Monday, January 15, 2018

How gerrymandered are we?

It’s a problem, right? I mean, look at this map of my home state:


Or, how about this map of the Detroit area:


And finally (and also to show it’s not just a Republican thing), this one of Maryland’s 3rd congressional district:


So, obviously there’s a problem here. But how to show it in a nice concrete, mathematical way?


Methodology

This one is actually pretty straightforward. Here’s what I did:
  1. Totally eliminated Louisiana (they have some weird system where that does not pit Republicans and Democrats against each other straight up)
  2. Calculated the percentage of Republicans in each house delegation (e.g., Mississippi, with 3 Republicans and 1 Democrat, comes in at 75%)
  3. Calculated the percentage of Republican votes cast in all house races (once again, Mississippi, with 681,000 Republican votes to 450,000 Democratic ones, come in at 60%)
  4. Subtracted the difference between the two, to see what kind of advantage dividing the state up into districts might have had (in Mississippi’s case, 15%)
  5. Ranked them all
So, not unlike comparing the electoral vote to the popular in the presidential election. But with this one being a lot more amenable to actual intervention.

Let’s see what we came up with …


"Dirty" Sweep

There are actually 11 states out there that are solely Republican or solely Democrat.  Now, some of them have only 1 congressional delegate to begin with, so we really can’t point to gerrymandering in those instances. I’ve gone ahead and shown those here, but with light shading to distinguish them:


A couple of things to note:
  • The Republicans have a distinct advantage, 11 to 6
  • They are strongest in the Plains and mountains, the Dems in the Northeast
  • Massachusetts wins the prize for 1-party state, with the highest number of delegates all of one party, at 9

Too True Blue

So, I think we can already see that this can go both ways. Now, here are the states that had 10% more delegates than they should have had, based on state-wide votes:


Note that I did eliminate the states with only 1 delegate (no chance of gerrymandering there).

Now, just to give you an idea of the strength of these differences, here’s a nice bar chart:


So, except for Nevada, pretty much the usual suspects, right?


Too True Red

And here’s what it looks like on the other side of the aisle:


And numbers wise:


No surprise there with the solid South, but how about all those states in the Midwest? 


Just Right

So, is there anyone playing fair out there? Luckily, there are a few. Here are the 10 states that were under 10%, positive or negative:


And here’s how that shook out exactly:


Pretty scattered around, no? Also, some of these seemed pretty obviously purple – Illinois and Colorado, in particular. I really am wondering, though, what Texas (on one side) and New York (on the other) are doing here.


Final Thoughts
  • Republicans win the gerrymandering sweepstakes, with a score of 22 states over the 10% threshold to 10 for the Dems
  • New Hampshire wins the most gerrymandered state award with a difference of -48% (it has 2 Democratic congressmen, though only 52% of the state voted for a Democratic candidate)
  • Arizona gets the least gerrymandered state award, with a difference of only 1% (Maine comes in 2nd with 2%)
  • There are other, much more detailed ways to looks at this (ways that I will leave to the professionals), but I did think this was rather interesting as a quick read

No comments:

Post a Comment