Someone here pointed out that The NYTimes' recent coverage disproportionately favors Trump to Clinton by a ratio of 15:1. If true, this could reflect the fact that, at the moment, exposure appears to be a liability to both candidates. For now, they both lean heavily on proxies: Pence for Trump, Obama for Clinton. If exposure is a liability to both, The Times' strategy to support Clinton is paradoxically well-served by disproportionately covering Trump. They are betting that, in the end, exposure will be Trump's undoing and that what worked in the primaries won't work in the general. Both candidates are well-known to the electorate. Both candidates have unusually high negatives. In the minds of most voters, the question is not, "Which is better?" but "Which is worse?". Hence the lopsided coverage.
5
I have a sinking suspicion many of the "undecideds" are intending to vote Trump in the privacy of the voting booth, but are too ashamed to admit it publicly. Further, there is evidence the momentum is presently with Trump. Every terrorist action in the US and overseas feeds anxiety and plays into Trump's hands. I have a horrible feeling the pollsters are going to be shown to have it wrong, very wrong, yet again.
8
I'm positive today the talk of the town of Day Oen will be Melania Trump's "inspiring" speech about her husband.
Isn't it amazing that she feels the same way about her hubby as Michelle Obama feels about Barack. What a coincidence. Who'd a thunk it?
DD
Manhattan
Isn't it amazing that she feels the same way about her hubby as Michelle Obama feels about Barack. What a coincidence. Who'd a thunk it?
DD
Manhattan
12
I suppose the FBI report hurt Clinton in the polls. The results of the lawsuit against Trump University are sure to hurt Trump as well. That said, as a Clinton supporter the race is too close for comfort.
My wife keeps saying don't worry America is not stupid enough or racist enough to elect Trump as its president. She reminded me that I didn't think America would elect a black man- twice, no less, and that I should give the American citizen more credit; and she's right.
But somehow this Trump thing has an eerie feeling about it because now is a perfect time for the RIGHT demagogue to come along and sweep America off its feet. I hope I'm wrong this time as well.
My wife keeps saying don't worry America is not stupid enough or racist enough to elect Trump as its president. She reminded me that I didn't think America would elect a black man- twice, no less, and that I should give the American citizen more credit; and she's right.
But somehow this Trump thing has an eerie feeling about it because now is a perfect time for the RIGHT demagogue to come along and sweep America off its feet. I hope I'm wrong this time as well.
15
We know all we need to know. I, for one, have pre-election burnout. Let's vote already.
6
Might help her if she communicates that she will actually do something about the trade issue instead of "taking a hard look". Overall, people are desperate for leadership, but she doesn't want to provide it.
4
I am not confused one iota by the polls. I follow them, but as an afterthought. They are snapshots. They are interesting for we political junkies. But other than Nate Silver's Five Thirty Eight (49 for 50 in '08; 50 for 50 in '12) and Ladbrokes, the London bookmakers who give you the odds if you were a betting man, than nothing else much matters.
The percentages showing Hillary beating Trump has been more consistent and is higher than were the odds at this point in time showing Obama beating Romney. Like Jack Webb, just the facts. ma'am.
Like any horse race, the closer we get to the finish line the more those polls will gel together and get more predictive. There are some things which will not change. The two candidates are so well known, so either hated or loved, there are hardly any Undecideds no matter what they say. And since 40 of 50 states will not budge all that matters are the 10 or so swing states which could be nail biters, or not.
Again, after Labor Day, as we approach the final stretch, then all will become more clear. For many last minute deciders, it will come down to whom they dislike the least, whom they think should never be president. Yes, it's a sad state of affairs when anger rules a voter's choice, but so be it. Politics has never been for the faint of heart. Some new to the game may be horrified. They shouldn't be. There's nothing new under the sun when it comes to the down and dirty aspects of politics.
DD
Manhattan
The percentages showing Hillary beating Trump has been more consistent and is higher than were the odds at this point in time showing Obama beating Romney. Like Jack Webb, just the facts. ma'am.
Like any horse race, the closer we get to the finish line the more those polls will gel together and get more predictive. There are some things which will not change. The two candidates are so well known, so either hated or loved, there are hardly any Undecideds no matter what they say. And since 40 of 50 states will not budge all that matters are the 10 or so swing states which could be nail biters, or not.
Again, after Labor Day, as we approach the final stretch, then all will become more clear. For many last minute deciders, it will come down to whom they dislike the least, whom they think should never be president. Yes, it's a sad state of affairs when anger rules a voter's choice, but so be it. Politics has never been for the faint of heart. Some new to the game may be horrified. They shouldn't be. There's nothing new under the sun when it comes to the down and dirty aspects of politics.
DD
Manhattan
5
Running scared is the best strategy. As a Democrat, I am very, very scared that Trump will win. The close polls motivate me to give money, to get out and campaign, and to do what I can to ensure Trump's defeat. I will be terrified until late into the night on November 8.
25
Some of the polls are, IMO, a joke. On NBC, Lester Holt sits there with a straight face and talks about "our Monkey Survey" poll. I am familiar with Monkey Survey and have even used it to poll a group I lead at times, but it is hardly a way to get a representative cross-section of the electorate. (Maybe Survey Monkey does well-crafted private polls?) I don't know if NBC is just really, really short of cash or whether online self-selected polls are the new normal.
I do know, though, in my 45 years of being a registered voter no real opinion poll has ever asked my preferences or opinions. I say "real" because I did have a Christian right-wing group call me one time at my office as a pastor and ask me to agree or disagree with statements like "I believe the gay agenda should be taught in schools."
That last brings up an important point: the way the questions are crafted and asked. Polls can get vastly different results by asking for the same information in different ways.
I do know, though, in my 45 years of being a registered voter no real opinion poll has ever asked my preferences or opinions. I say "real" because I did have a Christian right-wing group call me one time at my office as a pastor and ask me to agree or disagree with statements like "I believe the gay agenda should be taught in schools."
That last brings up an important point: the way the questions are crafted and asked. Polls can get vastly different results by asking for the same information in different ways.
22
I wouldn't take seriously any poll that has the word "Monkey" in its name.
8
It was a nice touch, Nate, completely ignoring the two third-party candidates who are included in most, if not all, polls. Keep pushing the two-party fight though. God-forbid you accept that candidates exist beyond Clinton and Trump.
3
For all intents and purposes they do not exist and have no hope of being elected, only spoiling the election of the real candidates. When 3rd parties can get a third of the vote, then they will be taken seriously.
25
Thomas, the most recent polls all show that Johnson and Stein are having no impact on the spread between HRC and DT.
10
When Mrs. Clinton just talks I find her approachable, intelligent and she sounds reasonable and competent. When she raises her voice for speeches she sounds shrill and preachy like fingernails on a chalkboard. I suspect her stump speeches hurt her more than help her; particularly to males,
She needs to slow down and do more talking and a lot less raising her voice as her whole delivery changes and not for the better. I suspect it hurts her in the polls.
She needs to slow down and do more talking and a lot less raising her voice as her whole delivery changes and not for the better. I suspect it hurts her in the polls.
15
Please do some introspective inquiry into what you are saying. Whenever a woman raises her voice it is shrill. She sounds strong and sure to me whatever level of voice she uses. Sounds like you took your ques right from Rush.
