Sexual Assault in the Military, Part I: Message Fratricide - How Our Own Words Undermine Our Efforts

By the Numbers

I am no statistician, so take these numbers with a grain of salt. I set out to compare the rates of sexual assaults reported in the military with those of American society as a whole, and what I found surprised me. You can find reports, estimates, and projections of all kinds on the subject. As evidenced by the Free Beacon article cited above, these numbers can be interpreted in any number of ways and support any number of potentially misleading conclusions. I chose to trust the numbers reported to Congress by the DoD Sexual Assault Prevention and Response Office (SAPRO) for the military, and the FBI Crime Statistics for society. The last data available for both comes from 2014.

In order to compare apples to apples, I am only considering unrestricted reports of sexual assaults by active duty victims, as opposed to the prevalence estimates DoD SAPRO published as a result of the RAND Corporation’s “Military Workplace Study.” I have limited the scope to unrestricted reports because only those can, and must, be reported to law enforcement in the same way civilian jurisdictions report to the FBI. There is no civilian equivalent to the military’s restricted sexual assault report, so the FBI has no mechanism or means to gather such data.

By extrapolating the prevalence rates among those service members who responded to the survey, and attempting to account for the number of such crimes that go unreported, the RAND study concluded that there were an estimated 18,900 sexual assaults on service members in 2014. While staggering, that number is a significant improvement over the 26,000 estimated by the same process to have taken place in 2012 (RAND conducts the study bi-annually). However, in 2014, there were only 4,768 actual reports made to military authorities by service members claiming they were assaulted while serving in the military.

The FBI does not provide estimates or extrapolations for society as a whole, so all they can report are those crimes that are reported to law enforcement in the jurisdictions from which they receive data. In addition to the raw totals, the FBI presents the data as a rate per 100,000 individuals in the reporting areas. For ease of comparison, I will do the same with the military’s figures. Additionally, and importantly, the FBI broadened the definition of reportable rapes and sexual assaults in 2013, following implementation of the National Incident Based Reporting System (NIBRS). Some jurisdictions reported under the new definition in 2014, while others retained the old definition.

While broader than the pre-2013 definition, the new FBI definition is still much narrower in scope than what DoD classifies as sexual assault. In other words, conduct that does not meet the FBI’s definition of rape or sexual assault, for reporting purposes, will often meet the military’s definition. By that measure, DoD’s relative numbers will virtually always be higher. I will discuss the terminology and definitions in more detail below. For comparison purposes, I am including rapes reported by the FBI under both definitions in 2014.

Now that I have gotten the methodology out of the way, here are the numbers:

While the reports received by the FBI do not cover the entire U.S. population, their methodology reveals that they receive reports from jurisdictions covering over 295 million people, which is fairly reflective of the population overall. (The 2010 U.S. Census reported a resident U.S. population of roughly 308.7 million people). In the jurisdictions reporting to the FBI under both definitions of rape and sexual assault, the rate of reported sex crimes was 34 per 100,000 individuals, or 0.03%.

According to DoD, there were 1,338,487 people serving on active duty in 2014. DoD received 4,768 reports of sexual assault for offenses that occurred while the victims were serving in the military (516 reports that year were for assaults that occurred prior to joining). 4,660 of those were unrestricted reports. Thus, 0.35% of the military population, or 348 per 100,000 service members, reported in 2014 that they were sexually assaulted while in service.

That is more than ten times greater than that reported by the general population, which is what might lead one to reasonably believe that the military has a serious sexual assault problem.

In 2014, 200,692, or roughly 15%, of the over 1.3 million active duty service members were women. However, in more than 80% of the sexual assault reports DoD received in 2014, the victims were women. Of the 4,768 reports in 2014, 3814 of them came from female victims. That’s roughly 2% of all women who serve, or over 4,000 women in 2014 alone. And those are just the ones who reported it.

DoD estimates that reported sexual assaults only account for approximately 25% of actual sexual assaults. If you accept the entirely reasonable premise that many military rapes and sexual assaults, 75% by DoD’s estimate, are never reported, the scope of the problem seems truly staggering. Indeed, this assumption is how RAND concluded that almost 19,000 sexual assaults occurred in 2014, bringing the estimated percentage of military victims to 1.4%, or 1,420 out of 100,000 service members. Again, however scientific they may be, these are estimates, and I would be wary of drawing too many conclusions based on them.

