The deVeber Institute
for Bioethics and Social Research

In 2001 an article was published in the Quarterly Journal of Economics in which the authors argued that legalized abortion reduces crime rates. The authors, economists John Dubner and Steven Levitt, examined the decrease in crime in the United States since the early 1990s and attribute it in large part to the legalization of abortion in 1973. The article immediately invoked heated debate, particularly because of its moral, social and political implications. Levitt later published the findings on the abortion-crime theory to a much broader audience in a chapter of his New York Times Bestseller, Freakonomics, which he coauthored with New York journalist Stephen J. Dubner. Freakonomics made the abortion-crime theory well-known and widely accepted among non-academics because of the simple commonsense way in which Levitt and Dubner presented their arguments and evidence. Although the book, and, consequently, the theory on abortion and crime, gained much popularity, many of the criticisms of the abortion-crime theory have gone unnoticed, leaving the general public with the sense that Levitt and Donohue’s original theory is indisputable. However, serious criticisms of the theory - many from highly acclaimed economists and academics - are numerous and, at best, cast much doubt on Levitt and Donohue’s findings if not discredit them all together.

Donohue and Levitt’s Abortion-Crime Theory

In their article Donohue and Levitt acknowledge that a number of factors may have contributed to the falling rates crime rates during the 1990s. However, they hypothesize that a significant contributor - if not the most important contributor - to the decline was the legalization of abortion. Legalized abortion, they purport, may “account for as much as one-half of the overall crime reduction” witnessed in the 1990s (Donohue and Levitt 2001:34).

Legalized abortion, Donohue and Levitt argue, lowers crime for two primary reasons: Firstly, it reduces the overall number of births for each cohort born after the legalization of abortion (Donohue and Levitt 2001:8). Simply put, if there are fewer people there will be fewer criminals. Secondly, abortion “reduces the number of people born in groups that would be considered high risk for developing criminal behaviour” (Donohue and Levitt 2001:10). High risk groups can be defined using factors such as the age of the mother, marital status, education level, and poverty. Levitt and Dubner succinctly write in Freakonomics: “Legalized abortion led to less unwantedness; unwantedness leads to high crime; abortion, therefore, led to less crime” (Dubner and Levitt 2005:139).

The first body of evidence that demonstrates why abortion has contributed to lower crime rates, according to Donohue and Levitt, is that the time in which the United States saw a decrease in crime coincides precisely with the time when those conceived during or just after the 1973 legalization of abortion would have been reaching their “peak ages for violent crime”, that being approximately 18-24 years of age (Donohue and Levitt 2001:3). Legalized abortion, they argue, eliminated many of the would-be criminals and, consequently, decreased crime because there were fewer criminals to commit crimes.

Donohue and Levitt also believe that a strong indicator of the link between abortion and a decrease in crime is that the states that had high abortion rates just after legalization also witnessed lower crime rates. In other words, the higher a state’s abortion rate, the lower the state’s crime rate (Donohue and Levitt 2001:4). Donohue and Levitt report that, “states with high rates of abortion have experienced a roughly 30 percent drop in crime relative to low-abortion regions since 1985” (Donohue and Levitt 2001:4). In states where more abortions were performed there was a significant decrease in crime fifteen to seventeen years later, precisely when those would-be criminals would have been reaching their peak crime years.

Donohue and Levitt believe that possibly the most convincing piece of evidence in favour of their abortion-crime theory is that, according to their findings, crime rates fell earlier in the five states that legalized abortion in the years prior to the 1973 Supreme Court Roe v. Wade decision before crime rates fell for the rest of the country. The five states that legalized abortion (at least in part) before the 1973 decision were New York, Washington, Alaska, Hawaii and California. If crime rates fell earlier in these five states, this demonstrates a correlation between legalized abortion and crime because, if there was no connection, all fifty states should have seen a decrease at the same time. According to Levitt and Donohue’s reasoning, the fact that crime rates fell earlier in the five early legalizing states demonstrates that there is a link between crime rates and the legalization of abortion.