14
The key to the election will be the vote of the undecided. Somewhere between 15-20% will not say to the pollsters who they plan to vote for.
3
After Brexit, we can see that some voters do not consider the consequences of their ballot. Angry people are not best suited to rational debate--nor do they make the most judicious choices.
31
"Modest but clear lead...". What a chuckle!
7
None of these polls other than the Quinnipiac poll can be believed. All ot the other polls except maybe Fox as Democrat bias built into them. It's the "herd rule" being put in place in the manner of Saul Alinsky, Karly Marx, and Leon Trotsky; that is create the illusion of a Clinton lead and people with follow suit as they want to support the winner. The facts are that the criminal enterprise that is Hillary Clinton, the lifelong congenital liar that she is, and the pandering to small isolated victims groups while failing to represent the broader wishes of the American People; that which she cannot do as a lifelong Saul Alinsky Communist. The Communist Party of America has taken over the Democrat Party, cleaning up the name by calling themselves "Progressives," to fool the American People. Anyone that votes for Clinton is voting for Communism and the end of the Republic. She will appoint up to three Supreme Court Justices to go with the two Obama appointed and the Republic is dead.
4
This is such a terrific piece of well-written satire!
Thanks, James.
Thanks, James.
9
The only poll that matters is the one that happens to show Trump in the lead? The same poll that predicted wins by Romney and previously McCain? And the only reason that this poll is more reliable is that.....it conforms to your wishes? That's the very definition of wishful thinking, James. Your portrait of Hillary is fictional, the stuff of the Right wing propaganda you mistake for news. A "congenital liar"? That means that she's been lying since before she was born. It's a silly phrase, but let's grant you a little poetic license. That being said, you should have no trouble coming up with a list of 10 actual lies from her 35 years in the public eye. Show us these lies, or stop saying "congenital liar". Unless you want to look silly.
21
Thank you comrade for your optimism.
6
It's no big surprise that the Polls are so contradictory---so are the candidates.
The only thing surprising, is how many Americans depend on them to make up their own minds when the choice is so abundantly clear.
Those who want, and demand PERFECTION, might as well realize it doesn't exist.....Not even Bernie Sanders could fill that tall order.
Want a racist endorsed by the Ku Klux Klan, who hates Obama, wants every American to own a gun, and plans to turn the U.S. Mint into his own private Bank??? -- Donald Trump is for you.
Fortunately, most Mexicans, Muslims, African-Americans, "Minorities", and enlightened individuals know what a conservative Republican President, Congress, and Supreme Court would do to the balance of this country.
And you don't need a Poll for that.
The only thing surprising, is how many Americans depend on them to make up their own minds when the choice is so abundantly clear.
Those who want, and demand PERFECTION, might as well realize it doesn't exist.....Not even Bernie Sanders could fill that tall order.
Want a racist endorsed by the Ku Klux Klan, who hates Obama, wants every American to own a gun, and plans to turn the U.S. Mint into his own private Bank??? -- Donald Trump is for you.
Fortunately, most Mexicans, Muslims, African-Americans, "Minorities", and enlightened individuals know what a conservative Republican President, Congress, and Supreme Court would do to the balance of this country.
And you don't need a Poll for that.
33
Those who want "perfection", N. Smith, would most definitely NOT expect that from Sen. Sanders! A coffee house radical, grumpy old man, useless Senator, vain, sore loser, Sixties Socialist, and that is just for starters! I deplore the center disappearing in American politics-they seems to be no more mainstream, just the hard-left on the Democrat side and the hard-right on the Republican. Please do not be naive about Socialism, or Bernie Sanders. Socialisn has had it merits, to be sure, but in the end it is not such an attractive system-and it is TOTALLY unsuitable to this country. Let us not make the mistake of being envious of people who "made good" in this country. If they didn't inherit their wealth or position, that usually means they worked harder, and are smarter than you and I. That is something to admire. Where is the old admiration for people who went "from rags to riches", or at least from middle-class to superwealthy? All I hear from the Bernistas is plain old envy- and bashing of the Clintons. Bill, for instance, came from a really poor background. He was an exceptionally smart guy- he was a Rhodes scholar! And he had a great knack for politics, debate, etc plus tons of charm. Do you? I see a real change in Americans like the Trump supporters and the Bernie admirers-and I don't like it. All I hear is envy, resentment, and boorishness. I think it is un-American!
8
Quinnipiac and Rasmussen consistently showed strong polling by Romney in 2012 and McCain in 2008. While we cannot know whet will happen in 2016, we know the election results for the two previous presidential contests. Ho hum. Don't be surprised not to be surprised in 2016.
www.endthemadnessnow.org
www.endthemadnessnow.org
10
Polls are worthless. People lie when asked. Don't believe any of them.
5
We could get a more accurate feel for a particular poll if they published the numbers of people polled according to area code or, even better, zip code. If they have been published, I don't recall ever seeing them.
3
One of the most annoying and under reported problems concerning the Democrat Convention is the terrible cost associated with being a delegate. Most of the Clinton delegates are established politicians, lobbyists and basically rich people who have an easy ticket to join the party. Bernie Sanders's delegates, many of them young and strapped for cash are forced to campaign for money in their communities to cover the horrendous costs. Delegates are required to all stay in the Convention Center's designated hotels that are at the very least $500.00 a night. No one is allowed to go in and out and buy groceries, so food has to be bought within the Center. I have been following some of these individuals and donating small amounts. It's all so ridiculously over expensive. The whole system is rigged from top to bottom. Unconscionable!
6
After the results of the election, those who got conned by this well-documented preposterous scam and voted for this buffon who is the embodiment of the 'ugly American', will have to live with the ridicule for the rest of their lives. If tragedy gets Trump elected and plunges us all into an economic catastrophe, if not worse with this famously vindictive bully armed with the temper of a 10 year old and leader of the most formidable armed forces in the world, then those who voted for him against the country's interest will only have their conscience and their eyes to cry with as they bear the consequences of their vote.
38
I fear you are in denial about Trump's recent show of strength in the polls. We have to face the horror of Trump's popularity. And we have to look at the polls that include Johnson and Stein, who will be on ballots in November.
Finally, the whole election outcome has great uncertainty because of the approximately 20% of voters who choose neither candidate. This is much larger than in 2012 or 2008.
Finally, the whole election outcome has great uncertainty because of the approximately 20% of voters who choose neither candidate. This is much larger than in 2012 or 2008.
8
For three months the President, Vice President, Elizabeth Warren, and the next VP are going to be out on the stump praising Hillary. (some in Spanish}. Morning Joe will not be able to try and destroy her with unsourced charges ; coyly delivered("fair and balanced") as they force guests for two hours to agree, or defend e-mail, speech fees, and C F charges. This while conducting a yr long infomercial for the Donald.
Meanwhile Trump will continue to repulse 60% of the electorate. He will waffle on such trivia as which "Muslim" "territories" will be chosen for his ban, will he torture, will he blackmail NATO. Also let's not forget the Army he wants to go house to house rounding up anyone without a Green Card. Meanwhile two trials are coming: his Baja Condo scam and Trump U.
Dr Kegel, Trump has never had "huge" popularity.
Look at the total number of votes Trump has received and compare that to the number he needs. Then look at Mitts numbers. Ask yourself "Where will he improve?"
