We Had Better Get Our Next Book Out: John DiNardo Is Getting Bored

As mentioned on MarginalRevolution, the economist John DiNardo has been quite busy over the last few years criticizing Freakonomics. He has written no fewer than three papers on the subject.

It’s too bad that he didn’t offer the standard academic courtesy of sending his criticisms directly to me before writing them up; if he had, I could have helped clarify a few things.

For instance, his recent American Law and Economics Review article, “Freakonomics: Scholarship in the Service of Storytelling,” devotes three or four journal pages to an alleged set of sins committed in a paper I wrote with Dan Kessler back in 1999.

These criticisms are based on an embarrassingly bad piece of research by Cheryl Marie Webster, Anthony Doob, and Franklin Zimring. (I apologize: the only version I could find of the paper is gated.) If DiNardo had bothered to read my reply to that piece, I suspect he would not have written what he wrote.

And if he ever does bother to read it, perhaps we will consider publishing an erratum. Although I guess that would get in the way of his storytelling.

DiNardo also spends two pages asserting that we misrepresented a paper by Christian Pop-Eleches on the impact of banning abortion in Romania. DiNardo writes that Pop-Eleches finds “virtually the opposite” of what we report in Freakonomics. A quick comparison of the two will make clear just how shamelessly we distorted Pop-Eleches’ arguments. In Freakonomics, we describe Pop-Eleches’ work as follows:

The cohort born after the abortion ban would do worse in every measurable way: they would test lower in school, they would have less success in the labor market, and they would prove much more likely to become criminals.

Here is the abstract of the version of the Pop-Eleches paper that we cited:

…controlling for composition using observable background variables, children born after the ban on abortions had worse educational and labor market achievements as adults. Additionally, I provide evidence of crowding in the school system and some suggestive evidence that cohorts born after the introduction of the abortion ban had higher infant mortality and increased criminal behavior later in life.

The introduction of the Pop-Eleches paper says:

This finding is consistent with the view that children who were unwanted during pregnancy had worse socio-economic outcomes once they became adults.

“Virtually the opposite?” Sounds to me like Freakonomics and Pop-Eleches are saying the same thing. While it’s true that we cited a yet-to-be-published version of Pop-Eleches’s paper (which was the most current version available at the time we were writing Freakonomics), the conclusions of the published version (which appeared in the Journal of Political Economy, and which I edited) were unchanged.

C’mon, John, you’re a top economist, and our book is 300 pages long. You must have better criticisms than that! When we publish our next book, give me a call beforehand, and I’ll tell you what’s really wrong with it.

lemmy caution

Here is the full quote from freakonomics:

Ceausescu's incentives produced the desired effect. Within one year of the abortion
ban, the Romanian birth rate had doubled. These babies were born into a country
where, unless you belonged to the Ceausescu clan or the Communist elite, life was
miserable. But these children would turn out to have particularly miserable lives.
Compared to Romanian children born just a year earlier, the cohort of children born
after the abortion ban would do worse in every measurable way: they would test lower
in school, they would have less success in the labor market, and they would also prove
much more likely to become criminals.

Rishi Garg

I agree with most of my friends here... its a person's work that speaks for itself and one need not fall to the same level as one's critics..


Two facts:

(1) In the working paper version at IDEAS, and in the published version at the JPE, Pop-Eleches says in the abstract: "Children born after the abortion ban attained *more* years of schooling and *greater* labor market success." This is what DiNardo cites in his paper.

(2) Levitt in the book says: "The cohort born after the abortion ban would do *worse* in every measurable way: they would test *lower* in school, they would have *less* success in the labor market,..."

Possible explanations. Either (1) Pop-Eleches wrote and published a paper with an abstract that contradicts the findings of his research or (2) Levitt chose apropriately a paragraph that suited his story-telling (and thus, DiNardo has a point) or (3) I am missing something important. What am I missing?


