Bad math, bad research: the truth about abortion and rape-related pregnancy

The numbers in a recent study are shocking and disturbing. Fortunately, they are not accurate.

As state legislatures continue to wrestle with where to draw the lines on abortion policy in the wake of the Supreme Court’s Dobbs decision, bad research threatens to misinform their work.

That’s certainly true in the case of a research letter published in JAMA Internal Medicine late last month. The piece claimed that more than 64,000 rape-related pregnancies have occurred in the 14 states that have banned abortion since the bans have gone into effect. The piece has garnered a flurry of media attention in recent days, and understandably so. The numbers are shocking and disturbing.

Fortunately, they are not accurate.

There are serious methodological flaws with the JAMA paper that led the authors to overstate the prevalence of rape-related pregnancy. The most obvious problem is a sizable math error the authors made when calculating the number of rape-related pregnancies, inexplicably inflating their result by 250 percent.

When calculating the number of rape-related pregnancies, the authors should have simply multiplied their estimate for the incidence of rape by the post-rape probability of getting pregnant. Instead, they took the curious additional step of multiplying that result by a ratio that appears to have no relevance at all — a number generated by dividing the lifetime probability of rape-related pregnancy as reported by the CDC by the rape-related pregnancy rate per victim as reported by another study.

There is no rational basis for this decision. Barring any population-wide shifts in health, stress, smoking or other factors that can affect fertility, the probability of getting pregnant on a randomly selected day will always be about the same. As a result, the per-rape probability of pregnancy does not need to be adjusted for the lifetime probability of rape-related pregnancy or the rape-related pregnancy rate per victim. In short, this necessarily biases the results.

There are also other problems with this study.

We know, for example, that only a very small percentage of abortions are sought for reasons of rape or incest. In fact, according to a study from the pro-abortion-rights Guttmacher Institute, only about 1 percent of abortions are sought for reasons of rape, and less than 0.5 percent of abortions are sought for reasons of incest.

The JAMA paper’s estimates for rape-related pregnancies, however, when converted to 12-month estimates, imply that about 22 percent of all abortions in these 14 states would have been sought for reasons of rape or incest, if not for the abortion bans. Given the data we have on abortions sought for reasons of rape, this statistic is simply not credible.

The study also fails to account for the many women who find themselves in the difficult situation of a post-rape pregnancy who have no desire to further their pain by aborting their unborn child. It is a mistake to conflate, as this research paper seems to do, the incidence of rape-related pregnancy with abortion, as many survivors of rape want to keep their child.

Add to this the fact that it is very difficult to get accurate numbers on the prevalence of rape. The research paper bases many of its numbers on the CDC’s National Intimate Partner and Sexual Violence Survey, which had a response rate of only 59 percent after eligible respondents were contacted. This of course, is not the fault of the authors of the JAMA paper. But we must realize that estimates can only be as good as the data we have available. Because the response rates on surveys asking about rape tend to be low, the incidence of rape may be under-reported or over-reported.

While the research paper’s estimates for the number of rape-related pregnancies may not be reliable, what is clear is that more needs to be done to combat the prevalence of sexual violence in America. But instead of talking about how to get more victims of rape into abortion clinics, we should talk about how to prevent rape from happening in the first place.

Policymakers should tackle this problem by passing broader and stricter penalties for persons who commit rape. They should take steps to tackle the epidemic of pornography that dehumanizes women as sex objects and destroys marriages. Policies aimed at stopping online child-groomers, sex-traffickers, stalkers and other persons likely to engage in sexual violence before they commit or facilitate sex crimes would also go a long way.

So no, there were not 64,000 rape-related pregnancies in states that have passed laws protecting the unborn. Instead of seeking ways to get more victims of rape into abortion clinics, we should do more deter rape through stricter penalties and effective enforcement. We should also seek further policy solutions that can help prevent these crimes from happening in the first place.

Jonathan Abbamonte is a senior research associate in the Center for Data Analysis at The Heritage Foundation, an educational non-profit think tank that defends the right to life for the unborn and condemns all forms of sexual violence.