Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
The Poll Perplex Returns
If you turn to Real Clear Politics today and look at the latest polls focused on a head-to-head conflict between Trump and Hillary, you will once again be perplexed. IBD/TIPP, which was the most accurate of the polls in 2012, has it all tied up. ABC/Wapo has Hillary ahead by a point. The LA Times has Trump up by six, and NBC/SM has Hillary up by seven. That is a thirteen-point spread. One of these last two polls is way off. Both may be. It could be all tied up.
If you go to the same site and look at the latest polls focused on a four-way race, IBD/Tipp, ABC/Wapo, and Rasmussen have it all tied up, and NBC/SM has Hillary up by six. The LA Times does not do a poll covering the race in this fashion.
This, too, would suggest a very close race. In normal years, the odds would be good that on 8 November we would see results more or less consistent with this. But this is not a normal year, and polling is as much an art as a science. Pollsters adjust what they learn in an attempt to make their sample reflect the public at large. When the public is in motion, when the old rules to do not apply, when Democrats are to an unusual degree apt to vote Republican and vice-versa, when turnout is exceedingly hard to predict, they are apt to stumble, and surprises can very easily present themselves.
Update at 4:13 p.m. EST. Real Clear Politics asserts on the basis of the polls that, if the election were to be held today, Hillary would win 273 electoral college votes and Trump, 265.
There is motion in the numbers that the pollsters are producing. Except in the case of NBC/SM, the motion is in Trump’s direction. In 1980, there was a sudden avalanche in Reagan’s direction at the end as voters pondered just how bad Carter had been and puzzled over the question whether Reagan was really a madman in the manner in which the mainstream media strongly suggested. That could happen this year. The polls show that the vast majority of Americans are aware that the mainstream media is in the tank for Hillary. That suggests that they are apt to discount what they are told by Pravda-on-the-Airwaves. And people do get annoyed when they think that they are being manipulated. So, who knows?
If you are a prognosticator, one thing to look for is signs of panic on the part of one campaign or the other. They all do internal polling, and they spend larger sums on it than do the outfits mentioned above. Knowing what is going on in the electorate is a life-or-death matter for the campaigns. They also poll individual states — especially, the ones that really matter — much more intensively than anyone else.
One final observation. Sean Trende, who is an able prognosticator, has an article out on early voting. He thinks it a poor predictor of final results. So, go figure.
Published in General
IMHO I think you have to look at some of these indicators:
African American vote – how big and for whom? Trump is doing well here with possibly 15% or more with low turnout.
Democrats voting for Trump – Trump is doing very, very well here, as much as 20-25% in some polls.
Independents swing where? Trump again is doing well with generally +10 to +20.
RCP is a lagging indicator. It tells you what polls revealed up to two weeks in the past.
I find the Remington Battleground County polls interesting. They have been polled right along and are canaries the the coalmine.
Here is Luzerne County, PA where Dems have won in the last four general elections and a state where Trump is polling behind:
http://axiomstrategies.com/abc/luzerne/
http://axiomstrategies.com/abc/
Well, WI was apparently ‘over’ but based upon the Presidential ads running and a steady stream of both Presidential candidates, their respective VP’s and surrogates – I’d argue this state could go either way.
This pretty reluctant Trump voter will happily go to vote for him, my wife won’t but will support the GOP down ticket. I’m working on it.
What is your African American number based on? In polls looking at the preferences of African Americans Trump was polling in at something like 3% if I recall. I doubt African American turnout would get so low as to boost that to 15%.
Quite frankly, my hope is that these surveys become so damned unreliable that the media just drops them. They serve no real useful function except to the candidates who change positions publicly for votes and then “reconsider” when it comes time to make policy.
I don’t understand this business of weighting the results to “reflect the public at large.” Using what criteria? Demographics? Past voting behavior? Party registration? Professed political leanings? Any of these potentially introduces more error than it corrects.
I’m no statistician, but when I took statistics back in my psych-student days, I learned that it’s all about a) sufficiently randomizing your sample and b) having a large enough sample. If the pot of soup is sufficiently stirred, and the ladle is big enough, you can rely on the ladle dipping out a representative sample.
If the pollsters aren’t getting sufficiently large samples, or sufficiently randomized samples, then they’re starting with bad data. Anything that follows from that, it seems to me, is of dubious value.
Nate Silver still gives Hillary 70.8% chance of winning, Trump 20.1%.
I think it is still important, as I’ve said on a couple of these threads, to look at the battleground states vs. the national poll. Same thing can work in both directions. If a candidate is really popular in California and New York, it will change the national polls but make no difference in the outcome. Same with Texas and any other dark Red state.
“NBC/Wall Street Journal poll currently has Trump polling at 7 percent amongst African-American voters, while Hillary Clinton is currently at 81 percent.”
SC poll with Trump getting 23% AA vote:
https://drive.google.com/file/d/0B4lhKxf9pMitUjZhbDJwVGQteTA/view
Well, Hillary is restarting ads in Blue states, and Trump is spending time in lean Blue states…
If, IF, true. Trump in a blowout.
But that might not be the big story, look at that poll from North Carolina. Trump is getting 23% of the DEMOCRAT vote, Hillary is getting 5% of the GOP vote. You can find similar results elsewhere.
Texas only seems like a dark red state because our state Democratic party is pretty terrible – same story as with Florida really.
There are plenty of Democratic voters here which is why things like Battleground Texas should be taken seriously.
NC is still the South how many legacy Democrats are left? It used to be in Southern States no one was a Republican. What was the case back in 2008 and 2012?
Also as to the point about the African American vote. I have been seeing reports indicating their turn out is down.
