Comment by ajtulloch | Hacker News Reader

Comment by ajtulloch | Hacker News Reader

Parent Full thread

e.g. https://en.wikipedia.org/wiki/Legalized_abortion_and_crime_e... or https://www.economist.com/finance-and-economics/2005/12/01/o...

> It was a good test to attempt. But Messrs Foote and Goetz have inspected the authors' computer code and found the controls missing. In other words, Messrs Donohue and Levitt did not run the test they thought they had—an “inadvertent but serious computer programming error”, according to Messrs Foote and Goetz.

> Fixing that error reduces the effect of abortion on arrests by about half, using the original data, and two-thirds using updated numbers. But there is more. In their flawed test, Messrs Donohue and Levitt seek to explain arrest totals (eg, the 465 Alabamans of 18 years of age arrested for violent crime in 1989), not arrest rates per head (ie, 6.6 arrests per 100,000). This is unsatisfactory, because a smaller cohort will obviously commit fewer crimes in total. Messrs Foote and Goetz, by contrast, look at arrest rates, using passable population estimates based on data from the Census Bureau, and discover that the impact of abortion on arrest rates disappears entirely. “I am simply not convinced that there is a link between abortion and crime,” Mr Foote says.

>In 2005 Levitt posted a rebuttal to these criticisms on the Freakanomics weblog, in which he re-ran his numbers to address the shortcomings and variables missing from the original study. The new results are nearly identical to those of the original study. Levitt posits that any reasonable use of the data available reinforces the results of the original 2001 paper.[10]

>The effect of legalized abortion reported by Donohue and Levitt (2001) is largely unaffected, so that abortion accounts for a 29% decline in violent crime (elasticity 0.23), and similar declines in murder and property crime. Overall, the phase-out of lead and the legalization of abortion appear to have been responsible for significant reductions in violent crime rates."

According to your link it doesn't appear to be shobby science.

That was not a rebuttal against the pointing out of the software error, but to the Lott and Whitley criticism.

The rebuttal to the Foote and Goetz criticism of a programming error is this:

> Donohue and Levitt subsequently published a response to the Foote and Goetz paper.[12] The response acknowledged the mistake, but showed that with different methodology, the effect of legalized abortion on crime rates still existed.

That's troubling to me. They had a hypothesis, and when the experiment didn't confirm that hypothesis, they simply tried a different methodology that might still show it. That feels like moving the goalposts to me.

It's also representative of the negative feeling I had all the time when reading their book: after all their praise of rigorous scientific methods and of not confusing correlation with causation, they follow that up with making those same mistakes themselves. (And I guess the reason that rubbed me wrong is because that is interjected with continuous praise all the time of Levitt. That in turn probably makes me more likely to criticise, so perhaps you should take my criticism with a grain of salt...)

Wow, you cherry-picked parts of that out and left out the part where they admit their science was wrong and say there is no way to prove/disprove. Re-read the last paragraph of that 2005 criticism section to see the authors admitting it was bad.

Sounds like you're the one cherry picking by focusing on only one part of the link the OP posted.

Reply on news.ycombinator.com