A New Row Over Pregnancy Caused by Rape


March 21, 2013

What percentage of raped women become pregnant? Answer: nobody knows.

Strike that. Many people claim to know, but the actual rate is hidden from us, though it can be estimated with considerable uncertainty.

Maybe the better question is, are pregnancy rates higher, the same, or smaller for raped women than for non-raped women? Answer: nobody knows.

Orthodox. Faithful. Free.

Sign up to get Crisis articles delivered to your inbox daily

Email subscribe inline (#4)

Well, we’re getting nowhere fast, because folks think they have this one pegged, too. So confident are they, that asking the question is seen by them as sacrilegious. Todd Akin (during his Senate run) learned this the hard way when he said that rates were smaller, which is not the desired answer.

Poor Akin paid for his apostasy, as you would have thought everybody knows. But Celeste Greig of California Republican Party Headquarters strangely didn’t, because in early March she unwisely answered the second question, too. Unlike Akin, she gave the true answer, which is that nobody knows. She said,

Granted, the percentage of pregnancies due to rape is small because it’s an act of violence, because the body is traumatized. I don’t know what percentage of pregnancies are due to the violence of rape. Because of the trauma the body goes through, I don’t know what percentage of pregnancy results from the act.

This is a good answer. Pregnancy rates for a session of intercourse are low no matter what, rape or non-rape. Rape is (nearly always) traumatic. Trauma by definition affects the body. It is plausible therefore to suggest that rape changes the likelihood of conception. She doesn’t claim to know by how much the likelihood is changed, if at all.

celeste greig rubioThe answer’s goodness is not interesting to MoveOn.org, and its sister site SignOn.org, whose permanently “outraged” members have begun a petition demanding Greig resign. They find Greig’s answer “offensive” (possibly the gravest of crimes) and an affront to science.

Science could not be reached for an interview, but papers on the subject made for interesting reading. The one exciting most interest is by Jonathan and Tiffani Gottschall from a 2003 issue of Human Nature, the journal devoted to publishing evolutionary theories of every aspect of human behavior. Jonathan Gottschall has a PhD in English and spends most days “bringing a Darwinian perspective to literary analysis” including “placing events in the Homeric epics, including rape, in evolutionary context.” Tiffani Gottschall is a PhD economist examining fertility of American Indians and other “demographic groups.”

In their review of the literature, they discovered reported rape-conception rates from 1% to 5%, with one paper boasting 10%. Their own work on a sample of 405 women from 1982 suggested 6.42% (not 6.41%, nor 6.40%, but 6.42%) is closer to the mark. They then fiddled with this number, “adjusting” it statistically, a trick examined below.

What science says, generously assuming the thoroughness of their literature review, is that the rape-conception rate is anywhere from 1% to 10%, more probably in the middle of that range. But “more probably” could mean the rate is 1% or even lower, just as it could mean it is 10% or higher. Therefore, to say that any one of these estimates is “the” correct one, is to say what is not warranted and is unscientific behavior. What the exact, actual, scientifically unambiguous number is, nobody knows, as claimed above.

The pair of Gottschalls make some interesting points, but some odd ones, too. What is the chance a women on birth control becomes pregnant when raped? Not high, but not zero, either, unless the form of birth control was sterilization. For the woman or the man. Yes, the man’s fertileness must be considered, which the Gottschalls do. They conjecture that rapists have beefier sperm, or that rapists produce more of the stuff so that once it finds its mark it crowds out competitors, or that rapists have developed keen fertility radars such that they preferentially target the fecund. That’s evolution for you.

Anyway, half a century ago few women used birth control. Now as many as 60% do, but even that number is dicey. Younger rather than older women naturally are more likely to employ contraception, and it’s younger women who are more likely to become pregnant. Women trying to have kids don’t use contraception. Women not trying for kids often insist their male partners use condoms.

