Taking Rape Statistics Seriously

face of a prisoner

Rape and sexual assault are acute forms of victimization. High rates of sexual victimization not currently addressed through criminal justice systems indicate serious problems for criminal justice systems. That’s particularly true for the U.S. criminal justice system, which currently provides an internationally extraordinary example of mass incarceration. Nonetheless, much U.S. work on sexual victimization is of poor deliberative quality. Scholarly research on rape and sexual victimization shows no appreciation for large anti-men gender bias in the justice system and the injustice of mass incarceration.

Consider, for example a Research for Practice report that the U.S. Department of Justice’s National Institute for Justice published in 2005. This report followed up on a major, government-funded study and a Congressionally mandated report. On the second page of the Research for Practice report, in a section entitled, “The scope of the problem,” the reader finds this text:

Just under 3 percent of all college women become victims of rape (either completed or attempted) in a given 9-month academic year. On first glance, the risk seems low, but the percentage translates into the disturbing figure of 35 such crimes for every 1,000 women students. For a campus with 10,000 women students, the number could reach 350. If the percentage is projected to a full calendar year, the proportion rises to nearly 5 percent of college women. When projected over a now-typical 5-year college career, one in five young women experiences rape during college.^

These are horrifying statistics. Struggling to think through these horrifying statistics, one could easily become baffled:

  • By what standard would anyone judge that just under 3 percent of all college women being victims of rape in a 9-month academic year is risk that “seems low”? Why does the text suggest that this is a low risk on first glance (“On first glance, the risk seems low”)? {The answer seems to be the authors’ prior expectation of a much higher share of victims.}
  • The figure “just under 3 percent” is just under 30 for every 1,000, or just under 300 for 10,000. Why does the text state 35 for every 1,000, and “could reach 350” for 10,000? {The answer is that the latter figure refers to crimes, and the former figure refers to women. That difference is not a translation of the percentage.}
  • Just under 3 percent for a 9-month academic year, scaled linearly, gives just under 4 percent for a full calendar year. Why does the text state “nearly 5 percent”? {The answer is that the “just under 3 percent” figure is actually for a reference period of 6.91 months, not a 9-month academic year.^}
  • The text scales 5 percent linearly over a “now-typical 5-year college career.” In 2005, about 40% of women enrolled in degree-granting colleges were enrolled in colleges granting 2-year degrees. In 2005, about 50% of women ages 25-44 who had some college education had a highest educational attainment of no college degree or a two-year degree. What percent of college women spend 5 years or more in college? {The answer is probably less than half.}
  • 5 percent scaled linearly across five years gives 25%, or one in four. Why does the text state one in five? {The answer is probably that the authors prefer the “one in five” statistic. The survey report published in 2000 projects the figure to “between one-fifth and one-quarter,” and adds in an endnote: “These projections are suggestive. To assess accurately the victimization risk for women throughout a college career, longitudinal research following a cohort of female students across time is needed.”^}
  • Why does the report ignore college men? The prevalence of sexual victimization of men is roughly equal to the prevalence of sexual victimization of women. Sexual victimization of men has been of almost no public concern.

Estimating objectively rape and sex assault via a crime victimization survey is conceptually and statistically difficult. But the issue here is different. The questions above don’t require subject-matter understanding and don’t concern complicated issues of statistical estimation. They involve basic thinking and factual coherence. Those qualities don’t seem to have mattered much in this important public research report. Given the size and functioning of the U.S. criminal justice system, that’s also horrifying.

The U.S. Department of Justice has sponsored and distributed research with deep anti-men gender bias. That has practical implications for the operation of the criminal justice system and the rise of mass incarceration.

Leave a comment (will be included in public domain license)

Your email address will not be published. Required fields are marked *