Crack Cocaine Epidemic

One crucial flaw in Donohue and Levitt’s analyses, argues economist Ted Joyce, is that it fails to take into account the rise and decline of the crack-cocaine epidemic that swept across the United States. According to Joyce, when this is taken into consideration, there is no evidence that supports a link between abortion and lower crime rates. As Joyce writes: “The study period coincides with the rise and decline of the crack cocaine epidemic, which many observers link to the spread of guns and the unprecedented increase in youth violence” (Joyce 2004:2). Joyce also believes the failure to properly factor in the crack-cocaine epidemic calls into question Donohue and Levitt’s evidence of the early decline and consistently higher rates of decline in crime rates in the five states that legalized abortion ahead of other states. Examining two of the largest cities and states that legalized abortion early, Joyce explains that “the data … suggest that New York City and Los Angeles were early sites of the crack markets. Not only are these the largest cities in the two largest states, but abortion became legal in both states roughly three years before Roe” (Joyce 2004:4). Thus, Joyce believes Donohue and Levitt’s findings are likely spurious because they do not account for the effects that the crack cocaine epidemic had on crime rates (Joyce 2004:4).

Aggregate vs. Disaggregate Crime Rates

Another mistake that Levitt and Donohue make in their calculations, according to economists John Lott and John Whitely, is that they use aggregate crime statistics, rather than disaggregate statistics; in order for Levitt and Donohue to verify their theory data that separates criminals into age categories needs to be used. This is essential, argue Lott and Whitely, because the crux of Donohue and Levitt’s argument rests on proving that crime rates fell within the age groups that would have been affected by the legalization of abortion. Lott and Whitely use the Supplemental Homicide Reports rather than the Uniform Crime Reports, used by Levitt and Donohue, because, as they explain, “they allow us to much more accurately disaggregate the number of murders committed by each age for each state” (Lott and Whitely 2007:308).

In using the Supplemental Homicide Report from the CDC, Lott and Whitely uncover some startling results. They find that rather than showing a decrease in murders starting with the youngest age category – those who would be the ones affected by the legalization of abortion – there is a larger decrease in murder amongst the oldest age group (Lott and Whitely 2007:310). In fact, rather than a decrease in crime among cohorts born after the legalization of abortion, there has been an increase in crime rates for this age category. Lott and Whitely explain their findings: “The murder rate changes appear to be more consistent with the theory that legalizing abortion increased (rather than reduced) murder rates. The murder rates for the two oldest age groups (26-30 and over 30 years of age) fell almost over the entire time period. The next two oldest age groups (16-20 and 21-25 years of age) both peak in 1993. Finally, the youngest age group peaks last in 1994” (Lott and Whitely 2007:310). Thus, according to Lott and Whitely, legalized abortion did not lower crime rates but rather led to an increase in violent crime, as revealed by states’ homicide records.

Increased Sexual Activity Means More Conceptions

In their book Freakonomics Dubner and Levitt admit that since the legalization of abortion there has been a dramatic increase in the number of conceptions (Dubner and Levitt 2005:139). One explanation for this increase in conceptions, according to journalist and author Ramesh Ponnuru, is that legalized abortion led to an increase in “careless conceptions” (Ponnuru 2006). With legalized abortion people became less concerned about using contraceptives or practicing abstinence. Thus, pre-marital sexual activity increased as well as the number of unwanted conceptions. Although along with more conceptions there were more abortions performed, “it stands to reason that some of those extra conceptions made it through to birth. Some kids, paradoxically, would not have been born if not for legal abortions” (Ponnuru 2006). Indeed, the number of out-of-wedlock births has risen dramatically since the legalization of abortion. Reporting on the number of children born out-of-wedlock, John Lott and John Whitely report in their article Abortion and Crime: Unwanted Children and Out-of-Wedlock Births that “between the 1960s through to the late 1980s…, there has been a tremendous increase in the rate of out-of-wedlock births. On average during 1965-69, only 4.8% of whites were born out of wedlock, rising to 16.1% 20 years later (1985-1989). For black, the number rose from 34.9% to 61.8%” (Lott and Whitely 2007:2). Far from reducing the number of births of unwanted or unplanned children, legalized abortion seems to have had the opposite effect where more children are born into unstable family situations, a high risk category Levitt and Donohue identify as a contributing factor in children developing criminal behaviour later in life.

