Chapter 1: Incidence and Prevalence of Sexual Offending

Introduction

Simple questions do not always have easy answers. For example, the answers to, "How many sex offenses are committed each year?" and "How great is an individual's lifetime risk of being a victim of a sex crime?" vary greatly depending on the source consulted. Even with the best sources of data, it is extremely difficult to estimate the actual number of sex crimes committed because of low levels of reporting. Sex crimes are not only often unreported, they are often unseen by anyone other than the victim and perpetrator. One group of researchers puts it aptly:

Among highly personal and sensitive behaviors and experiences, including other forms of interpersonal violence, rape and other forms of sexual violence are probably the most difficult experiences to measure. They are rarely observed and occur in private places (Cook et al., 2011, p. 203).

Nevertheless, statistics on the incidence and prevalence of sex crimes, as well as trend data, can provide important insight into the nature and extent of sexual violence that policymakers and practitioners can use to design and deliver more effective prevention and intervention strategies. This chapter presents empirically derived information that helps paint a portrait of what we currently know about the incidence and prevalence of sexual offending and victimization. It also describes the strengths and weaknesses of the available data so policymakers and practitioners can better assess and interpret the existing knowledge base.

Key Data Sources

(And What They Can Tell Us About the Incidence and Prevalence of Sexual Offending)

Creating a complete and accurate accounting of the extent of sexual offending is challenging. First, there is no single definition of sexual offending. Statutory definitions of sex offenses differ from jurisdiction to jurisdiction; a sex crime committed in one state might not be classified as a sex crime in an adjacent state. State laws differ on whether rape must involve physical force or threats of physical force, and so on. Even when using national standards, such as the categories reported by the 17,000 police departments submitting Uniform Crime Report (UCR) data to the Federal Bureau of Investigation (FBI), it is impossible for each officer in each department to use the same exact criteria to decide how to classify a crime. Comparing recorded crime and victimization statistics is also challenging due to the variety of reference periods. UCR data are reported on a calendar year basis while National Crime Victimization Survey (NCVS) data, also completed annually, are compiled based on reports of victimization in the 12 months prior to the time of the interview. Comparing victimization data from different sources is even difficult, as some sources measure lifetime victimization while others measure annual or college semester victimization. Finally, rate comparison can be problematic given the different ways in which the sample being studied is measured.

With these challenges in mind, following is a review of key data sources and what they reveal about the incidence and prevalence of sexual offending.

The FBI compiles its UCR from data submitted by law enforcement agencies across the nation. Law enforcement agencies reporting crimes to the FBI oversee approximately 93 percent of the total U.S. population (FBI, 2004). As part of the UCR program, the FBI collects data on 8 serious crimes¹ as well as arrest data for 21 additional crime categories. Prior to 2012, for the purposes of UCR reporting, the FBI defined forcible rape as the carnal knowledge of a female by force (including threats of force) and against her will.² Conversely, the NCVS definition of sexual assault measures the extent of sexual assaults against both men and women. Also, UCR does not count sexual assault, statutory rape without force, or simple assault.

Based on the definition of forcible rape used for UCR reporting prior to 2012, the FBI (2014) reported that 79,770 forcible rapes were reported to law enforcement in 2013, a rate of 25.2 crimes per 100,000 females in the United States. This is a 10.6 percent decrease from the 88,097 rapes in 2009, which had a rate of 28.7 crimes per 100,000 females in the United States. Slightly more than 4 out of 10 rapes reported to police in 2009, the most recent year such data were available, were cleared by arrest or exceptional means (FBI, 2009a). Overall, an estimated 21,407 arrests for forcible rape were made by law enforcement agencies in the United States in 2009 (FBI, 2009b). Offenders arrested for rape in 2009 were predominantly young, white, and (as expected) overwhelmingly male. Only 1 percent of the offenders arrested for rape in 2009 were female. About 15 percent of the nation’s rape arrestees in 2009 were under the age of 18, and 37 percent were 18–29 years old. Whites accounted for 65 percent of the rape arrestees, African-Americans accounted for 33 percent, and other races made up about 2 percent of the arrestees (FBI, 2009c).

National Crime Victimization Survey

NCVS was established by the Bureau of Justice Statistics (BJS) in 1973 to provide a source of information on the characteristics of criminal victimization in the United States. NCVS collects information on the frequency and nature of rape, sexual assault, personal robbery, aggravated and simple assault, household burglary, theft and motor vehicle theft. Murder is not included in NCVS as victim reporting is the method for collecting these data. A nationally representative sample of approximately 42,000 households is included in the survey. Each household is included in the survey for three years, and all individuals over age 12 in the household are interviewed. The initial interview is in person and subsequent interviews are conducted by phone. NCVS collects data on crimes reported to police as well as those not reported and assesses the victim's experience with the criminal justice system. NCVS gathers data on the nature and circumstances of the crime, such as where it occurred, when it occurred, and whether the victim knew the perpetrator.

Based on NCVS data, an estimated 284,350 rape/sexual assault victimizations³ occurred in the United States in 2014, a rate of 1.1 victimizations per 1,000 persons age 12 and over (Truman & Langton, 2015). NCVS data also indicate that most rape/sexual assault victims are female, white, and under age 30. Based on the 2010 survey, when victim gender information was most recently reported, more than nine out of every 10 rape/sexual assault victims in the United States were female (Truman, 2011). Further, based on the results of the 2008 national survey (the latest survey for which comprehensive rape/sexual assault victim demographic information is available), an estimated 63 percent of victims are white, 28 percent are African-American, and 9 percent are other races. In 2008, the rape/sexual assault victimization rate for African-American females was about three times higher than it was for white females. Among different age groups, people aged 16–19 and 20–24 had the highest rape/sexual assault victimization rates in 2008 — 2.2 and 2.1 per 1,000 persons in each age group, respectively. By comparison, BJS (2011) found that people aged 35–49 had an estimated rape/sexual assault victimization rate of 0.8 per 1,000 persons in the age group in 2008, and people aged 12–15 had an estimated rate of 1.6 (although the latter estimate is based on a small sample of cases) (Truman & Rand, 2010).

Although NCVS data provide valuable insights about the incidence and prevalence of sexual offending nationwide, BJS acknowledges —

The measurement of rape and sexual assault represents one of the most serious challenges in the field of victimization research. Rape and sexual assault remain sensitive subjects that are difficult to ask about in the survey context. As part of the on-going redesign of NCVS, BJS is exploring methods for improving the reporting of these crimes (Truman & Rand, 2010).

National Violence Against Women Survey

Sponsored by the National Institute of Justice (NIJ) and the Centers for Disease Control and Prevention (CDC), the National Violence Against Women Survey (NVAWS) was administered in the mid-1990s to assess the extent of violence against women in the United States. A nationally representative sample of 8,000 men and 8,000 women aged 18 and older were surveyed between November 1995 and May 1996.

NVAWS found that 17.6 percent of female and 0.3 percent of male respondents had been the victim of a rape at some time in their lives. Based on this finding, the authors estimate that almost 18 million women and almost 3 million men in the United States have been raped. Rape prevalence rates were the same for minority and nonminority women, but Native American/Alaska Native women were significantly more likely to have experienced a rape in their lifetime. Rape was more likely to affect younger women than older women, with more than half of female victims and nearly three-quarters of male victims being victimized prior to age 18. Overall, 86 percent of rape victims were female, while most rapists were male. Finally, female victims were significantly more likely than male victims to have been the victim of a rape by a current/former intimate partner and to be injured during the rape (Tjaden & Thoennes, 2006).

National Intimate Partner and Sexual Violence Survey

The National Intimate Partner and Sexual Violence Survey (NISVS) is an ongoing, nationally representative telephone survey that collects information about sexual violence, stalking victimization, and intimate partner violence among adult women and men ages 18 and older in the United States. CDC launched the survey in 2010, with the support of NIJ and the U.S. Department of Defense (Black et al., 2011; Breiding et al., 2014). NISVS data will be collected annually as long as funding for the survey is available.

