Sex Survey Discrepancy Tied to Prostitutes

Oct. 9, 2000 -- Everyone pretty much takes for granted the idea that men are more promiscuous than women.

Major studies of sexual behavior reveal heterosexual men usually report more sexual partners than heterosexual women.

The 1991 General Social Survey, for example, put the value at 47 percent higher for men. The National Health and Social Life Survey of 1992 said the rate for men was 74 percent greater.

But when researchers published these surveys, social scientists challenged the findings, saying the numbers did not add up. The total number of partners each sex reports should be equal, even if men had more sexual relations on average with fewer women. Yet in the surveys, the numbers were not equal.

To explain the discrepancy, researchers said men tended to inflate their conquests, while women underestimated their liaisons.

A new study published in the upcoming issue of the Proceedings of the National Academy of Sciences, says adding the visits of men to prostitutes helps even up the sexual partner accounting.

These earlier studies did not include enough “ladies of the evening,” the researchers say. When prostitutes are included, men and women have a similar number of partners: Men are still having more sex, but with a few chosen women.

Tracking Numbers of Prostitutes

The populations surveyed in the earlier studies included only households and failed to involve marginal living arrangements where prostitutes might live, explains Devon D. Brewer, a social scientist with the University of Washington in Seattle, and lead author on the study.

Rooming houses, military barracks, college campuses, motels, jails, prisons and homeless shelters were not included in these two national surveys, Brewer says.

To try to understand the differences detected in men and women in the two surveys, Brewer analyzed whether the number of prostitutes in the original studies was accurate. He found the original studies contained fewer prostitutes than were predicted for the overall population at that time.

The General Social Survey, for example, had among its respondents one prostitute who had approximately 60.5 partners in the previous year, Brewer says.

But Brewer found that survey should have included 1.67 prostitutes. A single prostitute would have had 694 partners per year, Brewer says.

“These numbers are all estimates,” Brewer says. “Very few studies give really reliable figures for the number of prostitutes in America and how many partners they have.”

Honest Answers

When Brewer reanalyzed the two surveys to include his new data, he found the average number of sexual partners of men and women appeared to equalize. The 74 percent figure for the National Health and Social Life Survey went down to 19 percent higher for men. The General Social Survey figure of 47 percent higher became 2 percent fewer partners for men.

The rate of sexual partners reported in the past five years also evened out, with his new analysis. Without Brewer’s prostitute data, men reported 240 percent more partners than women in the past half decade in the General Social Survey. With the prostitute data, the numbers were the same.

Brewer says his research is important because it says people are honest in what they report in surveys of sexual behavior. The problems weren’t due to misreporting but to sampling errors, Brewer says.

He concedes, however, his findings rely on three major studies estimating the number of prostitutes in the United States. “More research needs to be done to get accurate information about prostitutes and their behavior,” Brewer says.

Differences Between Men and Women

Commenting on the research, Norman Brown, associate professor of psychology at the University of Alberta, in Edmonton, questioned the prostitute data. “Brewer places too much importance on the small prostitute numbers, which also have reporting errors associated with it,” Brown says.

Brown has another explanation for the surveys’ discrepancies: men and women respond differently to questions requiring them to quantify their behavior. In his research, Brown has found men tend to give rough estimates of numbers, a strategy that leads to overestimation. Women tend to count, an approach resulting in underestimation.

But Brown adds, Brewer’s “Hooker Hypothesis” has some validity since the original surveys did not include as many people as possible.

“We are getting closer to understanding the factors contributing to the number of sexual partners men and women report,” Brown says.