Racially adjusted 2016 presidential vote by congressional district

[mieastwick]

https://fusiontables.googleusercontent.com/embedviz?q=select+col5%3E%3E1+from+1rkz_05Tptwe3B8BWqliZ5wl-Rj3Q2kyOebwzQYf8&viz=MAP&h=false&lat=38.533594380497526&lng=-91.35249365234375&t=1&z=5&l=col5%3E%3E1&y=2&tmplt=2&hml=KML

Inspired by the successful 2016 White and non-White vote by county project:
https://uselectionatlas.org/FORUM/index.php?topic=258343.0

The values in the map are residuals from a linear regression using racial percentages as measured by the Census in 2010 to predict the logit(two-party 2016 presidential vote) for each 2016 congressional district. These residuals were then transformed again by the logistic function to limit them so that the minimal possible residual is -.5 and the maximal possible residual is .5 . Congressional districts with values closest to zero are those with two-party presidential vote shares best predicted by the model.

The only obviously ridiculous aspect to this map is that, from the regression, Asians came out as a 98% Democratic group as they generally live in very Democratic Congressional Districts, so very Asian districts are much redder than they should be.

The most Hillary-voting districts relative to their racial demographics are WA-07 (Seattle), Washington, DC, and NY-12 (Upper East Side). The most Trump-voting districts relative to their racial demographics are HI-01 (due to the aforementioned Asian issue) and more realistically, TX-13 and TX-11.

It also appears Trump actually won the White vote in a Congressional District in Massachusetts (MA-03; the usual guideline for if Trump has won the White vote is a residual of about .147 .151 or lower). Romney must have won the White vote in this district as well, as Trump performed worse there than Romney. Romney also won the White vote in MA-08.

[100% pro-life no matter what]

Interesting- it looks like Trump won the white vote in every district in the South other than GA-5 (Atlanta) and VA-8 (Arlington), if you count that as the South.  There is one more- NC-4 (Durham) which is right on .147, so it may have gone either way.

[mieastwick]

Looking at New Hampshire, the guideline is more like at .151 , so the White vote in NC-04 most likely went for Trump, if very narrowly.

[100% pro-life no matter what]

Complete list of places where Clinton won the white vote:

GA-5 (Atlanta)
VA-8 (Arlington)
DC-AL
PA-1 (Philadelphia)
PA-2 (Philadelphia)
PA-14 (Pittsburgh)
NY-10 (NYC)
NY-7 (NYC)
NY-12 (NYC)
NY-13 (NYC)
NY-15 (NYC)
RI-1
MA-1
MA-2
MA-4
MA-5
MA-6
MA-7 (Boston)
MA-8
MA-9
NH-2 (rural NH)
VT-AL
ME-1 (Portland)
MI-12 (Detroit)
IL-4 (Chicago)
IL-5 (Chicago)
IL-7 (Chicago)
IL-9 (Chicago)
WI-2 (Madison)
WI-4 (Milwaukee)
MN-4 (Twin Cities)
MN-5 (Twin Cities)
CO-1 (Denver)
CO-2 (Boulder)
WA-2
WA-7 (Seattle)
OR-1
OR-3 (Portland)
CA-2 (includes part of San Francisco area)
CA-5 (Napa)
CA-12 (San Francisco)
CA-13 (Oakland)
CA-18 (Silicon Valley/San Jose)
CA-20 (Central Coast)
CA-28 (Los Angeles)
CA-33 (Los Angeles)
CA-37 (Los Angeles)
CA-40 (Los Angeles)

Even in California, it only seems that Clinton won the white vote in 10 of the 55 districts, all in the Bay Area or Los Angeles County.  Nationwide, Trump won the white vote in 390 districts and Hillary won it in 47.  The Hawaii districts are both listed as having an overwhelmingly Republican white vote, but I am skeptical of that due to the (possibly wrong) assumption about Asians in the original post.

Trump won the white vote in every district in the following states:

[mieastwick]

The thing with California, though, is that you quickly get into the Asian problem mentioned above, in which Asians came out as a nearly 100% Democratic group in the model due to them generally living in heavily Democratic districts, so the racially adjusted vote in heavily Asian districts is almost certainly too pro-Trump. The most obvious examples of this are CA-17 and CA-14, where it's easily imaginable Hillary actually did win the White vote, but we don't know that for sure because it's not easy to determine how Asians voted.

Also, MI-12 isn't Detroit, it's Dearborn+Ann Arbor. Very curious mix of Arabs, White Working Class, and the extremely well-educated.

Thank you for your remarks, ExtremeRepublican.

[nclib]

Great project, though not totally as accurate in the Deep South as in MS all CDs have a high black population (and MS blacks are as Dem as nationally if not more so), though whites in MS CDs are likely the most Republican in the nation, even moreso than TX-13 or TX-11.

Interesting- it looks like Trump won the white vote in every district in the South other than GA-5 (Atlanta) and VA-8 (Arlington), if you count that as the South.  There is one more- NC-4 (Durham) which is right on .147, so it may have gone either way.

Looking at New Hampshire, the guideline is more like at .151 , so the White vote in NC-04 most likely went for Trump, if very narrowly.

I live in NC-4 and the current district is more Chapel Hill and Raleigh than Durham. Durham whites are slightly more liberal than Raleigh whites, so HRC may have carried whites in the 2003-2012 NC-4 or the 2013-2016 NC-4. Either way it's very close. And good to see that we are the second most Democratic whites that are surrounded by the South.