The above statistics, in a vacuum, might lead one to conclude that you are ten times more likely to be raped or sexually assaulted in the military than you are on the streets of America (unless that street is located on a college campus). The chances may very well be increased. There is more than one member of Congress and advocacy organization who would like you to believe just that. However, it is not entirely the case, because, despite my best efforts, it is virtually impossible to compare apples to apples in this arena, after all.

What’s in a Name?

Author’s Note: It is impossible to discuss or understand the issues related to sex crimes without getting into graphic detail. If you are squeamish or uncomfortable discussing penises and vaginas, you might want to skip to the next section.

Words matter. It is a principle of almost every aspect of law, politics, and journalism that the words you use have meaning. Issues of the greatest national importance can sometimes boil down to a comma in the Constitution or poorly chosen words. Indeed, if you are former Supreme Court Justice Antonin Scalia, literally nothing in all of the law matters more than the specific words in the Constitution. His landmark opinion on the Second Amendment in District of Columbia v. Heller reads more like a lecture on grammar and etymology than a legal opinion. If you hear an attorney in court ask a witness, “wouldn’t you agree that words matter…that they have meaning,” you can rest assured the attorney is about to resurrect some previous contradictory statement of the witness and shove it unceremoniously, and not at all delicately, down that witness’s throat. It’s Trial Advocacy 101.

With that in mind, how can anyone expect to define or understand a problem as serious and attention-grabbing as “sexual assault” in the military when we cannot even define or understand the meaning of the term itself? How can you accurately compare apples to apples when we can’t even agree amongst ourselves on what an apple is, or more accurately, which flavor of apple should count for reporting purposes?

As I alluded to earlier, the FBI changed their definition of rape for reporting purposes in 2013, but some jurisdictions still report based on the old definition. As a result, some jurisdictions operate with a vastly broader category of rape and sexual assault than others. Prior to 2013, rape was defined as “carnal knowledge of a female forcibly and against her will.” That definition is 80 years old and reflects an extremely limited set of circumstances that could constitute the crime of rape. Let’s unpack it a little bit.

First, courts have interpreted the term carnal knowledge to be synonymous with sexual intercourse, which is defined as contact between the sexual organs of a man and a woman, usually involving penetration. So far, by definition, the crime can only be committed between a male and a female, a fact further reinforced by the inclusion of “of a female” following carnal knowledge. “Of a female” also connotes that only a man can commit the crime against a woman, and not vice versa, for a woman cannot penetrate another woman with an “organ,” as here defined.

Keep in mind, too, that the “organ to organ” contact required rules out many potential mechanisms of sexually assaulting someone. Forcing a penis into a mouth or anus would not be rape. Forcing a hand or finger into any orifice would not be rape. Likewise for any foreign objects, such as beer bottles or broomsticks. While most of us would probably agree that all of those things sound like rape, they would not be reported under the old definition.

The next key element to the crime is force. The carnal knowledge of a female must be accomplished by some degree of force. Force is understood, for these purposes, as the degree of power exerted on another sufficient to overcome their ability to resist. Thus, the assailant has to exert physical force on the victim. It could take many forms — striking, choking, restraining, hitting over the head, etc.-but there must be some form of force used for the act to qualify as rape. Thus, having sex with an unconscious or severely intoxicated woman would not be rape. Rendering a woman unconscious by drugging her drink and then having sex with her would not be rape. Threatening to kill or severely injure a woman or her loved ones if she did not have sex with you, by itself absent any attempt to actually harm, would not be rape. In none of those examples is actual, physical force employed to accomplish the carnal knowledge.

Finally, the carnal knowledge must be “against her will.” Put another way, it must be without her consent. Consent is a very tricky issue in rape and sexual assault. The only real evidence that a person did or did not consent to a sexual encounter is in that person’s head. You can make inferences based on the circumstances, but it almost always comes down to someone’s word (usually against someone else’s in the so-called “he said, she said” situation). What if, as in the examples above, the woman is highly intoxicated, and perhaps incapable of giving knowing consent, but does not appear as such to others? What if the victim did give consent, but doesn’t remember doing so, not an uncommon occurrence where alcohol is involved? What if a person has a physical age of 20, but a mental development age of 12? Can they consent?

