Back to the drawing board for our latest critics…and also the Wall Street Journal and (Oops!) the Economist.

Thanks to articles in the Wall Street Journal and the Economist, a working paper by Chris Foote and Chris Goetz that is sharply critical of John Donohue and me has gotten an enormous amount of attention.

In that working paper Foote and Goetz criticized the analysis underlying one of the tables in our original article that suggested a link between legalized abortion and crime. (It is worth remembering that the approach they criticize was one of four distinct pieces of evidence we presented in that paper…they offer no criticisms of the other three approaches.)

Foote and Goetz made two basic changes to the original analysis we did. First, they correctly noted that the text of our article stated that we had included state-year interactions in our regression specifications, when indeed the table that got published did not include these state-year interactions. Second, they correctly argue that without controlling for changes in cohort size, the original analysis we performed provided a test of whether cohorts exposed to high rates of legalized abortion did less crime, but did not directly afford a test of whether “unwantedness” was one of the channels through which this crime reduction operated. (Note: we didn’t claim that this particular analysis was a direct test of the “unwantedness” hypothesis. This last section of the paper was the most speculative of analysis of all that we did and frankly we were surprised it worked at all given the great demands it put on the data.) They found that once you made those changes, the results in our original Table 7 essentially disappear.

There is, however, a fundamental problem with the Foote and Goetz analysis. The abortion data that are available are likely to be quite noisy. As one adds more and more control variables (e.g. nearly 1,000 individual state-year interactions), the meaningful variation in abortion rates gets eaten away. The signal-to-noise ratio in what remains of the variation in measured abortions gets worse and worse. That will lead the measured impact of abortions on crime to dwindle. Because the analysis is carried out with a unit of analysis of a state-year-single year of age, the analyses performed are highly saturated with interactions: state-age interactions, age-year interactions, and state-year interactions. Together, these interactions account for over 99% of the variance in arrest rates and over 96% of the variation in the abortion proxy. It is an exercise which is incredibly demanding of the data.

In light of this, it seems uncontroversial that one would want to do the best one could in measuring abortion when carrying out such an exercise.

The abortion measure used by Foote and Goetz is one that is produced by the Alan Gutmacher Institute. The Alan Gutmacher Institute makes estimates based on surveys of abortion providers of the number of abortions performed per live birth in each state and year.

To proxy for the abortion exposure of, say, 19 year olds arrested in California in 1993, Foote and Goetz use the abortion rate in California in 1973. This is not an unreasonable first approximation (and indeed is the one we used in most parts of our original paper because it is simple and transparent), but it is just an approximation for a number of reasons:

1) There is a great deal of cross-state mobility. So many of the 19 year olds arrested in California in 1993 were not born in California. They were born in other states, or possibly other countries. Indeeed, I believe that recent figures suggest that over 30% of those in their late teens do not reside in the state in which they were born.

2) Using a date of 20 years earlier to proxy for the abortion exposure of a 19 year old induces an enormous amount of noise. If I am a 19 year old sometime in 1993, I may have been born as early as Jan. 2, 1973 (that would make me still 19 on Jan. 1, 1993) or as late as Dec. 31, 1974 (that would have me turning 19 on Dec. 31, 1993). Abortions occur sometime in advance of birthdays, typically about 13 weeks into a pregnancy. So the relevant date (roughly) of when those who are 19 in 1993 would have been exposed to legalized abortion is about six months before they were born, or July 2, 1972-June 30, 1974. While that window overlaps with the year 1973 (which is what Foote and Goetz use as their time period of abortion exposure), note that it also includes half of 1972 and half of 1974!

3) A non-trivial fraction of abortions performed in the United States, especially in the time when legalization was taking place, involved women crossing state lines to get an abortion. As a consequence, measuring abortions in terms of the state in which the abortion is performed (which is what the data Foote and Goetz use does), rather than the state of residence of the woman getting the abortion, induces further measurement error into their abortion proxy.

4) The Alan Gutmacher abortion numbers are, even by the admission of the people who collect the data themselves, far from perfect. Indeed, the correlation between these abortion estimates and another time series collected by the CDC is well below one, suggesting that even if problems (1)-(3) did not exist, there would be substantial measurement error. The correlation between the Alan Gutmacher measure and the CDC measure, not surprisingly, gets lower and lower the more control variables that are included. This is exactly what one would expect if the controls are taking the signal out of the abortion measures and leaving behind mostly noise.

