The common theme between these two papers and the chapter from Freakonomics is that they both focus on the relationship between the legality of abortion relative to crime rates and how they decreased. In the first paper, by Donohue and Levitt, they focus on why crime rates dropped in America in the 1990s and how abortion influenced if not caused the dramatic decrease. Both Freakonomics and the first paper focus on this area and look at the time when abortion was made legal, in 1973 as a result from Roe v. Wade. Basically what they find is fairly straight forward; legal abortion caused a drastic decrease, as much as 50%, in crime rates; however, not immediately after it became legal. Since people do not start committing crimes when they are babies or young children, there was a period of time that passed before observing these dramatic drops in the crime rates. Thus, in addition to studying how abortion affects crime rates, they also discovered that crime rates are highest amongst teenagers, from about fifteen to twenty years of age. This is because there was about an seventeen year lag from when abortion became legal and when crime rates started dropping. They also provide convincing arguments for why this is the case, and evidence with regressions. In addition, Levitt and Donohue account for several other possible causes of the sudden reason for the reduction of crime. However, they find that they did not have much of an impact upon crime rates relative to what they had been doing before 1990. This is primarily due to the fact that most of the regulations, laws, and other contributing factors to crime reduction had been in place before and after the drop, thus being fairly unimportant.
Although Donohue and Levitt do present a good argument for abortion and crime rates, the critique by Foote and Goetz offers some interesting points that they say Donohue and Levitt seemed to miss. The primary thing that they talk about is the way in which Donohue and Levitt ran their regression model and how they incorporated the data. They say that in order to look at something like crime rates, one cannot collect data from different states, cross-state data, and in this case different cities and then compare them. They suggest that collect data from each state individually, in-state data, and then analyze crime rates in cities amongst themselves. Then you can compare city to city rather than making larger, less accurate assumptions. In addition to this critique of their work, they also look at the variables in their regressions and find that by changing some of them, it makes the regression no longer significant. They reason that in all the regressions that Donohue and Levitt run, they use the amount of arrests per capita to prove their point, however, in one regression, they stray from this logic and simply use the total amount of arrests, which would completely change the output from the data. Thus, Foote and Goetz run the regression using per capita assets and find that the regression becomes insignificant, consequently nullifying the conclusions made by Donohue and Levitt.
Although after reading the first paper and comparing it to Freakonomics and then reading the second paper and finding out that it was mostly based off false conclusions was kind of annoying, however nonetheless interesting. Despite the fact that Donohue and Levitt’s point is technically based on a poor regression, they still do provide a very valid point; that the legalization of abortion do in fact decrease crime rates due to the basic principle of reducing the amount of people that could commit crimes several years down the road.