The waters muddy even further when you consider that reasonable mistake of fact as to consent is a defense to rape and sexual assault. If the assailant reasonably believed that the victim had consented, even if he was wrong, he is not guilty of rape. Apply that to a situation where a victim is intoxicated, where they are mentally disabled, or where they have consented to some acts, e.g. vaginal sex, but not others, e.g. anal sex, and you can see how the “against her will” portion of the definition makes the crime often difficult to prosecute.

Post-2013, this narrow definition gave way to a more all-encompassing one, presumably intended to more appropriately reflect changes in the law and what society wanted to consider criminal in this context. The new FBI definition of rape is “penetration, no matter how slight, of the vagina or anus with any body part or object, or oral penetration by a sex organ of another person, without the consent of the victim.” This definition fixes some of the problems I discussed above. Penetration is still required, as is lack of consent, but note the addition of “no matter how slight” following penetration. This effectively lowers the bar for what constitutes penetration by explicitly stating that it can be only the slightest of penetrations to qualify. This also raises some interesting anatomical questions. At what point have you slightly penetrated a woman’s labia or mouth or anus? It starts to blur the lines between a rape or sexual assault and another family of crimes that I will discuss in more detail below, sexual contact.

Most importantly, the definition has now been expanded to include all of the other means and mechanisms of penetrating any or all of the bodily orifices. Note, also, that force is no longer required. Penetration OF virtually anything WITH virtually anything without a victim’s consent is rape.

One would think that vastly expanding the definition of what is reportable to this degree would cause a serious spike in the reported rapes once the new definition went into effect. The results seem to be more moderate than predicted. The FBI reported more than 11,000 more rapes in 2013, the first year of the new definition, than in 2012, a 15% increase. Clearly, though, under these circumstances, more reports does not mean more actual rapes. We have just changed which flavor of apples we are counting. DoD has learned this lesson exceptionally well in the past few years, and have even managed to turn dramatic increases in the number of rapes reported by the force into a good news story in their messaging campaign on the subject — “Our program is working because more people are comfortable coming forward to report.”

If the FBI’s one-time change of definition muddied the waters for sexual assault reporting and tracking across the country, then DoD has dredged up the Mississippi River in trying to tackle this problem and convince Congress and the public that we have it under control. The military legal definitions of rape and sexual assault have not just changed multiple times in the last few years, but DoD’s own definitions are internally inconsistent, misleading, and confusing. The source of this problem lies in the fact that the definition of sexual assault for DoD SAPRO, and its subordinate service programs, is substantially different than the legal definition of sexual assault under the Uniform Code of Military Justice (UCMJ).

Go back and look at the EXSUM of the 2014 SAPRO report to Congress. In that one document alone, sexual assault is defined three different ways (See page 5, footnote 1; page 6, footnote 7; page 8, footnote 12). The RAND studies, from which the massive prevalence rates of estimated sexual assaults in the military are derived, used two different definitions in two different surveys. One asked about “unwanted sexual contacts” and the other about “sexual assault.” And yet, everywhere in the text of the document, DoD refers to it as sexual assault without agreeing, even with itself, on a standard definition of that term. The multiple definitions are so easily and frequently conflated that it is never clear exactly what we are talking about when we refer to the statistics. The UCMJ is passed into law by Congress, so there is little DoD can do, besides lobby, to change those definitions (and change they have, but that’s a subject for a later post). What DoD absolutely can control is the definition of sexual assault in their own directive.

“Department of Defense Directive 6495.01 defines sexual assault as intentional sexual contact characterized by use of force, threats, intimidation, or abuse of authority or when the victim does not or cannot consent. The crime of sexual assault includes a broad category of sexual offenses consisting of the following specific Uniform Code of Military Justice offenses: rape, sexual assault, aggravated sexual contact, abusive sexual contact, forcible sodomy (forced oral or anal sex), or attempts to commit these offenses.” -2014 SAPRO EXSUM, footnote 1.