What John Donohue and I have done (with fantastic research assistance from Ethan Lieber), is attempted to address as best we can these four problems with the abortion measure that Foote and Goetz are using. In particular, we do the following:

1) As we describe in our original paper on abortion, one can deal with cross-state mobility by using the decennial censuses to determine the state of birth for the current residents of a state (the results from carrying out this correction in our crime regressions are reported in Table 5 of the original 1999 paper). This is possible to do because the census micro data reports the state of birth and current state of residence for a 5% sample of the U.S. population. Note that the correction we are able to make is unlikely to be perfect, so it may not fully solve the problem, but it clearly moves us in the right direction.

2) Given that the window of abortion exposure that 19 year olds in 1993 spans from 1972 to 1974, the obvious solution to this problem is to allow abortions performed in 1972, 1973, and 1974 to influence arrests of 19 year olds in 1993. It is straightforward to work out roughly the weights that one wants to put on the different years’ abortion rates (or one can do it non-parametrically and let the data decide — the answers are virtually identical).

3) In order to deal with the fact that many women were crossing state lines to get abortions in the 1970s, we use the Alan Gutmacher Institute’s estimates of abortions performed on women residing in a state relative to live births in that state. (We were unaware of the existence of these better data when we wrote the initial paper, otherwise we would have used them at that time.) There is little question that measuring abortions by state of residence is superior to measuring them by where the procedure is performed.

4) The standard solution to measurement error is to perform instrumental variables in which one uses one noisy proxy of the phenomenon that is poorly measured as an instrument for another noisy proxy. (I recognize that most readers of this blog will not understand what I mean by this.) In this setting, the CDC’s independently generated measure of legalized abortions is likely to be an excellent instrument. Because there is so much noise in each of the measures, the standard errors increase when doing this IV procedure, but under a standard set of assumptions, the estimates obtained will be purged of the attentuation bias that will be present due to measurement error.

I think that just about any empirical economist would tend to believe that each of these four corrections we make to the abortion measure will lead us closer to capturing a true impact of legalized abortion on crime. So, the question becomes, what happens when we replicate the specifications reported in Foote and Goetz, but with this improved abortion proxy?

The results are summarized in this table, which has two panels. The top panel are the results for violent crime. The bottom panel corresponds to property crime.

Starting with the first panel, the top row reports the same specifications as Foote and Goetz (I don’t bother showing their estimates excluding state-age interactions because it makes no sense to exclude these and they themselves say that their preferred specifications include state-age interactions). We are able to replicate their results. As can be seen, the coefficients shrink as one adds state-year interactions and population controls.

The second row of the table presents the coefficients one obtains with our more thoughtfully constructed abortion measure (changes 1-3 above having been made to their abortion measure). With a better measure of abortion, as expected, all the estimated abortion impacts increase across the board. The results are now statistically significant in all of the Foote and Goetz specifications. Even in the final, most demanding specification, the magnitude of the coefficient is about the same as in the original results we published that didn’t control for state-year interactions or population. The only difference between
what Foote and Goetz did and what we report in row 2 is that we have done a better job of really measuring abortion. Everything else is identical.

The third row of the table reports the results of instrumental variables estimates using the CDC abortion measure as an instrument for our (more thoughtfully constructed) Alan Gutmacher proxy of abortions. The results all get a little bigger, but are more imprecisely estimated.

The bottom panel of the table shows results for property crime. Moving from Foote and Goetz’s abortion measure in the top row to our more careful one in the second row (leaving everything else the same), the coefficients become more negative in 3 of the 4 specifications. Doing the instrumental variables estimation has a bigger impact on property crime than on violent crime. All four of the instrumental variables estimates of legalized abortion on property crime are negative (although again less precisely estimated).

The simple fact is that when you do a better job of measuring abortion, the results get much stronger. This is exactly what you expect of a theory that is true: doing empirical work closer to the theory should yield better results than empirical work much more loosely reflecting the theory. The estimates without population controls, but including state-year interactions, are as big or bigger than what is in our original paper. As would be expected (since the unwantedness channel is not the only channel through which abortion is acting to reduce crime), the coefficients we obtain shrink when we include population controls. But, especially for violent crime, a large impact of abortion persists even when one measures arrests per capita.

The results we show in this new table are consistent with the impact of abortion on crime that we find in our three other types of analyses we presented in the original paper using different sources of variation. These results are consistent with the unwantedness hypothesis.

No doubt there will be future research that attempts to overturn our evidence on legalized abortion. Perhaps they will even succeed. But this one does not.


Steven D. Levitt has responded to the Foote and Goetz paper pointing out two errors he made in his abortion-cut-crime theory in a particularly striking way -- by introducing, at this late date, a whole new data set!