Now about that third party vote. Research shows that the electorate don't stay home for Presidential elections. The vast majority come home to the nominee of their party and vote. Nader was different because only one state separated the two candidates.
Soon, this race will require facts on both sides. Hillary's ideas are on her website. Trump's? So far diatribes, screes, hats and bumper stickers.
Meanwhile Trump will continue to repulse 60% of the electorate. He will waffle on such trivia as which "Muslim" "territories" will be chosen for his ban, will he torture, will he blackmail NATO. Also let's not forget the Army he wants to go house to house rounding up anyone without a Green Card. Meanwhile two trials are coming: his Baja Condo scam and Trump U.
Dr Kegel, Trump has never had "huge" popularity.
Look at the total number of votes Trump has received and compare that to the number he needs. Then look at Mitts numbers. Ask yourself "Where will he improve?"
Now about that third party vote. Research shows that the electorate don't stay home for Presidential elections. The vast majority come home to the nominee of their party and vote. Nader was different because only one state separated the two candidates.
Soon, this race will require facts on both sides. Hillary's ideas are on her website. Trump's? So far diatribes, screes, hats and bumper stickers.
19
I am amazed at the number of commenters ignoring the effect of (and maybe the actual existence of) the Electoral College. Sad to inform you all that it is still there and still determines the outcome of presidential elections in a way not always consistent with the wishes of the electorate.
6
The undecideds and 'other' respondents in all this year's polls are huge. At first I thought those folks would shift to Trump, but now I think they will remain undecided or 'other' and will simply fall out of the final vote. So this could be the lowest turnout presidential election in history ... thanks mainly to Trump.
4
A vote for Johnson or Stein is a vote for Trump so no mistake about that. A vote for Clinton is a vote for control of SCOTUS and overturning a lot of rubbish it's ruled on in the past (Citizens United, Voting rights Act). Impossible to get a constitutional amendment overturning Citizens United so it has to be done through SCOTUS.
For those who say Clinton's a liar: get over it. Sanders is/was no saint. He just never got vetted nationally so his clay feet are still hidden. Clinton in her 20s did more for minorities, kids, mothers, etc. than Sanders has done in his entire career. He talks a good game but he's a legislative nebbish because he's too pure to compromise. And he never would have had the guts to go to Beijing and make the speech she did in the 1990s. He's pretty much of an non-achieving legislator. In spite of all of her alleged sins and shortcomings she stands head and shoulders above Sanders in sheer accomplishments.
The real problem for Sanders is that if she loses because he's been so successful in demonizing her is that he will be held responsible for that loss and he will replace Ralph Nader as the most reviled, hated, despised person who's run for POTUS in our time. That's probably not quite the legacy Sanders is trying to preserve. So help your hero along and vote for Clinton.
For those who say Clinton's a liar: get over it. Sanders is/was no saint. He just never got vetted nationally so his clay feet are still hidden. Clinton in her 20s did more for minorities, kids, mothers, etc. than Sanders has done in his entire career. He talks a good game but he's a legislative nebbish because he's too pure to compromise. And he never would have had the guts to go to Beijing and make the speech she did in the 1990s. He's pretty much of an non-achieving legislator. In spite of all of her alleged sins and shortcomings she stands head and shoulders above Sanders in sheer accomplishments.
The real problem for Sanders is that if she loses because he's been so successful in demonizing her is that he will be held responsible for that loss and he will replace Ralph Nader as the most reviled, hated, despised person who's run for POTUS in our time. That's probably not quite the legacy Sanders is trying to preserve. So help your hero along and vote for Clinton.
49
On the contrary, Johnson is far more likely to attract people who would otherwise support Trump than those who would support HRC. There are plenty of Never Trump, Romney-type Republicans who would rather vote for Hillary than Trump.
2
You're not going to win Sanders' fans over by bullying them.
I voted for Sanders in the primary and will vote for Clinton in November. My objection to Clinton is not about domestic policy. She is a hawk and proud of it. The world is in crisis and she might well make it worse. Sanders lacks experience but his instincts were better and he could have found experts to advise him.
I voted for Sanders in the primary and will vote for Clinton in November. My objection to Clinton is not about domestic policy. She is a hawk and proud of it. The world is in crisis and she might well make it worse. Sanders lacks experience but his instincts were better and he could have found experts to advise him.
5
Instead of relying on polls why don't our politicians listen to us? Why don't they get out of their offices, walk on our streets, look at our housing, go to our schools, wade through the red tape for health care, or see what it's like to be an average American and then decide what the important issues are? I keep on hearing the same thing from the people I talk to every day: jobs, retirement, paying the bills, housing, affordable medical care, and the feeling that no one truly gives a hoot.
17
The people who don't seem to give a hoot are the political reporters, who cover the campaign. All you hear about what the candidates say and do is filtered through their interests, polls, gaffes, gossip.
You are right that Trump just flies in and gives his big rally then flies out, but both Bernie and Hillary spent the past year, in fact many years of their lives, out talking and listening to voters individually and in small groups, in addition to holding rallies
Here are Hillary's proposals, which do speak to the concerns and needs of the rest of us. She and people she reached out to devoted time and effort to put this together even though all the political press ever seems to ask her about are emails
https://www.hillaryclinton.com/issues/
You are right that Trump just flies in and gives his big rally then flies out, but both Bernie and Hillary spent the past year, in fact many years of their lives, out talking and listening to voters individually and in small groups, in addition to holding rallies
Here are Hillary's proposals, which do speak to the concerns and needs of the rest of us. She and people she reached out to devoted time and effort to put this together even though all the political press ever seems to ask her about are emails
https://www.hillaryclinton.com/issues/
14
Simple. Because that would be polling unscientifically and the pol would not know whether the folks he talked to represented his entire constituency of not. The people you talk to every day may or may not represent a cross section of your demographic.
4
To focus so obsessively on predicting the future as if the election were a ball game whose result we are gambling on is a degradation of democracy that has helped us think of election day in November 2016 as a reality show.
13
Your comment suggests that this is the first year there has been a focus on polls. It's been going on for decades. If the 2016 election feels like a reality show, it's because Trump and the media have treated it this way.
6
But the Upshot is the NYT feature dealing with polls and data. If we're it, then this is what we're reading, no?
2
You're commenting on an article by a guy who specializes in polls & statistics; there are others with a very different orientation...
1
I am getting really tired about the mis-representation of two important things.
1) Margin of error is computed in such a way as that the results should fall inside them 19 out of 20 times. This isn't happening by a long shot.
2) It matters not what percentage of potential voters favor which candidate for president. What matters is the count in the Electoral College. 51% and 99% in any state get you the same number of votes in all but a few states, such as Maine.
1) Margin of error is computed in such a way as that the results should fall inside them 19 out of 20 times. This isn't happening by a long shot.
2) It matters not what percentage of potential voters favor which candidate for president. What matters is the count in the Electoral College. 51% and 99% in any state get you the same number of votes in all but a few states, such as Maine.
4
1) The margin of error applies to both candidates' numbers. So if a poll shows 48 to 44 for clinton, that doesn't mean that 95% of the time the difference is 4 points. It means that 95% of the time clinton is within the range 44 to 52 and trump is in the range 40 to 48. So you could have a poll that shows Clinton 52 and Trump 40 (she's up by 12) or Clinton 44 and trump 48, she's down by 4. Down by 4 and up by 12 are within the 4% margin of error.
