Sexual Assault Experiences in the Active-Component Army: Variation by Year, Gender, Sexual Orientation, and Installation Risk Level

In February 2021, U.S. Secretary of Defense Lloyd Austin ordered the military services to take “immediate” action to address sexual harassment and sexual assault in the military, which included both a focus of efforts on “high-risk” military installations and the creation of a 90-day independent review commission to conduct an “immediate, impartial assessment” of the U.S. Department of Defense's (DoD's) efforts to prevent sexual harassment and assault (U.S. Secretary of Defense, 2021).¹ To guide its sexual assault prevention and response efforts, the U.S. Army asked RAND Arroyo Center to create descriptions of soldiers’ experiences of sexual assault in the active-component Army.

In this study, we provide the results of an exploratory, descriptive analysis of active-component soldiers’ sexual assault experiences. The analysis was conducted using the 2016 and 2018 Workplace and Gender Relations Survey of Active Duty Members (WGRA) (Breslin et al., 2019; Davis et al., 2017). All statistics presented in the findings are conditional on respondents being categorized as having experienced a sexual assault in the year prior to survey administration. This study is a companion to Calkins et al., 2021, which provides similar information on soldiers’ experiences of sexual harassment and gender discrimination in the active-component Army.

We provide information on

• the types of sexual assault behaviors experienced by sexual assault victims over the year prior to survey administration, during all sexual assault experiences
• perceived perpetrator intent and method of coercion involved in all sexual assault experiences over the year prior to survey administration
• the types of sexual assault behaviors involved in victims’ self-identified most serious experience of sexual assault over the year prior to survey administration
• characteristics (number, military status, gender, and relationship to the victim) of the perpetrator(s) of victims’ self-identified most serious experience of sexual assault over the year prior to survey administration
• time and place of victims’ self-identified most serious experience of sexual assault over the year prior to survey administration
• alcohol and drug involvement in victims’ self-identified most serious experience of sexual assault over the year prior to survey administration
• co-occurrence of victims’ self-identified most serious sexual assault experience with hazing, bullying, sexual harassment, and stalking over the year prior to survey administration.

Throughout the remainder of this article, unless otherwise specified, references to types of sexual assault behaviors, perceived perpetrator intent,² or method of coercion refer to the types of sexual assault behaviors involved in all sexual assault experiences over the year prior to survey administration, and references to any other characteristics of sexual assault experiences refer to characteristics of victims’ self-identified most serious experience.

We also compare the sexual assault experiences of the following groups of soldiers:

• sexual assault victims in the 2016 versus 2018 WGRA samples
• female sexual assault victims versus male sexual assault victims
• sexual assault victims who identify as heterosexual versus those of other sexual orientations
• sexual assault victims assigned to installations where they face a high risk of sexual assault versus sexual assault victims assigned to installations where they do not face a high risk of sexual assault, where high risk is defined as above-average risk of sexual assault for either male or female soldiers.

Findings

Sexual Assault Experiences Across the Active-Component Army

The most common type of sexual assault behavior, both for all sexual assault experiences and for victims’ self-assessed most serious experiences, is that someone intentionally touched private areas of the victim's body. Penetrative sexual assaults with a penis were the second most common type of sexual assault but were substantially less common than intentional touching of private parts. Nearly 90 percent of victims believed that the assault was committed for a sexual reason, and more than half indicated that the assault was meant to be abusive or humiliating (which can co-occur).

The typical perpetrator of victims’ most serious sexual assault experiences is a male enlisted member of the military acting alone. The perpetrator was most often a military peer of the victim, but slightly more than 40 percent of victims indicated that at least one perpetrator was the victim's supervisor or another higher-ranked service member. Slightly more than half of victims indicated that the perpetrator was a friend or acquaintance. Perpetrators who were strangers to the victim were uncommon, and assaults by spouses, significant others, or family members were comparatively rare.

