Monday, December 7, 2015

Bad Amateur Social Science

Many political arguments go as follows: “Social Problem X is much smaller in Country Y than it is in America. Therefore we should adopt Country Y’s policies.” Of course it’s a terrible argument. It’s a classic case of confusing correlation with causation. Often the speaker is literally comparing one single country to another (rather than comparing groupings of similar countries); this one-offing invites cherry-picking. Unfortunately this kind of amateur (pop-) social science is all too common.

To get this kind of argument right you have to compare likes-to-likes. Japan has less gun violence than the United States, you say? Do Japanese Americans have similar rates of gun violence to that of their home country? If so, you can’t credit the lower crime rates to gun policy. Norse countries have lower rates of poverty and child mortality than the United States, you say? What about Americans of Norse descent? Do these populations have similar rates of social issues to those of their home countries? Maybe it’s the deep cultural/institutional characteristics of these populations, rather than government policy, that drive the different outcomes. At any rate, it’s clearly not legitimate to simply point out a difference between two groups of people and attribute 100% of the difference to your favorite cause. 

This point is so obvious that it’s difficult to articulate it without sounding snarky or sarcastic, and yet somehow it needs to be said. People flout it all the time, including people who ought to know better.