NISVS is unique because it is the first ongoing survey designed to describe and monitor sexual violence, stalking, and intimate partner violence from a public health rather than crime perspective. Survey respondents are first asked about various health conditions to establish a health context for the survey. Then they are asked about victimization experiences using behaviorally specific questions. Research has shown that this health-based approach increases disclosure of violent victimization.⁵

NISVS also collects data on victimization involving sexual violence other than rape, control of reproductive health and other forms of sexual victimization that have not been measured in the past. It is also the first survey to provide national and state-level data on sexual violence, stalking and intimate partner violence.

Findings from the second year of NISVS data were published by Breiding and colleagues in 2014.⁶ Based on 12,727 completed interviews (6,879 women and 5,848 men), the 2011 survey found that nearly 1 in 5 women (19.3 percent) and 1 in 59 men (1.7 percent) have been raped in their lifetime. In addition, 43.9 percent of women and 23.4 percent of men reported being victims of other forms of sexual violence in their lifetime. Nearly one-half (45.4 percent) of female rape victims reported being victimized by an intimate partner, while a similar number (46.7) percent reported being victimized by an acquaintance. More than four in 10 (40.4 percent) female rape victims experienced their first completed rape before age 18.

The 2011 survey also found that 15.2 percent of women and 5.7 percent of men have experienced stalking victimization “in which they felt very fearful or believed that they or someone close to them would be harmed or killed” (Breiding et al., 2014, p. 6). Female victims were most often stalked by a current or former intimate partner, while men were stalked primarily by an intimate partner or acquaintance. Nearly one in four women (22.3 percent) and one in seven men (14 percent) reported experiencing severe physical violence⁷ perpetrated by an intimate partner, while nearly half of all women (47.1 percent) and men (46.5 percent) reported experiencing psychological aggression by an intimate partner.

In the 2011 survey, 1.6 percent of the female respondents surveyed reported being raped in the 12 months prior to taking the survey. That equates to an estimated 1.9 million women nationally. About one in 20 women and men (5.5 percent and 5.1 percent, respectively) reported being victims of sexual violence other than rape, while about 4.2 percent of women and 2.1 percent of men reported being stalked in the 12 months before the survey.

The 2011 NISVS also found that about 32.3 percent of multiracial women, 27.5 percent of American Indian/Alaska Native women, 21.2 percent of African-American women, 20.5 percent of white women and 13.6 percent of Hispanic women have been raped in their lifetime.

The 2010 NISVS produced the first national data on the prevalence of sexual violence, stalking and intimate partner violence victimization among lesbian, gay and bisexual women and men. Lesbians and gay men were found to have sexual violence victimization rates equal to or higher than those reported by heterosexuals, while bisexual women had significantly higher lifetime prevalence rates of rape and sexual violence other than rape compared to both lesbian and heterosexual women (National Center for Injury Prevention and Control, 2011a).

National College Women Sexual Victimization Study

The National College Women Sexual Victimization Study (NCWSV), funded by NIJ, surveyed a randomly selected, national sample of 4,446 women attending two- or four-year colleges during the fall 1996 semester (Fisher, Cullen & Turner, 2000). Students were asked via telephone about events that occurred that school year, a period of approximately seven months (Fisher, Cullen & Turner, 2000). Lifetime exposure to sexual victimization was not assessed.

Fisher, Cullen and Turner (2000) found that 2.8 percent of college women who responded to the survey had experienced either a completed or attempted rape during the semester. Of those who reported rape, 23 percent reported multiple rapes. As the study period included only one semester of college, the authors caution that over the years of the participants’ college experience, rates of victimization may be higher than reported for the time period under study. Further, they also found that many women did not characterize their sexual victimization as a crime. For the incidents categorized as rape by the researchers, 49 percent of the women responded "yes" when asked if they would describe the incident they experienced as a rape (Fisher, Cullen & Turner, 2000). The reasons survey respondents gave for not describing the incident as rape were varied, and included embarrassment, not clearly understanding the legal definition of rape, not wanting to define someone they know who victimized them as a rapist or because they blamed themselves for their sexual assault (Fisher, Cullen & Turner, 2000). Most victims knew their offender: The authors found that nine out of 10 offenders were known to their victim. Most often, the offender was a boyfriend, ex-boyfriend, classmate, friend, acquaintance or coworker (Fisher, Cullen & Turner, 2000). Campus rape victims were not likely to report the crime to police; fewer than 5 percent of completed and attempted rapes were reported. However, in approximately two-thirds of the cases, the victim did report the incident to another person, most typically a friend rather than a family member or college official (Fisher, Cullen & Turner, 2000).

Sexual Experiences Survey

The Sexual Experiences Survey (SES) was developed in the late 1970s to "operationalize a continuum of unwanted experiences that at the extreme reflect legal definitions of attempted rape and rape" (Koss et al., 2007, p. 359). Unlike the other sources described here, SES is neither an annual data collection instrument nor a one-time national sample. Instead, it represents a standard set of questions that have been used repeatedly by scholars, particularly in the public health field, to study unwanted sexual experiences. In some studies, the survey has been used in its original form, while in others researchers have customized the questions to suit their specific research needs. This self-report survey instrument was revised in 1987 and again in 2007. Its questions were modeled on the statutory definition of rape in Ohio. Separate versions of the survey assess victimization and perpetration of sex crimes. The survey tool is available in both short and long form, allowing for screening or for in-depth study. Respondents are asked about incidents since age 14 and in the past year, thus providing both annual results and lifetime rates of victimization.

Using SES questions, Testa and colleagues (2004) conducted in-person interviews with 1,014 women living in or near Buffalo, New York, between May 2000 and April 2002. Of the respondents, 38 percent indicated they had experienced sexual victimization since age 14. The most common experience reported by respondents was unwanted sexual contact. Of the respondents, 27 percent reported unwanted sexual contact, 17 percent reported a rape and 12 percent reported an attempted rape (Testa et al., 2004).

National Child Abuse and Neglect Data System

The National Child Abuse and Neglect Data System (NCANDS) has provided annual counts of incidents of child sexual abuse since 1992. NCANDS data are based on reports from participating state child protection agencies. NCANDS includes case-level data on the characteristics of screened-in reports of abuse and neglect made to the agencies, including the children involved, the types of maltreatment alleged, the disposition of the investigation, the risk factors of the child and the caregivers, the services provided and information about the perpetrators.⁸ The U.S. Department of Health and Human Services publishes an annual report — Child Maltreatment — that summarizes NCANDS data reported from the states. Each state has its own definition of child abuse and neglect based on federal law (Children’s Bureau, 2014). Child sexual abuse is reported annually along with incidents of neglect, physical abuse and psychological maltreatment. For 2014, NCANDS estimates that 58,105 children were victims of sexual abuse.

National Survey of Children's Exposure to Violence

The National Survey of Children’s Exposure to Violence (NatSCEV) is sponsored by the Office of Juvenile Justice and Delinquency Prevention (OJJDP) and supported by CDC. This is the first national survey to measure both past-year and lifetime exposure to conventional crime, child maltreatment, victimization by peers and siblings, sexual victimization, witnessing and indirect victimization, school violence and threats and internet victimization for children aged 17 and younger. (For more on “Internet-Facilitated Sexual Offending,” see Chapter 4 in the Adult section.) This study was designed by the Crimes Against Children Research Center at the University of New Hampshire. It attempted to measure children’s experience of violence in the home, school and community. Telephone interviews were conducted with respondents aged 10–17. For respondents aged 9 and under, their adult caregivers were interviewed.

Key research findings were that 5.6 percent of children had been sexually victimized in the past year and 9.5 percent had been a victim during their lifetime. Sexual victimization includes attempted and completed rape, sexual assault, flashing or sexual exposure, sexual harassment and statutory sexual offenses. In addition, 16.4 percent of youth aged 14–17 had been sexually victimized in the past year and 27.3 percent had been sexually victimized during their lifetime. This study found that children were often the victims of multiple types of violence. A child who was physically assaulted in the past year was found to be five times as likely to have been sexually victimized (Finkelhor et al., 2015).

