Precinct analysis: State Senate districts 2020

Congressional districts
State Rep districts
Commissioners Court/JP precincts
Comparing 2012 and 2016
Statewide judicial
Other jurisdictions
Appellate courts, Part 1
Appellate courts, Part 2
Judicial averages
Other cities
District Attorney
County Attorney
Tax Assessor
County Clerk
Fort Bend, part 1
Fort Bend, part 2
Fort Bend, part 3
Brazoria County
Harris County State Senate comparisons

Hey, look, we now have some 2020 district data. I found it all on the new Texas Legislative Council redistricting landing page. As of last week, when I went digging, only the State Senate and State House have 2020 data, so I’m going to spend a little time with them.

The 2020 State Senate election results by district are here. The first thing you need to know is that Joe Biden carried 15 of the 31 Senate districts. Here they are, in descending order of Biden’s percentage:

Dist    Biden    Trump   Biden%   Trump%
23    237,533   52,415    80.9%    17.8%
13    208,895   46,896    80.8%    18.1%
14    347,953  132,727    70.8%    27.0%
29    180,899   87,022    66.5%    32.0%
26    191,570   92,307    66.4%    32.0%
06    123,709   61,089    66.1%    32.6%
15    208,552  110,485    64.5%    34.1%
27    125,040   90,758    57.3%    41.6%
16    210,107  159,233    56.0%    42.5%
19    176,256  149,924    53.3%    45.3%
21    155,987  132,733    53.2%    45.3%
10    199,896  170,688    53.1%    45.4%
20    143,598  128,363    52.2%    46.6%
17    212,242  193,514    51.6%    47.0%
08    231,252  211,190    51.3%    46.9%

For the record, Beto carried the same fifteen districts in 2018. I’ll do a separate post on comparisons with other years, but I figured that was a thought many of you might have, so let’s address it here.

Only Biden carried the two Republican districts, SD08 and SD17. The range for other Democrats in SD08 was 46.4% (Chrysta Castaneda) to 48.1% (Elizabeth Frizell), and in SD17 from 46.5% (Gisela Triana) to 49.0% (Tina Clinton). Every Democrat got over 50% in each of the 13 Dem-held districts. This is consistent with what we’ve seen in Harris and Fort Bend Counties, where Biden outperformed the rest of the ticket by three or four points. For what it’s worth, we saw a very similar pattern in 2016, when Hillary Clinton ran ahead of other Dems, in some cases by quite a bit more. I’m thinking specifically of CDs 07 and 32, but there are other examples. My big question all throughout the 2018 cycle was whether those voters who voted for Clinton but otherwise generally voted Republican downballot would be inclined to vote for more Democrats that year, and judging by the results I’d say the answer was mostly Yes. We’ll have to see what happens this time around.

I’m sure you’ve noticed the lower-than-expected percentages in the Latino districts. SD20 is Chuy Hinojosa, and he won re-election by a 58.5% to 48.5% margin. SD21 is Judith Zaffirini, and she cruised 60.1% to 39.9%, while our old friend Eddie Lucio took SD27 64.8% to 35.2%. You may recall that in an earlier post on the Latino vote in 2020, one factor put forward for Trump’s better-than-expected performance was incumbency. As you can see, these incumbent Dems all ran comfortably ahead of Joe Biden. Now take a look at SD19, where Roland Gutierrez knocked out incumbent Pete Flores with a seemingly unimpressive 49.9% to 46.7% score. However much stock you put in the overall hypothesis, I’d say Flores’ incumbency helped him here. Not enough, thankfully. As for the two urban districts, SDs 06, 26, and 29, I’ve discussed SD06 before, so I’ll skip it. SD26 is basically on par with 2016, while SD29 slipped a bit from then but improved by a little bit over 2012. Again, I’ll get into more detail in a subsequent post.

Where Democrats really improved is in the whiter urban and suburban districts. SD14 was always a Democratic stronghold, but it really punched above its weight in 2020. No Republican district generated as many votes for Trump as SD14 did for Biden, and only one Republican district had a wider margin for Trump. We Dems maybe don’t appreciate Travis County as much as we should. I’ve discussed SD15 and how it went from a solid Dem district to a powerhouse in 2020. Look at SD16, which was a Dem takeover in 2018, and marvel at how Mitt Romney won it in 2012 with 57% of the vote. This is the kind of voting behavior shift that should have Republicans worried, and as you’ll see there’s more where that came from. Similar story at a lesser scale in SD10, which Trump carried in 2016 by a fraction of a point.

