29

In the book Freakanomics, the author speculated that due to the legalization of abortion in the US, crime rates fell because the people that would normally go out and commit these crimes, well they were never born in the first place.

Aside from the evidence and arguments used in the book itself, are there any evidence to support or repute this?

In his speculation of the possible reasons for the crime drop which was strongly correlated with legalizing abortion (with the appropriate delay) (drop also happened when the consensus among experts was that we were headed for a dramatic rise in crime), this was the rough chain of logic.

  1. Unwanted children are the most likely to grow up to become criminals
  2. Unwanted children are the most likely children to be aborted
  3. Legalizing abortion increases the aborting of unwanted children and thus reduces the number of children in the category defined in #1.
Oddthinking
  • 140,378
  • 46
  • 548
  • 638
Samuelson
  • 3,781
  • 2
  • 29
  • 44
  • 10
    That was not his actual claim. His actual claim was that the legalization of abortion had a strong correlation. He speculated on why that might be but (as I remember) did not definitively say that there was a causative affect. – Russell Steen May 12 '11 at 14:07
  • 1
    Aside from the evidence used in the book? So we're required to have read it, then, to be able to answer your question? Regardless, I think this is self evident; legalizing things is always the quickest route to lowering crime. If you were to legalize driving under the influence, your "number of persons arrested for drunk driving" statistics would drop rather dramatically as well. – David Hedlund May 12 '11 at 14:08
  • @Russell : Ok I've editted the question to say the author "speculated" – Samuelson May 12 '11 at 14:12
  • @David : Yes, I found the argument quite interesting. I'm interested in hearing if there has been any other studies into his theory. – Samuelson May 12 '11 at 14:13
  • @Stickman -- Added some details from the book in hopes that it's a more answerable question now. – Russell Steen May 12 '11 at 14:18
  • 4
    @David : But would the legalization of driving under the influence cause say "number of persons arrested for murder" to drop? He's suggesting that because people could now abort their kids instead of not raising them properly, these kids never had a chance to grow up and commit crimes. – Samuelson May 12 '11 at 14:19
  • Around the same time that we legalized abortion we also did away with leaded gas, which was literally proven to cause mild brain damage in most of the population. It's really hard to prove correlation vs causation but from what little I've seen it seems more likely the decrease in violent crime was caused by removing a pollutant known to cause people to struggle to control violent impulses. Sadly I don't have a high enough quality study comparing the two hypothesis to link to as an answer, but suffice to say there is at least one valid counter argument to the reason for decrease in violence. – dsollen Oct 14 '22 at 15:26

2 Answers2

27

According to The Impact of Legalized Abortion on Crime, in THE QUARTERLY JOURNAL OF ECONOMICS Vol. CXVI May 2001 Issue 2 by John Donohue III and Steven Levitt (one of the authors of Freakonomics):

We offer evidence that legalized abortion has contributed signicantly to recent crime reductions. Crime began to fall roughly eighteen years after abortion legalization. The ve states that allowed abortion in 1970 experienced declines earlier than the rest of the nation, which legalized in 1973 with Roe v. Wade. States with high abortion rates in the 1970s and 1980s experienced greater crime reductions in the 1990s. In high abortion states, only arrests of those born after abortion legalization fall relative to low abortion states. Legalized abortion appears to account for as much as 50 percent of the recent drop in crime.

The authors posit the basis for this, in the conclusion:

According to a recent National Academy report, there appears to be “a causal and adverse effect of early childbearing on the health and social and economic well-being of children; this effect is over and above the important effects of background disadvantages” [Institute of Medicine 1995, p. 58]. Moreover, unintended pregnancies are associated with poorer prenatal care, greater smoking and drinking during pregnancy, and lower birthweights. Consequently, the life chances of children who are born only because their mothers could not have an abortion are considerably dampened relative to babies who were wanted at the time of conception. The drop in the proportion of unwanted births during the 1970s and early 1980s appears to be the result of the increasing availability and resort to abortion.