In addition, the majority (61 percent) of reported past-year peer victimizations (including assault, bullying, sexual victimization and property crime) occurred at school (Turner et al., 2011). Emotional bullying by peers was most likely to occur at school, while sexual assault and rape were most likely to occur elsewhere (Turner et al., 2010). Considering only serious violent events (rape, sexual assault, robbery and aggravated assault), the rates for these crimes at school for adolescents aged 12–18 are lower than those occurring away from school. In 2008, the serious violent victimization rates were four per 1,000 students at school and eight per 1,000 students away from school (Robers, Zhang & Truman, 2010). Children living in households with lower incomes have higher rates of exposure to sexual and physical assault than those living in households with middle and high incomes (Crouch et al., 2000).

Developmental Victimization Survey

A precursor to NatSCEV and conducted by the same researchers, the Developmental Victimization Survey (DVS) was a random-digit-dial survey of households conducted in 2003. The survey sample consisted of 2,030 children aged 2–17 within the households surveyed. One child from each household was randomly selected (the child with the most recent birthday). Telephone interviews were conducted directly with children aged 10–17, while a caregiver was questioned regarding children aged 2–9. DVS uses the Juvenile Victimization Questionnaire, which is widely known and used to screen individuals for incidence of violence (Finkelhor et al., 2005). DVS results indicated that one in 12 children in the sample (82 per 1,000) had experienced a sexual victimization in the sample year, including 22 per 1,000 who experienced an attempted or completed rape and 32 per 1,000 who experienced a sexual assault (Finkelhor et al., 2005).

National Incidence Studies of Missing, Abducted, Runaway and Thrownaway Children–2

The National Incidence Studies of Missing, Abducted, Runaway and Thrownaway Children–2 (NISMART–2) was an estimate of the number of missing children based on surveys of households, juvenile residential facilities and law enforcement agencies. Data were collected via telephone interviews in 1999 and have been reported in several publications.

Key research findings were that an estimated 285,400 children were victims of a sexual assault, for a rate of approximately 4.1 victims per 1,000 children in the United States. Of these, an estimated 141,400 children were victims of a rape (anal, oral or vaginal penetration) and 60,400 experienced an attempted rape. Eighty-nine percent of victims were female and 95 percent were assaulted by a male. Eighty-one percent of victims were aged 12–17. Victimization of whites and African-Americans was proportionate to their presence in the general child population. Hispanics constituted 9 percent of victims and 16 percent of the U.S. child population. Seventy-one percent of child sexual assault victims were victimized by someone they knew or recognized by sight, 18 percent were victimized by a stranger and 10 percent were victimized by a family member (Finkelhor, Hammer & Sedlak, 2008).

National Women's Study

The National Women’s Study (NWS) is a three-year longitudinal study of a national probability sample of 4,008 adult women in the United States aged 18 and older. Three waves of interviewing were completed: at the time of the initial study and at one and two years after the initial interview.

The results indicated that 13 percent of women reported being the victim of at least one completed rape in their lifetime. Based on this, it was estimated that 12.1 million women in the United States have been the victim of one forcible rape in their lifetime. In addition, .07 percent of women surveyed reported having been raped in the past year, equating to 683,000 adult women in the United States. Of those who reported being raped, 56 percent reported one rape and 39 percent indicated they were raped more than once (with 5 percent uncertain how many times they were raped). Twenty-nine percent of the rapes occurred when the victim was younger than age 11 and another 32 percent occurred when the victim was between 12 and 18. In total, 61 percent of rapes were committed against a female victim younger than age 18. In terms of the relationship between the perpetrator and the victim, 22 percent were raped by a stranger, while the vast majority of rapes were perpetrated by an intimate partner, family member, friend or neighbor (National Victim Center & Crime Victims Research and Treatment Center, 1992).

Other Data Sources

Drug-facilitated, Incapacitated and Forcible Rape: A National Study

For the Drug-facilitated, Incapacitated and Forcible Rape (DAFR) national study, Kilpatrick and colleagues (2007) conducted a national telephone survey of 5,000 women aged 18–86, including approximately 3,000 who represented all women in the United States and 2,000 women attending college.

The research results indicated that for women of all ages, an estimated 18 percent had been raped during their lifetime. This translates into an estimated 20 million victims of rape out of 112 million women in the United States. In looking at past-year victimization alone, the authors estimated that more than 1 million women had been raped. Of those who reported that they were raped, 16 percent indicated they had reported the crime to law enforcement. The study found that victims of drug-facilitated or incapacitated rape were less likely to report the crime to the police than victims of forcible rape (Kilpatrick et al., 2007).

National Survey of Adolescents

The National Survey of Adolescents consisted of interviews of 4,023 adolescents (aged 12–17) on various topics that included victimization history. The survey used random-digit dialing and stratified sampling techniques to identify households that had a telephone, an adolescent aged 12–17 with a parent or legal guardian, and both a parent or guardian and an adolescent who spoke English or Spanish.

Results of the survey indicated that 8.1 percent of those responding had a history of sexual victimization. Native American adolescents had the highest prevalence rate of sexual victimization (15.7 percent), compared to 13.1 percent for African-Americans, 10 percent for Hispanics, 6.7 percent for whites and 6.5 percent for Asians. Adolescent females were at greater risk of sexual assault than males (13 percent compared to 3.4 percent). Seventy-four percent of victims reported knowing the perpetrator prior to the sexual offense. Finally, only 13 percent of victims reported the sexual offense to the police (Kilpatrick, Saunders, & Smith, 2003).

Adverse Childhood Experiences

The Adverse Childhood Experiences (ACE) study was conducted with members of the Kaiser Permanente Health Plan who had visited the San Diego Health Appraisal Clinic. Data were collected from 17,337 study participants between 1995 and 1997. Unlike some of the other studies described above, the ACE study surveyed adults about a variety of previous childhood experiences: psychological, physical and sexual abuse during childhood; substance abuse; mental illness; violence against the respondent’s mother; and criminal behavior in the household. The results indicated that 20.7 percent of the sample experienced childhood sexual abuse, including 24.7 percent of women and 16 percent of men (Felitti et al., 1998).

Youth Risk Behavior Survey

The Youth Risk Behavior Survey (YRBS) is a national school-based survey conducted biennially by CDC. In 2013, 13,583 questionnaires were completed in 148 schools. This survey monitors several categories of health-risk behaviors among youth and young adults, including violence. Both public and private schools in the United States with students in grades 9–12 are included in the sampling frame for this survey. YRBS includes violence measures for dating violence, rape and bullying. The 2013 study indicated that 7.3 percent of the sample had been subject to forcible sexual intercourse, with 10.5 percent of females and 4.2 percent of males reporting such an incident. This number did not change significantly from 2011 (8 percent) to 2013. In addition, 10.4 percent of students who dated reported being victims of sexual dating violence during the year prior to the survey, including 14.4 percent of females and 6.2 percent of males (Kann et al., 2014). (For a discussion of adult “Sex Offender Risk Assessment,” see Chapter 6 in the Adult section.)

Survey of Inmates in Local Jails

BJS’s Survey of Inmates in Local Jails (SILJ) has been conducted periodically over approximately the past 40 years and consists of interviews with a national sample of jail inmates. Based on the 2002 survey of nearly 7,000 jail inmates, it is estimated that the most serious offense for 3.4 percent (nearly 21,200 inmates) of the 623,492 jail inmates in the United States was for rape (0.6 percent) or another type of sexual assault (2.8 percent) (James, 2004). This survey and the 2004 Survey of Inmates in State and Federal Correctional Facilities (SISFCF) both provide information about the prevalence of sexual offenders within local, state, and federal correctional/detention facilities.⁹   

Survey of Sexual Victimization

BJS’s Survey of Sexual Victimization (SSV) has been conducted annually since 2004 on juveniles in detention facilities and adults in corrections facilities, in response to the Prison Rape Elimination Act of 2003 (PREA; P.L. 108-79). The SSV is a summary of allegations of sexual victimization by other inmates or staff reported to correctional authorities. In 2012, there were 865 allegations of sexual assault in state juvenile systems and 613 in local, private and Indian Country facilities. The rate of allegations in the state juvenile systems has more than doubled from 19 per 1,000 in 2005 to 47 per 1,000 in 2012. The rate of allegations in local or private facilities increased from 7.2 per 1,000 in 2010 to 13.5 per 1,000 in 2012. From 2007 to 2012, there were nearly 9,500 allegations of sexual assault in state, local or private facilities, 55 percent of which were youth against youth (25 percent of which were substantiated) and 45 percent (10 percent of which were substantiated) were staff against youth allegations (Beck & Rantala, 2016).