And then we have the two Republican districts that Biden carried. Both were battlegrounds in 2018, and I think the closeness of the race in SD08 was a genuine surprise to a lot of people, myself included. That’s a district that has shifted enormously, but it’s got more company than you might think. To understand that better, let’s look at the districts that Trump won, as above sorted by the percentage that Biden got.

Dist    Biden    Trump   Biden%   Trump%
09    161,000  166,632    48.3%    50.0%
25    256,178  302,919    45.1%    53.3%
07    188,150  232,201    44.1%    54.5%
05    199,253  250,002    43.4%    54.5%
12    211,292  270,287    43.2%    55.2%
11    161,818  232,156    40.4%    58.0%
02    138,917  208,774    39.4%    59.2%
18    161,933  271,898    36.8%    61.9%
22    128,415  253,102    33.2%    65.4%
04    142,522  281,331    33.2%    65.5%
24    126,340  257,861    32.3%    65.9%
30    121,646  329,601    26.5%    71.9%
01     92,593  265,715    25.5%    73.3%
28     76,925  222,872    25.3%    73.3%
03     77,364  294,559    20.6%    78.4%
31     59,684  229,768    20.3%    78.2%

Biden came within less than six thousand votes of taking a 16th Senate district, which would have been a majority. SD09 was Beto’s nearest miss for a sixteenth as well, though he came a little closer. The top five here for Biden are the same for Beto, with SDs 05 and 07 flipped; indeed, all of these districts are more or less sorted in the same way for both years.

I will have more numbers in the next post to show just how much movement there’s been, but in the meantime feel free to look at the 2012 district results and see for yourself just how uncompetitive these district used to be. The 2011 Senate map gerrymander was extremely effective, until all of a sudden it wasn’t. The Republicans will have some challenges ahead of them this fall.

There is of course some spare capacity for the Republicans to use, but it’s not as simple as it looks. Here’s the current map, to illustrate. None of SDs 01, 28, or 31 is anywhere near a Democratic stronghold. SDs 03 and 30 do border on Dem areas, and of course those other three districts can be sliced and diced to siphon off some Dem support, but it’s not quite that simple. For one thing, shifting the center of gravity in these districts from their rural centers towards the urban and suburban parts of the state means that their rural constituents – the Republican base – get less attention and power. They also increase the risk of a primary challenge from an opponent in a higher population area. I think playing defense will be a more urgent priority for the Republicans – they may try to carve out a more amenable South Texas district to capitalize on the Latino shift, but it’s not clear how persistent that will be, and there are still Voting Rights Act protections in place to guard against that, however tenuously – but maybe they could take a shot at Sen. Powell in SD10. As with the Congressional map, it’s a question of their risk tolerance as well as their appetite for gain. We’ll know in a few months.

Related Posts:

This entry was posted in Election 2020 and tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

5 Responses to Precinct analysis: State Senate districts 2020

  1. Greg Shaw says:

    No matter how Democrats try to put him there, Trump will not be on the ballot in 2022. It will be a good chance to see if 2018 was the “new Texas normal”(politically), or (as I suspect) Orange Satan was the greatest GOTV gift that Texas Democrats could have ever gotten.

  2. Kibitzer says:


    Data like these should delight political scientists, who are good at explaining what has happened, but less adept at predicting future election outcomes. Not to mention correctly.

    As for retrospective election data, the inherent beauty is that it actually represents political behavior, not self-reported past behavior or mere voting intentions or current preferences solicited from a small sample of respondents, though not at the individual level. Actual voting data is valuable even at the aggregate level, and — unlike polling data — it actually decides election outcomes. So this is — empirically speaking — high-quality “hard” data on mass political behavior that actually counts. The same cannot be said of sundry opinion polls and surveys, which are additionally always afflicted with sampling issues, question-wording effects, and miscellaneous errors.


    It’s easily seen why the predictive value of such data is problematic despite its excellent quality.