Approximately two-thirds of victims’ most serious experience of sexual assault occurred at a military installation. Slightly more than half of victims indicated that their most serious experience occurred during a required military activity, the most common of which was while at work during duty hours. Victims were approximately equally likely to be assaulted while at work, while in their or someone else's home or quarters, and while out with friends or at a party. Forty percent of victims indicated that they were drinking alcohol during their most serious sexual assault experience, and 42 percent indicated that the perpetrator was drinking alcohol during their most serious sexual assault experience, with 25 percent of victims indicating that the perpetrator either bought or gave them alcohol.

Sexual assaults often co-occur with sexual harassment—and, less often, with stalking—by the same perpetrators. Measures of co-occurrence with hazing and bullying are not comparable across years, but in 2018, 17 percent of victims described their most serious experience as hazing and 24 percent described their most serious experience as bullying.

Table 1 provides a high-level overview of the findings for respondents’ most serious sexual assault experience, focusing on broad categories of characteristics of these assaults (e.g., personal relationship between the perpetrator and victim, time and place of the assault). For each characteristic category and group of soldiers being compared (e.g., men versus women) the table indicates whether we observed statistically significant differences across groups with a solid circle in the appropriate table cell. Blank cells in the table indicate that no significant differences were found. In the sections that follow, we focus only on those categories where we observed differences that may be of practical importance for policy or prevention efforts. Importantly, there may be other differences in different groups of soldiers’ sexual assault experiences that we cannot identify because of small sample sizes; this is often true of the comparisons for sexual orientation, especially for men.

Differences by Survey Year

There were few statistically significant changes in the circumstances surrounding sexual assault across survey years. Of the few statistically significant differences across survey years, none are of substantive importance for policymaking (i.e., indicate a needed change in the primary focus of current prevention efforts).

Differences by Gender

Male and female sexual assault victims in the active-component Army tend to have experienced different types of sexual assault behaviors during all sexual assault experiences over the 12 months prior to the survey date, and there were differences in perceived perpetrator intent and methods of coercion as well. Among soldiers who were sexually assaulted during the year prior to survey administration, and among all sexual assault experiences over the 12 months prior to the survey date, female victims were more likely than male victims to experience completed sexual assaults, where they were penetrated by someone else, or attempted penetrative sexual assaults. Male victims were more likely than female victims to experience sexual assaults where they were made to penetrate someone else. Among soldiers who had been sexually assaulted, nonpenetrative sexual assaults were equally common for male and female victims. Male victims were more likely than female victims to indicate that they believed that the perpetrator intended to abuse or humiliate them, whereas female victims were more likely than male victims to indicate that they believed that the perpetrator had a sexual reason for their actions. Female victims were also more likely than male victims to indicate that they were sexually assaulted while asleep, unconscious, or incapacitated by drugs or alcohol.

The typical perpetrator of soldiers’ most serious sexual assault experience was similar for both male and female victims: a male enlisted member of the military acting alone. However, men were more likely to indicate multiple perpetrators; at least one female perpetrator; nonmilitary perpetrators; and, among military perpetrators, perpetrators who were commissioned officers. There were also gender differences in the relationships between victims and perpetrators. Although the typical perpetrator for both male and female victims was a military peer, male victims were more likely than female victims to indicate that at least one perpetrator was their subordinate or someone they managed. Female victims were more likely than male victims to describe the perpetrator as a friend or acquaintance, whereas male victims were more likely than female victims to indicate “none of the above” (i.e., someone they knew other than a family member, current or former spouse or significant other, or friend or acquaintance) on the personal relationships item—a category that might include some military coworkers. Men were also more likely to indicate “not sure” on the personal relationship item.

Male victims were more likely than female victims to indicate that their most serious sexual assault experience occurred during required military activities, particularly while on temporary duty (TDY), during official military functions, during military combat training other than basic training, and while at work during duty hours. Female victims were more likely than male victims to indicate that their most serious sexual assault experience occurred while in their or someone else's home or quarters. Female victims were more likely than male victims to indicate that they and/or the perpetrator were drinking alcohol during their most serious sexual assault experience, and male victims were more likely than female victims to describe their most serious experience as hazing.