Trend Data

Historical data on the incidence of sexual assault can provide important insight about trends over time. Data from law enforcement as well as from victimization surveys suggest that sexual assaults, much like other types of crime, have substantially declined over the past 10 to 20 years.

According to the FBI, the number of forcible rapes reported to the police fell 14 percent between 1990 and 2009, from 102,555 to 88,097. The number of rapes reported to police per 100,000 U.S. residents also fell during that time (from 41.1 to 28.7), a decline of 30 percent (FBI, 2009). Data on sexual assault victimization surveys follow a similar pattern. According to NCVS, the number of rape/sexual assault victimizations for those aged 12 and older in the United States fell by more than 30 percent between 2002 and 2011, dropping from 349,810 to 243,800 over the nine-year period. Overall, the estimated number of rape/sexual assault victimizations fell by more than one-third in 2011 (from about 383,000 in 1990 to 243,800) (Rennison, 2000; Truman & Planty, 2012). It should be noted that the number of rape/sexual assault victimizations increased to 300,170 in 2013, and then decreased by 5.3 percent to 284,350 in 2014 (Truman & Langton, 2015). Finally, data from NCANDS indicate that substantiated cases of child sexual assault in the United States have also fallen in recent years, dropping by 53 percent between 1992 and 2006 (Finkelhor, Hammer & Sedlak, 2008).

Limitations of the Data

No single definition of sexual offending is used across data sources.

Although increased scholarly attention has been paid to sexual victimization and victimization surveys in the past two decades, and notable improvements regarding the reporting of sexual assault crimes have recently been made in the national UCR program, much remains to be done to develop standard definitions of sex crimes and to measure victimization in a way that elicits self-report of a traumatic experience many victims may choose not to discuss. As Cook and colleagues (2011, p. 210) point out, the field “remains hampered by the lack of a standard definition of rape and its components of act, tactics and non-consent.” What is known about victims and offenders is based on an incomplete picture of the true extent of victimization. Studies of victims rely on self-report, resulting in dramatic undercounts of victimization. What we know to date is that sexual victimization is far more common than existing sources indicate and that more needs to be done to develop a credible literature on the extent, causes and consequences of sexual victimization.

Summary of the Data

Information on the incidence and prevalence of sexual offending in the United States can be obtained from a diverse range of sources. Some sources, such as the FBI’s UCR program, focus on sex crimes reported to the police and the offenders arrested for those crimes. Others, such as NCVS, focus on victims of sex crimes. Some sources collect and report data on a regular, ongoing basis. Others do so only periodically or on a one-time basis. A wide range of methods are used to collect incidence and prevalence data as well. Despite these differences, the available data provide important insight about the extent of sexual offending in the United States, along with the characteristics of victims and known offenders.

The FBI’s UCR statistics indicate that in 2013, 79,770 forcible rapes were reported to law enforcement (FBI, 2014). Arrestees for forcible rape are typically young, white males (FBI, 2009c). Based on NCVS data, 284,500 rape/sexual assault victimizations are estimated to have occurred in the United States in 2014 (Truman & Langton, 2015). NCVS data also indicate that most rape/sexual assault victims are female, white and younger than age 30 (Truman, 2011). Based on NCANDS estimates, nearly 58,105 children were victims of sexual abuse in 2014 (Children’s Bureau, 2014).

NVAWS found that 17.6 percent of women and 0.3 percent of men had been the victim of a rape at some time in their life, meaning that almost 18 million women and almost 3 million men in the United States have been raped (Tjaden & Thoennes, 2006). Data from NWS indicated that 13 percent of women reported being the victim of at least one completed rape in their lifetime. Based on this, it was estimated that 12.1 million women in the United States have been the victim of one forcible rape in their lifetime (National Victim Center & Crime Victims Research and Treatment Center, 1992). Trend data indicate that the number of forcible rapes reported to the police fell 14 percent between 1990 and 2009, and that the estimated number of rape/sexual assault victimizations fell by more than one-third between 2002 and 2011 (Rennison, 2000; Truman & Planty, 2012).

Although these data provide insight into the incidence of sexual offending and victimization, the gap between sexual victimizations and sex crimes reported to police, and the characteristics of victims and perpetrators, they must be interpreted in light of their limitations. Differences in the methods used to collect data as well as when the data were collected can render the comparison of statistics from certain sources difficult and sometimes meaningless. Users of the data must also recognize that quantitative statistics on sexual offending and victimization lack precision. An accurate accounting of sexual offending and victimization is virtually impossible because so many sex crimes and victimizations are hidden from public view. Although the available data can help policymakers and practitioners better understand incidence patterns and trends, efforts to enhance existing data systems and improve both the quality and comparability of the data are needed.

Underreporting of Sex Crimes

At least 16 different data sources report on sex crimes and victimization.

One of the greatest challenges to developing an accurate estimate of the incidence and prevalence of sexual offending is the fact that not every victim will disclose the incident to law enforcement, and many will also not disclose the incident to a researcher during a survey. Research has clearly demonstrated that many sex offenses are never reported to authorities. For example, NCVS data suggest that only about 25 to 33 percent of rapes or sexual assaults have been reported to police over the past 15 years, with some between-year fluctuations (Bachman, 1998; Truman & Langton, 2015; Truman & Planty, 2012). In addition, Tjaden and Thoennes (2006) found that only 19 percent of women and 13 percent of men who were raped since their 18th birthday reported the rape to the police. Several studies of victims have shown that the likelihood that a victim will report sexual assault to law enforcement decreases with the victim’s age (Kilpatrick, Saunders & Smith, 2003). Finally, NWS results indicated that 84 percent of victims did not report the rape to the police. Of the 16 percent who did report the rape to the police, 12 percent did so within 24 hours of the rape and 4 percent did so more than 24 hours after the rape (National Victim Center & Crime Victims Research and Treatment Center, 1992).

Larcombe (2012, p. 483) argues that “the attrition of sexual offenses … both before and after reporting to police, ensures that the minority of cases that secure a conviction for a sexual offense are not reflective of the most common or injurious forms of sexual violence experienced by women and children.” Citing two Australian studies on police and prosecutorial discretion (Statewide Steering Committee to Reduce Sexual Assault, Victoria, 2006; Lievore, 2004, as cited in Larcombe, 2012, p. 482), she points out that cases “clearly interpretable as violence” and not involving what appears to be “potential sexual partners” are more likely to proceed to conviction. Examples include cases involving male and younger victims and victims whose perpetrators were strangers, particularly when force, threats and weapons are used and result in physical injury to the victim and when verbal resistance is offered. Further, Larcombe (2012) reports that women aged 15–24, those with a psychiatric disability or a mental health issue and indigenous women are less likely to see their cases result in conviction. In terms of the offenders, those who have a prior history of criminal behavior (particularly violent and sexual offenses), are indigenous and have a disability (intellectual or psychiatric) are more likely to have their cases proceed to conviction (Larcombe, 2012). Given the evidence that sex crimes and sex offenders that are identified by authorities and processed in the criminal justice system are not representative of sexual crimes and perpetrators overall, Larcombe (2012) suggests that policies, practices and research need to consider attrition dynamics and their implications. Although the studies cited by Larcombe may have limited applicability because they used Australian samples, research on attrition dynamics and characteristics using American samples likely would strengthen our understanding of the incidence and prevalence of sexual victimization in the United States.

Special Populations and Related Topic Areas

The vast majority of victims do not report sex crimes.

Some studies have focused on a number of special populations and related topic areas regarding the incidence and prevalence of sexual victimization.¹⁰ Several of these areas are addressed below: stalking, sexual offending on college campuses and sexual offending against individuals with disabilities, members of the military and Native Americans in Indian Country.