That's what the graph in the article shows, too.
2) True, but the likelihood of losing the EC while winning the popular vote is quite low, and it is much easier to poll nationally than it is state by state. However, the state by state polling that we do have is consistent with the national polling.
That's what the graph in the article shows, too.
2) True, but the likelihood of losing the EC while winning the popular vote is quite low, and it is much easier to poll nationally than it is state by state. However, the state by state polling that we do have is consistent with the national polling.
10
Never the less
1) If the MOE is 4 points and the candidates numbers don't fall in that range, the estimate is poor. This is a zero sum game.
2) The EC is the only thing that matters, so why not report about it?
Lazy, lazy, lazy, that is why.
1) If the MOE is 4 points and the candidates numbers don't fall in that range, the estimate is poor. This is a zero sum game.
2) The EC is the only thing that matters, so why not report about it?
Lazy, lazy, lazy, that is why.
Clinton's weaknesses are more than offset by her dedicated efforts for women, minorities and children. Trump on the other hand has been cheating the public for example Trump University, dissembling facts and self serving. A loathsome human being. Those who want to vote for Gary Johnson should pay attention to his policies and the possible Ralph Nader effect in electing GW.Bush over Gore. The choice is between Clinton and Trump. Trump represents the worst of America.
51
Hmmm, never mind the fact that the Clinton Foundation which Hillary was a board member of until she resigned that post just before announcing her run for president, took millions in contributions from Saudi Arabia which is a country known for it's abuse of women's rights. As for her claiming to be a champion in defending women of sexual abuse or harassment. . . what about the women her husband took advantage of? Hillary discredited them and destroyed their careers in defense of her husband. I just can't believe her or her stated positions, they will change depending on the person or situation she is facing.
3
Just a quick comment about one of Hillary's weaknesses... "those damned e-mails".
If I understood correctly, almost all the e-mails Hillary sent were recovered. Even pieces of "deleted" e-mails were seen by the F.B.I.
My comment ignores whether some were secret.
The accusations I hear from Mika B., Joe, and Fox is that she was hiding a criminal organization involving the sale of access, money laundering, and other nefarious acts using the Clinton Foundation. They've never been clear about the charges but that seems to be why people wear "Hillary for Prison" T-shirts.
Has anyone noticed that not one e-mail contained criminal activity?
Wouldn't that be note worthy? You'd think they would have found at least one out of 33K that was damaging and you know if one exists, we'd be subject to hours and hours of attacks.
No! You can't say she found and deleted all the "bad" ones. We already saw that she and her lawyers were not 100% accurate in their choices. Remember the 122 and the 8?
A reasonable person would have to believe that though she was wrong to use Bill's server, she's not a crook.
If I understood correctly, almost all the e-mails Hillary sent were recovered. Even pieces of "deleted" e-mails were seen by the F.B.I.
My comment ignores whether some were secret.
The accusations I hear from Mika B., Joe, and Fox is that she was hiding a criminal organization involving the sale of access, money laundering, and other nefarious acts using the Clinton Foundation. They've never been clear about the charges but that seems to be why people wear "Hillary for Prison" T-shirts.
Has anyone noticed that not one e-mail contained criminal activity?
Wouldn't that be note worthy? You'd think they would have found at least one out of 33K that was damaging and you know if one exists, we'd be subject to hours and hours of attacks.
No! You can't say she found and deleted all the "bad" ones. We already saw that she and her lawyers were not 100% accurate in their choices. Remember the 122 and the 8?
A reasonable person would have to believe that though she was wrong to use Bill's server, she's not a crook.
8
Yet again Nate illustrates the ridiculous inequity of the all or none electoral college vote in 48 of the States. A Democrat living in Texas has no incentive to vote in the Presidential election, even if he or she lives in a heavily Democratic stronghold an has a Democratic representative in Congress. Texas has gerrymandered itself into the next-worst thing to a one-party state. Presidential election polling data from Texas is completely meaningless.
22
As a Democrat living in Texas and Arizona before that, I have always felt an incentive to vote not just in Presidential elections but all elections. Where the gerrymandering is felt particularly strongly is in elections for congressional representatives and in state elections. The first time I voted in Texas in 2014 was the first time I felt that my vote didn't really matter. Of course, I have since adjusted my attitude and gotten out to vote on all election days.
14
True! Losing an election is one thing, losing by default is another. One is sad, the other is stupid. Even if you are the only person who votes a certain way, it shows something.
1
As a life-long Democrat now living in Texas, I switched parties to vote in the latest Republican primary ... against spawn-of-the-devel Ted Cruz, which doesn't mean that I would ever support Trump. It's just that Cruz is the kind of person who would want to sell you autographed photographs of Jesus. Even if your vote doesn't count because of gerrymandered districts, you can still have an effect on the election. Remember "A vote for Clinton is a vote for control of SCOTUS and overturning a lot of rubbish it's ruled on in the past (Citizens United, Voting rights Act)."
8
More educated good.
9
Most people haven't the slightest clue about polls. National polls are completely meaningless. They are the bright, shiny object of election years.
7
Back to school, Mike. They've been used and have been mostly accurate since WW2. There have been some notable boners over the years (the Dewey election victory) but most of them have been the result of polling bias in the questions or sampling mistakes. They also drive TV programming.
4
The graph "A Stable Vote, but ...": the labeling is severely deficit, making the interpretation difficult. Vertical axis and the line colors [?!].
8
Just remember, these are the same poling methods that said the Brexit vote would never succeed, and how did that turn out?
I don't pay attention to the polls, I will vote how I want, likely for Gary Johnson. The other three candidates are not for me especially Trump and Clinton, they will both do or say anything to be elected.
I don't pay attention to the polls, I will vote how I want, likely for Gary Johnson. The other three candidates are not for me especially Trump and Clinton, they will both do or say anything to be elected.
4
??? Brexit polls showed the vote was going to be close. Was it close?
5
You are perfectly within your rights to vote for anyone that you want to. The problem with the presidential election before us is that the choice is Trump or Clinton. Gary Johnson, for all that he may be a better candidate, has zero chance of winning but what he does have is a chance of sucking enough votes away from Clinton to make sure that Trump wins.
You referenced the Brexit poll. In the days afterwards many people said that they voted exit as a protest vote against the establishment, not believing that the result would be a vote for and exit, but casting enough votes to swing the result.
You referenced the Brexit poll. In the days afterwards many people said that they voted exit as a protest vote against the establishment, not believing that the result would be a vote for and exit, but casting enough votes to swing the result.
4
You didn't label the y axis well on the second graph
2
Nate: I used to consider your predictions to be gospel. But, as you admit in this latest column, they are becoming dramatically less reliable. You mention but do not give enough attention to how polling is conducted in the age of 21st century media. Polls that rely on home telephone contact are clearly going to skew to old people who have nothing to do but answer the landline, even if they don't recognize the number that is calling. This is just the tip of the iceberg in the area of unreliable instruments. I wish you would put a temporary halt to predicting (it has damaged a number of progressive causes when you are wrong...i.e. last year's governor's race in MD). How about more columns on how impossible it has become to predict winners based on political polls?
2
Geez, pretty ageist of you to claim that older people have nothing to do but answer the landline even if they don't recognize the number. I know lots of people over the age of 65 and (1) they have very active interesting lives, and (2) they never answer their landlines unless they recognize the number.