Foote and Goetz showed that Levitt arrived at his conclusion due to technical incompetence, and that objective analysis of Levitt's own data shows no impact of legalizing abortion on crime. So, Levitt now introduces a new data set, which he claims provides less noisy data on abortion rates by states, than the one he and Foote-Goetz worked with. Unsurprisingly, Levitt claims this new data set proves he was right all along, even though his original data said, when analyzed correctly, that he was wrong.

Is his new data set really better? Is it worse? Did Levitt botch up his analysis again? Who knows? I'm sure it will take months, at least, for objective analysts to look it over. The last dataset, now junked by Levitt, took 6 years to be debunked.

And is that the last word in data sets? I strongly doubt it. For example, America's most dangerous criminals were performing "selective post-natal abortions" on each other at an unprecedented clip in the gang wars of the early 1990s. AIDS was also taking a toll on criminals then. Levitt hasn't adjusted for how many criminals died during this period. Nor has he adjusted for "object lesson" impact of the sorry end of so many criminals in the early 1990s had on their younger brothers, who grew up to be better behaved.

But, from a marketing standpoint, in terms of preserving the value of the Freakonomics brand name, Levitt has put a marker down that his true believers can refer to to ward off their Doubts.

We're now way, way out in how many angels can dance on the head of a pin territory. If Levitt really is explaining close to half of the huge decline in crime that occurred in the 1990s, as he has claimed, the evidence shouldn't be so fragile that it collapses when somebody else stares at it hard and Levitt has to throw away his old data and replace it with a new set of data that nobody has seen before. It should pass a few reality checks.

When I studied marketing models in MBA school a few eons ago, the professor constantly pointed out that the true test of the statistical analyst is creating robust models. You can always fiddle around with historical data and variables long enough until you obtain a high correlation coefficient and a high degree of statistical significance and declare victory. But that's not a robust model and it's not going to be much use in making business or policy decisions.

Economist Roehlano Briones writes on his Go Figure blog:

"Extraordinary claims require extraordinary evidence" is a phrase popularized by Carl Sagan - in turn derived from Hume's examination of miracles-claims. Now the original abortion-crime hypothesis is far from alleging a miracle. It is however extraordinary as it implies that causal mechanisms of crime originate from circumstances prevailing at the time of birth. Moreover, the claim that the behavior of eliminating live births is skewed against this causal mechanism (that is, abortion does not neutrally eliminate future crooks and law-abiders on a 50:50 ratio).

The issue remains, as it has since Levitt and I debated in 1999, who should have the Burden of Proof on his shoulders.



Mr. Sailer's interpretation seems tenuous to me. He claims that Levitt's "original data said, when analyzed correctly, that he was wrong." At most, though, Levitt's original data weren't strong enough to prove his case. They certainly didn't prove it wrong (i.e., demonstrate that abortion had no effect on crime). For someone sensitive to the issue of burden of proof, Sailer is careless in his rhetoric.

Sailer also doesn't seem to appreciate the way that science asymptotically approaches the truth. As primitive data are replaced by better information, theories should be able to withstand more exacting scrutiny. Demanding that new data never be used to bolster a position means cutting off this process. It would be different if Levitt were using private data that couldn't be analyzed by third parties, or data that mysteriously disappeared in a computer crash or something.

Remember that this particular critique of Levitt's work doesn't claim that he was wrong, merely that his data weren't good enough to do the job. Introducing better data is a fair response, if that data is in fact useful.

Sailer thus fails to make a convincing procedural claim about how Levitt responded to the critique, and he doesn't make any substantive claim about the new data at all. Sailer should remember that science isn't a war in which one side is right and the other side is populated by scoundrels. It is instead a process in which honest criticism is met with a thoughtful response. No one calls foul when someone else introduces better data into this process.



Mr. Sailer:

Extraordinary evidence? How would one describe that? Would four approaches supporting a hypothesis count?

It seems to me that Dr. Levitt was introducing a new data set to illustrate a point: as you remove the "noise" (to use his terminology) even that data set supports his claim. Besides, ask yourself: why reintroduce what has already been written when you can use even more data to support your claim? Your lack of understanding on that level and your criticism of it is disturbing. In fact, if your reading comprehension is a representation of all MBA graduates, I would consider the MBA to be a worthless degree. Fortunately, I know the truth to be otherwise.

