Other Race/Ethnicity

[Builder Refused]

I was going through exit polls of midterms on Wikipedia to compare Asian and Hispanic vote patterns for the incumbency bias. I noticed in the most recent midterm exit poll this group actually split R+13. So, checking the source on CNN, it's a bit wonky like Native American +21 which I guess is due to too small a sample. So, they constitute 1% of that 3%, based off that who's the other 2% in other?

[Del Tachi]

I think "other" is checked by a lot of Middle Easterners, non-White Latinos (i.e., Haitians and Filipinos) and South Asians. 

[Oryxslayer]

I think "other" is checked by a lot of Middle Easterners, non-White Latinos (i.e., Haitians and Filipinos) and South Asians. 

Let's actually go on a bit of a deep dive here. Georgia tracks voter registration by self-described race/ethnicity. The Groups are White, Black, Asian, Hispanic, Native Americans, Others, and Unknowns. The Other category here is actually rather small, so we will merge it with Unknowns which in this situation is more than five times larger and more befitting of a national Other response category. Overall, we are looking at ~12% of the state's RV.

The county with the largest amount of registered voters in this category is Forsyth. This makes sense with the above statement, given the high number of pan-Asian migration into those suburbs. The results overall do have a small correlation with Census Asian respondents, but not a significant one: White Rurals still have 8%+ respondents. There is also a correlation with Black voters, but once again minimal, the correlation vanishes in the Black Belt. Growth is a strong correlator which again makes sense, since those who recently got registered are unlikely to have a full demographic profile in the system, but shrining areas still have plenty of respondants.

What we need to accept therefore is that Others covers all these things: minorities who don't want to give away information the Government, Libertarian or Privacy types for the same reason, new voters, mixed-race voters, and voters of Demographic groups not included in the census catergories.

[kwabbit]

Just based on the actual likely demographics of the other vote, it should be a Democratic group. Seems like choosing other makes someone much more likely to be Republican, in the same way that people who say "I don't see race" are often conservative. I imagine that the GOP wins a huge margin among Whites who choose other.

Before the election, Insider Advantage (which is kind of dubious overall) had Republicans getting huge margins in the other vote in PA. That might be because Whites who choose this option are more conservative and there is a smaller Black population in PA than GA. The same attention wasn't drawn to IA other crosstabs in GA.