Stalking

Stalking was first defined as a crime in 1990 by the state of California. Since that time, all 50 states and the District of Columbia have passed a law against stalking. State statutes define stalking behavior rather differently, however, and no single legal definition of stalking applies across all states. Even for a victim, defining stalking can be difficult, as behaviors that often appear as a part of stalking (e.g., gifts, notes and visits) are not in themselves criminal. Rather it is the nature of the behavior — unwanted attention, unwanted gifts, persistent or threatening notes, harassing visits and so forth — that defines the act as stalking. In some states, stalking laws are invoked for verbal threats, while in other states the threat must be written or implied by the conduct (Klein et al., 2009). In some states the perpetrator must act in a way that makes the victim fearful, and in other states it is sufficient if the perpetrator acts in a way that would make a reasonable person fear the behavior (Fox, Nobles & Bonnie, 2011). As Tjaden (2009, p. 263) points out, “Nearly 20 years after the first stalking law was enacted, many policymakers and practitioners still are unclear about what constitutes stalking.”

With this confusion among state laws, it is not surprising that there has been no standard definition used by researchers in studying the crime of stalking. Fox, Nobles, and Bonnie (2011) found that the four major national assessments of the extent of stalking all used different questions, making comparisons across the studies problematic. A few of the questions are similar, but none are exactly the same. Fox, Nobles, and Bonnie (2011) also point out that the major studies have failed to publish reliability estimates for their scales, reducing the ability to generalize beyond the sample population. It also should be noted that the sampling frame used in each of the four national stalking victimization studies is different. NVAWS, NCVS and the National Intimate Partner and Sexual Violence Surveillance System (NIPSVSS) sample the general population, while NCWSV samples only female college students (Fox, Nobles & Bonnie, 2011). In reviewing 56 peer-reviewed assessments of stalking, Fox, Nobles and Bonnie (2011) found that 55 percent of the studies examined a college-age population while 45 percent examined the general population; they also found that many studies of stalking rely on subpopulations, such as college students or nonprobability samples.¹¹

Tjaden and Thoennes (1998) conducted the first-ever national study of stalking within NVAWS and found that 8 percent of women and 2 percent of men have been stalked in their lifetime. This telephone survey of 8,000 men and 8,000 women found that most stalking victims were female (78 percent) and most perpetrators were male (87 percent). In most cases, stalking involved perpetrators and victims who knew each other — only 23 percent of female victims and 36 percent of male victims reported being stalked by strangers. Stalking was in many cases the continuation of a violent relationship — 81 percent of women who were stalked by a present or former spouse or partner had also been sexually assaulted by that person and 31 percent had been raped by that person (Tjaden & Thoennes, 1998). These findings demonstrate the importance of looking at domestic violence, stalking and sexual assault as a connected constellation of behaviors, given that the risk to the victim increases with the presence of these factors. In addition, Tjaden and Thoennes (1998) found that about half of stalking victimizations (55 percent for women and 48 percent for men) were reported to the police. In one out of five cases reported to the police, the victim indicated the police did not take any action. Only 24 percent of women and 19 percent of men who reported a victimization to the police indicated that their stalker had been criminally prosecuted (Tjaden & Thoennes, 1998).

In 2006, NCVS included a Supplemental Victimization Survey to assess the extent of stalking. Approximately 65,000 men and women completed the survey between January and June 2006. This survey estimated that 2.4 percent of the population experienced stalking or harassment in the year prior to the study (Baum et al., 2009). The study did not seek to assess lifetime victimization. Women were at greater risk of stalking victimization than men (3 percent of women reported being a victim of stalking compared to 1.7 percent of men). Stalking and harassment rates for those aged 18–24 were significantly higher than for other age groups, with the rate of victimization decreasing with age. For those who reported stalking, many reported being victimized by persistent offenders — 46 percent of stalking victims experienced at least one unwanted contact per week, and 11 percent of victims said they had been stalked for five years or more (Baum et al., 2009). Most offenders were known to their victims (73 percent), and more than half of stalking victims lost 5 or more days from work due to their victimization (Baum et al., 2009). One-third of women and one-fifth of men reported stalking or harassment to law enforcement. However, many other victims did not categorize their experience as stalking. Researchers asked questions about seven types of harassing or unwanted behavior and classified a respondent as a victim of staking if he or she had experienced at least one of the behaviors on at least one occasion and felt fear as a result. Respondents who experienced the behavior but did not feel fear were categorized as victims of harassment. Researchers did not use the term “stalking” until the final question. Of those whose experiences were classified as stalking, 60 percent reported that the experience was “not stalking” (Baum et al., 2009). This finding raises the issue of definitions and terminology, and underscores the complexity of accurately providing a picture of the national experience of victimization in general and stalking in particular.

In a further analysis of 2006 NCVS data (n = 1,159), 31 percent of stalking victims were found to have engaged in informal help-seeking behavior (family or friends) while 29 percent engaged in formal help-seeking from law enforcement. Factors identified that influenced the type of help-seeking behavior utilized included perceived seriousness, fearfulness, and whether the offender was known. It was further noted that informal help-seeking may also lead to formal help-seeking, and seeking help was influenced by how the victim viewed the victimization (Reyns & Englebrecht, 2014).

Finally, results of the National Intimate Partner and Sexual Violence Survey (NISVS) highlighted earlier in this report also include survey results related to stalking. Stalking is receiving increased attention in the literature and policy development, and further research is needed.

Sexual Offending on College Campuses

College campuses have frequently been used by researchers at universities seeking convenience samples for small studies. College campuses have also become of interest to researchers and policymakers in order to better understand the unique risks for young people during their likely first experience of living without parental supervision.

In a special study on the victimization of college students, BJS found that students experience both violent crime and serious violent crime at lower rates than nonstudents of the same age (Hart, 2003). Campuses themselves may provide some protection, as BJS found: “The number of off-campus victimizations of college students was over 14 times greater than the number of on-campus victimizations” (Hart, 2003, p. 1). The only category of violent crime for which the rates were not lower on college campuses was rape. Unlike robbery, aggravated assault, and simple assault, rape was reported at the same rate for those on campuses and same-age nonstudents. Several studies further examine rape and sexual victimization on college campuses.

In examining lifetime exposure to sexual violence, higher rates of rape victimization tend to be found for college women. Kilpatrick and colleagues (2007) conducted a national telephone survey of drug-facilitated, incapacitated and forcible rape¹² that included 2,000 women attending college. The study found that 11.5 percent had experienced rape during their lifetime. When looking at past-year victimizations, they found that 5.2 percent of college women were raped. Of those reporting rape in this study, about 12 percent of the crimes were reported to law enforcement. Victims of drug-facilitated and incapacitated rape were less likely than victims of forcible rape to report to the authorities (Kilpatrick et al., 2007).

A larger and more recent study found that similar levels of college women reported being sexually victimized in their lifetime. McCauley and colleagues (2009) interviewed a national sample of 1,980 college women and found that 11.3 percent reported having been sexually victimized at some point in their life.

The link between alcohol or drug use and sexual victimization has been studied, with consistent findings of a strong connection. Mohler-Kuo and colleagues (2004) surveyed nearly 24,000 women between 1997 and 2001 in the College Alcohol Study. They found that 4.7 percent had been raped and of those, 72 percent were intoxicated at the time the incident occurred (Mohler-Kuo et al., 2004). In a survey of approximately 300 female college students, Lawyer and colleagues (2010) found that 29.6 percent of the respondents reported a drug-related sexual assault or rape, and 5.4 percent reported a forcible sexual assault or rape. Voluntary incapacitation (via drugs or alcohol) preceded 84.6 percent of drug-related assaults, and involuntary incapacitation preceded 15.4 percent of drug-related assaults. The majority of drug-related assaults (96.1 percent) involved alcohol consumption prior to assault (Lawyer et al., 2010).

