Facebook Twitter Tumblr Close Skip to main content
A Project of The Annenberg Public Policy Center

Abortion Distortions

Senators from both sides make false claims about Roe v. Wade.


As President Bush considers exactly whom to nominate to succeed Justice Sandra Day O’Connor on the Supreme Court, Senators Barbara Boxer and Rick Santorum both have distorted some facts about the effect of Roe v. Wade.

Boxer, a Democrat, claimed that repeal of Roe “means a minimum of 5,000 women a year will die” from illegal abortions. But that’s a 69-year-old figure dating to a time before penicillin and the birth-control pill. Experts say nowhere near that many women were dying from abortion complications even in the years just before Roe made abortions legal nationwide.

On the other side of the abortion debate, Republican Santorum says that suicides by women, and also crime, “got worse, much worse” after Roe. But in fact, the female suicide rate is one-third lower now than in 1973. And the Justice Department’s annual survey on crime victimization shows a 69 percent drop in property crime and a 53 percent drop in violent crime since Roe.


 Boxer’s False Statistic

On July 5, Sen. Boxer claimed that overturning Roe v. Wade would cost the lives of more than 5,000 pregnant women a year. That might have been true before the invention of penicillin and the birth control pill, but it’s not true now. The best evidence indicates that the annual deaths from illegal abortions would number in the hundreds, not thousands.

Boxer made the claim to support her position that the repeal of Roe would be the sort of “extraordinary circumstance” that could justify use of the filibuster to stop the confirmation of a nominee to the Supreme Court. The Associated Press quoted her this way:

Boxer: It means a minimum of 5,000 women a year will die. So all options are on the table.

But Boxer was just wrong.  The figure comes from a 1936 study by Dr. Frederick Taussig who estimated that abortion claimed the lives of 5,000 to 10,000 women a year. It is impossible to know if his figures are accurate, given that no reliable records exist on the total number of illegal abortions that occurred, much less the number of deaths. Taussig extrapolated the data from trends in New York City and Germany.

His estimate is at least plausible. Women had few means to prevent unwanted pregnancies, and illegal abortions were often performed in less than sanitary settings. Furthermore, penicillin wasn’t in use until World War II, and not widely available to the civilian population until after the war ended in 1945. And Enovid, the first oral contraceptive, wasn’t available until 1957. But whether Taussig’s estimate was accurate or not, the conditions of the 1930’s don’t apply today.

From the 1940s through the 1960s, in fact, the best available evidence shows a dramatic decline in abortion-related deaths occurring even before the first states liberalized abortion laws in 1967. The Journal of the American Medical Association quotes official estimates from the National Center for Health Statistics showing an 89 percent decrease in abortion-related deaths by 1966. That is based on counting the number of death certificates that listed complications from abortion as the cause of death. The numbers reported for any given year are assuredly low since doctors could easily misstate the cause of death to protect the family. Still, these are the only figures that allow comparisons over time. There’s no reason to think that the rate of under-reporting would vary from one year to another, and so little reason to doubt that a steep downward trend took place long before Roe was decided.

Christopher Tietze, one of the leading experts on abortion trends, wrote in 1969 that it was plausible that 5,000 women a year died from abortion in the 1930s, but concluded that it cannot be anywhere near the true rate now.” He said that, although the 235 formally listed on death certificates in 1965 was too low, “in all likelihood it (the actual number) was under 1,000.” An abortion statistics expert at the Guttmacher Institute, Stanley Henshaw, is studying abortion rates during the first part of the century. Though his data collection is unfinished, Henshaw concurred that Tietze’s estimate of fewer than 1,000 deaths is “reasonable.”

Boxer would have been correct to say that some increase in deaths of pregnant women would result should abortions be made illegal. But the number is much lower than she claimed. In 1972, the last year before Roe v. Wade legalized abortion nationwide, CDC counted only 39 deaths from illegal abortions based on surveys of health care providers, medical examiners’ reports, state and national records, and news reports. However, Henshaw said it’s difficult to quantify the number of deaths that could result today if Roe were overturned. For one thing, it is not clear how many states would actually make abortions illegal again. And Henshaw noted it is unlikely that the numbers of deaths would be as high as they were before 1973 due to medical advances and emergency services available today. In any case, Boxer’s 5,000 figure was nearly 70 years out of date, and clearly wrong.