Perhaps, since you are so knowledgable in this field, having taken a graduate marketing class aeons ago, you could analyze the robustness of Dr. Levitt's models? Since he is a professor at the University of Chicago--which is considered the very best school for Economics--the burden of proof moves to you and other critics to prove him wrong, not to Dr. Levitt. Dr. Levitt has already been published on a peer-reviewed journal, which means that other Economists--I assume that they are among the best--concur with his conclusions. Until you do the same, or until you show more than a few rhetorical points that appear hostile at best (though they are interesting points) your case remains weak.

This is not to say that Levitt is right or wrong. I'm just explaining the perspective of an economist.



Dr. Levitt's other avenues of evidence have already been debunked. For example, Dr. Levitt's crucial "wantedness" theory -- that aborted children would have had lousier upbringings if they'd been forced to be born -- relies on European studies because American studies came to the opposite conclusion. For a theory of American crime, Dr. Levitt should have used American studies. CCNY Economist Ted Joyce wrote:

“Second, analysts, I being one, have tended to overestimate the selection effects associated with abortion. A careful examination of studies of pregnancy resolution reveals that women who abort are at lower risk of having children with criminal propensities than women of similar age, race and marital status who instead carried to term. For instance, in an early study of teens in Ventura County, California between 1972 and 1974, researchers demonstrated that pregnant teens with better grades, more completed schooling, and not on public assistance were much more likely to abort than their poorer, less academically oriented counterparts (Leibowitz, Eisen, and Chow 1986).

“Studies based on data from the National Health and Social Life Survey (NHSLS) and the National Longitudinal Survey of Youth (NLSY) make the same point (Michael 2000; Hotz, McElroy, and Sanders 1999). Indeed, Hotz, McElroy, and Sanders (1999) found that teens who abort are similar along observed characteristics to teens that were never pregnant, both of whom differ significantly from pregnant teens that spontaneously abort or carry to term.

“Nor is favorable selection limited to teens. Unmarried women that abort have more completed schooling and higher AFQT [the military's IQ test for applicants for enlistment] scores than their counterparts that carry the pregnancy to term (Powell-Griner and Trent 1987; Currie, Nixon, and Cole 1995).

“In sum, legalized abortion has improved the lives of many women by allowing them to avoid an unwanted birth. I found little evidence to suggest, however, that the legalization of abortion had an appreciable effect on the criminality of subsequent cohorts.”

No, the black box of the state level analyses has always been Dr. Levitt's ace in the hole, simply because it's so much more work for outsiders to dispute.

But, you can see the state level analyses are shakey just thinking about it in simple human behavior terms.

Levitt notes that something like 30% of older teens aren't living in the states where they were born. That should raise some questions about state-level analyses, especially when they don't agree with the national level analysis, as these don't. If movement between states was random, that shouldn't cause too much of a problem for his methodology.

But what if crime itself is a driving force in causing people to move? Indeed, local crime rates have a major impact on who moves, when they move, and where they move.

What if, say, people in socially liberal states like New York (that had lots of abortion earlier in the 1970s and tended to have the crack wars hit first in the late 1980s) who worried that their children might be crime-prone tended to move to socially conservative states in the hopes that a God-fearing environment would keep them on the straight and narrow. Maybe it worked some time, but an influx of crime prone young people would likely drive up the local crime rate.

And what if people who were less worried about the crime proneness of their children due to the family's affluence or educational level or whatever were more likely to move to socially liberal states like NY?

That would foul up Levitt's state-level analyses severely by creating a spurious correlattion. It would look to him like socially liberal states with high abortion rates were benefiting from abortion, but instead they were just shedding crime-prone youths and acquiring crime-resistant ones. I'm not saying it happened, but it might have.

Once again, let's keep our eye on the burden of proof.


General Disarray » Blog Archive » The Levitt/Donahue-Foote/Goetz abortion and crime debate

[...] A reply by Steven Levitt, arguing that the proper specification supports his argument if a better dataset is used [...]


I would like to see the citation for your assertion below.
*******************************************For example, Dr. Levitt's crucial “wantedness” theory—that aborted children would have had lousier upbringings if they'd been forced to be born—relies on European studies because American studies came to the opposite conclusion
I was also interested in the study you cited on who had more abortions, "good" vs. "bad" teenage mothers
A careful examination of studies of pregnancy resolution reveals that women who abort are at lower risk of having children with criminal propensities than women of similar age, race and marital status who instead carried to term. For instance, in an early study of teens in Ventura County, California between 1972 and 1974, researchers demonstrated that pregnant teens with better grades, more completed schooling, and not on public assistance were much more likely to abort than their poorer, less academically oriented counterparts (Leibowitz, Eisen, and Chow 1986).
You did provide a list of authors for that study and I was quickly able to find the citation, (Demography 23 (1):67-77). This means that I can find out why and how this effect occurred just as soon as I get to a decent library. Thanks.



