Experts Question Accuracy of New Rape Statistics

New government crime statistics showed the number of rapes dropped by a third last year, from 141,070 in 1999 to 92,440 in 2000. But many women's advocates and criminologists don't trust the numbers or the reporting methods.

Bonnie Campbell

(WOMENSENEWS)-Everyone wanted to take credit last week when the Justice Department reported a 15 percent drop in overall violent crimes-and a stunning 34 percent reduction in the number of rapes in the year 2000. But many people are not convinced.

Attorney General John D. Ashcroft said it showed a “betterment of the quality of life.” Democratic Sen. Joseph R. Biden Jr. of Delaware claimed Clinton-era anti-crime programs had been a huge success.

Now, however, criminologists and women's groups increasingly are questioning the accuracy of the National Crime Victimization Survey, especially the supposed drop in rapes.

The National Crime Victimization Survey reported the lowest level of violent crimes since the Department of Justice started the annual sampling in 1973.

Surveyors questioned 160,000 Americans, half of them women, to determine how many people were victims of crimes, including rape.

“I don't believe it at all,” said Bonnie Campbell, who until recently oversaw the Violence Against Women Office in the Department of Justice. “I have not seen one thing that would indicate that.”

Women Say Low Official Figures for Rape Don't Fit the Reality

“It just doesn't fit together,” said Karen Baker, project manager for the Pennsylvania-based National Sexual Violence Resource Center. “It doesn't make sense.”

Karen Baker“I haven't heard from anyone that ‘Oh, nothing's happening,” or ‘We're sitting here with nothing to do,'” said Mary Beth Carter, of the California Coalition Against Sexual Assault.

There's good reason for skepticism:

Only three weeks ago, on May 30, the FBI announced that in the year 2000, the number of violent crimes-murder, rape, robbery and serious assaults-had slightly increased, not decreased. And the FBI's Uniform Crime Report said the number of rapes during the year had shown a small increase.

Equally confusing was last year's National Crime Victimization Survey, also known as NCVS, which showed that rape increased by 20 percent in 1999. The FBI's Uniform Crime Report, or UCR, said the number of rapes went down 5 percent that year.

“I distrust the rape statistics in the NCVS and the UCR,” said Carnegie Mellon University professor Alfred Blumstein, a frequently cited expert on crime in the United States.

“We shouldn't have a whole lot of confidence in the way they are measuring rape,” said University of Pennsylvania criminologist Lawrence Sherman.

Reported Attacks Rise When Definition Includes Spousal and Date Rape

In reality, rape is far different from other crimes measured in the Justice Department surveys. There is little debate about the definition of murder or robbery. But by merely defining rape more broadly, to include date and spousal rape, the reporting of attacks can increase dramatically.

As awareness of rape increased from 1960 to 1973, when the National Crime Victimization Survey began, the number of attacks that police classified as rapes more than doubled.

It's unclear how much of the change was an increase in real crime and how much involved the change in the definition of rape. Some police, for example, are just beginning to acknowledge that date rapes and spousal rapes actually occur.

Last week's report was not just an occasional statistical anomaly.

In the past year alone, the Justice Department has published a variety of studies with estimates of rape in the United States varying by huge rates.

And it's not just a matter of statistics. The crime numbers are a key ingredient for determining the allocation of millions of dollars in government funds for crime fighting and victim assistance.

“It drives policy and funding decisions, and we absolutely must have an accurate picture of crime,” said one ranking official in the Justice Department's Violence Against Women Office.

“There's an old adage in Washington that until somebody counts the problem, nobody does anything about it,” said criminologist Sherman at the University of Pennsylvania.

The statistics also can affect policy.

To Boost City Image, Some Police Departments Hid Rape Complaints

In an ill-conceived effort to make Philadelphia look safer, for example, police hid hundreds of sexual assault complaints each year, leaving victims without protection as suspected rapists remained on the street. The rape complaints of prostitutes and homeless women were routinely shelved.

The city of Atlanta this year uncovered a similar pattern of detectives hiding rape complaints in secret files.

Officials responsible for the National Crime Victimization Survey have been scrambling to explain the supposed drop in rapes and other crimes in 2000.

