Thursday, November 3, 2016

Pols and Polls - the good, the bad, and the ugly

Summary for Nov. 3: Some forecasts for a Clinton win are optimistic: 98% (Huff) and 86% (Upshot). Both have been more favorable than Nate Silver’s 538 model - now forecasting a 67% chance of a Clinton win, 3.2% advantage in the popular vote, and 295 vs. 242 electoral votes. One reason for optimism is that Clinton leads in key states (Huffpollster). Moreover, compared to Obama’s standing at this point in 2012, Clinton is looking good. In 2012 Obama’s 1.5% lead was good enough for a 91% forecast from Silver’s 538 model; Clinton’s 6% lead in the HuffPost chart prompts the 98% forecast (Huff Post model - see below). Oh, and about those shadow Trump supporters? “Donald Trump has insisted for months that polls are failing to capture the breadth of his support because some of his backers won’t admit it to a pollster over the telephone. He’s wrong.” Read about why the shadow voters are shadows only in the last item below.

Huffpollster: FORECAST UPDATE - Clinton still has a 98 percent chance of winning the presidency. Democrats are up to a 71 percent chance of getting 51 or more seats in the Senate; Republicans’ chances have dropped to 5 percent. The shift in the Senate comes from polls showing increasing leads for Democratic candidates in Illinois, Pennsylvania and New Hampshire. Wisconsin and Indiana will be races to watch ― Democrats maintain the lead in both states, but their margins have been decreasing in recent weeks. [Presidential forecast, Senate forecast]

NYT Upshot gives Clinton a 86% chance of winning.

New York Times/CBS poll has a Clinton with a narrow lead

… Most voters say their minds are made up and late revelations about both candidates made no significant difference to them, according to the latest New York Times/CBS News poll released Thursday.

Five days before Election Day, the margin between the candidates is narrow, with 45 percent of likely voters supporting Mrs. Clinton, the Democratic candidate, to 42 percent for Mr. Trump, the Republican nominee. The difference is within the poll’s margin of sampling error. Gary Johnson, the Libertarian candidate, has the support of 5 percent of likely voters, and the Green Party nominee, Jill Stein, takes 4 percent.

[Of note:] At this point in the 2012 campaign, Mr. Obama and Mr. Romney were deadlocked in polling averages, and Mr. Obama went on to win the election by a four-point margin.

POLLS SHOW CLINTON STILL LEADING IN KEY STATES - HuffPollster: “New polls from CNN/ORC in four states started the day on a mixed note for Clinton. Two of those showed Trump leading by 6 points in Nevada, and by 5 points in Arizona. Both are more pro-Trump than most other recent polls have been in those states, but Arizona is a traditionally red state, and Trump’s lead in Nevada has wavered back and forth a few times this year. CNN/ORC showed Clinton up 2 points in Florida, and by 4 points in Pennsylvania ― both well in line with recent trends. Quinnipiac University released a set of four polls Wednesday afternoon, giving Clinton the edge by 1 point in Florida, 3 points in North Carolina, and 4 points in Pennsylvania. The only bad news for Clinton in that set: She trails by 5 points in Ohio…. No poll on the HuffPost Pollster chart has ever shown Trump leading in [Wisconsin]. Marquette didn’t change that trend on Wednesday ― the poll showed Clinton leading by 6 points. The pollsters did note some effects from Friday’s FBI announcement about more emails pertinent to Clinton’s private server, but she had a strong enough lead that those effects didn’t change the outcome.” [HuffPost]

Nationally, Clinton is in a better place than Obama was in 2012 - More from HuffPollster: “In theHuffPost Pollster chart, Obama widened his lead to just 1.5 percentage points over Mitt Romney nationally ― but that was good enough for Simon Jackman’s HuffPost model and Nate Silver’s FiveThirtyEight model to give Obama a 91 percent chance of winning. Obama’s odds in the states were good enough that he was predicted to get 332 electoral votes (which is what he got). Clinton is currently about 6 points ahead of Trump in the HuffPost Pollster national chart, and our predictions give her 341 electoral votes. Assuming the polls haven’t become fundamentally more error-prone than in 2012, a 98 percent certainty of her election makes sense ― especially when compared with Obama’s 91 percent certainty with only a fraction of that lead. Maybe there are reasons to think 2016 is more uncertain than 2012 was; it certainly has been an unconventional year. But we’re trusting that pollsters have captured opinions reasonably well on average.” [HuffPost]

FEW TRUMP SUPPORTERS ARE TOO EMBARRASSED TO ADMIT THEIR SUPPORT - Steven Shepard: “Donald Trump has insisted for months that polls are failing to capture the breadth of his support because some of his backers won’t admit it to a pollster over the telephone. He’s wrong. According to a new study commissioned by POLITICO and conducted this past weekend by Morning Consult, a hidden army of Trump voters that’s undetected by the polls is unlikely to materialize on Election Day. The study… showed only a slight, not-statistically-significant difference in their effect on voters’ preferences for president…. Overall, the POLITICO/Morning Consult study released Thursday showed Clinton with a 5-point lead among voters interviewed over the phone, 52 percent to 47 percent. The race was modestly closer in online interviews: Clinton’s share of the vote ticked down a single point, to 51 percent. Trump’s support inched up a point, to 48 percent…. A previous Morning Consult study late last year found a Trump ran about five points better among Republicans and GOP-leaning independent voters in online interviews than he did over the phone. But that same effect is not carrying over to the same extent in the general election.” [Politico]

No comments:

Post a Comment