Why does Sailer obsess so much over such a benign theory. Some people find the evidence compelling, some people don't. It's not all that interesting a conclusion in the first place.


I am not qualified to discuss in any detail either the data sets or their analyses, but Levitt and Donohue's result with respect to abortion alone is logical. I won't insult it by saying that it is so obvious that it is trivial, because it is an important result. I'm sorry it is so controversial, because that interferes with addressing the crime problem. Repeated studies by sociologists, psychologists, social workers and criminologists over the years have supported -- dare I say proven? -- the abortion/crime reduction link beyond a doubt. It makes sense that "unwanted" children would suffer more deprivation and would lack the support and love given to wanted children, thereby leading to more crime committed by those unwanted children relative to the rest of the population. Other than the obvious (ie, a view that abortion is bad, or that he is racked with jealousy) I don't understand Mr. Sailer's point in pursing this. Sometimes the obvious answer is the right one, even when using complex, noisy datasets.



Proof only has value when people can see it. With certain papers and journals covering only the rebuttals, aren't you worried you're losing "popular opinion"?

For most scientists, peer opinion matters more. But you seem to have focused yourself on popular opinion, which begs the question:

How are you getting out the message that you are still right, even though others loudly said you were wrong?


For all the hemming and hawing about how science "asymptotically approaches the truth" - our Freakonomist didn't go on to the Daily Show and qualify his abortion cuts crime theory by saying - "I'm only taking an asymptotical towards the truth, here - let us wait for better data sets before we get excited" - Nope - he rode the wave of high praise generated by his theory - which in some minds provides some moral purpose for abortion and therefore sticks it to those awful Red-staters. Crimefighters unite.


The Daily Show is not the venue for academic discussion; Levitt had maybe 2 minutes to talk about it and he wasn't going to bore the audience with econometrics.


This entire episode reminds me again of why data is so hard to come by in the social sciences. When intelligent people have to decide between studying things people care about (like abortion), and stuff people don't (stars and starfish) many choose the easier path. If your wrong in physics, you join Newton and Einstien. If your wrong in social science, you join a political party.


Foote and Goetz used your data, fixed errors and omissions in your modeling (all of which I agree with) and found that your hypothesis was no longer supported. I suggest that you accept that criticism without attacking them for using the same data that you used!

Your overall abortion/crime hypothesis may be valid, but either way I consider it absurd for the rebuttal to repeatedly imply that the data problems were somehow the fault of Foote and Goetz. Foote and Goetz used YOUR DATA. If you now question that data, then you should take the blame and not act as though the data problems are due to your critics. I am inclined to question your work even more having read the way the rebuttal seems to shift blame.

You should also post the Stata code that you used to create the "corrected" abortion data. I'd like to try it out and I don't think your post provides anywhere near sufficient description for replication.



A group of students presented an economics project about factors that might influence violent crime in LA and NY. I didn't expect them to try to re-test your hypothesis but I was shocked they didn't mention it. I made a comment about it, keeping in mind that I attend a Christian university. The group told me their reason for not mentioning abortion was exactly that: fear of coercive action by the school. No further comments.
One thing they did right though was to conclude using regressions that adding police force does not necessary translate into lower crime.
My professor also mentioned that you Mr. Levitt is bound to win a Nobel prize someday. I agree with him. I love your work.


Why is the abortion cut crime controversy important?

Obviously, both Dr. Levitt and I have a lot of ego bound up in it. It's his signature theory and I was the first critic who pointed out that, as persuasive as it sounds in the abstract, readily available national crime figures showed that exactly the opposite of what he predicted had actually happened. (Here's our 1999 debate in

But, it should be important to disinterested observers as well for what it says about

- Two major issues: abortion and crime

- And about the shortcomings of the academic process, intellectual life and the "serious" media in America. How did a theory that badly failed to predict the past become Conventional Wisdom so rapidly?

- But I'm perhaps most interested in this: Why didn't this sensible sounding theory work in reality? Why did the serious violent crime rate and homicide rate go through the roof precisely among those groups most severely culled by legal abortion during the 1970s?

(I'm not saying that social scientists won't ever find that legalizing abortion had some sort of effect, either positive or negative, on serious crime rates. I'm just saying that the effect, whatever it was, was dwarfed by the cumulative impact of other effects, and couldn't have been close to Dr. Levitt's grandiose claims that legalizing abortion accounts for close to half of the crime decline in the 1990s.)

