I particularly like his time series analysis which estimates the various demographic flows and trends in the states in question. Trende writes:
Let’s use one of my favorite tools: Nate Silver’ demographic calculator from 2013. Silver’s algorithm freezes the vote shares of demographic groups at 2012 levels (though you can alter them), then makes projections in states based upon current projections of population growth (which you can also alter).
So I took Silver’s projection, assumed nothing changed electorally, and also assumed that immigration reform with a path to citizenship did not pass. Then I simply marched through the years, waiting for Georgia, Texas or Arizona to flip.
I waited a long time. Arizona finally flips in 2036, and Georgia flips in 2048. Texas never does.
