Get more FiveThirtyEight
In 2017, dozens of women, as well as several men, made claims of sexual misconduct against powerful men in Hollywood, the media, the tech industry, and government. But sexual assault is a pervasive phenomenon, and these high-profile revelations barely scratched the surface. The National Crime Victimization Survey, released last month, estimates that people in the U.S. experienced over 320,000 incidents of rape and sexual assault in 2016.1The basic unit of the NCVS is a “victimization,” which is a crime affecting one person or household. If one person was the victim of the same type of crime on three occasions, for example, that would count as three victimizations. If two people were affected by a single crime, that would count as two victimizations. That works out to 1.2 such assaults per 1,000 people age 12 or older.
The annual household survey, conducted by the Bureau of Justice Statistics,2First interviews are typically conducted in person, with subsequent interviews conducted either in person or by phone. captures data on a number of crimes, including rape and sexual assault, that respondents experienced during the previous six months. It tends to capture much more information than sources like police statistics, since many crimes are never reported to the police. The survey helps us understand who these victims are and how they respond to being attacked.3We’re only using the numbers from 2016 for our analysis. Last year, the NCVS underwent a routine sample redesign to account for population changes identified in the 2010 census. For this reason, the most recent data can’t be compared to data from previous years.
Among survey respondents who said they’d been victims of sexual assault or rape, the majority of incidents were reported by women and girls (84 percent), while 16 percent were reported by men and boys.
Most research on sexual violence focuses on male perpetrators and female victims, though the NCVS estimates that men were victims of over 51,000 incidents of rape or sexual assault last year. Some research suggests that men face a greater stigma in reporting rape and sexual assault than women, though men identified themselves as victims of predatory sexual behavior in several high-profile cases in 2017.
“I think men are victimized far more than we realize, because when they come forward, people question their masculinity and sexuality,” said Callie Rennison, a professor for the School of Public Affairs at the University of Colorado Denver.
Sixty-six percent of the incidents of rape and sexual assault included in the NCVS survey were reported by non-Hispanic white respondents, 15 percent were reported by Hispanic respondents, and 13 percent were reported non-Hispanic black respondents.4We used the NCVS measure of race and Hispanic origin to capture victimization by both race and ethnicity. That tracks fairly closely with the racial and ethnic breakdown in the U.S., which is 62 percent non-Hispanic white, 17 percent Hispanic and 12 percent non-Hispanic black (according to the Census Bureau). However, much of the existing literature on sexual violence underrepresents racial and ethnic and minority groups. The current #MeToo movement has faced criticism for failing to include or highlight the voices of people of color.
“What’s largely left out of the conversation about the impact on victims is the impact on people of color and minority groups,” said Jennifer Long, CEO of AEquitas, a national organization that helps prosecutors improve how they approach cases involving domestic violence, stalking and human trafficking.
Because researchers typically treat victims of sexual assault as a cohesive group, according to Scott Berkowitz, founder of the Rape, Abuse & Incest National Network, there have been few attempts to understand how these crimes affect subgroups of the population.
Numbers may not add up to 100 due to rounding and answers classified as “unknown” in the NCVS data. Respondents with household income between $25,000 and $34,999 were grouped with the unknown responses because the sample size for this group was too small to be reliable.
Some 44 percent of victims who participated in the survey said their household income was less than $25,0005The Census Bureau set the federal poverty level at about $24,400 for a family of four in 2016. — while only 22 percent of households nationwide earn less than $25,000 a year, according to the Census Bureau. When you compare the lowest- and highest-income groups, the difference in victimization rates is stark: People with household incomes of less than $7,500 reported a victimization rate of 4.8 incidents per 1,000 persons age 12 or older, which is 12 times the rate reported by those with household incomes greater than $75,000 (0.4 per 1,000).
Research has shown an international link between poverty and sexual violence around the world. The Centers for Disease Control and Prevention lists poverty and lack of jobs as community risk factors for sexual violence. And people with low incomes, who have less access to resources, are more vulnerable to sexual assault; research by advocacy groups suggests that perpetrators are more likely to target victims who are less likely to report what happened.
“In low-income communities, there’s a real risk of a cycle that reinforces itself,” Berkowitz said. “That means that those abusers in those areas are more likely to be able to continue without repercussions.”