In our science of rape-pregnancy rates, how do we account for the women who are raped who are on contraception? How do we account for age? How do we compare studies through time, when the proportion of women using contraception has been changing so much? Women who were once on contraception but now off have more difficulty conceiving, so we’d have to factor that in, too. The demographics of the country have changed also, and these matter in calculating fertility rates.

Comparing conception rates between raped and non-raped women isn’t easy. Statistics for non-raped women are often presented as chance of getting pregnant over a month or year of having unprotected intercourse. That’s because the number of times a couple has intercourse over a month or year isn’t well documented. (Anecdotal reports suggests couplings decline with age.) Rapes are usually single events.

A lot of older women are engaged in fertility enhancement. If these women are raped, perhaps the chance of becoming pregnant is larger for them than for women aging naturally. Women are now having their first kids older, which also lowers chances of conception. A certain, and shifting, proportion of rapes are of girls who have not reached menarche. No pregnancies there.  And women are only fertile for a portion of time. Raped women have forced intercourse at any time in their cycles, while non-rapes women avoid intercourse off season.

Sometimes (as the Gottschalls remark), a woman is uncertain of the paternity of her child: could have been the rapist, could have been her man. They only cite one statistic for “per-incident consensual pregnancy rate for women in their twenties”. This was between 2% and 4%, extracted in a book written in 1966 by the very pro-abortion Lawrence Lader. He also wrote, in 1971, Breeding Ourselves to Death, which is a less jolly companion to Paul Erhlich’s The Population Bomb.

Some rapes are more physically traumatic than others. It’s not unbiological to suggest that, considering only cases of rape, increasing trauma affects pregnancy rates. Nobody knows whether or not it does, because nobody has done a study.

We could, and should, go on, but the point has been, or at least should have been, made. Nobody knows what the exact pregnancy rate is for non-raped women, nor for raped women. Comparisons are fraught with uncertainty, and that’s assuming the data we have collected is pure and useful.

It isn’t always. The Gottschalls made use of the National Violence Against Women survey, which was sponsored by the National Institute of Justice and the CDC. Eight-thousand women were called and asked not if they were raped but “has a man or boy ever made you have sex by using force or threatening to harm you or someone close to you?” (“sex” was later spelled out). There were 405 females between 12 and 45 who reported just one incidence (those saying more than one were excluded) occurring mainly between 1977 and 1979.

They scientifically acknowledge that women could have lied in answering the unknown and mysterious phone surveyor (in both directions, with nobody knowing the exact effect), and that rapes resulting in pregnancies are more likely reported—by how much nobody knows. The real magic in their paper happens when they statistically “adjust” their 6.42% rape-conception statistic to “account” for contraception. Did they do this by asking the raped women if they were on contraception? No sir, they did not.

They instead looked in 1982′s Statistical Abstract of the United States and teased out (with error) percentage of pill and IUD use by age, and then with fingers crossed assumed the (self-reportedly) raped women of 1977-1979 were on average identical. Curiously, the mathematical technique they employed to do the “adjustment” was not published, but since the answer was pleasing, no explanation by the reviewers was apparently demanded. Their final number, based on this small sample and the unknown fiddling, is 7.98% (and not 7.97%, etc.).

This last number is subject to wide and vast uncertainty, around which any competent statistician would have included a plus-or-minus range. None appears in the Gottschalls’s paper. And anyway, it’s just wrong. A raped woman on contraception doesn’t (usually) get pregnant. There is no adjusting to do.


  • William M. Briggs

    William M. Briggs is a consultant and adjunct Professor of Statistics at Cornell University, with specialties in medicine and the philosophy of science. He blogs at wmbriggs.com.

Join the Conversation

Comments are a benefit for financial supporters of Crisis. If you are a monthly or annual supporter, please login to comment. A Crisis account has been created for you using the email address you used to donate.

Editor's picks

Item added to cart.
0 items - $0.00

Orthodox. Faithful. Free.

Signup to receive new Crisis articles daily

Email subscribe stack
Share to...