6
I think you are confusing Nate Cohn with Nate Silver.
4
No vote for Clinton. A liar and not to be trusted. Trump is a joke.
2
Staying home is a vote for Drumpf. You will be blamed, and you should blame yourself.
8
So? What do you plan to do throw your vote away on one of the little 3? We saw where that got us (GWB). Presidential elections in this country for better or worse are between the two major party candidates (I'll grant you some rationale on that topic of "what should be").
4
That's fine, but either Trump or Clinton will be the next president, so you can just stay in denial, but if you think there's not a significant difference, you're just not paying attention.
3
What surprises me is not that Hillary is ahead in the polls, but that as many people are supporting Trump and his mentality as there are. This country is a lot sicker than anyone could have imagined before Obama was elected, and Trump is doing a service to us all, in a sense, by being the mirror through which we can see it.
49
Agree 100 % with what you say, Mr Ferrugia! It is scary to think of all those voters out there who don't seem to mind that Donald Trump is an inexperienced buffoon, with a volatile temper, who changes his opinion on important issues every other day, and who is a racist and misogynist to boot!
14
Indeed! Hear hear.
6
Here's an informal poll result for what it's worth: there isn't a single Romney voter I've talked to in my extended circle of contacts who who plans to vote for Trump. Not one.
39
Same here, GSL. My circle of friends consists mainly of Republicans-this is the South, after all-a few Independents, and a few Democrats. I myself lean towards the Democrats on economics, our social agenda, basically the domestic front, and lean heavily towards Republicans in matters of Foreign Policy. But the Republican candidate this year is unspeakably vile, so none of my Rep. friends will vote for him, nor will I. As David Brooks wrote in his column a while ago: "No, not Donald Trump, not now, not ever!!" Amen to that!
3
Just curious, who do they intend to vote for?
4
Yep, European Liberal, there is no better testament to the superior Republican foreign policy chops than the Iraq war.
1
The polls usually do not include the Gary Johnson/Bill Weld Libertarian ticket, which is a credible alternative with more Executive experience than the parties in the polls. Given the need to reach 15 % in multiple to enter the debates, these candidates should be included. Wouldn't the debates then be enlivened if the included Johnson who, for instance, supports the Trans Pacific Partnership and would explain why?
3
Most polls I've seen do count third and fourth parties. You say the Libs are a credible alternative, I say no. The top one percent would get even richer with those people in charge. I do like their thoughts on social issues, though. Bernie would have been perfect, but Hillary is pretty dang good too.
11
That's not true. Many polls include Johnson and some include both Johnson and Stein. Neither breaks 10% and Stein usually cannot break 3%. The margin between Clinton and Trump is pretty similar in polls where third party candidates are included and those where they are not included.
Look at poll aggregates like Huffington Post or Real Clear Politics for the polls with third party candidates.
Look at poll aggregates like Huffington Post or Real Clear Politics for the polls with third party candidates.
10
Real Clear Politics - http://www.realclearpolitics.com/epolls/latest_polls/ - lists all of the poll, and shows that most pollsters are conducting polls that include Johnson and Stein, as well as polls that do not. Currently Johnson is averaging 8.3 percent, while Stein is averaging 3.7 percent, making it unlike that either will be included in any national debates.
1
If the Hillary/Trump total equals 100 than the poll is bogus. This goes for any presidential election 100% of vote never goes to two candidates.
We KNOW that third party candidates are going to have a strong showing. Also polls don't include undecided. They don't publish the demographics, nor how many people actually responded. We can tell that the number of people that actually respond is quite low from the massive variance in the daily polling.
A nation of 300 million doesn't have that kind of swing in political affiliation. Judging from the variance I estimate that less than 500 actually respond.
This is assuming that all things are equal, that there are no dirty tricks going on ;)
We KNOW that third party candidates are going to have a strong showing. Also polls don't include undecided. They don't publish the demographics, nor how many people actually responded. We can tell that the number of people that actually respond is quite low from the massive variance in the daily polling.
A nation of 300 million doesn't have that kind of swing in political affiliation. Judging from the variance I estimate that less than 500 actually respond.
This is assuming that all things are equal, that there are no dirty tricks going on ;)
Picture this after Hillary wins and takes Senate. Supreme Court finally "packed" by Democrats. This will cement Roe vs Wade. It will end Citizens United. It will interpret the constitution to mean militias a.k.a. Nat'l Guard can have machine guns not individuals. It will end Gun Show loop hole. It will reinstate Civil Rights Act and voting rights. We will control gerrymandering forcing Congressman. no longer "fed" by Large donors to court voters in their districts. Sound radical ? According to almost every poll, all of this has huge popularity among all voters.
That's just part of what we Democrats can do.
I believe if Hillary looks in the camera and just using Supreme Court explains how easily all this can be done,this CHANGE in all our lives will result in a Democratic landslide. The following quarter she can borrow a few hundred billion (interest rate cries for this) and fix all our certified "dangerous" bridges, resulting in steel sales , cement sales, Carpenters, electricians, laborers. And remember when we put money in middle class pockets, they buy stuff which creates more jobs and more pockets and so on...and so on...
We are so close. Sounds wonderful doesn't it? Actually it is a description of this country when I was a boy. Vote the Democratic ticket straight down to City Council and we can relax for a change.
That's just part of what we Democrats can do.
I believe if Hillary looks in the camera and just using Supreme Court explains how easily all this can be done,this CHANGE in all our lives will result in a Democratic landslide. The following quarter she can borrow a few hundred billion (interest rate cries for this) and fix all our certified "dangerous" bridges, resulting in steel sales , cement sales, Carpenters, electricians, laborers. And remember when we put money in middle class pockets, they buy stuff which creates more jobs and more pockets and so on...and so on...
We are so close. Sounds wonderful doesn't it? Actually it is a description of this country when I was a boy. Vote the Democratic ticket straight down to City Council and we can relax for a change.
71
Hear Hear! Right on; that's the message she needs to voice.
8
I've never been contacted by a pollster in my entire life and have a landline phone, cell phone, etc. I am a well-educated, professional, female who is appalled at the way I have been mis-characterized in the media for being pro-Hillary. I am not for a liar who has misused her power throughout her life to advance one person -- Hillary. From her phony foundation to her Bengazi travails, she is not the person I want to lead this county. Guess I'll never hear from a pollster.
10
Right there with you, Ellen!
1
Trump will not take you anywhere either.
I will choose her over a racist demagogue.
I will choose her over a racist demagogue.
8
Not really sure you know how polling or data significance works.
17
Here's a disturbing statistic. Just for fun, I just performed two word searches on the NY Times homepage as of 7-18-2016 at 11:47 a.m. EDT.
Search #1: "clinton" -- one hit
Search #2: "trump" -- 15 hits
So much for being even close in equal coverage. And the Times is not alone in this, of course, but this is getting ridiculous. One would wonder if the Democratic Party even HAS a candidate!! Sheesh.
Search #1: "clinton" -- one hit
Search #2: "trump" -- 15 hits
So much for being even close in equal coverage. And the Times is not alone in this, of course, but this is getting ridiculous. One would wonder if the Democratic Party even HAS a candidate!! Sheesh.
82
Very disturbing indeed. Would the same press overlooking his major flaws during a campaign also overlook his monumental screw ups, crazy policies and disastrous economic and social programs?
