Much of my research over the past dozen years or so is about how an understanding of differences across geographic entities (regions, cities, states) is important for understanding how the national economy works. I realize that a normal human being would find such a claim to be tautological, but macroeconomists are not normal human beings. Even so, I think I made some headway over the years. It's good to see, therefore, that my former St. Louis Fed colleague David Andolfatto has seen the wisdom of my approach (here and here).
(Unless you're interested in the arcana of economics, there's no need to read any further)
It would have been nice, however, if David had set himself apart from others (including Nobel prize winners) who make the mistake of calling the plot of unemployment and vacancy rates a Beveridge curve. The Beveridge curve is a locus of unemployment and vacancy rate combinations under a given search technology. The scatterplot of unemployment and vacancy rate combinations connected by lines should be called something like a Beveridge plot. David seems to be aware of this, but he should not perpetuate the error of many famous macroeconomists before him who do not seem to understand the difference between shifts of a curve and movements along it. Here's a paper looking at subnational Beveridge curves that manages to do this.