Further, when looking at stalking behavior on college campuses as part of NCVS, Fisher, Cullen, and Turner (2000) found that 13 percent of college women had been stalked. This rate of victimization is much higher than that reported by Tjaden and Thoennes (2006). The key difference in the studies is the age of the sample population. While Tjaden and Thoennes (2006) studied all ages, Fisher, Cullen and Turner (2000) studied college students exclusively — an age group at higher risk for stalking.

In a study of NCVS sexual victimization data of college age females (aged 18-24) from 1995-2013, college students had a 20 percent reporting rate (6.1 per 1,000) while non-students had a 32 percent rate (7.6 per 1,000). Eight of 10 victims knew the offender prior to the offense across both groups. Fifty-one percent of students were sexually victimized during leisure activities away from home, while 50 percent of non-students were sexually victimized during other activities at home (Sinozich & Langton, 2014).

Finally, as part of a systematic review of studies on prevalence findings from sexual assault on college campuses from 2000-2015 (n = 34 studies), Fedina et al. (2016) found the range of completed rape rate was .5-8 percent, unwanted sexual contact rate was 1.8-34 percent for females and 4.8-31 percent for males, and incapacitated rape rate was 1.8-14.2 percent for females and 1.9 percent for males. Overall, the composite sexual assault range was 6–44.2 percent for females and 1.4–3.2 percent for males. Variations in findings are attributed to differences in research designs, sampling strategies and characteristics, measures used, behavior definitions, time frames and the presence or absence of provided examples. The authors noted that “the majority of research has sampled predominately White, heterosexual, female students attending four-year colleges and prevalence findings are limited to this population” (Fedina et al., 2016, p. 15).

In summary, college students seem to be at unique risk for sexual violence and warrant the increased attention paid by policymakers in the form of sexual violence prevention and intervention programs. Tharp and colleagues (2015) further suggest that such efforts must not only focus on rape but also on unwanted sexual contact and verbal sexual harassment. Finally, a summary of the research related to bystander interventions to prevent sexual violence on college campuses described this strategy as effective and one worthy of further consideration (Fenton et al., 2016).

Sexual Offending Against Individuals With Disabilities

The rate at which individuals with disabilities are victimized is not well understood. Until mandated by law, no national statistics on this population were gathered in the United States. The few studies that have been conducted are mainly outside the United States or are exploratory in nature. Petersilia (2001, p. 658) described the state of literature in the field as “… not a scientifically rigorous literature, consisting mostly of anecdotal evidence, data from convenience samples and nonrandom program evaluations.”

Key issues for individuals with disabilities include challenges reporting crimes and being believed or taken seriously when they do report crimes. This problem is documented in Sorensen (2002), who refers to crime victims with disabilities as “invisible.” Another challenge is repeat victimization. In a Canadian study of 162 individuals with cognitive disabilities, Sobsey and Doe (1991) found that 80 percent of those who had been sexually assaulted had been victimized more than once, while 49.6 percent had experienced 10 or more sexual assaults.

In a survey (n = 352) of adult family caregivers and advocates (84.2 percent of the sample) and school-age child victims of sexual assault who have developmental disabilities (15.8 percent of the sample), respondents reported behaviors including having clothes pulled off or down (41 percent of the sample), having genitals exposed (40.4 percent of the sample), and forced intercourse (30.1 percent of the sample). Finally, 75.6 percent of respondents who reported victimization noted multiple incidents (Caldas & Bensy, 2014).

In one of the few studies specifically designed to gather data from individuals with cognitive disabilities, Wilson and Brewer (1992) surveyed 174 individuals at a sheltered workshop in Australia. The study found that the rate of sexual assault was 10.7 times greater in the sheltered workshop than for the general population (Wilson & Brewer, 1992). Further, Wilson and Brewer (1992) found that rates of victimization were greater for individuals living in institutions. The Crime Victims with Disabilities Awareness Act became law in 1998 to “increase public awareness of the plight of victims of crime with developmental disabilities, to collect data to measure the magnitude of the problem, and to develop strategies to address the safety and justice needs of victims of crime with developmental disabilities” (P.L. 105–301, October 27, 1998). NCVS now includes statistics on the rate of victimization of individuals with disabilities.

In examining the data collected as part of the 2008 NCVS, Harrell and Rand (2010) found that the rate of violent crime against individuals with disabilities is twice that of individuals without disabilities. When adjusted for age, the rate of victimization for individuals with disabilities is two to three times higher than it is for individuals without disabilities for each type of violent crime measured (rape/sexual assault, robbery, aggravated assault, and simple assault). When measuring rates of victimization for rape/sexual assault, the age-adjusted victimization rate for individuals with disabilities is more than twice that of individuals without disabilities. Individuals with disabilities were slightly less likely than individuals without disabilities to offer resistance to the offender during a crime. Within measured categories of disability (hearing, vision, ambulatory, cognitive, self-care, and independent living), individuals with cognitive disabilities had the highest risk of victimization (Harrell & Rand, 2010).

These data are limited in that they do not include those residing in institutions. A significant number of individuals with disabilities reside in institutions, particularly those with severe disabilities. In addition, the format for the NCVS interviews may limit the ability of individuals with hearing or cognitive disabilities to participate. Individuals with cognitive disabilities may have difficulty understanding the terminology used in the interview questionnaire, and individuals with hearing impairments may not be able to participate in telephone interviews. Finally, when proxy interviews are allowed for individuals who are not able to answer for themselves due to cognitive or communication challenges, it may lead to underreporting of victimization. The proxy responder may not know about the victimization, or could even be the perpetrator of abuse. In these cases, the crime would go unreported (Harrell & Rand, 2010).

The issue of sexual offending against individuals with disabilities is receiving more attention today than in the past, yet both the rate and characteristics of sexual victimization involving individuals with disabilities is not well understood. Clearly, more and better data are needed to determine the extent of sexual offending against this population.

Sexual Offending Against Members of the Military

Depending on the population studied and the definitions used, the extent of sexual offending against members of the military varies widely. Studies have produced estimates suggesting that as few as 4 percent and as many as 78 percent of armed forces members have been the victim of a sex offense (Bastian, Lancaster, & Reyst, 1995).

Unfortunately, there has been little consistency across studies in the methodologies, sample population characteristics, definitions of sexual offending, and the wording of questions used to determine if a sex offense has occurred. In response to congressional reporting requirements, the U.S. Department of Defense (DoD) now collects data on sexual offending against active-duty service members. However, this data collection effort — in place since 2002 — only addresses active-duty personnel. A major shortcoming of the data collected by the department is the reluctance of service members to report acts of sexual misconduct. Scholars are beginning to look at the experience of military personnel both during active military service and after, but far more work is needed to understand the incidence and prevalence of sexual offending against this population.

The Navy’s Tailhook convention scandal in 1991 and the cases of sexual harassment and sexual assault at the Army training camp in Aberdeen, Maryland, in 1996 drew public attention to how women are treated in the military.

The department undertook a study of sexual harassment among active-duty military members in 1994 and published the results in 1995. Results of this study showed that 55 percent of women and 14 percent of men reported one or more incidents of harassment at work during the prior year. When asked about unwanted “sex-related attention” at work or off duty, 78 percent of women and 38 percent of men reported they had been harassed in the prior year (Bastian, Lancaster & Reyst, 1995). Researchers asked about five types of unwanted sex-related attention: sexual assault, sexual coercion, unwanted sexual attention, sexist behavior and crude/offensive behavior. Table 1 shows the results for women and men respondents.

Many service members did not consider the experiences they reported to be sexual harassment. Although 78 percent of women and 38 percent of men reported experiences that fell into the five categories presented in table 1, only 52 percent of women and 9 percent of men indicated that their experiences constituted sexual harassment (Bastian, Lancaster & Reyst, 1995). Of those who indicated sexual harassment in the survey, 24 percent reported their experience to someone else, including 40 percent of women and 17 percent of men (Bastian, Lancaster & Reyst, 1995). For those who reported the experience to someone else, the vast majority were not investigated, with only 14 percent of women and 4 percent of men indicating that the harassment was being investigated. Further, 10 percent of women and 7 percent of men were encouraged to drop their complaint, and 23 percent of women and 16 percent of men indicated their report was not taken seriously. Of those who reported the incident, 21 percent of women and 12 percent of men indicated that a supervisor or coworker was hostile after the complaint was reported (Bastian, Lancaster & Reyst, 1995).