    First and foremost in the current situation, we don’t know how the new districts are going to be drawn. There is little use speculating, and since Democrats are not going to have much say in the process, there doesn’t appear to be any good reason for Dems to propose a new map of their own that represents their best-case scenario, and their notions of fairness. It would only help the Republican map-makers, who want to protect their own incumbents and other fellow-partisans, to finetune their gerrymandering schemes to thwart the Dems’ plan.

    That said, there are serious problems in using tallies from the last election to make projections about the next one that complicate matters even if all the district lines stay the same.


    Most  notably are turnout differences between presidential and gubernatorial election years, so we are not even talking about the same aggregate of voters. And self-selection among registered voters (i.e., decisions to vote or sit out a particular election) may make an even bigger difference in the respective parties’ share of the vote in 2022 to the extent the participation level and outcome in 2020 was significantly influenced by a desire to get rid of Trump at all costs. And this may not bode well for Democratic candidates strategizing now. 

    Trump allies may be on the ballot in the mid-term elections, such as Paxton or Bush in the AG’s race, regardless of who wins in the GOP primary, but that’s not quite the same as having Trump on the top of the ticket and the specter of more Trumpism in and from the White House.

    To the extent some otherwise lukewarm Texas voters were primarily motivated to participate in 2020 to prevent Trump from getting re-elected, that won’t be a reason to head to the polls in 2022. People in that group could end up abstaining in 2022 without thinking much about it. By default, if you will.


    Projections could alternatively be  based on mid-term results in 2018, but that will be four years prior, and some changes at the aggregate level will necessarily have happened since then. On a statewide basis, voters will have died, and youngsters will have reached voting age. And that demographic replacement (assuming voter registration of newly enfranchised young voters to take the place of the deceased) doesn’t even reflect in-migration and out-migration, or acquisition of citizenship by existing residents who then register to vote, for that matter.  

    And then there is the question whether the Dems’ good showing in 2018 was specific to that year, or even specific to Beto O’Rourke (“Betomania”). Can that excitement be duplicated if Beto decides to run? Probably not. Nor is the challenge to Senator Ted Cruz really equivalent to taking on the sitting governor, especially on the reasonable assumption that Abbott will again be the GOP’s nominee.

    Finally, as for top statewide races, we don’t know who will be running against whom. That weakens the value of past voting data even further, even if it is otherwise useful to ascertain the geographic distribution of support for (or potential for support for) Democratic candidates. There might even be an independent candidate for governor, which would make the dynamics in an otherwise polarized political match-up much more interesting. And even less predictable. 

    At the state level, of course, the total tally matters, and is not going to be affected by redistricting, except to the extent that the manner the new districts are drawn has a significant effect on turnout in non-competitive districts. But this would be difficult to estimate. Once the districts are redrawn, there won’t be any historical data yet to go by. Not yet.


    In any event, the dynamics of statewide races are fundamentally different from races in legislative districts. Every vote counts regardless where in the state it is cast. It may be worth making that clear to voters in gerrymandered districts, where safe incumbents may not even have much incentive to campaign vigorously. Especially not in general elections.

    Even if the general election outcome in a noncompetitive district is a foregone conclusion, each voter’s choice for governor and other state officials still counts as much as anywhere else in the state.

    Outreach to would-be supporters in safe districts may be worthwhile as part of a state-level electoral strategy even if the local representative is in a safe seat and has little incentive to work hard to garner additional votes to increase the win margin. And that logic also applies to minority party members in such gerrymandered district. They may have no clout in the district race and zero chance to affect the outcome, but their vote for a statewide candidate could still make a difference if the state-level race is competitive.

  3. Joel says:

    Holy cow, kibitzer, you don’t really think a single person is reading all that, do you?

    You could get your own blog for no one to read.

    Or you could do something more useful for humanity, like cleaning out under the seats in your car.

  4. asmith says:

    While they picked up SD10 and SD16 in 2018, they should have put more early resources into senate districts 8, 9, 17. They are already ruing the day.

    The GOP has to shore up 8,9, 17, and maybe 25. They will tinker with SD10 and dramatically alter SD16. I don’t know if they will crack 16 and give portions of it to Paxton, West, and Hall and draw a new SD16 in Collin going to East TX, but I wouldn’t put it past them.

  5. Pingback: Texas blog roundup for the week of May 24 – Off the Kuff

Comments are closed.