a thread on why I think the open seat TX-22 is somewhere between a toss-up and shaky tilt R territory

a few things to know here— TX-22 voted for Cruz by about half a percent last cycle, and relative to the state it's moved bluer every presidential race since the 90s. the big shift happened in 2016, when it went from Romney+26 to Trump+8 (somewhere between GA-7 and GA-6 level)

since tx-22 was drawn as a "Fort Bend county without the POC" district starting back in 2003, Fort Bend casts about 70% of the vote, with Harris and Brazoria counties contributing the remaining third

still, the Ft. Bend portion isn't homogeneous; there's a pretty substantial exurb belt along its western edge, and the northern 'triangle' of the county is noticeably less diverse and more conservative than the rest— if we count them separately, it gives us 5 regions to work with

the turnout distribution has been pretty flat over the past 10 years— all 5 regions are booming, but the Harris Cty. part of the district was built out in the '80s, so it lost some ground to other areas, especially north Fort Bend

Beto did well here, and that's pretty clear, but there are some noticeable things in the background, like how even though Cornyn did ~12 points better than 2014 than Cruz in 2012 statewide, his advantage was only 6-8 points here

additionally, the fastest growing part of the district, North Ft. Bend, has swung the hardest (growth is probably why— but that's still not where you want to be hemorrhaging support if you're the GOP)

on the presidential level, 2016 is as blatant as 2018, but just as Cornyn slightly (but cautiously) didn't match his gains here with his statewide result, Romney did about 2.5% better here while doing 4% better than McCain statewide

as a whole, the picture is pretty clear: Rs have lost ground almost everywhere over the past 2 decades, while yanking what's left out of the 5% cast in the rural part of the district

This gives you a number of trends to work with— the issue is: which one? Was Beto an outlier? Is 2008 relevant anymore? What about 2002? The answers to each define the trajectory going forward into 2020, so my answer is: look at all angles.

If you take the long view— looking at this district as slowly lurching towards the left, you would expect Biden to need to lose Texas by about 3-4% to carry TX-22. If you think it's more of a sudden left-turn, then Biden could lose the state by 6-9% and still take TX-22 with him

On the other end of things, if you view 2016/2018 as an abomination in time, and believe that things will quickly revert to the blessed era of Rick Perry and Kay Bailey, then Biden needs to do really well statewide, even outright win Texas. However, that's pretty far-fetched.

This is where polling comes in: Trump's lead in the state is only about +2.6 in the RCP average, +1.6 if one looks back over the past year entirely— both of which would throw TX-22 well into the *tossup* range of projections laid out earlier.

If Trump can't clear Biden by three points or more, which (as it stands) isn't the case, then TX-22 is in danger. I think there are also some good signs for either party here, starting with the Republicans.

Firstly, Trump will probably pull ahead in Texas; it's doubtful he outdoes Hillary, but a 6-7% win seems pretty plausible. Secondly, the HD-28 special that they (handily) won casts some doubt on there being any secret D strength lingering around this area.

On the other hand, if you're a Democrat, you get the opportunity of running against the pretty heavily right-wing, Trump-loving Troy Nehls in a rapidly diversifying part of the state. Lastly, the reddest part of the district always casts a smaller share of the vote during

presidential years than in midterms (more in '02 than '04, '06 than '08, '10 than '12, and '14 than '16), which could provide the crucial kick if it comes down to the wire, as it very well might.

conclusion? it's gonna be close folks. →tossup