Dr. Levitt's theory is an example of a "selectionist" theory, which I typically find more appealing than the alternative "constructionist" or "environmentist" or "educationist" theories, even though the latter are vastly more socially respectable and politically correct.

An example of selectionism vs. educationism might be: Why do Harvard graduates make more money on average than U. of Houston graduates? The educationist perspective is that Harvard gives you a much better education and you then use all the additional knowledge you learned from your famous professors at Harvard to make tons of money, and if only the Houston students were allowed into Harvard, they'd get rich too. The selectionist perspective is that the big difference between Harvard and Houston is that Harvard's admissions committee selects smarter and more promising high school students than the U. of Houston's.

Dr. Levitt's abortion-cut-crime theory is a classic selectionist theory -- if we abort more of the poor and black and unwanted than we abort of the affluent and white and wanted, then we ought to have less of what the poor and black and unwanted tend to do more of on average, such as commit crimes.

This is straightforward logic. Dr. Levitt didn't invent it. Back in the 1980s, my inlaws used to tell it to me after they had a few drinks. Howard Stern used to make an anti-black joke about the effects of abortion on the crime rate.

I always found this selectionist logic fairly plausible. I tend to lean toward selectionist explanations in general. I didn't have an opinion on whether this particular selectionist theory would prove true or not, but I certainly wouldn't have ruled it out.

But then this abortion-should-cut-crime theory seemed to fade out of public consciousness in the 1990s, probably because the violent crime rate among black youths born after legalization shot up so high in 1990-1994. During the enormous 1992 LA riot, for example, it would have seemed pretty silly to say, "Thank God Ronald Reagan signed the California legalized abortion law in December of 1969, or things would really be getting out of hand in South Central!"

But, then the violent crime rate started to drop, especially among black teens born in the early 1980s (long after abortion, by the way), and by 1999 the academic and media elites had largely forgotten the demographic details of who was doing all the shooting and dying in 1990-1994 (especially because the press tries to play down the large differences in crime rates among ethnic groups: e.g., African-Americans are imprisoned at a rate over 30 times that of Asian-Americans, but when was the last time you read this interesting fact in the press?)

So, when Dr. Levitt and Dr. Donohue's draft paper was leaked to the Chicago Tribune in 1999, America's intellectual defense were weakened. Abortion-cut-crime was a dormant theory in many minds, and by 1999, many had influential people had forgotten why they had stopped believing it. And because you are not supposed to think in selectionist terms, the punditarian was unarmed intellectual combat with this selectionist theory.

As someone who leans toward selectionism in general, I've personally found it fascinating and educational to find out why selectionism didn't work in this case. As I said in our debate in 1999: "It's ironic, but because I've been arguing for years that genetic diversity affects society, I was one of the few to notice in this particular case that crime has risen and fallen not because we are aborting the poor and black and unwanted, but because of that staple of genteel liberal commentary, Cultural Forces (e.g., crack)."

- The most surprising thing I've learned from this about legalized abortion is, despite the enormous political tumult over it, just how pointless it mostly proved. Legalized abortion turned out to be reminiscent of Homer Simpson's toast: "To alcohol! The cause of, and solution for, all of life's problems." Legal abortion is a major cause of what it was supposed to solve -- unwanted pregnancies. Levitt himself notes that following Roe, "Conceptions rose by nearly 30 percent, but births actually fell by 6 percent …" So, roughly speaking, for every six fetuses aborted in the 1970s, five would never have been conceived except for Roe!

- I also found out that legal abortion didn't work the way I (and Dr. Levitt) assumed it didn't. Within each demographic group, all else being equal, it tended to be more popular among the more ambitious women, the ones who would have gotten their babies into a higher class and a better peer group if they had been born. Legal abortion's impact on the black middle class was particularly strong. It seems to have hollowed out the black middle class, and made it proportionally smaller compared to the black underclass.

Dr. Levitt and Dr. Fryer are working on why black progress stopped in the later 1980s. If they are frank with themselves, they should definitely investigate the impact of legal abortion on reducing the birthrate of the black middle class relative to the black underclass. This may be a major reason why underclass social norms (such as gangsta rap) came to dominate black popular culture.



Let me make a few points about the meta-discussion.

In 1999 Sailer did make interesting counterarguments to Levitt. I also see the empirical studies he now cites as intresting contributions. But except his in 6 years he has not added anything to the actual abortion controversy.

When Sailer writes about his core issues, such as immigration, IQ and race relations, he is a well reasoned and logical person, whether or not you agree with him. He is even original, perhaps having some economic “rent”, as very few other smart people dare openly write (and think) about such controversial issues.