Since 2002, the department has been required by law to conduct a quadrennial Workplace and Gender Relations Survey of Active Duty Members. This survey assesses the extent of “unwanted sexual contact”¹³ between military service academy members (Rock et al., 2011). Data from 2014 found that 8.2 percent of women and 1.1 percent of men experienced some form of unwanted sexual contact during the previous year, down from 12.4 percent and 2 percent respectively in 2012 (DoD, 2015). In the same study, 48 percent of women and 10 percent of men experienced some form of sexual harassment, which, compared to 2012, is a decrease from 51 percent of women and unchanged for men. These statistics may underestimate the extent of unwanted sexual contact, as “8 of 10 sexual assaults in the military go unreported” (Whitley, 2010, p. 1). The study also found that there were 28 complaints of sexual harassment at military service academies for the academic program year 2014–15, up from 20 in 2013–14; additionally, there were 91 sexual assault reports in 2014–15, up from 59 in the prior year (DoD, 2016a).

The DoD also completes an annual summary of the number of sexual assaults and sexual harassment complaints reported by military service members (DoD, 2016b). There were 6,083 reports of sexual assault involving service members in 2015, a 1 percent decrease from the number reported in 2014. In addition, there were 657 sexual harassment complaints, of which 56 percent were found to be substantiated.

In a national cross-sectional study of women serving in the military from 1971 to 2002 (Vietnam era to Persian Gulf era), Sadler and colleagues (2003) found that 79 percent reported sexual harassment during their military service, 54 percent reported unwanted sexual contact, and 30 percent reported one or more completed rapes (Sadler et al., 2003). Of those who had been raped, 14 percent indicated they had been gang raped. The study found that rape occurred more frequently on base, often in the barracks. The definition of rape used in this study was that adopted by the American Medical Association and the American College of Obstetricians and Gynecologists.

In examining the extent of sexual offending against veterans of military service while they were still active, recent attention has focused on military sexual trauma (MST), defined by the U.S. Department of Veterans Affairs as sexual assault or harassment during military service. Recognizing the importance of providing services, department hospitals now provide free services to survivors of MST, regardless of their eligibility for any other department services (Center for Women Veterans, 2011). The department has also supported research to understand the extent and consequences of MST.

Since 2002, the Veterans Health Administration has implemented universal screening for MST for veterans returning from Afghanistan and Iraq. The first national, population-based study of veterans accessing administration services after returning from Afghanistan or Iraq shows MST in 15.1 percent of women and 0.7 percent of men (Kimerling et al., 2010). This study was a cohort analysis of the medical records of 22,000 women and 143,000 men. High rates of postdeployment mental health conditions were found among all patients, and those with MST were significantly more likely to have a mental health diagnosis. This study may underestimate the extent of MST due to underreporting. Burnam and colleagues (2009) documented the stigma associated with seeking help among Afghanistan and Iraq war veterans. Examples of the stigma associated with help-seeking for MST are shame, desire to maintain unit cohesion, and fear related to reporting a fellow service member with whom the victim may continue to work.

Underreporting of incidents of sexual offending is a serious problem. A U.S. Government Accountability Office (GAO) survey found that half of service members who had indicated they had been sexually assaulted in the prior year did not report the assault (GAO, 2008). The office also found that even after being trained, 13 to 43 percent of military personnel were not sure how to report an assault (GAO, 2008). Sadler and colleagues (2003) reported that one-third of military women were uncertain of specific steps to take to report a rape and only 26 percent of victims reported their rape to a superior officer. It has been suggested that the range of reasons for not reporting include the victim’s fear that he or she will not be believed, shame, and fear of retaliation, being negatively judged, being revictimized, having to continue to serve alongside the perpetrator, and being prosecuted for other problematic or illegal behavior that occurred at the time of the attack, such as underage drinking, adultery, and so forth (DoD, 2004; GAO, 2008; Mullins, 2005; U.S. Air Force, 2002). Sadler and colleagues (2003, p. 5) found that one-fifth of victims did not report their attack because they believed that “rape was to be expected in the military.”

In a study with a convenience sample of 196 female veterans, 72 percent reported that they had experienced sexual abuse (Himmelfarb, Yeager & Mintz, 2006). This study differs from others in that it asked participants about sexual abuse that occurred during their childhood, during adulthood but before military service, during the time of their military service, and after their military service. Many other studies of MST among female veterans focus on the time of military service or solely on adulthood, and therefore are not comparable. This study was conducted at a Veterans Administration hospital in Los Angeles. The sample is not representative of the ethnic makeup of female veterans as a whole, or of those in other regions. The study participants volunteered to be in the study and may differ in their level of MST from those who chose not to participate. Also, the sample may not be representative of female veterans as clinical populations typically report higher levels of trauma than nonclinical populations (Himmelfarb, Yeager & Mintz, 2006).

The issue of sexual offending against members of the military has received significant attention in the media in recent years, and DoD has placed a renewed emphasis on prevention and intervention policies and practices. Still, more study on the extent, nature and dynamics of sexual offending involving members of the military is warranted to determine future policy directions.

Sexual Offending Against Native Americans in Indian Country

Native Americans comprise 0.9 percent of the U.S. population, or 2.9 million. Across the 565 federally recognized tribes (per the Bureau of Indian Affairs), there is no single, standardized repository of crime data to measure the extent of sexual offending against victims in Indian Country; thus, estimates of sexual offending in Indian Country are imprecise. Even though none of the existing estimates regarding the extent of sexual offending or victimization in Indian Country are precise, the available data consistently indicate that Native American women experience violent victimization and sexual victimization at significantly higher rates than other women in the United States.

NCVS’s most recently published results do not provide victimization rates for Native Americans.¹⁴ This is due to the reduction in the overall sample size for NCVS that recently occurred and the impact it had on the size of subsamples for certain groups, such as Native Americans. In short, Native Americans are no longer sampled in sufficient numbers to provide valid statistics for the group. To have an accurate national estimate of victimization against Native American women, NCVS would need to return to its original sample size.

In NVAWS, conducted in 1995 and 1996, 34 percent of Native American women reported a victimization of rape at some point in their life — the highest victimization rate of any racial or ethnic group and nearly twice the national average for all ethnic groups. The number of male rape victims cannot be estimated from this source due to the low numbers reported (Tjaden & Thoennes, 2006). The data sample for this survey included telephone interviews with a total of 88 Native American women and 105 Native American men. Caution is necessary when generalizing about 2.9 million Native American women and men from this sample of 193 individuals.

In their 1998 research, Tjaden and Thoennes found that lifetime victimization rates for stalking are higher for Native American women than for women of any other ethnicity examined in their study. For Native American women, the lifetime rate of victimization for stalking was 17 percent, while the rate for the study population as a whole was 8.2 percent (Tjaden & Thoennes, 1998). Similarly, Native American men reported higher lifetime victimization for the crime of stalking, with a rate of 4.8 percent compared to the 2.3 percent lifetime rate for men in the study population as a whole.

Available estimates of the extent of victimization are consistent in indicating high levels of victimization among Native American women. Few large-scale studies exist to describe the nature of victimization of Native American women. However, the studies that do exist point to a need for further research.

Hamby found that Native American women were twice as likely as women of other racial and ethnic groups to say that police would not believe them or would blame them if they reported a rape (Hamby, 2008). Hamby also found other barriers for Native American victims in seeking help from law enforcement, including prejudice, conflict between Western and Native values, language barriers, and poverty. Getting help is also complicated by jurisdictional issues if the crime takes place in Indian Country, as often it is not clear which law enforcement entity has jurisdiction to prosecute the crime. For crimes that occur on Indian lands, jurisdiction for handling the investigation may fall to federal, state, or tribal law enforcement, depending on whether the perpetrator is Indian or non-Indian as well as on the nature of the crime (e.g., felony, misdemeanor).

