The level national playing field

The national map is surprisingly fair overall. That's a start for further reforms.

Now that the Congressional election is starting to settle out, it’s looking like Republicans will take control with a tiny majority. Contrary to some of what you’re reading, their win is quite fair - in a national sense. The easiest way to see this is with a simple scatter plot.

This plot shows the outcomes of national Congressional elections since 1946. Each point represents one election. Black dots show 1946-2010, and they cover the gray shaded area. Red dots show 2012-2020, the decade following the Great Gerrymander of 2012, a remarkable accomplishment by Republicans. Because of this pattern, many of you may have gotten used to thinking of gerrymandering as an irresistible force that helps Republicans. But this year, that was not true.

The green dot indicates a preliminary result: the likely seat share, a range of 219-222 seats, plotted against the average district-by-district vote share, currently a 1.7-point Republican margin. The dot is right in the middle of the shaded gray zone. In short, this year’s election falls within a pattern that has stood since World War II.

Let me now gently debunk two contrary ideas that are floating around out there.

RealClearPolitics has a running total in which they show Republicans winning the national vote by over 4 points. However, that total leaves out a substantial part of California, which is not nearly done counting. The plot above, which averages the percentage margins on a district-by-district basis, is likely to come closer to the final vote count.

Roll Call has a headline, “Redistricting helps Republicans in close fight for House control.” That’s half true. If Democrats maintain control, which is highly unlikely but technically still possible, one could just as well say “Redistricting helps Democrats in close fight for House control.” Redistricting consists of dozens of actions, which in a sense of raw power, add up to generate an overall effect.

Compared with a party-blind drawing process, Democrats were helped in Illinois, Maryland, and North Carolina. Republicans were helped in Texas and Ohio. And new commission-based processes in Michigan, Colorado, and Virginia (in this case rescued by the state Supreme Court), as well as split control and a court battle in Pennsylvania, took those states off the table as targets for gerrymandering.

Putting it all together, by several measures the total effect was close to zero. The median district nationally is 1.1 points more Democratic compared with the national average margin. Here’s another way to see it: 221 out of 435 districts were above average for Democrats. In short, Democrats had a slight advantage.

Note that before the election, I suggested that Republicans would have about a 2-point advantage in terms of average-median difference. Instead the advantage swung the other way - barely. The average-median difference can fluctuate depending on particular conditions, such as unusual patterns of high performance in some districts. I conjecture that Democrats were more enthusiastic in the districts they needed. Whatever the case, the average-median difference is tiny compared with 2012, when it was 7.3 points.

The bottom line: Thanks to the hard work of reformers and activists (and self-interested actions by both parties), this year’s Congress will be majoritarian. The vote was closely divided, and so will the people’s House.

Now, some concerns.

The overall balance does not overcome the fact that there were many local offenses. For example, several racial gerrymanders have been allowed to stand. From a pure power standpoint, these were cancelled to a degree by other offenses. It doesn't excuse them individually. 

To step back, I would characterize some major criteria for good districting as: (1) fairness to the parties, (2) competition in states and districts, and (3) representation for diverse racial and political communities. There are much better ways to meet these criteria than what we saw in the last decade - and a lot of work to do in the years ahead.