The definition is circular and contradictory on its face. Sexual assault is sexual contact, but includes UCMJ offenses including sexual assault. It’s a perfectly good definition, but not for the word it is trying to define. Ironically, the key to solving the puzzle is in the definition itself. The term it more appropriately and accurately defines is “sexual offense.”

As discussed regarding the FBI definitions, and as is more starkly clear under the UCMJ, rape and sexual assault are penetrative offenses. Under the current version of Article 120 of the UCMJ, as amended in 2012, rape and sexual assault require penetration, however slight, of the vulva, anus, or mouth of another person by the penis or any other body part or object. This fairly closely matches the FBI’s post-2013 definition (notice the absence of any mention of consent, however, except in a very specific version of sexual assault).

Sexual contact, on the other hand, only requires touching. Specifically, touching or causing another to touch certain specified body parts with the intent to abuse, humiliate or degrade, or touching any body part with the intent to arouse or gratify sexual desire. This can range from a pat on the rear-end as someone walks by to a shoulder massage to shoving your hands down someone’s pants. It all comes down to why you did it, i.e. intent.

All of these things taken together, along with attempts under Article 80; these same offenses against children under Article 120b; forcible sodomy under Article 125; and indecent viewing, recording, and broadcasting, forcible pandering, and indecent exposure under Article 120c, are properly defined as sexual offenses, or more commonly sex crimes. Convictions for the overwhelming majority of these offenses will land you on a sex offender registry.

Here we get to why words matter so much in this context. If Congress and the public hear that there were 26,000 sexual assaults in the military in 2012, the first image that jumps to mind is of one of the most violent, violating, degrading acts one human can do to another. The FBI rightly defines rape, and by extension sexual assault, as violent crimes in its statistics. The very core of our mental and physical autonomy and integrity is attacked in crimes like these, not to mention the actual physical injuries that can result. According to our own messaging, in 2012, 26,000 people reported that they were victimized in that way while serving in uniform.

By that definition, it is no wonder that people are horrified and outraged when they hear such a number. And no wonder the percentage of service members who have been victimized under this definition is so much higher than the rest of society. Remember, the FBI only reports penetrative offenses. The military is reporting every crime of a sexual nature as sexual assault, when in fact the majority of them do not meet this definition.

None of this should be taken to mean that sexual contact and other non-penetrative offenses, and the victims of them, are less deserving of our efforts to fix the problem. Sexual harassment and abuse deserve the same attention and energy towards resolution as do the penetrative offenses. However, there is a qualitative difference between a sexual assault and a sexual contact or indecent exposure, and that difference can shape perception. Conflating the two has only made the perception problem worse for the military.

Moreover, the terminology and the numbers we are using do not reflect reality. They exaggerate the amount of sexual violence we are actually facing and, thus, the scope of the problem and our capacity to deal with it. This is where DoD has lost the messaging and public relations battle surrounding this issue. Words have meaning, and we have failed to define the narrative by ensuring shared understanding of the problem we face.

The Numbers, Revisited

Now that we have better defined what it is we are trying to measure, let’s unpack the statistics further and see if we cannot find the actual flavor of apples we need to be talking about.

Of the 4,660 unrestricted reports received in 2014, 48%, or 2,237, were for penetrative offenses (rape and sexual assault), substantially similar to those offenses reported by the FBI under either definition. That reduces the overall percentage of service members who reported they were raped or sexually assaulted in 2014 to 0.17%, or 167 per 100,000. That is less than half of what it initially appeared above, but still more than 5 times higher than the civilian rate.

However, remember that the perception among Congress and the public is that the core of the problem is service members victimizing other service members. Of those same 4,660 unrestricted reports for which we have the necessary data (636 cases were excluded due to insufficient demographic data), only 63%, or 2,528, had a service member victim and a service member subject. Another 14%, or 575, were service member victims and unknown subjects that may or may not have been service members. Thus, worst case scenario, 0.2% of service members, or 232 per 100,000, were victimized by other service members. Best case, 0.19% or 189 per 100,000.

