That’s because you have to account for independents. Sometimes, especially when they are angry at the incumbent party, they swing huge towards the opposition. Usually this is a midterm phenomenon, but ABC is seeing in their poll a rather large swing: +16 Trump among independent voters. If ABC’s polling is accurate, all Trump needs to win is for Republicans to finish coming home.
Like 2012, though, then we get into arguments about people who self identify as “independent” and what that really means these days. For example, I can easily see an alt right voter who doesn’t think the Republican Party goes far enough on racial issues describing himself as such.
That’s overthinking it though. It’s not unusual for the winner of the election to lose independents by 5 points or so. Losing them by double digits is fatal.
Nate notes today that Obama at this time had a 2 point lead or better in enough states to give him 332 electoral votes. Clinton’s lead is 2 points or more in only enough states to give her 272:
That’s why Obama’s smaller lead in the national vote was more secure than Clinton’s is now, along with the undecideds problem. Obama could afford to lose some states. Clinton can’t even lose one at this point where she’s currently ahead.
ABC polling today has it down to +7 Trump among independents.
Support among independents collapsing? 
FWIW Clinton’s support among Blacks ranges (of the two-party vote) from 93% in the most recent ABC tracker to 97.6% in the most recent CBS/NYT poll (83%C to 2%).
asahi’s claim of Clinton being weaker than Obama among African Americans is just not supported by the data available.
I’d be shocked if Clinton got as many African-American votes as Obama did. She should have lower turnout, and lose a couple of points compared to Obama.
“Should”?
None of us “should” be caring about "should"s. But please do place your prediction of what you think will be in this predict turnout and share by demographic group thread.
The polling is what it is. Of the two party vote Clinton is running as good or better as Obama did in 2012. Whether she should or not. (Really though, do you think that Trump “should” do better with Black voters than Romney did?)
Will turnout be off? Personally I doubt to any significant degree.
Your own predictions are what i said!
Regarding the above discussion about demographics, what I’m curious about (when I’m not in a corner bawling my eyes out in a fetal position) is: Will the increased number of Hispanic voters–seemingly motivated to show up at the polls due to Trump’s demonizing rhetoric–make up for the (presumed) lower enthusiasm/turnout on the part of black voters when compared to 2008/2012?
Oh addie? The CBS/NYT poll has independents at T+4.
choie, my thoughts on the Black voter side of that question already stated, allow me to offer some comment on the Hispanic voter side. There has been huge increases of Hispanic voters voting early. That could be just the same Hispanics who voted last time taking advantage of early voting this time but it could be a sign of increased turnout overall.
That Washington Post/Univision poll runs at 67% Clinton to 19% Trump, or put another way 78% of the two party vote polled for Clinton. Obama got 69% of the Hispanic vote.
+4 and +7 are more consistent with a slight Clinton lead. Independents +16 would surely sink her.
Just found out this great update page on 538 with one or two poll updates every hour including the pollster ratings and an updated forecast. Probably the best way to keep up with the insane avalanche of polls we are getting.
Anyway the highlight of the day was undoubtedly the massive release of Ipsos state polls. On the whole good news for Hillary especially with respect to her blue wall in the two most vulnerable states:NH +6, CO +8
Also a remarkably +12 in Nevada which may be a data point backing the early-voting analysis showing her leading in a big way.
After eight years, there are more of them. You may be familiar with the procedure.
Sorry, haven’t seen Storks yet.
If the issue were just the relative appeal of Clinton vs Obama, you’d be right. The difference here is Trump, who has so blatantly been courting the white nationalist contingent that most black voters will vote for Clinton to stop him from gaining power. This is also the force driving the Hispanics to vote early.
We’ll see. As much as the CW says that people often vote against a candidate rather than for one, I believe that the only thing that drags unmotivated voters to the polls is a candidate who inspires. I just don’t think they are going to be turning out at Obama levels and it’s possible that Trump could do better than Romney among African-American voters. Or at least for Clinton to do worse(since there are 3rd party options available).
We don’t have enough good data in some states, which is the problem. What is absolutely clear, though, is that Clinton’s once safe lead is now no longer safe. In terms of the polling data as it applies to the electoral college vote, the race is getting into the danger zone.
Both candidates have opened up the map a little. Clinton’s map and her options for winning have been varied for some time. The difference now is that Trump seems to finally have different paths to victory himself.
Not a fan of the Ipsos poll as it fluctuates too much. One week it will show Clinton up by 15; the next it’ll show her virtually tied. Maybe the trick is to average weekly polling data over a 2 or 3 week period.
It’s an interesting contrast, in 2012 state polling was excellent so Silver’s model tended to prefer it to the national polling, which showed a much closer race. And which also led a lot of Democrats to believe that their electoral map was almost unassailable. But this time around the state polling is not good, so Silver’s model has to project from the national polling, which means more uncertainty.
Vox posted an article about how aggregators may have made polling more predictive, but it also made individual pollsters stop wanting to waste their money. So we’ve got less polling in 2016 than in 2008 and 2012.
And that’s another thing that I wonder if the other aggregators are failing to take into account. But we’ll find out. You don’t just have to judge Silver by his probabilties, but also by his state forecasts. Silver has Clinton at 272 electoral votes, Wang at 316. That’s a rather wide difference to judge them by even if Clinton wins.
Keep an eye on Georgia next Tuesday.
I had written it off as a Trump state and no doubt, given the history of voting there, Trump will probably still claim it. But a little closer analysis reveals to me why Georgia could actually be interesting after all, and it comes down to two factors: rapidly changing demography, and education.
-
Most of us are familiar with the fact that Georgia’s demographics are changing. This is one reason why it has been ground zero for vote suppression efforts in various voting precincts. But it’s not just African Americans who are moving to GA; there are a growing number of Latinos there as well. Metro Atlanta dominates Georgia’s population and it is pretty diverse.
-
I did not know this but Georgia actually has a pretty educated population as well. Most of us think of the backwoods and places like Forsythe County. But Georgia ranks in the top half of states with highest percentage of people with graduate degrees. And we know that Trump tanks with those who like to ‘better themselves.’
The South in general has become a great place for people and businesses to move to. Low taxes, cheap real estate, warm weather attracting graduates to fill those jobs.