Editorial Context: National Election Pool (NEP) is what AP, CNN and the major news services rely upon for exit polling data. It found AAPIs had a huge swing from 2012 and were the group that Republicans made the largest gains with, coming out 52D-48R (revised by them to 50D-49R.) This is slightly controversial as it doesn’t capture the Limited English Proficiency AAPIs in our population. I will also add that some of the debate around how pollsters got 2014 so wrong, some say is because pollsters pegged too closely to the US Census American Community Survey and therefore did not capture the midterm audience accurately (tends to be less diverse and older than voters in presidential years.) Also it should be noted that this in-language poll was taken pre-election.
Here are a few links about how and why the polls were so wrong. Also any errors/ ommisions in typing were mine as I was taking notes on the call. For pretty graphics look at the presentation. -Caroline
Asian American Decisions did an in-language election eve poll of 2014 AAPI voters found here: http://vote.18mr.org/static/AAD_EE_Presentation.pdf
All findings from www.asianamericandecisions.com & Taeku Lee
1150 voters in 6 languages versus
National Exit Poll was of 129 voters in English, margin of error could be as high as 9%.
2-1 margins in favor of Dems 61-34%.
Although AAPIs have certain progressive values, most AAPIs do not identify as Democrats.
Texas Gov race: Asian Ams went 48-48 for Davis and Abbott.
VA Sen: AAPI 68-29 for Warner. Independents went 65-30 for Warner. Vote margin for Warner was estimated 25-30k votes, larger than Warner’s margin of victory.
Economy/ Jobs 32% (less prominent than 2012-2008)
Education Schools 22%
40% of AAPIs indicated healthcare was either the most important issue or one of the top issues.
Immigration was lower down the list. 22% said immigration was not important in voting.
Dem ID 97%
Republican ID 83%
58-33 (15 pts higher than general population via Gallup)
Dem vote share went from 73% in 2012 to 66% in 2014
45I 14R 41D
40I 22R 39D
Partisan split on Obamacare
52 favorable 37 unfavorable
60 support 26 oppose
Rep ID 47 support (more than oppose)
Rep ID: 57 support 32 oppose
Indep ID 67-24 support
Dem ID: 86-10 support
Rep ID: 38-48 oppose
Ind ID 56-30 support
1 in 4 rely on ethnic media
1 in 6 rely on Asian language ethnic media
40% of sample was in language polling
56% were not contacted about voting
Contacted about voting by:
22 D only
11 R only
39 both parties
21 community org
33% said they were undecided or refused to give a response
2 not heard of
13 don’t know
GOP hopefuls have high unknowns, unfavorables.
25 fav 46 unfav 11 unheard 18 don’t know
Christie had 43% unfavorables.
Q & A:
Immigration sounds like it’s less important as an election issue than for the Latino community. Not single most important issue. 2014 is not a big change for AAPIs.
High no opinion rate that shows up in most polls. Party ID high portion of AAPIs either ID as independent or say they don’t know how to answer that q. Also for 2016 a high undecided portion. Tend not to share vote choice and partisanship in phone surveys until close to closing days.
Immigration linked to pathway to citizenship, were there other options provided for immigration reform? Big change between 2008 and 2012 in terms of big increase in AAPI support for immigration reform with pathway for citizenship. in 2008 it was 1 in 3. 2012 increased to majority. 2014 still a majority.
Since this is partnership between Asian American Decisions and Latino Decisions, any additional info on Young Kim & Janet Nguyen, 2 AAPI Republicans who won over Latino Dems in Orange County & implications?
Oversamples CA TX VA representative of AAPIs throughout the state. Didn’t ask horserace questions but will look into more info.
If you are only tapping English-only AAPIs, then that’s only one slice. NEP weighting moved from 52-48 D then they moved it to 50-49. This survey is pegged to Census weights. Still waiting to see why NEP they changed the weight. NEP is low sample size and they haven’t designed a way of getting an accurate picture of how AAPIs voted.