Assuming that the percentage of service member on service member reports applied across the gamut of possible sexual offenses, i.e. there were not drastically more service member on service member rapes than service member on service member contacts, we can calculate the number of service member on service member penetrative offenses as 1,409 (63% of 2,237). Now it is 0.11% of the total force, or 105 per 100,000 service members. If we assume that all of the unknown subjects in the reports were service members, the number rises to 1,722 reports of penetrative offenses (0.13% or 129 per 100,000).

This is the same flavor apple that the FBI is reporting. Thus, instead of the risk of being sexually assaulted in the military being ten times greater than society as a whole, as originally thought, it is actually about three times greater. Again, that is still way too high and way too many, and there may very well be many more penetrative offenses that go unreported. However, it puts the problem into a much less catastrophic perspective than if we take the numbers reported by DoD at face value.

If 80% of reports across the board are made by women, then 1,127 of the penetrative offense reports were made by females. That is 0.56% of the total female population of the military, or 562 female service members per 100,000. The FBI does not break down their numbers by gender, and the proportion of civilian society that is female is much higher than in the military, so I am unable to make an accurate comparison to the number of female civilian victims per capita. The number is much more frightening given the relatively small percentage of the force that women make up, and remains strong evidence that the military still has a long way to go in correcting its culture of misogyny.

The remaining report demographics cover non-service member subjects victimizing service members (5%), which DoD has limited ability to affect, and service member subjects victimizing non-service members (18%). The latter DoD does have greater ability to affect, and they are certainly cause for concern, but they do not rightly apply to the narrative of service members victimizing each other. Thus far, this analysis has been victim-centric — we are trying to determine how many service members are being victimized. Addressing sexual assaults by service members on non-service members would have to take a subject-centric approach, perhaps a topic for another day.

“We cannot, however, simply prosecute our way out of this problem.”

Finally, I will address the other side of the coin when it comes to sexual assault tracking and reporting — namely, what happens with all of these reports? The prosecution and conviction rates for sexual assault in the military appear to be as staggeringly bad as the reporting rates are staggeringly high. Indeed, the perceived failure by the military to adequately prosecute and hold offenders accountable has further motivated prominent members of the Senate to introduce legislation to heavily revise the military justice process in order to fix this perceived shortcoming.

GEN Odierno’s quote above, and Senator McCaskill’s response to it, illustrate the divergent points of view on the utility of prosecution in combatting sexual offenses. GEN Odierno, and the military in general, believe that the problem cannot be solved solely through punishing offenders. While offenders should most certainly be punished, and justice sought for victims wherever posisble, the change must be cultural.

If the goal is to prevent or eradicate sexual victimization in the military, then, by itself, prosecuting and punishing offenders after they have offended cannot ever solve it. Instead, DoD has focused its efforts on raising the stigma and likely consequences associated with the crime to such high levels that service members will not commit the acts in the first place. Part of this effort does come from prosecution insofar as the likely consequences are often determined by courts-martial, but at its core, the goal is to change people’s minds about the acceptability of sexual violence before the possibility of becoming an offender even occurs. The Not In My Squad initiative is the latest iteration of this effort.

Senator McCaskill, herself a former prosecutor, does seem to believe that you can “prosecute your way out of this problem.” She goes on to make the point that prosecution can be effective against sexual predators, for whom rape and sexual assault are not about sex, but about dominance and violence.

Dominance and violence are certainly prominent parts of military culture, perhaps even the raison d’etre of the military as an institution. But prosecution will not change that. As I said, by the time you get to prosecution, it is way too late to prevent the sexual assault. In order to get at the root of the problem, it must begin long before with indoctrination of service members into a culture of dignity and respect.

Any prosecutor will tell you that these cases are hard, each one is different, and they all have to be considered on their own merits. Sex crimes prosecutions frequently lack much in the way of physical evidence. Most sexual assaults do not involve significant physical violence, so readily apparent injuries are not usually present. SANE examinations cannot distinguish between consensual and non-consensual sex. They can only tell you if sex occurred. Occasionally, if the victim reports the crime immediately, and has not showered, brushed her teeth, changed, or washed her clothes before going to a hospital, DNA can be obtained. For most sexual assault prosecutions, this is the exception rather than the rule. Even where DNA is found, it is only useful if you have a known individual’s DNA with which to compare it. If a victim can’t identify her attacker or give a useful description, then DNA does you no good.