Differences by Sexual Orientation

Sexual minorities (i.e., those who did not identify as heterosexual) in the Army are more often female than male, so an examination of experiences by sexual orientation only would likely be biased because of the different gender balance of individuals in the two groups. We therefore present results regarding sexual orientation separately by gender of the victims. Because of small sample sizes, we had to aggregate active-component soldiers who were sexual minorities into relatively large groups according to their response to the WGRA's sexual orientation question. For women, these groups are (1) heterosexual; (2) lesbian, gay, bisexual, or other (LGBO); and (3) no response (NR) and prefer not to answer (PNA). For men, these groups are (1) heterosexual and (2) all others (LGBO, NR, and PNA).

Differences by Sexual Orientation for Active-Component Women

We found few differences between heterosexual and LGBO female victims’ sexual assault experiences, but we found several differences between the sexual assault experiences of (1) NR and PNA female victims and (2) heterosexual and LGBO female victims.

Heterosexual female victims were more likely than LGBO female victims, who were more likely than NR/PNA female victims, to indicate that they experienced a nonpenetrative sexual assault where they were made to touch private areas of someone else's body during any of their sexual assault experiences in the year prior to the survey. Furthermore, heterosexual female victims were more likely than LGBO female victims and less likely than NR/PNA female victims to indicate that they believed that the perpetrator intended to abuse or humiliate them.

Female sexual assault victims who did not or preferred not to respond to the sexual orientation question (i.e., NR/PNA victims) were more likely than heterosexual or LGBO female victims to indicate that the perpetrator(s) of their most serious sexual assault experience fell into the personal relationship category of “none of the above” (i.e., someone the victim knew, but not a family member, current or former spouse or significant other, or friend or acquaintance). We believe that this category includes military coworkers who the victims did not consider friends. A similar pattern exists for whether the victim would describe their most serious experience as bullying. Additionally, LGBO and NR/PNA female victims were more likely than heterosexual female victims to indicate that their most serious sexual assault experience occurred at a military installation, and LGBO women were statistically significantly more likely than heterosexual and NR/PNA women to indicate that their most serious experience occurred while being intimate with the other person.

These results indicate that the female sexual assault victims who most often experienced assaults that were meant to abuse, humiliate, or bully were not the LGBO female victims, but NR/PNA female victims. Although this seems important, we cannot say definitively who the NR/PNA female victims are (in particular, whether they are sexual minorities), which makes it difficult to determine the actual policy implications.

Differences by Sexual Orientation for Active-Component Men

We found several important differences by sexual orientation in male victims’ sexual assault experiences. Because of the very small sample of LGBO male victims in particular, there may be some important differences by sexual orientation in male victims’ experiences that we are unable to identify.

LGBO, NR, and PNA male victims were more likely than heterosexual male victims to indicate that they experienced all penetrative or attempted penetrative sexual assault behaviors during any sexual assault over the year prior to survey administration, and they were less likely to indicate that they experienced having someone touch private areas of their body. LGBO, NR, and PNA male victims were also statistically significantly more likely to indicate that the perpetrator used or threatened them with violence to make them comply.

We found several differences in the characteristics of perpetrators. LGBO, NR, and PNA male victims were more likely than heterosexual male victims to indicate that the perpetrators were a mixed-gender group; that at least one perpetrator was their supervisor, another higher-ranked member of their chain of command, or a stranger; and that they were not sure of the number, gender, military status, pay grade, and/or their personal relationship with the perpetrator(s) of their most serious experience. LGBO, NR, and PNA male victims were less likely to indicate that all perpetrators were female and that at least one perpetrator was a military member, and they were as likely as heterosexual male victims to indicate that all perpetrators were male.

LGBO, NR, and PNA male victims were more likely than heterosexual male victims to indicate that their most serious experience occurred during basic, officer, or technical training; official military functions; and while on TDY.