“The measure of rape is always the most challenging,” said Lawrence Greenfeld, the acting director of the Bureau of Justice Statistics. “We recognize that nothing is perfect.”

Greenfeld said the national rape statistics in the victimization survey were based on responses from about 80 women who said they had been sexually assaulted. He said that to find that limited pool of 80 rape victims, surveyors questioned about 80,000 women. Any effort to increase the number of victims, and possibly to improve the accuracy of the study, would be exceptionally expensive, he said.

“It's just very expensive to increase that,” Greenfeld said. “It's not like going out and measuring the temperature.”

Greenfeld said that, in addition to problems caused by the small group of rape victims interviewed each year, other factors prevented rapes from being accurately counted in the victimization survey.

Nearly One-Fourth of All Rape Victims Are Younger Than 12

The survey, which has been measuring crime since 1973, involves interviewing victims at home. It only questions victims aged 12 and older. Greenfeld said that in nearly one-fourth of all rapes, the victims are younger than 12.

And young victims may skew the survey in another way, Greenfeld conceded.

The survey is designed to ask identical questions of all the people interviewed. The survey's questions about sexual assault avoid a level of explicitness that might be traumatic for a child who is 12 or 13 years old.

The recent Justice Department study of college students, however, found that the women might be more willing to tell interviewers about rapes when detailed and explicit questions were asked.

Telephone Surveys Miss Low-Income People Who Can't Afford Phones

And economics is a factor.

Poor people generally are more frequent victims of violent crimes than middle-class Americans. But during the past two years, the survey has been using more and more telephone interviews, rather than face-to-face contact with people being surveyed. And poor people have fewer telephones.

“A lot of victims don't have telephones, so we keep knocking on doors,” said Greenfeld of the Bureau of Justice Statistics. “The highest rates are at the lowest economic rung.”

Last week's survey estimated that in raw numbers of attacks, the incidence of rapes in the United States dropped to 92,440 last year from 141,070 in 1999.

When Greenfeld was asked if he really believed that rapists had attacked a third fewer women last year than in 1999, he insisted that aside from the survey, “I have no other way of knowing.”

Greenfeld's assistant, Michael Rand, one of the Justice Department officials who oversaw the study, characterized the 34 percent drop in reported rapes as “our best guess,” but then quickly corrected himself to call it “our best estimate.”

Both Greenfeld and Rand said that because of the small number of incidents of rape, about 80, recorded by the victimization survey, the best way to use the survey was to view trends over several years.

“It's always hard to put too much emphasis on an annual event,” said Rand. “It's important to take a longer view.”

Conflicting Official Statistics May Point to Flawed Rape Stats Over Time

“Looking at six or seven years, you can learn a heck of a lot,” said Greenfeld. “Take a number of years and look at that slice of data.”

But the longer-term rape reporting included in last week's victimization survey may raise even more questions.

It said that in the seven years between 1993 and 2000, rape and attempted rape dropped by 62 percent. The FBI for those same years reported a 14 percent reduction in rapes and attempted rapes.

It may simply confirm that the rape statistics have been flawed over the long term.

“My fear has been for a long time that the failure to deal with the culture around rape and sexual assault has made these numbers somewhat irrelevant,” said Campbell, the former director of the Violence Against Women Office who now is in private law practice in Washington.

Despite last year's supposed 34 percent drop in attacks, Campbell said, women's groups should have no doubts about the reality of rape in the United States.

“It certainly doesn't mean that our work is done,” she said.

Mark Fazlollah is an award-winning reporter for the Philadelphia Inquirer.

For more information, visit:

Federal Bureau of Investigation:

Bureau of Justice:

Other articles by Mark Fazlollah at the Philadelphia Inquirer:

Other articles by Mark Fazlollah on Women's Enews:

Detroit Police Admit Rape Arrest Stats Were Wrong (April 27, 2001):

Women Activists Monitor Philadelphia Rape Squad (Feb. 19, 2001):

City, National Rape Statistics Highly Suspect (Jan. 8, 2001):

Police May Ignore Rape Complaints to Boost Image (June 15, 2000):