And finally:

These estimates suggest that legalized abortion is a primary explanation for the large drops in murder, property crime, and violent crime that our nation has experienced over the last decade. Indeed, legalized abortion may account for as much as one-half of the overall crime reduction.

Obviously there is bias in the source I've given because it's the same author as the claim being analyzed. However, the paper spells out in better detail the actual claims Levitt has made, and further it provides the basis and references for those claims — The evidence put forward in the paper supports the conclusions in the book.

Clearly, analysis by individuals other than the author of the book would better illuminate how well supported the conclusions are, but I wasn't able to find any right away.

Brian M. Hunt
  • 17,999
  • 13
  • 99
  • 176
  • "*adverse effect of early childbearing*" - I guess that explains the Old Testament. I don't agree that having children later is good: I know several mothers who had children in their late 30's, all of these children are Autistic or have learning disabilities. These are in stable and well-off homes. –  Oct 24 '15 at 19:32
  • 1
    See discussion [here](https://www.econjobrumors.com/topic/was-levitt-ever-proven-wrong-on-the-link-between-abortion-and-crime) on whether the conclusions in the paper stack up - the original paper had quite a few problems – Nils Gudat Jan 05 '17 at 12:52
-4

A well-documented news article by John Lott, author of "More Guns, less crimes", can be found on the Fox News website here .

It is widely inspired by this paper : Abortion and Crime: Unwanted Children and Out-of-Wedlock Births from the same author and John E. Whitley. Its conclusion and its alternative analysis of the Donohue-Levitt results tends to show there is no significant influence of the legalization of abortion on the crime rates.

Furthermore, a study by G. Akerlof et al., An Analysis of Out-Of-Wedlock Births in the United States can imply, if we link the "investment in human capital" received by a child to his/her likeliness to commit crimes (as implies the Donohue Levitt study), the exact opposite as the D-N results. This is linkable (I am simplifying) to a correlation between the increase in the number of abortions after Roe vs Wade and the increase in out-of-wedlock birth rates.

