Maybe I need to drink more.
Seriously, I’m having trouble articulating the problem with trying to assess the difference between recent electoral trends in a district with long-term ones. As I see it, there are three major traps involved: the two obvious ones would be favoring one over the other; but the third one is more subtle. Can you really define any trend for any geographical-based metric? After all, Western society has become increasingly atomized over the last sixty years, to the point where in many places you have to work at it to know your neighbors.
It’s a bit of a puzzler, and the annoying bit is that reconciling the quandary will affect at most maybe two lines in a particular post.