# Precinct analysis: The median districts

This is a straightforward post, with a simple answer to an important question. We know that Joe Biden carried 74 State House districts and 15 State Senate districts. How much better did he need to do to get a majority in each chamber? Daily Kos calls this the “median district”. In this context, that means the data for the 76th-most Democratic House district, and the 16th-most Democratic Senate district. The idea is to see how far off the Dems were from being able to win those districts and thus claim a majority in each chamber.

We’ll start with the State House. The table below gives the data for the median district in each of the last three Presidential elections for the Presidential race, the Senate race (2012 and 2020 only), and the Railroad Commissioner race:

```
Year    Dist      Dem      GOP   Tot D
======================================
2012   HD138   39.29%   59.16%      54
2016    HD54   43.58%   50.50%      65
2020    HD54   48.85%   48.98%      74

2012    HD97   38.35%   58.88%      54
2020    HD92   46.04%   51.12%      68

2012    HD97   36.16%   59.58%      54
2016    HD66   37.77%   54.46%      56
2020    HD31   46.52%   50.55%      68
```

In 2012, the 76th-most Democratic district was HD138, in which Barack Obama received 39.29% of the vote to Mitt Romney’s 59.16%. This is a polite way of saying that the 2011 gerrymander was super effective, and the Democrats weren’t within hailing distance of winning half the chamber. The last column shows the total number of districts carried by the Democratic Presidential candidate. In 2012, this closely mirrored the total number of seats that the Dems actually won, which was 55. One Democratic-held seat was carried by Romney – HD23, the Galveston-based district won that year (and for the final time, as he declined to run again) by Craig Eiland. As you may recall from previous analyses, that district has trended away from the Dems ever since – in 2016, it was won 56-41 by Trump, and in 2020 it was 57-41 for Trump. Obama carried zero Republican-won seats – the closest he came was a 52-47 loss in HD43, another district that has moved farther away from Dems over the decade. He came within six points in three Dallas districts that Democrats now hold – HDs 113, 107, and 105. Like I said, an extremely effective gerrymander. Also a consistent one, as Paul Sadler and Dale Henry won the same districts Obama did, no more and no less.

Until it wasn’t, of course. The cracks began to show in 2016, when Hillary Clinton carried 65 districts, though Dems still only won 55 of them overall. HD23 fell to the Republicans in 2014, but Dems earned their first flip of the decade (*) by taking HD107, which as noted above was one of the closer misses in 2012. The nine GOP-won districts that Hillary Clinton carried were HDs 113, 105, 115, 102, 112, 114, 138, 134, and 108. Seven of those are now Democratic districts, with six flipping in 2018 and one (HD134) flipping in 2020.

Note how Clinton ran ahead of other Dems as well. Perennial candidate Grady Yarbrough picked up only HD105, and that by a 45.9 to 44.6 margin (there was a lot of third-party voting in that extremely unappealing race), and it was the same at the judicial level. You may recall this is why I was more guarded in my optimism about 2018 initially – I had some doubts about what the Clinton/GOP voters would do their next time out.

We know how that turned out, and we know how Biden did, as well as how MJ Hegar and Chrysta Castaneda did in 2020. Look at how the median district shifted over time. In 2012, the 76th district was more Republican than the Presidential race was, at each level. In 2016, the median district looked a lot like the Presidential race, and to be honest a lot like the RRC race as well; Wayne Christian defeated Grady Yarbrough 53.1 to 38.4, a bit closer than the median but not far off. In 2020, at all levels, the median district was closer than the statewide race was. Republicans outperformed their baseline in the House, and they needed to because by this point their vaunted gerrymander had completely failed them. I have to think this is something they’re giving serious thought to for this time around.

Here’s the same data for the State Senate districts:

```
Year    Dist      Dem      GOP   Tot D
======================================
2012    SD08   36.60%   61.67%      11
2016    SD09   41.75%   53.09%      12
2020    SD09   48.30%   50.00%      15

2012    SD08   35.94%   61.05%      11
2020    SD09   45.40%   51.70%      13

2012    SD08   33.34%   62.19%      11
2016    SD08   36.19%   55.94%      11
2020    SD09   44.60%   51.60%      13
```

It’s a similar pattern as above. In 2012, Mitt Romney carried SD10, which Wendy Davis won in a hard-fought race. In 2016, Hillary Clinton carried SD16 by a 49.9 to 45.3 margin, and just missed in SD10, losing it 47.9 to 47.3; she also came within a point of SD17. The median district was a little friendlier to the GOP in 2016, but in 2020 as with the House it was closer than the corresponding statewide race. Again, the once-solid gerrymander buckled at the knees, aided in large part by the suburban shift. Dems also managed to hold onto all of the red-shifting Latino districts, while Biden dropped two of them in the House.

What does any of this mean going forward? I have no idea. I’m seeing map proposals for Congress that are pretty brutal, but who knows what we’ll get in 2022, and who knows how population growth and the shifts in suburban and (mostly rural) Latino areas will affect things. Texas is a more challenging state than the likes of Wisconsin or Michigan to control over an entire decade precisely because it changes so much in that time. Republicans will have some opportunities for gain in 2022, but they also have a lot of vulnerabilities, and their best defense may be to just try to shore up everything they now have. The choices they make, based to some degree on their level of risk tolerance, will be fascinating to see.