Santorum’s Overreaching

Republican Senator Rick Santorum of Pennsylvania claims in a new book that a number of social ills got “much worse” after Roe was decided in 1973. He’s clearly right about some, but wrong on at least one, female suicide, and possibly on another, crime.

The book is It Takes a Family: Conservatism and the Common Good, released on July 4. It devotes several chapters to abortion, and one of them includes this argument :

Santorum, p. 250: Back before 1973, there were all sorts of claims in favor of legal abortion. Legal abortion would lead to less domestic violence, since young women would not be forced into unhealthy and inappropriate marriages. Fewer desperate women would commit suicide. There would be fewer out-of-wedlock births. There would be fewer divorces. There would be fewer children in poverty, less crime, and less child abuse, since all children would be wanted and grow up in stable families. None of this happened. Not a single social ill improved as a result of legal abortion: in fact, they all got worse, much worse.

Santorum is right on some things: The percentage of children living in poverty is up, according to the Census Bureau. It was 14.4 percent in 1973 and 17.6 percent in 2003, the most recent year on record. The birth rate among unmarried women aged 15-44 has increased as well. It went from 24.5 per 1,000 unmarried women in 1973 to about 44 most recently, according to the Center for Disease Control’s National Center for Health Statistics. And Santorum has a strong case regarding divorce: the number of divorced persons has risen from 3 percent of the adult population before Roe to 10 percent most recently. The divorce rate (per 1,000 population) rose for several years after Roe and didn’t dip back below pre-Roe levels until 1999, according to Census Bureau figures.

Santorum may also be right about child abuse and domestic violence. We could find no reliable statistics on either that allow comparisons with 1973, and Santorum’s Senate staff did not respond to several requests from us to say where he is getting his information. As things stand, we consider those claims unverified.

But not all the social problems Santorum cited have gotten worse since the Supreme Court decided Roe v. Wade. And at least one has actually gotten better.

Suicide Rates

Santorum says suicides by “desperate women” got “much worse” since 1973. Actually, the suicide rate for women has dropped by one-third since Roe was decided. According to the Center for Disease Control, the rate was 6.5 per every 100,000 women in 1973, and had fallen to 4.06 by 2001, the most recent year on record.

As seen in this chart, it is true that the female suicide rate went up after Roe , but only slightly — by 0.4. It peaked in 1977 before plunging.

And those numbers refer only to the raw rate of suicides per 100,000 women. The National Center for Health Statistics also publishes an “age-adjusted” suicide rate. Statistics show that women in their 40’s are more than twice as likely to commit suicide as women in their 20’s, and the age-adjusted rates attempt to cancel out changes in the overall rate that might be due simply to a greater concentration of women in the population who have reached a suicide-prone age. The age-adjusted figure offers even less support for Santorum, however. It shows an even more dramatic decrease since Roe – a decline of 41 per cent, compared to a decline of 34 per cent for the unadjusted rate. Either way, Santorum was way off.


Santorum’s claim that crime has gotten worse since Roe  is also doubtful, though here the picture is a bit muddled.

According to the FBI’s Uniform Crime Reporting (UCR) Program the overall rate of crimes reported to police is actually 2 percent lower now, though it rose after Roe and is still higher for some categories of crime. But according to an annual survey conducted by the Department of Justice, the number of people saying they were crime victims has shown a huge and steady drop since 1973. The survey may be the more accurate, since it attempts to capture the large number of crimes that go unreported to police.

Santorum would be correct to say that the FBI’s crime rate rose in the first years after Roe, continuing a trend that had been evident for more than a decade. But it plunged starting in 1992 and was about 2 percent below the 1973 level in 2003, the most recent year on record. And it is true that violent crimes reported to police are still 14 percent higher than in 1973, although reported crimes against property are down 3 percent, according to the FBI’s statistics.

By another official measure, crime decreased dramatically since 1973 in both categories. The Bureau of Justice Statistics’ (BJS) National Crime Victimization Survey shows that crime rates in 2003 are at the lowest levels recorded since the survey’s inception, which coincidentally was the year Roe was decided.According to this annual survey, the number of people saying they were victims of property crime dropped steadily after Roe  to, most recently, about 69 percent below 1973 levels. Meanwhile, the number of people who say they were victimized by violent crime has decreased by 53 percent since 1973 . Furthermore, by this measure the rate of violent crime was actually stable between 1973 and 1977, the first several years following Roe.

