In an opinion piece in the July 15, 2013, Washington Post, state Sen. Wendy Davis gave reasons "why I stood up for Texas women" in a filibuster against abortion restrictions that drew national attention.
Among them: "Each year, about 25,000 American women ... become pregnant through rape or incest," wrote the Fort Worth Democrat.
A Democratic effort to exclude victims of rape and incest from the Texas restrictions eventually proved futile, as did Davis’ filibuster; the measure was signed into law by Gov. Rick Perry on July 18, 2013.
We remained curious about Davis’ rape-or-incest statistic.
Davis spokesman Rick Svatora told us via email that the number came from a media fact sheet posted online by NARAL Pro Choice America, an abortion rights advocacy group formerly called the National Abortion and Reproductive Rights Action League. The fact sheet says, "Each year, approximately 25,000 women in the United States become pregnant as a result of rape."
There are no incest pregnancy statistics on the sheet, and we found none in other sources. Dean Kilpatrick, director of the National Crime Victims Research and Treatment Center at the Medical University of South Carolina, told us by phone that although rape and incest often are mentioned together in abortion laws, there is little research on incest, a difficult area to study for reasons that include underreporting.
By email, Svatora provided two articles on rape-related pregnancy:
A 1996 Medical University of South Carolina analysis of the federally-funded 1990-92 National Women’s Study of 3,031 women over 18. (Kilpatrick worked on both the study and the analysis.) The analysis concluded that among rape victims aged 12 to 45, 5 percent of rapes resulted in pregnancies
,and, using the study’s figure of 683,000 "forcible rapes" in 1990, estimated that "there may be 32,101 rape-related pregnancies annually among American women older than 18 years."
A 2000 article in the American Journal of Preventive Medicine, using the 1996 analysis’ estimated rate of 5 percent pregnancies from rape, assumed a 50 percent drop in rapes since 1992 to reach an estimate of 25,000 pregnancies from rape in 1998.
The NARAL Pro Choice America fact sheet attributes its 25,000 figure to the 2000 article, and a spokeswoman for the group, Samantha Gordon, said via email that is its most recent estimate.
The U.S. Centers for Disease Control and Prevention don’t keep data on pregnancy due to rape, nor does the U.S. Department of Justice, spokeswomen told us via email, while spokesman Greg Phillips at the American College of Obstetricians and Gynecologists said the 32,000 figure is "the best estimate we have now."
Kilpatrick said reasons why the 1996 analysis he worked on is still cited include the National Women’s Study’s large population sample and the fact that more recent research has tended to support its 5 percent rate.
The Rape, Abuse and Incest National Network, a Washington, D.C.-based charity, reached a much lower figure by doing the same calculation with an estimate from the Justice Department’s 2005 National Crime Victimization Survey. On its website, the network takes the survey’s annual average of 64,080 rapes committed in 2004 and 2005 and applies the 5 percent pregnancy rate to reach an estimate of 3,204 pregnancies a year from rape.
That’s ten times less than the 1996 estimate of 32,000. The biggest reason for that difference: According to federal statistics, rapes -- like other violent crimes -- declined significantly in the U.S. during those years.
Crime experts have said the decline shown by federal data that began in the 1990s was part of a long-term pattern. They cited falling crack cocaine use, a strong economy and other societal factors, with some citing tougher police strategies.
A March 2013 federal Bureau of Justice Statistics report, "Female Victims of Sexual Violence, 1994-2010," compiled statistics from the National Crime Victimization Survey showing that average annual completed rapes dropped 65 percent from 402,000 in 1995 to 143,300 in 2010.
Using the method the 1996 and 2000 researchers did -- multiplying the rape counts by the 5 percent pregnancy rate -- we calculated that there could have been 20,100 pregnancies from rape in 1995 and many fewer, 7,165, in 2010.
On the other hand, Kilpatrick disagreed with the results of the federal survey, which he said misses most rape cases in large part because its questions are framed in the context of crime and can leave doubt about exactly what’s being asked. An Interviewee, he said, might respond "no" to the question "Have you ever been raped?" but answer yes when asked if she had ever experienced specific behaviors such as a man forcing his penis into her vagina.
He and his colleagues estimated higher numbers of rapes in the 1990-92 survey and in a 2007 follow-up, estimating in both studies that only about 17 percent of rapes were reported. In fact, their data showed a 61 percent increase from 1990 to 2006, from 683,000 to 1.1 million rapes.
Kilpatrick noted that a 2010 CDC report found a similar total -- 1.3 million women raped in the previous 12 months. His work indicated the percentage of women raped has not changed significantly since the 1990s, he said, while the U.S. population increased.
Applying the 5 percent pregnancy rate to a rounded estimate of 1 million rapes yields his 50,000 figure.
We also checked FBI-collected crime statistics, mindful that they include only rapes reported to local law enforcement agencies. In 2011, according to the FBI’s most recent finalized rape statistics, 83,425 forcible rapes were reported to law enforcement. Multiplying that by the 5 percent rate suggests 4,171 pregnancies resulting from reported rapes that year.
If that FBI figure represented about 20 percent of rapes committed, in line with Kilpatrick’s research, the total would be 417,125 rapes. Applying the 5 percent pregnancy rate suggests about 20,900 pregnancies could have resulted from rape in 2011.
To sum up, we found or calculated these estimates cited from 2010 through 2012, all taking into account unreported plus reported, all based on an assumption of 5 percent chance of pregnancy:
3,200 -- From national anti-sexual-violence organization.
7,165 -- Our calculation with federal data for 2010 released in 2013.
20,900 -- Our calculation with FBI reported-rape data for 2011, adjusting for assumption that only 20 percent of rapes were reported.
32,000 -- Based on 1990 study but still in use by the obstetricians and gynecologists’ group.
50,000 -- Kilpatrick’s estimate based on 2007 and 2010 research showing approximately 1 million rapes.
By phone, Svatora noted that Davis’ number fell within that range.
Davis said, "Each year, about 25,000 American women ... become pregnant through rape or incest." She offered no backup for the incest part of the statement, and we found none.
Estimates of rape-related pregnancies per year vary from 3,200 to 50,000. Davis’ number, despite dating from 1998, fits inside that range, but there is too much disagreement for us -- or Davis -- to pin any one statistic as the definitive number of rape-related pregnancies per year.
We rate her statement as Half True.