The “Real Time Economics” blog in the Wall Street Journal has an interesting post matching state voting preferences (determined by current polls) to changes in home prices, changes in personal income, and the unemployment rate.
You can do a quick sort online, and a few points jump out right away…
- The six states with the lowest unemployment rate are all strongly Republican. Four of the six with the highest rate are voting Democratic.
- 10 of the 12 with the greatest gain in income are going Republican, while 10 of the 12 with the lowest income gain are voting Democratic.
- The home appreciation category is really interesting. 18 of the 20 states with the greatest losses in home prices are going Democratic. 16 of the 20 states that best maintained their home’s values are leaning Republican.
It’s tempting to look at that last category and translate the pain felt by homeowners into a desire for a new political ideology. Not so fast. The vast majority of the states (in those batches of 20) are voting the same as they did in 2004, so really it was a case of the blue states getting hit the worst by the recent housing downturn.
Using their data, I converted the polling data to a linear scale, and did some simple correlation analysis. The coefficient of correlation was roughly 25 percent for both the categories of unemployment rate and change in personal income. It rose to 45 percent for the home price change category.
I thought it might be interesting to look at the difference between states’ voting from 2004 to 2008 and created a category for the ‘delta.’ I was hoping for some big insight, but the results were disappointing. The correlation coefficients were actually less, so the big takeaway might be that there is a lot of inertia. Changes in economic conditions may take longer to make a large impact.
My spreadsheet is here for your own analysis.
Of course, the United States’ political landscape is narrowly divided into Red and Blue states, so even a small shift in voter sentiment is translated into a sea change.
Best, Bert