A corollary is, when exactly will the media take the same approach of intense or even negative scrutiny of Trump as they apply to Clinton?
A corollary is, when exactly will the media take the same approach of intense or even negative scrutiny of Trump as they apply to Clinton?
56
Surprise, corporate owned media puts their efforts towards click baits and ad revenue.
6
Why don't you try it again on the first day of the *Democratic* convention?
3
Part of the reason why the polls are contradictory and confusing is because Trump and Clinton are not the only two options, and they aren't the only candidates for whom people will be voting in the fall. Gary Johnson and Jill Stein are both viable third party candidates that collectively possess at least 20% of the of the popular vote in the polls that include them. That's 20% of people who aren't going to vote honestly (or at all) when only presented with the two options of Clinton and Trump in a poll. Gary Johnson and Jill Stein need to be included on all polls, and then you'll start to see accurate numbers.
5
Not sure what you mean by "viable." If by viable, you mean people *can* vote for them, then yes, they're viable. I believe a more widely accepted definition of the term would be a candidate who can win. Neither Stein nor Johnson has any chance at taking even a single electoral vote, much less winning. Not very viable.
10
Natalie....please see this as a Respectful disagreement. I know that there are a number of greenies and no-gov-libbies (and confess to have some of that same genetic material in my persona) but when push-meets-shove in the ballot baox, most want our vote to COUNT. Granted that Nadar did steal enough from Al Gore to put W's hand on the Bible the following January 20, he was a quantity known for 40 years BEFORE the Floida vote. So I see those 20% dwindling to, say, 3 or 4% in Nov 2.
4
No, they haven't got a shot at anything right now other than distracting people from facing the dire aftermath of a Trump win. They have to be at a place to get enough votes to win - and aren't. I'd prefer Stein, but - unless Hillary is hit by lightening and taken out of the race between now and November - she is the flawed, but better, option.
2
l will never ever vote for TRUMP, l am not crazy about Hillary either so l am going to write Gen. Powell name or may be Barnie
3
Akin Caldiran, you can tell yourself you're voting for Powell or Bernie, but we all need to recognize the real-world consequences of our actions. The only vote that can help keep Trump out of the White House is a vote for Hillary. Failing to cast that vote is an assist to Trump.
83
That is a vote for Trump.
10
and a copout. Voters need to be the adults in doing due diligence in casting our vote. We have the right to vote, except in Kansas, and a parallel responsibility to vote for the best candidate, not for us personally, but for the nation's future. Clearly, that means a vote for Hillary Clinton.
42
The fact that Trump's success depends on non-college educated white voters is empirical evidence our schools are failing America and do not deserve a passing grade.
70
Or, you could say that WE have failed to support our public education system, which started down hill with Reagan in California. I was there.
15
So was I! Reagan turned California from a leading edge, high-achieving state into something more disappointing and bifurcated. He destroyed a lot of the achievements of the 'good government' crowd that had toiled through the late 1800s and the first half of the 20th C. Then Reagan moved on to repeat the destruction for the entire U.S.
18
Indeed they are, JA Barry. In math, for instance, our high schools are trailing 17th place after 16 other Western countries! My husband and I put our kids in private schools-big sacrifice, no vacations or buying a home, all our money has gone to their education. But I feel we made the right choice. On the other hand, don't forget that Americans are pushed to excel, whether at school or in sports or volunteer work. Some of the European schools may be better, but on the other hand they are ok with students getting B- or C's, as long as they are "passing the grade." Happens in high school AND in universities, law and medicine for instance. There are two sides to this issue.
As others have noted, the HRC-Trump poll needs to be "balanced" by a sense of how the congressional races are going. Without a Democrat majority, HRC's presidency would face the same fate as Obama's.
While the NYT continues to hustle on behalf of Team Hillary, that is only part of the story. Love or loathe her, vote for her with enthusiasm or as the lesser of two evils, the success or failure of her presidency depends on Congress. The past 8 years have demonstrated that.
While the NYT continues to hustle on behalf of Team Hillary, that is only part of the story. Love or loathe her, vote for her with enthusiasm or as the lesser of two evils, the success or failure of her presidency depends on Congress. The past 8 years have demonstrated that.
26
I have an advanced degree and work in the IT space. I support him.
3
David - That's not really something to brag about.
10
The problem with these polls is that they only sometimes include Gary Johnson, and sometimes the same poll will first ask who people prefer between Clinton and Trump, and then ask a separate poll between Trump, Clinton, Johnson, and sometimes Stein. This skews the results, and actually has an effect on how people vote. The Libertarian Party will most likely be on the ballot in all 50 states, but will at least be on enough states to get 270 Electoral College votes. He should be on every poll!
6
I am simple person, I don't need complicated and contradictory polls to confuse me. It is very clear to me that pundits, political commentators and the press including reputed newspapers have hired journalists who spin the news to give their candidate a lift. Polls are inherently useless with 4 months left for the November election. Things are happening every day that makes people change sides and the final decision on election day will decide who they will vote for under a set of conditions on that day. Even the Sanders supporters may change sides and not follow the Bernie's herd to Hillary. Last year around this time, I would have said with certainty that Trump is a joke with no chance of ever winning the white house. Fast forward a year later, I am not so sure about the sure thing Hillary. Nation needs a significant change and under the current set of domestic and global conditions, Trump and Pence look like a serious option in the best interest of USA to consider.
8
Take a look at articles in conservative economic publications like The Economist about the disastrous implications of Trump's policies (such as they are). Not the kind of change most of us would like to see.
21
We definitely need change and the first step is to get rid of the Republicans, who are preventing it from happening.
16
The problem with the GOP convention are the horde of reporters looking for a sliver of discontent, idiotic gaffs, slip of methane gas or a bold blunder. So, the writing is already on the wall (or already written) and as reported are an idiotic hord of right wing nuts trying to control the morality of America notwithstanding the secret efforts to lord over and control the pelvic floor of women everywhere.
12
Its in Hilary's advantage that polls in general show a close race and some show trump edging ahead. If she were projected to lead by a large margin, then some who dislike her, but dislike trump more, will conclude that she has it in the bag and will either not vote or cast their vote for a protest candidate. The democrats cannot afford to fall asleep at the wheel on this one, lest they risk experiencing a Clinton version of the Brexit vote, where the polls had it precisly wrong opposite on the eve of the eve of the polling.
33
Yep. As a Clinton supporter, I want the poll close right to end, to put the fear of God in Dems to get out and vote.
16
Can you do a column on which demographic can Hillary make the most gains in? and how?
It seems white men is a lost cause for her, and no matter what she did, even if she ran on Donald Trump's ticket, she'd lose the white men vote.
It seems she can't improve on her black support, which is not only solid, but about as high as it is for Obama.
So that leaves women, hispanics, and the young (or any such combination). The question then is how? Elizabeth Warren as VP? Would the gains in women and the young be offset by the flight of even more white male votes? Or a young dynamic Latino VP?
It seems white men is a lost cause for her, and no matter what she did, even if she ran on Donald Trump's ticket, she'd lose the white men vote.
It seems she can't improve on her black support, which is not only solid, but about as high as it is for Obama.