The vast majority of the time, these cases come down to people’s word, and witness testimony, especially about what may have been the most traumatic event in the person’s life, is notoriously unreliable. Add in alcohol, drugs, and/or the vagaries of human memory, and you can see how reaching “beyond a reasonable doubt” becomes difficult.

A word here about the standard of proof: Most people have heard the term “beyond a reasonable doubt” in relation to criminal prosecutions, but I wonder how many really understand what it means. As legal standards go, it is very, very high; the highest possible, in fact. As a judge instructs a jury (“panel” in military courts), beyond a reasonable doubt is not “a mathematical, or 100%, certainty,” but it is awfully close. Rather, beyond a reasonable doubt means that every reasonable explanation for what happened is excluded except for that of guilt. In other words, if there is any reasonable explanation for what happened besides “a rape occurred and the accused (military-speak for “defendant”) committed it,” then the accused must be found not guilty. Combine this burden with the difficulties of proof in these types of cases and prevalent misconceptions about how rapists and victims are “supposed” to act, and a high conviction rate becomes less and less likely.

Furthermore, sex crime investigations, for various reasons, usually take a significant amount of time to complete. In 2014, the average was four and a half months. Witnesses, especially victims, often give multiple statements to address inconsistencies and other questions that develop throughout the course of the investigation. As a result, stories frequently change over time. Sometimes these changes are minor, and sometimes they are not so minor. Either way, they provide ample fodder for attorneys to attack the credibility of the witness. In the absence of physical evidence, credibility is often all you’ve got, and cases often turn almost entirely on whether or not the judge or jury believes the victim.

So how does the military stack up to the civilian world when it comes to prosecuting sexual violence?

On this question, it becomes even more difficult to compare apples to apples. For one thing, military law enforcement classify and submit their reports based on the most severe offense believed to be alleged at the time the report is made, or at least, very early on in the investigation process. What may sound like a sexual assault at first often turns out not to be upon further investigation. If, for instance, a victim alleges that an assailant touched her vagina, that could mean there was penetration (remember “however slight”), and it is initially investigated as a sexual assault. However, it later turns out that the victim still had her underwear on, which makes penetration highly unlikely (though not technically impossible), and the case becomes a sexual contact case. The case would still be reported as a sexual assault, albeit unfounded (16% of reports were unfounded in 2014). So to say that there were 1,409 reports of penetrative service member on service member offenses does not mean all 1,409 necessarily turned out to be legitimate penetrative offenses.

The definition problem comes into play here, as well, because, when analyzing data on disposition of the reports, SAPRO is still using the flawed definition of sexual assault that includes all of the sexual offenses. (They at least make an effort to differentiate by listing them as “Sexual Assault Offenses” but that does little to alter the perception that they are all sexual assaults.) If a case is reported as a sexual assault under SAPRO’s definition, but it only meets the legal definition of a sexual contact or a physical assault (touching without the requisite intent), the accused will only be charged with a sexual contact or a physical assault, and is potentially more likely to be dealt with by some disciplinary action less than court-martial.

Thus, when examining the numbers, it looks like a sexual assault went un-prosecuted or received a “slap on the wrist,” when in fact it never had a chance of being prosecuted as a sexual assault in the first place because it didn’t meet the legal definition. Nearly a third of reports of sexual assault (SAPRO definition) in 2014, for which there was sufficient evidence to take disciplinary action, were substantiated as something less than a sexual offense (by either definition).

Furthermore, does the sexual contact or other lesser offense really merit a court-martial? Remember, a broad range of things can be classified as a sexual contact, and it all boils down to the intent with which the person was touched. A smack on the rear-end or the fondling of a breast are certainly inexcusable and must be dealt with, but do they really merit sex-offender registration (a requirement if you are convicted of sexual contact at court-martial) and jail time?

Commanders have many administrative and non-judicial tools at their disposal when it comes to addressing misconduct in their ranks, and they are absolutely right to tailor their response based on the circumstances of each individual case. The misclassification of the report as a sexual assault makes it easy for opponents to point to DoD’s own numbers to say that people accused of “sexual assault” (SAPRO definition) got a slap on the wrist via discharge, non-judicial punishment, or reprimand when in fact that was the appropriate level of disposition for what the report ultimately was.