But sometimes when he moves outside this area the quality decreases. He stops applying the same empiricism and demand for cold logical arguments that he (rightly) complains are lacking in for example the IQ question. He instead mostly just uses bits and pieces of evidence that suits his argument, ignores the rest and than spend the rest of the time psychologising about his opponents.

An example: He is ardently anti-Iraq war, so he has writes that the US invaded Iraq because the neocons watched war movies and got excited, or that “The idea of beating up on a sure loser like Saddam may have especially appealed to GWB because of the President's personal qualities.”. Or maybe it is Bushes cultural roots as a Scotch-Irish. Or maybe.. You get the picture.

With the current discussion Sailer is putting most of his ammunition making the case that the abortion thing having “made” Levitt, asking him to let it go now that he is successful.

“So, you BSed your way through. And it paid off. You are now the most glamorous economist in the world this side of Alan Greenspan. You're now a brand name: the Freakonomist!”

Levitt had 9 articles published in the top five journals when the abortion one was published. I don't know if you realize how the hierarchy in economics works (white males). Many economist never publish *any* article in the top 5 journals in a lifetime. He had 9 in his 30s, none of them about abortion. He has since published 9 more (10 inkl. abortion paper). Currently he lists 54 articles published in total, most in top 20 ranked journals(!).

In the hirrarchical system Levitt operates the Freakonomics stuff is worth very little, more probably than not frowned upon or seen as a distraction. Not admitting wrong data would be a whole other issue. Do you understand his incentive strukture better now Mr. Sailer?

I don't know Levitt, but I know his reputation and I know how economics works.

Drop this mental image you have some guy manipulating data to sell a popular story. The way these people work often a paper starts of in one direction, and after getting ripped appears by the rooms with sphingolipid storage mutations. Levitt (and others like him) have published papers with one result

and than when new data arrives published papers with the opposite results.

Levitt may well be wrong, but not for the reasons you seem to think.

Relying heavily on prescribing motive to your opponents is not impressive or convinscing, as someone dealing with race and IQ probably should know.

Now in order to return the favor let me offer further psychologising about Steve Sailer:

Sailer is used to debate IQ with liberals and immigration politics with neocons. In both cases the opponents are heavy on emotionional and moral arguments and light on facts and logic (note that I said politics, not policy). So it's enough for him to find a nice graph, quickly win the debate and have the opponents walking away muttering, beaten but often not convinced. Still it works.

It's not as easy here. As he points out Levitt uses devilish tools of econometrics, which you need months to understand and years to master. So Sailer was stock, having to end the debate because he could go no futher. But with a suspicion something was wrong. Until now that is, when others Wizards with econometrics criticize Levitt.

Aha! The suspicious is confirmed! He knew something was fishy, and now the debate is over. Just give it up. And look, instead of giving up he introduces a NEW dataset, as if we are so east to fool! Just what someone would do who has something to hide.

Come on. It just doesn't work this way.

Since you consider yourself an empiricists I hope Sailer notes a few facts:

1. “if Levitt really is explaining close to half of the huge decline in crime that occurred in the 1990s”

Levitt explicitly writes: “legalized abortion … account for 25–30 percent of the observed crime decline in the 1990s.”

2. “Dr. Levitt didn't invent it.”

Nor does he claim he did, writing “a hypothesis that to my knowledge was first articulated in Bouza (1990)”

3. Just because YOU think the abortion story is all what his book is about doesn't mean it so. Sure it may be the chapter that gets most attention from left of center media. But clearly the story and the buzz of the book was never one particular theory, it was Levitt himself and the “odd” way he used economics. This was evident from the NYT story that started the whole thing. That's why the book is called “Freaconomics”, not “The case for free choice”.

To be honest most people who read it take the abortion story as they take all of the rest, fascinating but with a grain of salt. The plebs are not trained to think deterministically about social theory anyway, many of them still think the case between prisons and crime is also a theory.

The only way this story will clear up is through discussions among trained economist. To Sailer this I a Black Box, which he neither understands or trusts. But to me it is almost unthinkable that Levitt would be proven wrong academically but maintain his position publicly. Sure this is how it works when it comes to crime, IQ and many similar issues, but not University of Chicago econometrics.

For what it matters Levitt may be proven wrong. But assuming the question is already settled and questioning his motives is not serious.

PS. Having said all this I advice anyone who isn't overly politically correct to visit

And read probably the most insightful film reviews in America.