Yves
  • 327
  • 3
  • 8
  • 1
    I don't see any correlation between abortion rates and murder rates according to that paper. The 0.5% to 7% is a huge margin, and all of the graphs point to pre-existing trends that were not altered by abortion legalization. Second, that article is pure junk from any skeptical standpoint. It makes completely unsubstantiated claims and provides no references (abortion = higher unprotected sex rates, single parents give less attention to offspring, said attention deficit affects behavior, does not delineate between "umarried" and co-habiting...). – MCM Sep 08 '12 at 14:56
  • Which paper are you talking about ? In my opinion, the article is what you can expect from the Fow News website, but it is well referenced, and the data in the paper of John Lott (not the author's conclusions) is presented in a rather clear and objective way. – Yves Sep 08 '12 at 15:44
  • I didn't see any unexplained choice or claim, even if they're not convincing. John Lott's conclusions are questionable, but the data he provides is useful to build an opinion, whether you're convinced or not of his "demonstration". I would say he worked in a scientific way. – Yves Sep 08 '12 at 16:01
  • The Whitley paper. The data is presented in a clear fashion, but I'm not convinced they drew the conclusions from the data. To me it looked like their data revealed there wasn't any correlation - which I think is supported by the correspondingly large variance in their conclusion (0.5% to 7%). – MCM Sep 08 '12 at 16:30
  • 1
    Yves: The unexplained claims: Once abortion became widely available, people engaged in much more premarital sex, and also took less care in using contraceptives...In practice, however, many women found that they couldn’t go through with an abortion, and out-of-wedlock births soared... Abortion also eliminated the social pressure on men to marry women who got pregnant...All of these outcomes — more out-of-wedlock births, fewer adoptions than expected, and less pressure on men “to do the right thing” — led to a sharp increase in single-parent families... All completely uncited. Useless. – MCM Sep 08 '12 at 16:35
  • It doesn't matter if he has an explanation if the explanation goes uncited. Science isn't built on good arguments, it's built on facts inferred from data. Heck, he even references his own paper that he conducted with Whitley and didn't acknowledge his involvement. The article is almost the opposite of science. Intentionally deceitful and utter rubbish, and knowing the author's strong bias it puts Lott's work with Whitley in a very spurious light. – MCM Sep 08 '12 at 16:39
  • My mistake on the last comment. He did acknowledge his involvement. However, looking a bit further, I found something even worse: What Lotts said about the Foote/Gooetz study: "Work by two economists at the Boston Federal Reserve, Christopher Foote and Christopher Goetz, found that, when the test was run correctly, it indicated that abortion actually increases violent crime." What the study actually said: "With these modifications, we have shown that DL’s own data **reveal no evidence for a difference in criminal propensities**..." Lotts is a worthless source. – MCM Sep 08 '12 at 16:47
  • I won't spend my time defending Lotts, but I couldn't find any unexplained bold claim or unreferenced claim in the paper about premarital sex (4 occurrences, 2 concerning his own argued point of view) and even as the world's worst scientist, it would still be his job to interprete the phenomenoms he sees, a research paper is not only a sum-up of other's views or a list of data. His results are not clear and even contradictory, but his methodology make them only easier to question. – Yves Sep 08 '12 at 17:55
  • On the article, it is useless for anyone serious, but it's easy to read, and contains lots of links to slightly more serious stuff. That's why I put it along with a reference to the paper. I don't know any serious person who would expect to find something scientifically exploitable on the Fow News website. – Yves Sep 08 '12 at 18:00
  • His explanations do not necessarily reflect reality, and that's a problem. Also, I'm pretty sure all of the evidence for Skeptics is supposed to be 'serious.' What I, as the reader, saw was a link to a clearly biased news article with an author who misrepresents the findings of others, and a link to a study (performed by the same author as the biased article) whose conclusions are questionable in context of the study alone. I think you missed the point of an "Answer" if your hope was for people to sift through all of the other links to make up their mind. – MCM Sep 08 '12 at 18:27
  • I clearly introduced the article as what it was, even if I really think it can be useful through its references, linked or not (From a Skeptics methodology point of view, should I have listed them in the answer ?) and I find your description exaggerated. I mentioned the article here so that readers can have an idea. I find useless to be given some papers (these are not studies) to read without some insight on what the author wants or who he is, especially in social science. I know no research results in social science from which one can't say that they "do not necessarily reflect reality". – Yves Sep 08 '12 at 19:27
  • And on the other side, most studies need an interpretation to build answers to a question. So of course a link to this article doesn't build an answer (Should I delete this link ?), but I think it is useful to do it. – Yves Sep 08 '12 at 19:33
  • I'm not sure how I exaggerated anything. Lotts clearly misrepresented another person's work, is biased towards a conclusion, and the study he and Whitley performed may have provided adequately respectable data - but the conclusion is questionable. We will simply have to disagree here. I do not think either link is a good source for information in regards to the question asked. – MCM Sep 08 '12 at 20:46
  • A lot of issues: 1) please include relevant citations and avoid comments like "it's got lots of tables"; 2) "out of wedlock", seriously... XIX century morality; 3) criminality is not even mentioned in the title; 4) "more guns, less crimes" is hardly relevant, and makes me seriously doubt the credentials of this person (but I may be wrong! The point here is that you are presenting irrelevant data to make this chap seem more authoritative). – Sklivvz Sep 09 '12 at 09:18
  • I noted your comments. 1) You're certainly right. I wanted to point out the fact that the raw data used in the paper are per se useful to answer the question. 2) "Out-of-wedlock" is an expression used in statistics and I didn't see any moral judgment there. I precisely included the reference to the author's book to show the possible political bias of his work. It isn't a yes/no answer to the indirect question (did abortion legalization make crime rates drop). The paper contains evidence to show there is no positive influence. It concludes with a neg. influence, but with very weak arguments. – Yves Sep 09 '12 at 10:30