17.2.13

NOT GERRYMANDERED.

An assignment of electoral votes by Congressional district suggests that the Republicans are a closer approximation to a national party than are the Democrats.  Now comes Michael Barone, with a finer-grained analysis of the construction of Congressional districts.
Democratic voters, particularly in the two Obama elections, tend to be clustered in black, Hispanic and gentry liberal neighborhoods in major metropolitan areas and (this makes a difference in state legislative districting but not so much in congressional districting) university towns. Democrats win districts in such areas by huge margins, with upwards of 70% of the vote in many cases. Republican voters are pretty evenly dispersed outside these clusters.
Left for further research: whether the presence of the constraint in the voting rights laws that requires some fraction of the districts be majority-minority augments the concentration of Democrats. Absent that requirement, more Congressional districts might be contested, and the House and to some extent the Senate less fractious.

Mr Barone doesn't solve that problem, but a related problem offers food for thought.
Proof [c.q.] comes from this “Electoral Reform Map” which divides the United States into 50 equal-population states. The folks who did this want to reduce the power of small states in the Electoral College. They used algorithms in order to draw the states and then smoothed out the boundaries, keeping major metro areas together in one state (or more than one when they have enough population). It’s a pretty neat looking map. And each state would have the same number of electoral votes.
It's also a pretty politically infeasible map: imagine running the State Patrol in one of the Trans-Mountain states.  The simpler way out might be for Congressional districts to be drawn on an equal-area, equal-population basis wherever possible.

No comments: