GA-SurveyUSA: Trump +4

[cMac36]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

[Fmr President & Senator Polnut]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

I can't find 2012, but the 2008 exit poll had the electorate as 30% black.

[cMac36]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

I can't find 2012, but the 2008 exit poll had the electorate as 30% black.

A-ha.  In sample of 500 then 150 will be Black.  Difference between 14% and 7% would be 9 Respondents?  Thnx

[Adam Griffin]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

To be specific (since GA actually records both voter registration and turnout figures by race), the black share of the electorate was:

2008: 30%
2010: 28%
2012: 30%
2014: 29%

[Fmr President & Senator Polnut]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

To be specific (since GA actually records both voter registration and turnout figures by race), the black share of the electorate was:

2008: 30%
2010: 28%
2012: 30%
2014: 29%

Looking at the 08 exit polls, there are three potential gold-mines for Clinton.

College-educated whites, McCain beat Obama here by nearly 50% and they're almost 1/3 of the vote.

White women - McCain also beat Obama here by almost 50% and I'd be extremely interested to see how Clinton does with people who identify as moderates (because we all know 'Independent' as a political identifier means next to nothing now).

[Adam Griffin]

The best thing about all these close GA polls is they probably mean Clinton is up meaningfully in NC.  Still think Trump wins GA by 5ish in the end.

I don't know. It doesn't necessarily mean that she's winning. At best, it'd probably be a tied race there if we go off of the past three elections.

The funny thing is that after all of that investment (at this point, hundreds of millions of dollars in the aggregate), organization and whatnot poured into NC in 2008 & 2012 (not to mention 2014), it has only moved to the left one point more than GA since 2004 (NC was 2 points more Democratic than GA then; 3 points more Democratic in 2008 & 2012). I also doubt that sheer demography is going to send NC careening too far away from that tracking this cycle, either: if anything, GA's demographics have been shifting faster than NC's.

Of course, Democrats made the decision to invest in NC most likely because of a greater number of white Democrats, liberals and college graduates there at the time. They've been largely mobilized; GA is still an untapped gold mine when it comes to the college-educated and unregistered voters, and I bet it'd move at least one more point to the Democratic relative to where it is now if an equivalent amount of investment were injected.

[dspNY]

The best thing about all these close GA polls is they probably mean Clinton is up meaningfully in NC.  Still think Trump wins GA by 5ish in the end.

I don't know. It doesn't necessarily mean that she's winning. At best, it'd probably be a tied race there if we go off of the past three elections.

The funny thing is that after all of that investment (at this point, hundreds of millions of dollars in the aggregate), organization and whatnot poured into NC in 2008 & 2012 (not to mention 2014), it has only moved to the left one point more than GA since 2004 (NC was 2 points more Democratic than GA then; 3 points more Democratic in 2008 & 2012). I also doubt that sheer demography is going to send NC careening too far away from that tracking this cycle, either: if anything, GA's demographics have been shifting faster than NC's.

Of course, Democrats made the decision to invest in NC most likely because of a greater number of white Democrats, liberals and college graduates there at the time. They've been largely mobilized; GA is still an untapped gold mine when it comes to the college-educated and unregistered voters, and I bet it'd move at least one more point to the Democratic relative to where it is now if an equivalent amount of investment were injected.

Obama won NC by 0.4% and lost GA by a little over 5 in 2008
Obama lost NC by 2 and GA by 8 in 2012.

GA is about 6 points right of NC, not 2 or 3. So a slim deficit in GA is a likely lead in NC. If Dems made the proper investment then GA could vote more like NC

[Adam Griffin]

The best thing about all these close GA polls is they probably mean Clinton is up meaningfully in NC.  Still think Trump wins GA by 5ish in the end.

I don't know. It doesn't necessarily mean that she's winning. At best, it'd probably be a tied race there if we go off of the past three elections.

The funny thing is that after all of that investment (at this point, hundreds of millions of dollars in the aggregate), organization and whatnot poured into NC in 2008 & 2012 (not to mention 2014), it has only moved to the left one point more than GA since 2004 (NC was 2 points more Democratic than GA then; 3 points more Democratic in 2008 & 2012). I also doubt that sheer demography is going to send NC careening too far away from that tracking this cycle, either: if anything, GA's demographics have been shifting faster than NC's.

Of course, Democrats made the decision to invest in NC most likely because of a greater number of white Democrats, liberals and college graduates there at the time. They've been largely mobilized; GA is still an untapped gold mine when it comes to the college-educated and unregistered voters, and I bet it'd move at least one more point to the Democratic relative to where it is now if an equivalent amount of investment were injected.

Obama won NC by 0.4% and lost GA by a little over 5 in 2008
Obama lost NC by 2 and GA by 8 in 2012.