George A. Akerlof and Janet L. Yellen put forward an interesting explanation as to why legalized abortion may have led to an increase in sexual activity, conceptions and unwanted child births. According to their theory “the legalization of abortion reduced a women’s ability to withhold premarital sexual favors from men. Women who are willing to obtain an abortion are more likely to engage in premarital sexual activity without a promise of marriage should pregnancy occur. However, other women who are unwilling to obtain an abortion face competition from women who are willing to obtain an abortion as men ‘seek satisfaction elsewhere’ (Akerlof and Yellen 1996:296-97)”. Thus, far from reducing the number of children born into single-parent families, it has led to an increase. Akerlof and Yellen also raise the salient point that legalized abortion made “shotgun” weddings unnecessary. Prior to the legalization of abortion, if a girlfriend was found to be pregnant there was far more pressure on the father to marry the woman and take responsibility for the child. With abortion easily obtainable, fathers of children are less likely to feel pressured into marrying the mother because if she was unwilling to have an abortion, there was less impetus for him to make the sacrifice of marrying her and caring for the child. Akerlof and Yellen state: “The fact that the birth of a baby is now a choice of the mother has implications for the decisions of the father. The sexual revolution, by making the birth of the child the physical choice of the mother, makes marriage and child support a social choice of the father” (Akerlof and Yellen 1996:281). Legalized abortion, therefore, had the unintended effect of fewer fathers taking responsibility for their children, leaving women who choose not to have an abortion to raise the child alone.

Different Methods Yield Different Results

Responding to Donohue and Levitt’s initial paper, economists Christopher L. Foote and Christopher F. Goetz note two problems with the calculations and overall findings. Using different methods, they point out, “give different answers” (Foote and Goetz 2008:15). First, they note, there was a coding error with the final regressions Levitt and Donohue use. They note: “The missing errors would have absorbed variation in arrests on the state-year level, ensuring that the abortion coefficient was identified using within-state comparisons only” (Foote and Goetz 2008:2). Once the error with the state-year-age regressions has been corrected, according to Foote and Goetz, there is “little evidence for a selection effect of abortion” (Foote and Goetz 2008:15). The second mistake Foote and Goetz claim Levitt and Donohue made in their calculations is the way in which they conclude that crime differed between states because of the abortion rates in each. Foote and Goetz explain, “Crime in New York will be determined by different factors than crime in Utah so it isn’t surprising that the crime rates in the two states diverge over a given period” (Foote and Goetz 2008:2). They contend that the “best way to determine whether abortion has a causal effect on crime is to compare two people who are in a similar environment today, but who had differing probabilities of being wanted at birth” (Foote and Goetz 2008:2). Taking these mistakes into account and recalculating the abortion-crime link, Foote and Goetz found that there was “no compelling evidence that abortion has a selection effect on crime” (Foote and Goetz 2008:16).

Concluding Comments

A theory such as the one presented by Donohue and Levitt has far reaching social, political and moral implications and, as such, needs to be rigorously debated and researched. The purpose of this article is to demonstrate that, although the abortion-crime theory gained much attention and popularity, Donohue and Levitt’s findings are not indisputable and should not be taken as fact. The critiques of their conclusions need to be given as much attention and consideration as the findings and arguments originally put forward. Indeed, much of the research done after Donohue and Levitt’s article was published in 2001 disproves the abortion-crime theory and casts much doubt on whether such a link exists at all.

REFERENCES

Akerlof, George A. and Janet L. Yellen. “An Analysis of Out-of-Wedlock Childbearing in the United States.” Quarterly Journal of Economics. 111: 277-317.

Donohue, John J. and Steven D. Levitt. “The Impact of Legalized Abortion on Crime.” Quarterly Journal of Economics 119: 249-275.

Foote, Christopher L. and Christopher F. Goetz. “The Impact of Legalized Abortion on

Crime: Comment.“ The Quarterly Journal of Economics. 123: 407-423

Levitt, Steven D. and Stephen J. Dubner. 2005. Freakonomics. New York: Harper Collins

Joyce, Ted. “Did Legalized Abortion Lower Crime?” The Journal of Human Resources. 39: 1-28.

Lott, John R. Jr. and John Whitley. “Abortion and Crime: Unwanted Children and Out-of-

Wedlock Births.” Economic Inquiry. 45: 304-324

Ponnuru, Ramesh. 2006. The Party of Death. Washington: Regnery Publishing, Inc.