Bachman and colleagues (2010) found that rapes involving Native American women are more severe than rapes committed against other women. Analyzing archived data from NCVS, Bachman and colleagues (2010) found that 94 percent of rapes reported by Native American women involved physical assault, compared to 74 percent of rapes reported by non-Native American women. Fifty percent of Native American women rape victims were physically injured during the rape, compared to 30 percent of non-Native American women rape victims. Finally, more than three times as many rapes of Native Americans involved weapons — 34 percent compared to 11 percent (Bachman et al., 2010).

Hamby and Skupien (1998) conducted in-person interviews with 117 women living on the San Carlos Apache Reservation and found that in their current relationship, 75 percent had experienced physical violence from their partner and 62 percent had been injured by their partner. In a study of 341 women who visited health clinics located on the Navajo Nation Reservation, Fairchild, Fairchild and Shirley (1998) found that 42 percent had been physically assaulted and 12 percent had been sexually assaulted by a partner in their lifetime.

Comparisons across these studies are difficult as the sample sizes, sampling methods, study methods and definitions used are different. Many of the studies use convenience samples, which may make the results less generalizable to the broader population. Further, there may be differences in the experience of Native American women in rural areas and urban areas, yet this has not been studied. Also, methods of data collection differ. For example, Tjaden and Thoennes used a telephone survey with random-digit dialing to select participants; Fairchild, Fairchild and Shirley (1998) used in-person interviews conducted among medical clinic populations; and Hamby and Skupien (1998) used in-person interviews but recruited volunteers via media outreach.

Amnesty International found that Native American victims seeking help at their local health facility may not get the help they need, as facilities often lack rape kits or the specialized training needed to preserve evidence for use at trial. The organization reported that 44 percent of Indian Health Service facilities lack personnel trained to provide emergency services to respond to sexual violence, and 30 percent lack the basic protocols for treating victims (Amnesty International, 2007).

Similarly, the scarcity of resources in the criminal justice system in Indian Country is also a challenge. The Bureau of Indian Affairs reported that more than 30 Indian reservations had violent crime rates that exceeded the national average in fiscal year 2008. Many of these reservations have law enforcement staffing shortages that require a handful of officers to cover geographically large areas. For example, according to the Senate report accompanying the Tribal Law and Order Act of 2009 (Report 111–93), the Wind River Indian Reservation in Wyoming (with a violent crime rate that is more than three times the national rate) has only six or seven officers to patrol 2.2 million acres of reservation land. With two or three officers on duty at any given time, each could be responsible for covering 1 million acres (U.S. Senate, 2009).

Sexual assault has a significant impact on members of the Native American community. Despite the limitations of the available data, sexual victimization appears to occur disproportionately among Native American women, and resources for preventing and responding to sexual offenses in Indian Country appear to be inadequate and fragmented. While more research and better data collection systems are needed to document and understand sexual offending and victimization in Indian Country more thoroughly, there is little question that the problem of sexual offending against Native Americans warrants greater attention. As a result, the Attorney General’s Advisory Committee on American Indian and Alaska Native Children Exposed to Violence (2014) emphasized the need to address issues of violence, including sexual violence, against Native American children. Finally, the National Institute of Justice (NIJ) has developed a comprehensive research program to address violence against Native American women, and are testing a survey instrument administered by Native American researchers via computer assisted personal interviewing (Crossland et al., 2013).

Summary

Although credible conclusions are difficult to make given the limitations of the available data, statistics from several key sources suggest that the incidence of sexual offending may be declining. UCR data on sex crimes reported to the police, NCVS data on sexual victimizations, and NCANDS data on substantiated reports of child abuse and neglect all present a consistent picture of declining incidence over time. Whether this pattern can be explained by factors other than a true decline remains subject to debate, but the convergence of key indicators and other empirical evidence suggests that the true decline hypothesis should be further examined and not dismissed (Finkelhor & Jones, 2004). Keeping in mind the limitations of the data, policymakers are encouraged to monitor key indicators of incidence over time and to work with researchers to better understand the factors influencing patterns in the data, including the roles of various policies and practices designed to prevent, treat, or otherwise intervene in sexual offending behavior.

Knowledge Gaps and Recommendations for Future Research

With so many inconsistencies and uncertainties in the available data, identifying the most important knowledge gaps and priorities for future research is a daunting task. Nevertheless, to improve our understanding of the incidence and prevalence of sexual offending, several of the most pressing issues that warrant examination are as follows:

• Rates of disclosure. Evidence indicates that victims sometimes choose not to disclose crimes that have been committed against them. In some cases, it may be to save themselves from re-experiencing the trauma of the event. The SOMAPI forum participants identified the need for additional research concerning the ways the criminal justice system contributes to underreporting and the steps that can be taken to address the problem and improve support for victims.
• Victim perception of the crime. With so many victims not labeling a sexual victimization as a crime or a rape, further study is needed to help identify the factors within the victimization experience (e.g., offender manipulation, posttraumatic stress disorder) that shape victims’ perceptions of the offending behavior.
• Wording of questions. Some evidence suggests that the way in which questions are worded in a victimization survey will influence reported levels of sexual violence. For example, Cook and colleagues (2011) reported that when the question uses the tactic of leading with a behaviorally specific description of an unwanted sexual act rather than a question about the sexual act, greater rates of victimization and perpetration are described by both men and women. This merits further investigation.
• Prevention. The literature on what works in preventing sexual abuse is neither complete nor rigorous. More study in this area could provide insight into how best to allocate scarce resources.
• Vulnerable populations. More research is needed to understand the extent and nature of sexual victimization of individuals in vulnerable situations, including children in schools or youth programs, young adults on college campuses, the elderly, individuals with disabilities, and those living in rural and hard-to-reach areas (including Native American and Alaska Native women and men). Individuals in settings such as these may have limited ability to protect themselves or seek help after victimization.

Without valid data on the nature and extent of sexual victimization, policymakers and practitioners are more likely to rely on anecdotes, opinions or stereotypes rather than facts when developing prevention and intervention strategies. Gaining a better understanding of the extent and nature of sexual victimization will help policymakers and practitioners develop responses that are both more effective and more responsive to the needs of victims.

Given the current state of our knowledge base, there is an acute need to both improve and expand our data on the incidence and prevalence of sexual victimization. In particular, work should be done to enhance the comparability of incidence and prevalence data from different sources and time periods. Currently, methodological variations — including differences in the ways sex crimes and victimizations are defined — make comparisons across data sources and time periods challenging. Agencies responsible for administering data collection efforts should actively seek opportunities to implement common and consistent data collection methodologies when possible. Funding for such efforts and for the expansion of data collection is critically needed.

Sex offenders do not typically self-report sex crimes.

There also is an acute need to learn more about the underreporting of sex crimes. Steps should be taken to create an environment in which victims feel appropriately supported and protected in the criminal justice and service delivery systems. Many of the barriers to reporting have already been identified through research, but SOMAPI forum participants acknowledged the need for further study in this area. In addition, policymakers must be committed to making the types of changes within the criminal justice and service delivery systems that are needed to overcome reporting barriers. Just as importantly, steps should be taken to help ensure that victims are not re-traumatized when reporting any victimization to authorities or when supporting the prosecution of perpetrators.

An accurate accounting of sexual offending is virtually impossible because so many sex crimes are hidden from public view.

Finally, it is important to acknowledge that even when sex offender management approaches are designed and delivered based on scientific evidence, hidden offending presents significant challenges. (For more, see Chapter 8. “Sex Offender Management Strategies,” in the Adult section.) Given the number of sex crimes that go unreported, the number of sex offenders that have never come to the attention of authorities, and the disproportionate attrition of certain sex offenses and sex offenders within the criminal justice system, any perception or expectation on the part of the public or policymakers that sex offender management professionals working in the community are providing victims and the public with protection against all sex offenders is unrealistic. Simply put, there are many unidentified sex offenders who are not being managed within existing systems and much reoffending that is not accounted for in the management process. Therefore, practitioners must be up front about these limitations and expectations for sex offender management on the part of the public, and policymakers must be tempered accordingly.