I’m going to look at inequality across the urban-rural divide.

RMarkdown is giving me fits so I’m just going to post an update while I build my script behind the curtain. Sorry about this.

I’ll use income data at the census tract level. American Factfinder seems to think I’m looking for data set S1901 from ACS. I pulled ACS five-year data for four states — New Jersey, New York, Connecticut and Pennsylvania.

Then I’ll introduce equivalent data from the 2000 census and look at the per-capita or median change, from 2000 to 2016, in incomes at the tract level.

Next will come the urban-versus-rural consideration. I’ll find or introduce that indicator and do some uncontrolled (bivariate) and controlled comparisons of change in incomes across that divide. An econometric matching process, possibly geographically weighted (to try and compare metro and rural areas close to one another), might help. The literature will help me identify what basket of features to control for.

Finally, I’ll replicate things after settling on a standard, agreed-upon measure of inequality.

The first thing after that is just to take the temperature (that measure) regarding income.