As for the numbers, there are three tiers to this analysis. First, how many credible reports have sufficient evidence to consider prosecuting? Second, how many do we actually prosecute? And third, of the ones we do prosecute, how many result in convictions?

Once you remove the reports where DoD did not have jurisdiction over the offenders (civilians and foreign nationals), where the victims declined to participate in the investigation or the evidence was otherwise insufficient to prosecute, and where the statute of limitations had expired or the victim or subject died before the case could be resolved, there were 1,550 cases of sexual assault (SAPRO definition) with sufficient evidence for commanders to take disciplinary action in 2014.

Of those, 318 (21%) received non-judicial punishment, 111 (7%) were administratively discharged, and 123 (8%) were subject to other administrative action, which usually means a written reprimand. The substantial majority (998; 64%) had court-martial charges preferred against them (“prefferal of charges” is the military term for officially charging a service member with a crime at court-martial).

Disposition information was only available for 861 of those cases upon the completion of the 2014 SAPRO report. The rest of the cases bled over into the following year. 176 (20%) had charges dismissed (40 of those still received non-judicial punishment), and 97 (12%) were allowed to resign or be discharged in lieu of court-martial, leaving 588 (68%) to go to trial.

By contrast, while hard numbers on nationwide civilian prosecution rates are not available, the widely accepted, or at least often repeated, estimate of civilian reports that are referred to civilian prosecutors for charging is 17%. This would seem to indicate that DoD prosecutes four times as many allegations of sexual assault than civilian prosecutors. But this is misleading, once again, because of the infinitely broader definition of sexual assault that DoD uses.

Of those 588 cases that were taken to court-martial, 74% were convicted of at least one charge, sexual or otherwise (DoD does not track how many who are charged with a sexual offense are convicted of a sexual offense). 73% of those went to jail, and almost 60% received a punitive discharge (Bad Conduct or Dishonorable). The rest (154; 26%) were acquitted of all charges. Thus, the numbers do not bear out that the military is somehow unable or unwilling to hold offenders accountable.

Admittedly, the conviction rates in the civilian sector are probably higher. In the civilian world, something like 95% of all cases result in guilty pleas. There is no reason to believe this doesn’t apply equally to sex crimes prosecutions. Having worked with many civilian prosecutors’ offices, both state and federal, I have learned that the likelihood of winning at trial plays a far greater role in their decisions to prosecute than it does in the military. It is also likely that the military will charge allegations that civilian prosecutors have already declined.

We take more cases to court-martial, many of which, I can say from experience, civilian prosecutors would never charge in the first place because the likelihood of winning was so low. This was true even before the issue garnered such wide media and congressional attention. Is it any wonder that we lose more of them?

It is also worth noting that judging the seriousness with which the military takes this problem by the number of people we convict is flawed. That is not how systems of justice are supposed to work. Indeed, the standard of proof is as high as it is precisely because it is supposed to be exceedingly difficult for the government, civilian or military, to subject people to such severe consequences. That is what due process of law is all about. It is only after we are as absolutely sure as we possibly can be that you committed a crime that we can impose the most severe penalties known to our society. Anything that falls even a little bit short of this high bar must not result in a conviction, or it makes it too easy for the government to abuse its power. A tie, or even a close call, under these circumstances, most emphatically goes to the runner.

It is telling that a quarter of the time (154 out of 588), even when we think there is solid evidence to support prosecution, we still get it wrong and the trials result in acquittals. An acquittal does not necessarily mean that a guilty man got off, any more than it necessarily means that an innocent man was exonerated. The justice system gets it wrong all the time.

Nor can we be certain that the other 434 who were convicted actually did what they were accused of doing. A judge or jury may have said so, but only those people who were involved in the crime know for absolutely certain. It is entirely possible that legitimately innocent people had such a body of evidence, however incorrect, stacked against them that prosecutors were able to obtain a conviction anyway. Trials are less about the truth than they are about the weight of evidence. As the intrepid defense lawyer in the landmark military justice film, A Few Good Men, said, “it doesn’t matter what I believe, it only matters what I can prove.”