Reporting Query System, and academic research. Data on the reserve component is available only from 2010 to present. The annual number of suicides in the Air Force Reserve, Air National Guard, Navy Reserve, and Marine Corps Reserve are, in each, so few and unpredictable that calculating a rate is avoided here. National data is adjusted for ages 17-60 and reflects a male/female ratio of 85:15, roughly that of the U.S. military. Special thanks to We Are Flint for chart design.

Image: U.S. Air Force, Airman 1st Class Corey Hook