The Oil Drum

I have always had a  very hard time taking seriously any sort of state-by-state ranking on such completely subjective things as 'cohesiveness' or 'quality of life'. It is really BS disguised as science.

Try comparing quality of life in Nigeria with quality of life in the Netherlands. I assure you that you will find very real differences. The differences between US states are no less real, just smaller. Or are you really saying that there is no difference between California and Mississippi?

On several ocassions in the past I have been involved in this sort of thing, and they are typically done by using some sort of matrix scoring system to make it look objective, but in reality the way the various factors are weighted introduces the bias that was there from the very beginning. In other words, these things generally come up the way the people doing the ranking wanted them to come out before they went through the formal exercise.

So have I. One should expect some correlation between what we think we'll see and what our measurements say, else why make the measurements at all? This is what theory formation and hypothesis testing is all about. If you don't like the way things are measured, then go get your own measurements and present them for others to critique.

Social scientists labor under two problematic conditions --the inablity to conduct experiments and 'squishy' variables that are based on difficult-to-measure concepts. Yes, it's hard, and it's often wrong, but social scientists at least make the effort to apply, however imperfectly, the scientific method.  

I think the main flaw in a state-by-state ranking is inherent in the very idea that somehow a 'state' is something having the same level of reality as a  mountain range or a sea. People forgot that a state is really nothing more than an agreement by the general populace to govern the area inside these lines on a map seperately from the area inside those lines on a map. I think some people need to be reminded that you can't see any state boundary lines in a satellite photo of the US.

Yes, but collective agreements on governance are often more important to everyday life than something 'real' like a mountain range. Legal slavery, for instance, was an important distinction between US states before 1865, no? If you look here you can see that since 1945 alone political violence has killed an estimated 24,083,800 people worldwide. That's quite a lot for something that isn't real, isn't it?

In my view the main divide is between urban and rural. A resident of New York City and a resident of Chicago will generally have much more in common with each other then the New York City resident has with an upstate New York farmer or the Chicago resident has with a downstate Illinois farmer.

So, state-by-state is a very flawed way of looking at such things as cohesiveness or quality of life. At the risk of oversimplification, the single most important factor determining your quality of life is how much money you have, not which state you live in.

I agree on both points. Rural/Urban is probably the most important societal cleavage and that is always better to try to get as disaggregated data as possible. Unfortunatley, that is often very difficult to get. You have make do with what you have. Insofar that some states are more or less rural that will usually show in the statistics. However, in the map above, rural/urban doesn't seem to be a terribly great indicator of social cohesiveness. The Dakotas and the South are very rural, yet are on the extremes.

How did Putnam rank the states? What was his method?