So that leaves women, hispanics, and the young (or any such combination). The question then is how? Elizabeth Warren as VP? Would the gains in women and the young be offset by the flight of even more white male votes? Or a young dynamic Latino VP?
19
With all due respect, you would be making a big mistake to underestimate the Black vote.
Sanders did -- and look at what happened.
Sanders did -- and look at what happened.
1
These polls consistently show that a higher education correlates with voting Democrat. That fact alone speaks volumes and helps explain why Republican politicians are so anti-science. They have to be in order to stay in office. Scary.
143
The only people I know who A. have a land-based phone, B. don't have Caller I.D. to filter out robocalls, and C. would actually take the time to answer a pollster are A. old, B. white, and C. unemployed or retired. Which is to say old, white, and on a limited income.
So it's hard for me to see how a land-line based poll is anywhere near accurate, since it disproportionately draws from the classic Trump voter.
So it's hard for me to see how a land-line based poll is anywhere near accurate, since it disproportionately draws from the classic Trump voter.
64
I work from home, and there are many calls per day that I don't answer because the caller is not identified. If I saw a call from a known polling organization, I would certainly answer, because I'd like to put in my two cents worth. I wonder if polling organizations identify themselves? If not, they are making a mistake, in my opinion.
28
I agree. I rarely answer an phone call unless it is someone I know. Even it said pollster on it (some do say this) I still do not answer. I don't know if it a real legitimate pollster or someone trying to sell me a time-share in Florida. I probably answer one in ten phone calls.
I would like to see more analysis of the bets being made on the election. These seem to me to be better indicators of how people vote. That is, people vote where their money goes if they are betting.
I would like to see more analysis of the bets being made on the election. These seem to me to be better indicators of how people vote. That is, people vote where their money goes if they are betting.
6
I'm old, white and retired but I do screen Caller ID so the only pollster I would talk to is one who left a polite non-computerized message. i was able to talk to a lot of voters when making calls in 2012 but I'm hoping the Democrats in this area have a new plan for me to help get out the vote this year. Emails would be good if they started talking issues and the need for voters instead just asking for money. We old retired white voters have more time than money so give us something to do!
1
Polling organizations may be biased and poll number is manufactured. If poll taken in Staten Island in New York City , it will be pro Trump and taken in Manhattan will be po Hillary. Personally I do not believe Quinnapac and Rasmussen polls.
5
I respected Mr. Silver's work, but I have less and less faith in polling in general. I was polled once on a landline and it was a surprisingly in-depth inquiry that, I believe, resulted in a very good picture of my voter profile. Because this type of conversation is essentially impossible to conduct in this age, polling has become less and less reliable... From my private seat-of-the-pants canvassing of friends, neighbors and casual acquaintances, I have Trump getting clobbered in the general.
21
This is Nate Cohn, not Nate Silver. Nate Silver used to work here. Now he runs 538.com.
17
The polls have actually been very accurate this year. As Nate Silver himself noted, he failed to pay attention to the polls when he discounted Trump, because he simply couldn't believe a man like Trump could win the Republican nomination. I have to say, I find it hard to hold that error against Silver. In any case, there's been little wrong with the polls; I wish I could say the same for the Republican electorate.
7
I have volunteered in many campaigns, and I know how important get-out-the-vote operations are. Clinton has a highly organized campaign , and I expect campaign headquarters to open up here in the Lehigh Valley after the convention . A well organized campaign is an excellent thing to experience, because if you go to volunteer, you will be told what to do and how to do it. In 2012,, I was struck by the invisibility of the Romney campaign. My preference is to canvass, and I saw almost no evidence that any Romneyites were doing the same, although my populous suburban area votes 50-50 as a rule. BIG advantage Clinton. People who volunteer are passionate and follow the news, and I expect a strong turn-out of volunteers. They came over here from New Jersey and New York in 2008 and 2012,, and I expect the same thing to happen this year.
As an aside, when is there going to be any action involving the totally illegal solicitations for donations sent by the Trump campaign to foreigners? At first it was thought to be an error based on faulty email lists, but I read yesterday that the solicitations have continued apace. Can you imagine the uproar if the Clinton campaign was caught doing this?
As an aside, when is there going to be any action involving the totally illegal solicitations for donations sent by the Trump campaign to foreigners? At first it was thought to be an error based on faulty email lists, but I read yesterday that the solicitations have continued apace. Can you imagine the uproar if the Clinton campaign was caught doing this?
73
Lucy:
Don't I recall stories about Hillary doing just this in 2008? Should we also think about the Clinton Foundation for a minute while we're thinking of foreign money?
Don't I recall stories about Hillary doing just this in 2008? Should we also think about the Clinton Foundation for a minute while we're thinking of foreign money?
3
magpie Bill Clinton was investigated in 1996 for soliciting foreign contributions, but was not fined. Hillary was not investigated for soliciting foreign contributions for her 2008 bid for president. Whatever we may think the Clinton Foundation is, it is not an organization in place to raise money for political candidates.
However, the Trump campaign got its fingers caught in the cookie jar, in regards to soliciting campaign funds from foreign citizens.
However, the Trump campaign got its fingers caught in the cookie jar, in regards to soliciting campaign funds from foreign citizens.
11
Money from the Clinton Foundation does not go to Hillary's campaign fund. Read their annual report if you want to see where that money goes: https://www.clintonfoundation.org/about/annual-financial-reports
12
Given the high number of robocalls and scams these days, I never answer either my landline or cell if I don't recognize the caller. I often wonder how this might impact political polls.
If, say, antiestablishment types are less likely to answer their phones, that could skew the polls, especially if the polling firms are perceived as part of the establishment. This could be a particular concern this year as this demographic is less likely to support Clinton.
I received several calls from Quinnipiac a few years ago and eventually answered, but only because I've studied polling and recognized the name on my Caller ID as a legitimate polling firm.
If, say, antiestablishment types are less likely to answer their phones, that could skew the polls, especially if the polling firms are perceived as part of the establishment. This could be a particular concern this year as this demographic is less likely to support Clinton.
I received several calls from Quinnipiac a few years ago and eventually answered, but only because I've studied polling and recognized the name on my Caller ID as a legitimate polling firm.
5
Polling is much like ballistics
Math, calculation, heuristics
What goes up must come down
In an arc toward the ground
Of lies, damned lies, and statistics!
Math, calculation, heuristics
What goes up must come down
In an arc toward the ground
Of lies, damned lies, and statistics!
8
If only they weren't as reliable as they are, you might have a lyrical point. But, alas....
2
Rob, actually I find statistics properly derived from adequate data sets to be quite reliable. Polls -- not so much.
1
Yes it would SORT OF look the simulation in the graph. But if it actually looked as smooth as the "simulation" suggests that would be a a fake graph -- which it is.
Are polls normally distributed? How could they be when each pollster use a slightly different methodology in sampling practice, slightly different sample size, and slightly different way to determine likely voter? Moreover, a sample of polls is not actually a random sample!
Much better idea for teaching about the "noise" aka systemic variation would be to explore W E Deming's insights into variation. Better to just do the red bead experiment to teach about variation and use a control chart for polls to understand systemic variation vs assignable cause variation.
Are polls normally distributed? How could they be when each pollster use a slightly different methodology in sampling practice, slightly different sample size, and slightly different way to determine likely voter? Moreover, a sample of polls is not actually a random sample!