Levitt deserves to respond as he considers appropriate and the assumption that there are posh academics above the fray is responsible for the misrepresentations of the work of professionals. If a critic tries to use a professional's name to get himself noticed and published, then he ought to do it right or expect a response. Why shoot the messenger?

Jirka Lahvicka

Based on the abstract of Pop-Eleches's work (I do not have access to the full version), I think that it really seems misrepresented in Freakonomics.

After the ban of abortion in Romania, educated women (who had tended to have abortions before) had more children with understandably better outcomes. It does not really matter that that the outcomes were worse after controlling for background variables - the overall impact of the abortion ban on the Romanian society was positive (at least according to this study).

However, based on the study, you can also conclude that the impact of abortion ban would be negative in any society where abortion is more widespread among lower-class women (that may be the case in the US, but apparently was not the case in Romania).


Either way I'M pumped for the next book too!
Give 'em hell!

Keith Weintraub

I suggest a "stone circle of fire death match".

jack sparrow

I too agree with Levitt, its his blog and his post is clear and straight forward though a little mean, but hey therez no need to play nice for the heck of it too.


To #28: Ehm, no. It looks like Levitt selected the statement from Pop-Eleches which better conformed to his "story", which is precisely DiNardo's claim (have you read it?).

Further, how exactly does the finding that the cohorts born after the introduction of the abortion ban had worse outcomes (controlling for parents' background) prove that it is because of unwantedness? It just proves that for whatever reason those cohorts did worse irrespective of their family background. Or do we want to conclude that ALL kids born after the abortion ban were unwanted???

Rob Oxoby

Yeah, it must be rough having someone criticize your work without contacting you. I know just how you feel Steve:




well that is what Eleches's paper says:
born after the abortion ban attained more years of schooling and greater labor
market success. This is because urban, educated women were more likely to have
abortions prior to the policy change, and the relative number of children born to
this type of woman increased after the ban. However, controlling for composition
using observable background variables, children born after the ban on abortions
had worse educational and labor market achievements as adults."


I believe the historic and social circumstances were not properly taken into account in your study. So what happened: after Ceausescu banned abortions in 1968, more educated women had children. That much Mr Eleches's study shows. However, you have to remember that the communist education and economic structure was completely rigid at handling more people. For example, the number of university places was pretty much fixed, regardless of the fact that the new generation had twice as many children.
SO of course, university admission became so much more competitive. The children of high educated women (who had also resources to pay for private tuition, usually required for the very hard exams to get in), simply took the place of the others. University admission, to give just one example, was incredibly tough in the late communist period, since the number of places was completely fixed in advance centrally and did not move (in medicine there were 30 candidates for each place, in engineering over 10 etc.).

SO of course, the children of less well off families had a much more difficult task to get into university in the late 80's - very early 90's than beforehand simply because there were more children from better families available to compete, not because they were less wanted by their own families!

I think that, in an free market society, the doubling or trebling in demand for more university education and good job opportunities would have led to the creation of more of those opportunities. In a completely rigid communist system, this was simply not possible. This is the reason why people of the same family wealth and education background (i.e. after statistical control) did worse: there were simply much more talented and better-equiped competition around, for the same number opportunities, and NOT because they were somehow less wanted or neglected by their families.

But I really think that you are a bit on thin ice to claim this tells you anything about the US case. In the US, most of the unwanted children are born in poor single parent families. It is NOT a question that the society refuses to provide a high school and university place for them.

With all due respect, I believe you are a bit abusing the power of regression analysis here. I do not claim to be an economic expert, though I did also study economics as a second study and I did use regression analysis (which is not exclusively an economics tool!) for other purposes. But I believe that you cannot control for whole range of social and historical effects with such a simple, statistical technique. What success in life means in Romania in the late communist period (and the chaos that followed in the early 90's, after the collapse) differs from the meaning of life success in the US on so many levels that such a straightforward statistical comparison seems, at least to me, meaningless.