LGBO, NR, and PNA male victims were more likely than heterosexual male victims to indicate that they were drinking alcohol or did not know whether they were drinking alcohol at the time of their most serious sexual assault experience. They were also more than twice as likely as heterosexual male victims to indicate that they believed they had been drugged and more than three times as likely to indicate that they did not know whether they had been drugged. Finally, LGBO, NR, and PNA male victims were twice as likely to indicate that they were stalked both before and after their most serious sexual assault experience by the same perpetrator(s) and more likely to describe their most serious experience as hazing and/or bullying.

Differences by Installation Risk Level

Installation risk level is defined as an installation's average rate of sexual assault (separately for women and men), where installations with rates above the overall Army average (excluding the Pentagon and military academies) are designated as “high risk.” Like the results for survey year, few significant differences were observed between high-risk and non–high-risk installations, and of those differences that do exist, none are of substantive importance for policymaking (i.e., none indicate a needed change in the primary focus of current prevention efforts).

Policy Implications

Sexual Assault Prevention Training Should Be Aligned with Victims’ Experiences

The first major policy implication of our research is that sexual assault prevention training materials should emphasize the most common behaviors and scenarios outlined in this study, as we previously recommended for sexual harassment and gender discrimination prevention training. The approach to sexual assault prevention in the Army is, in many ways, geared toward addressing the sexual assault of heterosexual women. However, our findings show that there are large differences in the circumstances surrounding sexual assault in the Army by gender and sexual orientation. Sexual assault prevention training materials should be expanded to incorporate the experiences of men, sexual minorities, and others whose sexual assault experiences differ from those of heterosexual women.

We found that the sexual assault experiences of soldiers at high-risk and non–high-risk installations are broadly similar. We therefore conclude that there is no need to tailor the content of training materials to individual installations. For example, a program that uses vignettes as discussion prompts does not need to generate different vignettes for low- and high-risk installations. However, our results are not meant to imply that the overall prevention approach at high-risk installations should look exactly like the approach at non–high-risk installations. For example, a high-risk installation might decide to invest proportionately more prevention resources than a lower-risk installation.

More Data Are Needed on the Experiences of Sexual Minorities in the Army

The second major policy implication of our research is that there is a crucial lack of data on the sexual assault experiences of sexual minorities in the Army. We found suggestive evidence that sexual minority male victims experienced more penetrative sexual assaults, more violent sexual assaults, and more assaults meant to haze or bully than heterosexual male victims. We also found some evidence that NR/PNA women experienced more assaults meant to abuse, humiliate, or bully than heterosexual women or LGBO women. However, given the relatively small number of soldiers who identified with a sexual orientation other than heterosexual in the combined 2016 and 2018 WGRAs, and our resulting need to aggregate soldiers of differing sexual orientations, it is difficult for us to truly examine these soldiers’ experiences in depth. This lack of detailed information on the unique sexual assault experiences of sexual minorities hampers the effort to prevent and respond to sexual assault in the Army.

The Army should further investigate the sexual assault and other potentially discriminatory experiences of sexual minority soldiers by including sexual orientation as a sociodemographic variable in existing administrative data and in survey data collected in the future. If sexual orientation were treated as other sociodemographic characteristics (e.g., gender, race/ethnicity) are in administrative data, it might be possible to use that information as part of analyses and as a sampling and weighting variable in future data-collection efforts. Such treatment of sexual orientation data will require a change in DoD-level policy, and the Army alone may not be able to implement the recommendation. Such a change would also require a thoughtful examination of how to protect sexual orientation data and the confidentiality of service members’ identities. Sexual minorities are not a monolithic group, and if the current restrictions on collecting sexual orientation data remain in place, both DoD and the Army will continue to be limited in their ability to determine whether there are differences between the experiences of sexual minority soldiers and their majority peers. Also, if these restrictions remain, it will ultimately limit the Army's ability to learn about the best way to prevent sexual assaults and serve victims’ needs.