GA is about 6 points right of NC, not 2 or 3. So a slim deficit in GA is a likely lead in NC. If Dems made the proper investment then GA could vote more like NC

I'm not talking about margins: I'm talking about the share of the vote for Democrats. NC was roughly 2 points more Democratic in 2004 and 3 points more Democratic in 2008/2012. At the end of the day, though, margin/DPI is mostly a case of "six versus a half-dozen".

[Fmr President & Senator Polnut]

I'd really be interested in seeing polling out of SC.

[heatcharger]

I'd really be interested in seeing polling out of SC.

This. Though South Carolina doesn't have the same growth in the tech sector where lots of white-collar minorities are coming into the state like NC and GA do. I'd think Trump is probably doing fine, maybe up 9.

[HillOfANight]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

To be specific (since GA actually records both voter registration and turnout figures by race), the black share of the electorate was:

2008: 30%
2010: 28%
2012: 30%
2014: 29%

Those numbers are probably understated. Landmark expects 32% black in 2016 (though they estimate it was 33.5% in 2012)/

A lot of voters are "unknown" or don't fill in that category. Voter file research indicates it's mostly black/biracial/nonwhite voters
http://www.peachpundit.com/2014/10/03/new-poll-deal-carter-headed-runoff/#comment-389632
http://www.peachpundit.com/2014/10/27/dueling-early-voting-data/#comment-392120

[dspNY]

I don't see Trump getting 14% of the black vote, but I think +4 is about right.

In the subdemos within these polls, what is Black vote in GA, 20% of total electorate?, the difference between 7% and 14% is a handful of respondents, make little of it yes.

To be specific (since GA actually records both voter registration and turnout figures by race), the black share of the electorate was:

2008: 30%
2010: 28%
2012: 30%
2014: 29%

Those numbers are probably understated. Landmark expects 32% black in 2016 (though they estimate it was 33.5% in 2012)/

A lot of voters are "unknown" or don't fill in that category. Voter file research indicates it's mostly black/biracial/nonwhite voters
http://www.peachpundit.com/2014/10/03/new-poll-deal-carter-headed-runoff/#comment-389632
http://www.peachpundit.com/2014/10/27/dueling-early-voting-data/#comment-392120

No, I think the 28-30% numbers are correct. We could get 32% this November but it would take a massive turnout operation

[DrScholl]

I'd really be interested in seeing polling out of SC.

That's what I was thinking about too.

[Adam Griffin]

To be specific (since GA actually records both voter registration and turnout figures by race), the black share of the electorate was:

2008: 30%
2010: 28%
2012: 30%
2014: 29%

Those numbers are probably understated. Landmark expects 32% black in 2016 (though they estimate it was 33.5% in 2012)/

A lot of voters are "unknown" or don't fill in that category. Voter file research indicates it's mostly black/biracial/nonwhite voters
http://www.peachpundit.com/2014/10/03/new-poll-deal-carter-headed-runoff/#comment-389632
http://www.peachpundit.com/2014/10/27/dueling-early-voting-data/#comment-392120

Yes, that's probably about right. At any rate, exit polls in 2008 & 2014 seemed to show comparable numbers; they're not perfect, but self-identification via exit polling is worth something. At the end of the day, the numbers as measured through this method should remain mostly proportional to one another from cycle to cycle, so you can still compare it through this method.

I guess it depends on how the cookie crumbles statewide and what methods specifically Rountree used to calculate these figures. They probably have access to a more refined voter file, but even then...there can be hiccups unless the information is collected directly from voters.

I've noticed that there is a correlation between the percentage of Latinos and Asians in a county and the percentage of voters who are Unknown/Other.  Just eyeballing data in the past, I came to the conclusion that in most counties where the non-white, non-black population is below the state's average, the unknown/other figures are below the state's average; in counties where non-white, non-black people make up more of the population, the number tends to be above the state's average.

My county, for instance (Whitfield), had something like 10% unknown/other via SoS the last time I checked, which is higher than most counties. It also has a 32-33% Latino population. You'll see similar trends in places like Gwinnett and Hall where there are larger Latino/Asian populations, but not so much in many of the blacker and more rural counties in Georgia.  

This may in fact have nothing to do with race directly, and rather, reflects the relatively recent registration growth of Latino and Asian voters (as was mentioned by Rountree, GA no longer requires race when registering to vote). I don't know the figures off the top of my head, but I'm pretty sure that a clear majority of these non-white voters weren't registered or voting a few years ago.

[HillOfANight]

When early voting season starts, Georgia publicly hosts the early voting files, and you can see names and addresses of people. If you have the $, you can buy another file to link race, etc.

Anyway, it was interesting to see that Korean sounding names were like 3/4 voting in the Democratic primary, while Vietnamese were the opposite. Mark and other groups like him probably "guess" the file that way, in addition to supplemental research (maybe direct voter contact in some cases).