Over half (59 percent) of the incidents that the NCVS logged in 2016 were reported by victims who lived in urban areas, while 28 percent were reported by victims who lived in suburban areas and 13 percent by victims living in rural areas. But while sparsely populated rural areas accounted for the lowest percentage of all reported incidents, people in rural areas reported the second-highest rate of rape and sexual assault among the three types of areas: People in urban areas reported 2.1 rapes and sexual assaults per 1,000 persons age 12 or older, compared with 1.1 such incidents per 1,000 people in rural areas. The rate was lowest in suburban areas, at 0.6 incidents per 1,000 people.6Using urban as the comparison group, the differences in rates are statistically significant at the 95 percent confidence level.
Urban areas tend to have more total crime, in part because they simply have more people. But sexual violence against people in rural areas brings with it its own set of challenges. Reports from advocacy groups and independent researchers have found that victims in these areas can be physically isolated from resources, and that social norms can make it difficult for victims there to find help. Because people in small communities are more likely to know one another, it can be harder for victims who report crimes to the police to remain anonymous to their friends and neighbors, which can make victims more reluctant to come forward, said Berkowitz, in part because the police officer taking the victim’s report may know the victim, the perpetrator, or both.
In the majority of the rapes and sexual assaults reported to the NCVS, the victim knew their attacker. Last year, in 39 percent of reported attacks, the victims said the offender was a well-known or casual acquaintance, and in another 33 percent of attacks, the victim said the offender was someone they had an intimate relationship with, including a current or former spouse, girlfriend or boyfriend. Only 19 percent of incidents involved victims who said the person who assaulted them was a stranger.
Advocates have been working for years to dispel the myth that sexual assault is predominantly carried out by strangers, and research has confirmed that most victims of such crimes know their attackers. Similarly, it’s a myth that sexual violence typically involves a weapon; in fact, only 11 percent of the cases of rape and sexual assault reported in the NCVS involved a weapon.
“It’s still a perception that has to be overcome, not just by society, but starting with the victim,” said Long. “It’s constantly reinforced that it’s not real rape if it’s someone you know and you weren’t physically injured outside of the rape. The victim is left feeling very confused and looking to blame themselves.”
Only 23 percent of incidents of rape and sexual assault were reported to the police, according to the NCVS, which makes these crimes the least likely to be reported of any type of crime recorded in the survey.
Fear of not being believed, self-blame and concerns about how the justice system will handle the incident are among the reasons that sexual violence is generally underreported. In previous years of the survey, when victims gave a reason for not going to the police, they most commonly cited fear of reprisal or the fact that they considered the event a private matter as their reason for not reporting.7These findings are from data that cover 10 years of BJS survey results, from 2006 to 2015. The sample sizes on the annual numbers are so small that they should be looked at in aggregate, according to Grace Kena, a co-author of the BJS’s 2016 Criminal Victimization report, which summarizes some of the NCVS results. Because of the survey redesign, data for 2016 would not be comparable to data from previous years, so 2016 was not included in this aggregate.
Victims told the NCVS that they received assistance from a victim-services agency, which Bureau of Justice Statistics statisticians say is defined as any agency other than law enforcement that helps victims of sexual violence, after just 11 percent of incidents of rape or sexual assault.
Measuring the true prevalence of rape and sexual assault is tricky. In addition to the notorious problem of underreporting, many surveys use different methodologies for counting victims and different definitions of what constitutes sexual violence, which can make it impossible to compare results.
The NCVS in particular faces criticism for undercounting victims, and it has the same limitations that other household surveys do — the survey relies on respondents to report their own experiences to an interviewer, which increases the potential for underreporting. However, the NCVS is conducted every single year, using fairly consistent methods, and, as many researchers point out, there’s no perfect way to assess rates of sexual violence.
Christopher Krebs, a researcher with RTI International who has worked with the BJS in the past, said one flaw of the survey is its language. Right now, survey-takers are asked, among other things, whether they’ve experienced “rape” or “sexual attack” as part of the process of screening for respondents who should be asked more detailed questions about sexual violence. “By using terms like rape and sexual assault, some people believe what you end up doing is triggering respondents’ memories to think about incidents that fit their understanding of what those terms mean,” said Krebs. But “what they conjure up in their minds are definitions that don’t necessarily fit the legal definition.”
Researchers say adding questions about behavior would help respondents identify their experiences with sexual violence. Lynn Langton, chief of the Victimization Statistics Unit at BJS, said the department is planning to launch a redesign of the survey in the next few years, which may include changes in how interviewers ask about sexual violence. But researchers and advocates say that until society encourages more women and men to come forward, reported numbers of these crimes will likely always be incomplete.
“It’s important to get a sense of the numbers because it helps support resources,” said Long. “But it’s also important to understand that even in a perfect survey, someone might not disclose.”