The overall polls are one thing; the polls that reflect the electoral college are far more important. And if I’m not mistaken, Hillary has enough strength in those battleground states even today. Trump would have to pull a surprise in NC, PA, VA, or a combination of a couple states to even have a chance. As of the moment, any movement in his direction isn’t enough.
But now they’re talking about a Reagan-style last minute wave, which my instinct ascribes to wishful thinking. Who knows?
I sincerely think the best rationally-expected outcome at this point is that Hillary wins, but by a terribly thin margin, where the GOP barely keeps the Senate, and the House stays solidly Republican. The GOP control of the Congress is crucial, because it means that the committees maintain their investigations, the FBI keeps investigating, and for all the “historic” significance of a woman being elected, the Democrats can’t stop the GOP from blocking Hillary’s every move. In fact, the Democrats will avoid the White House instead of fighting for Hillary, because they won’t want that Clinton Stench on them (it’s so hard to launder out!).
The S&P 500 market predicts…
NC is Purple. Until recently the governor was almost always a Democrat. We had a Democrat Senator. My congressman was a Blue Dog. States been trending a little more Red recently, but heavy migration from NY, NJ, OH, Penn etc into NC.
I wanted to “like” this twice.
The criteria is developed by each research firm. For demographic weighting, many pollsters use Census data. Where the data is available, pollsters often use a model of likely voters that incorporates past vote history and other variables predictive of voting behavior.
Unfortunately, it’s not possible to get a theoretically pure random sample because some people just can’t be reached. Even with live phone interviews, which is the time-tested method, there are response biases. You tend to get more older voters and more males than females, to name a few. Some weighting is necessary just to compensate for that known bias. Quotas can also be used to ensure your sample comes back relatively balanced.
You’re right to be suspicious but it should be for different reasons. Personally, I think the criteria should be to reflect likely voters using the most accurate predictors. Usually the weighting criteria isn’t made public because of its proprietary nature, so take it with a grain of salt.
The surveys are not to tell us the truth. They are a way to sell a narrative, which means coverage, advertisers, etc.
The media is a business. Entirely fictitious surveys would be pressed into service if there were no actual one. And that may be what is happening right now.
I think that’s right. I’d say the same goes for Montana. We’ve always considered Montana to be deep red, but there has been a massive influx into places like Missoula, Helena, Kalispel, Bozeman of liberals from out of state. They are beginning to outnumber the conservatives, although the state is still generally red. Of course, even conservatives in Montana favor things like farm subsidies, which are anti-conservative (though, the economic ignorance and pro-America, which is badly mistaken in economics, attitude, will likely tilt people in Trump’s favor).
I just filled out my WA state ballot. Very depressing. I voted “no” on everything, picked every Republican candidate down-ballot, made a few choices for judges who I personally know, and ignored all of the “non-partisan” offices with liberals running against other liberals. I already know that virtually every issue will go the opposite of the way I voted.
Sure, I recognize that it’s probably not possible to get a representative sample. I’m just not convinced that this business of weighting based on broad criteria makes things better, because it builds in too many assumptions and potential biases.
What if the criteria we’re weighting on aren’t the ones that are important? To make up a ridiculous example, suppose what really matters is hair color? Or shoe size? Or whether you lived in Pittsburgh during the 1970s? No matter how sophisticated our data manipulation is, it can’t adjust for every possible imbalance, because we don’t even know what most of them are. We can’t know what factors are important, so if we pick the ones we think are important and then manipulate the data based on those assumptions, all bets are off.
And, of course, weighting your sample to match the population at large assumes that you have accurate data about the population at large, which you might not. Another unknown, and another potential source of error.
I’m not suggesting there’s a better way. I’m simply saying polls are given way too much credibility nowadays. I suspect we’ve reached a point where polling with any real accuracy is simply not possible.
Who’s got the other 9.1%? SMOD?
So 45% have their minds made up on Clinton, 45% on Trump and 10% are flakes who flip back and forth and who will ultimately decide our future.
Just a hunch, but I think the Trump surprise will come in Colorado. It is, after all, home to Tom Tancredo.
Correct, the media is a business governed by the laws of supply and demand. As long as there are political junkies like us anxiously clicking links and visiting web pages with articles about the latest poll results, we’re generating demand, and the media will keep on commissioning polls (of dubious quality) to meet that demand.
The samples are not all that large. But by doing everything that you mention above, they get results that generally track reality (as revealed on voting day). But this will only work if the categories you mention still predict something.
Right. But the battleground poll needs to be recent — very, very recent . . . given the fact that the ground is moving beneath our feet.
It is possible to get a representative sample, just not possible to get a truly random sample. The quotas and weighting are what gets you from the former to the latter.
I hear what you’re saying but I think the problem is you’re arguing at a higher level of abstraction than what is applicable in the actual experience of the research. That is, it’s actually the case that we do know, generally, which variables predict vote behavior. In fact, my boss developed a model back in 1992 that he’s used quite effectively over the years. To be sure, there are cycles that are anomalous but even that doesn’t happen often. With that being said, there is an ongoing refining process as technology and new data collection methods develop.
The public polling you see commissioned by media companies is one part of the industry. The polling done for campaigns, PACs, and other interest groups is a different animal and involves different methods and objectives. I’d argue it’s more accurate than what you see released to the public.
Amen — especially that last remark about the polling doe for campaigns, PACs, and other interest groups. They desperately need accurate polling, and they are willing to pay what it takes.
So, to take an example, the fact that both candidates are campaigning here in Michigan shows you that their pollsters believe that Michigan is in play.