Since I flog the book Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do incessantly, I thought I should point to the companion weblog. Why do I bring this book up so much? Well, I think the posts will tell you why. For example, look at this map which shows where the rich voted more than 50% Kerry. Or this post which illustrates the fact that religion polarizes the elite much more than the lower classes. Of course, Andrew Gelman is not a god, but the quantitative and relatively explicit methods he uses means that critiques, extensions or refutations are much easier to produce and follow for interested lay persons than when it comes to the arguments of more qualitative thinkers.
This does not mean that qualitative thinkers are wrong of course, but, it is difficult to evaluate their claims quite often because one does not have the dense implicit data set with which one can use to cross-reference their assertions. Samantha Power and Fareed Zakaria can have a very fruitful discussion, but an outside observer without their backgrounds might have a hard time making any informed assessment of which side they feel is generating a more accurate model of reality. Rather, I think most people will likely just go with their “gut” strongly informed by their normative preferences.*
Addendum: Since many readers have a biological background, a quick analogy might make what I’m trying to say clear: compare Systematics to the way it was before and after the rise of Cladistics. I was once told by a phylogeneticist that the greatness of the new method was that it dispensed with a great deal of “Because I said so!” arguments. And yet friends who have worked in phylogenetics have also complained that hardcore Cladists are often excessively dismissive of any other methodology. Similarly, I think qualitative thinkers still have a role to play as a complement to the quantitative ones in the human sciences (verbal arguments often precede formal ones).
* This is not to say that that can’t happen with quantitative models, it does all the time. But the clarity and precision of the initial arguments mean that differences are more pointed and transparent.
Comments are closed.