1. I gave the rest of "Freakonomics" -- especially the parts where it gently pushed the envelope about what can be said in respectable society about IQ, heredity, and race (although not in combination) -- a generally positive review at:

2. As I've said before, the reason I am likely to spend more time pointing out flaws in the thinking of really smart folks like Dr. Levitt, Malcolm Gladwell, Jared Diamond, Michael Barone, Karl Rove, Gary Becker, Richard Dawkins, Claude Steele, the Thernstroms, Charles Krauthammer, Stephen Jay Gould, Francis Fukuyama, Tooby & Cosmides, L.L. Cavalli-Sforza, etc. instead of the usual cast of idiots most pundits denounce is because ... they are smart enough to know better.

Over the years, I've had a gratifying impact on some important thinkers, including one very big name who has, under my influence, largely reversed his views on the questions that most concern me. He shall remain nameless, of course, because guilt by association has returned to being a favorite mode of argumentation in America.

3. Dr. Levitt plays two different roles -- the economist in scholarly journals and the Freakonomist in the media spotlight. In the scholarly role, he is exposed to review by his peers. When he makes statements that turn out not to be empirically supported, eventually he gets called on them.

The problem comes when he switches to his Freakonomist role and writes in his bestseller a description of the abortion-cut-crime debate that leaves out all the facts that emerged over the previous half decade that would induce skepticism in a disinterested observer.

Now, everybody who enters the public lists tries to put the best polish on his case. Now, one concern is that Dr. Levitt is bringing his credibility as an economist to his very different role as the Freakonomist, which gives his more tendentious statements believability that nobody in their right mind would attribute to, say, my obiter dicta about the President's personality that the gentleman above complains about. But, that's minor.

The real problem is the pathetic state of public policy discourse in this country, as seen by the wildly credulous reviews of "Freakonomics" in the prestige press, which promoted the abortion-cut-crime theory into Instant Conventional Wisdom. How many reviewers and feature writers bothered to spend 15 minutes on Google to find out that there was indeed a definite empirical controversy over the validity of Dr. Levitt's most celebrated theory? How often was criticism of his theory smeared as purely moralistic? Thomas Roesser in Dr. Levitt's hometown paper, the Chicago Sun-Times, was an honorable exception to the near universal gullibility in the more prominent media outlets.

Here's a comparison of Roesser's analytical review to Jim Holt's New York Times review, which simply consisted of a lengthy summarization of Dr. Levitt's argument, with no doubts creeping in whatsoever:

Of course, Dr. Levitt also enjoyed in the NYT a credulous column by an op-ed columnist, and a second rave book review, this one after he had been hired to write a column for the NYT!

4. Public policy discourse in America is evolving away from debate and toward multiple, uncommunicative consensuses. Every shade of opinion now has their own websites where they can mutually admire and reinforce each other's pre-existing beliefs, and where dissenters are denounced as "trolls."

Personally, I believe debate is good for America. (I arranged for Dr. Levitt to debate me in Slate in 1999. When his publicist recommended he debate me again this spring to promote his book, he said he didn't have time.) But most of all, I find the contemporary rage to banish "trolls" to be extremely boring.



I agree with pretty much all of what Teller said. Steve is an academic economist at a top department with a lot of papers published in top journals. While he is getting a lot of publicity with his book, and I think having fun with it, my impression is that he is the sort of guy who will tire of it pretty quickly and stick with the paper writing, rather than someone like Paul Krugman who to some extent gave up his research career to write columns about how all republicans are evil and nasty and all democrats are pure and good (I'd say he's half right - but that's me). Assigning motives to Steve that center on publicity or popular acceptance for his views is not likely to be right. I suspect he'd trade a lot of book sales for a "well done" in regard to a paper from any one of several of his Chicago colleagues.

Two other remarks: First, there is an academic review of Freakonimics on its way to the Journal of Economic Literature, the journal of the American Economic Association that publishes such things. It will be interesting to see how that turns out.

Second, bringing in new data is good, not bad. Singer's remarks that imply that this is a bad thing or somehow underhanded are just silly. There are a lot of data out there, and sometimes one finds better data on something after having first used another data set to write a paper. Also, sometimes new data arrive over time. Both are good things, not bad things.

Jeff Smith


Brad DeLong's Website

More Family Planning, Fewer Unwanted Children, Less Crime

Steve Levitt and John Donahue are unconvinced by the critics of their "more wanted and fewer unwanted children means happier families and less crime" hypothesis. In large part what is going on is "normal science"--Foote and Goetz are inducing Levit...


What do Chris Foote and Chris Goetz think of this reply?