A warning on nationwide vote totals and House elections

As election day approaches, I'm just trying to get a bunch of these warnings up, and one that seems particularly important is a point that nearly everyone gets wrong about majority status in the House of Representatives. How many votes does it take for the Democrats to get control of the chamber? You'll hear a bunch of different numbers, like the nationwide percentage that the Democrats need to take in order to win the chamber, and you'll also hear things about how it's possible to win more total votes in all House elections combined and still not have a majority in the chamber, just like it's possible to win the presidency without a "popular vote" majority because of the electoral college.

Well, I've ranted about how that argument is bullshit regarding the electoral college, and it's bullshit here too.

Here's the problem with trying to aggregate votes across House elections. There are actually a bunch of districts in which the incumbent runs unopposed! Yes, that happens. Why? The other party knows they'll lose. But, suppose they expect to get 30%, so they don't find anyone to bother running, and they get 0% instead. Calculate the nationwide vote total, and there are 0 votes from that district, not because the party didn't have any support, but because nobody from that party was on the ballot. You're getting a screwed up estimate. Care to guess which party is in that boat a lot? The GOP. Why? Which party is more likely to have 30% of the vote in some hypothetical district? The GOP. In a "majority-minority district," which is a district in which a majority of the population is of a racial or ethnic minority. So, you get a bunch of Democrats running unopposed, bringing down the GOP vote total nationwide, making it look like the Democrats perform better than they have, and so you look at the Dem-GOP vote totals nationwide and say, "hey! Look how many more votes the Dems got, nationwide! And they still lost! That's fucked up!"

Part of that is the number of times the Democrats ran unopposed. Then, you've got just weak candidates running in the heavily slanted districts when the disadvantaged party runs someone. Running a weak candidate is little different from running nobody at all. Then, you've got varying levels of turnout across districts and states because of varying levels of competitiveness in the associated elections, and that will screw up your aggregation. You can't aggregate votes across districts like that.

Partisan gerrymanders are real. They aren't as bad as you probably think, and mostly they occur naturally because Democrats cluster in urban areas. However, when someone shows you nationwide vote totals for House elections, if they aren't even bothering to make adjustments for unopposed incumbents, ignore them.