How can the FBI’s statistics show an increase in violent crime while the Bureau of Justice Statistics survey shows it dropping by half? There’s reason to believe that persons are simply more likely now than in 1973 to report certain crimes to police.

The survey collects responses from a statistical sample of the population, and is not an actual count of crime reports. It attempts to measure the large number of crimes that were never reported to police, as well as those that were. Even now, half of all violent crime goes unreported. Furthermore, the increase in the FBI’s violent crime rate is due entirely to big jumps in the numbers of reported rapes and aggravated assaults, which may simply reflect that women are less likely to keep silent about such crimes than they were 30 years ago. The FBI’s murder rate, meanwhile, has dropped more than 39 percent since 1973. That’s one category of violent crime that is almost always reported, since a dead body is hard to ignore.

In any case, even using the FBI’s crime statistics, there is less crime today overall than there was in 1973, contrary to Santorum’s claim.

Santorum would have been correct to say that many of the arguments for legalizing abortion proved to be unfounded. But he doesn’t have the facts to include female suicide and crime among his examples.

These false claims by Boxer and Santorum show that whoever is named to the current Supreme Court vacancy, truth is already a casualty in the confirmation fight.

– by Jennifer L. Ernst & Matthew Barge

Correction, Aug. 22, 2005:  In our original article we said the poverty rate for children was 15.1 percent in 1973. Actually, it was 14.4 percent. So the rise in child poverty after Roe has been even more pronounced that we originally stated.

Footnote, Aug 22, 2005: Many of our visitors have written to fault us for failing to observe that changes in such things as suicide rates or child poverty rates don’t constitute evidence that Roe was or was not the cause of those changes. We agree entirely. However, Santorum didn’t claim Roe was the cause, he said the other side did. We have no quarrel with his logic, only some of his facts.



Justin M. Norton, “Boxer: Filibuster To Block Anti-abortion Supreme Court Candidate,” The Associated Press, 5 July 2005.

Christopher Tietze and Sarah Lewit, “Abortion,” Scientific American: January 1969, 220(1).

Frederick J. Taussig, M.D., F.A.C.S., Abortion: Spontaneous and Induced, (St. Louis: C.V. Mosby Co., 1936) 25-28.

“Induced Termination of Pregnancy Before and After Roe v. Wade,” Journal of the American Medical Association, 268 (Dec. 1992): 3231-3239.

Suzanne White Junod, Ph.D, “FDA’s Approval of the First Oral Contraceptive, Enovid,” Update, Food and Drug Law Institute: (July-August 1998).

Ricki Lewis, Ph.D, “The Rise of Antibiotic-Resistant Infections,” FDA Consumer Magazine, September 1995.

“Abortion Surveillance – United States, 2001,” Morbidity and Mortality Weekly Report, Center for Disease Control: 26 Dec. 2004.

Criminal Victimization, 2003,” National Crime Victimization Survey, Bureau of Justice Statistics, September 2004

Criminal Victimization, 1973-1995,” National Crime Victimization Survey, Bureau of Justice Statistics, April 1997

“National Crime Victimization Survey Violent Crime Trends, 1973-2003,” Bureau of Justice Statistics,   Web site, 12 September 2004

“National Crime Victimization Survey Property Crime Trends, 1973-2003,” Bureau of Justice Statistics,  Web site, 12 September 2004

“Crime in the United States,” Federal Bureau of Investigation, Web site, undated

Michael R. Rand and Callie M. Rennison, “True Crime Stories? Accountin for Differences in Our National Crime Indicators,” Chance Vol. 15, No.1, 2002

Reducing Suicide: A National Imperative,” Institute  of Medicine , 2002

Leading Causes of Death 1900-1998,” Center for Disease Control, Web site, undated

Age-Adjusted Rates for 69 Selected Causes by Race and Sex Using 200 Standard Populatino: United States, 1968-78,” National Vital Statistics System, Center for Disease Control/National Center for Health Statistics, undated

Age-Adjusted Death Rates for 72 Selected Causes by Race and Sex Using Year 2000 Standard Population: United States, 1979-98,” National Vital Statistics System, Center for Disease Control/National Center for Health Statistics, 2002

Center for Disease Control, Web-based Injury Statistics Query and Reporting System, Web site, accessed July 2005.