I just now discovered that I have 4 hours left to do a rather huge and important project on my econometrics class - if I fail to do it I might have to take another semester, which will be a huge pain in the ass.
I am supposed to make a simple linear regression (OLS) model, which would normally be a piece of cake - however, there are two main factors that make it quite difficult:
1. the model should be based on at least 40 observations
2. among the independent variables there should be one binary variable
The model should be at least loosely connected with economics, so no weather patterns etc., and I need to score some extra points with this guy, so I decided to do something with macro-economical data. I normally love doing time series analysis, but in this case I honestly cannot think of anything meaningful - because of the first requirement I'd have to use monthly data (40 yrs = almost certain structural breaks), so cross-sectional data it is.
My first choice in cross-sectional data, as a huge eurofag would be EU, but it is still too small god damn it... That's why I think it would be best to choose the US. And here is where you guys, especially the Americans come into play:

TL:DR
Does anyone have any ideas for a model based on some cross-sectional, macro-economical data concerning individual states, that would include the Democratic-Republican alignment (there will be other factors as well of course, but I need the binary variable to be significant), that would make sense?