Much better idea for teaching about the "noise" aka systemic variation would be to explore W E Deming's insights into variation. Better to just do the red bead experiment to teach about variation and use a control chart for polls to understand systemic variation vs assignable cause variation.
3
Polling data at this point in the race is suspect. The people who aren't paying attention right now will eventually start paying attention or vote based on the latest sound bite so the polls could be way off so there is no reason for Clinton supporters to be dancing in the aisles.
Stay calm. Fasten your seat belts. The debates, guaranteed to be the most watched in history, will decide the election. And they won't be, you can bet, decided on the biased media spin coming at us fast and furious. People will judge for themselves who wants to Make America Great Again.
Re: the polls. The polls reflect a huge difference in the amount of advertising dollars being spent so far. Clinton is flooding the country with negative ads. Trump's opponents in the primary spent millions on negative ads. And yet Trump is within the margin of error. With a decent Democratic candidate he would be down 20 percentage points given the amount of negative advertising being lobbed at him. Trump has a lot of things going for him and is not a slick, well-prepared, predictable politician so don't be so quick to write him off. Ordinary people seem to sense that he says what he means and not what some focus group or highly paid groups of policy advisers tell him to say. Educated people are more easily fooled by the 'right' words.
Stay calm. Fasten your seat belts. The debates, guaranteed to be the most watched in history, will decide the election. And they won't be, you can bet, decided on the biased media spin coming at us fast and furious. People will judge for themselves who wants to Make America Great Again.
Re: the polls. The polls reflect a huge difference in the amount of advertising dollars being spent so far. Clinton is flooding the country with negative ads. Trump's opponents in the primary spent millions on negative ads. And yet Trump is within the margin of error. With a decent Democratic candidate he would be down 20 percentage points given the amount of negative advertising being lobbed at him. Trump has a lot of things going for him and is not a slick, well-prepared, predictable politician so don't be so quick to write him off. Ordinary people seem to sense that he says what he means and not what some focus group or highly paid groups of policy advisers tell him to say. Educated people are more easily fooled by the 'right' words.
2
I think if educated people were "more easily fooled" we'd have shorter life spans and less to eat. Not to mention doctors might be bleeding us for a fever and think stinky smells, not germs, spread disease. I'm very leery of folks that suggest education is a detriment.
As for Trump, I think his standing makes it clear that any major candidate enters the contest with 40% of the electorate in their corner. Nick Confessore made the point in a Twitter post, that the billion about to spent on the presidential race is really just targeting the 3% of the population that can be swayed. Because we're talking about Trump, I think that number is higher than 3% this year - and what the polls probably show is they're voting for Hillary. I guess we'll know in a few months.
As for Trump, I think his standing makes it clear that any major candidate enters the contest with 40% of the electorate in their corner. Nick Confessore made the point in a Twitter post, that the billion about to spent on the presidential race is really just targeting the 3% of the population that can be swayed. Because we're talking about Trump, I think that number is higher than 3% this year - and what the polls probably show is they're voting for Hillary. I guess we'll know in a few months.
7
"Ordinary people seem to sense that he says what he means " . Please, can we have a candidate that does not require a oujia board to figure out what the heck he really means. The oblique and hyperbolic statements Trump consistently makes are intended to simply pander to his supporters
13
Looking at the most recent Washington Post/ABC News poll showing Clinton leading Trump 42-38 among registered voters, I see a few other fascinating answers the media doesn't necessarily want to highlight:
Respondents would prefer someone with experience in politics vs. someone outside the political establishment by a margin of 55-41. So much for the desired narrative by many in the media that 2016 is inherently an anti-establishment election year.
Respondents think Clinton has a better personality/temperament to be President than Trump by a margin of 59-28. This seems like a hugely important and telling difference.
Respondents think Clinton has better judgment than Trump by a margin of 54-35.
Regardless of who they're voting for, 57% of respondents rate Clinton as being qualified to be President, while only 37% rate Trump as being qualified. 60% think he's unqualified.
56% of respondents approve of President Obama's job performance.
As a Clinton supporter, I'm very encouraged by these unhighlighted poll answers. But I'm also acutely aware they're only a snapshot, and complacency could just as easily lead to a loss in November.
Respondents would prefer someone with experience in politics vs. someone outside the political establishment by a margin of 55-41. So much for the desired narrative by many in the media that 2016 is inherently an anti-establishment election year.
Respondents think Clinton has a better personality/temperament to be President than Trump by a margin of 59-28. This seems like a hugely important and telling difference.
Respondents think Clinton has better judgment than Trump by a margin of 54-35.
Regardless of who they're voting for, 57% of respondents rate Clinton as being qualified to be President, while only 37% rate Trump as being qualified. 60% think he's unqualified.
56% of respondents approve of President Obama's job performance.
As a Clinton supporter, I'm very encouraged by these unhighlighted poll answers. But I'm also acutely aware they're only a snapshot, and complacency could just as easily lead to a loss in November.
115
Thanks Nate. It is this type of expert analysis, completed with unfettered honesty, that we need to pay attention to. There is undeniable education gap in the US and this is one of the key reasons that polling important national issues makes accuracy difficult.
I am surprised that there are so few comments on this article.
I am surprised that there are so few comments on this article.
6
Maybe we should be gratified that the trolls seem to be avoiding this article. I was going to say they have no way to debunk fact-filled, well-researched stories, but Lord knows the truth has never stopped them before this cycle.
9
Actually as a liberal I find these polls to be encouraging. IF people get out to vote.
As only landlines are polled anyone with VoIP or a cellphone is not polled. These non-polled people tend to skew younger, more technologically savvy, better educated AND liberal. The slice being polled is likely one of the worst case scenarios. Which is to say older, less educated and more conservative. And less likely to vote for HRC.
If HRC is still doing well in this environment, she should do well overall.
IF people get out and vote. That is the real key. You can poll anyone but if they don't vote what they think is not a factor.
disclaimer: I work in Business Intelligence where recognizing unforeseen factors that skew the data can be significant.
As only landlines are polled anyone with VoIP or a cellphone is not polled. These non-polled people tend to skew younger, more technologically savvy, better educated AND liberal. The slice being polled is likely one of the worst case scenarios. Which is to say older, less educated and more conservative. And less likely to vote for HRC.
If HRC is still doing well in this environment, she should do well overall.
IF people get out and vote. That is the real key. You can poll anyone but if they don't vote what they think is not a factor.
disclaimer: I work in Business Intelligence where recognizing unforeseen factors that skew the data can be significant.
26
UNfortunately as Brexit has shown us these non-polled people don't seem to be very good about actually voting. Let's hope they paid enough attention to what happened as a result to keep the lightning of ignorance from striking again here in November.
4
Soemthing is seriously wrong w/Quinnipiac polls lately.....They had Toomey +10 & Rubio +17 in their respective senate races.....Neither of those results will ever happen, they may win, but it would be by about 1-2%....Rubio lost in the Florida GOP primary by 20% to Trump, yet we are to believe he's winning by 17% in a senate race?.....
Goes to show, if your modeling and/or sample size is all wrong, you can get any result you want...I'm sure I could come up with a poll showing Clinton leading in the deepest of red states like Oklahoma, Idaho, Wyoming, etc if I oversample Democrats or use a statistical model that might have worked in the 1964 election but isn't relevant for 2016...
Goes to show, if your modeling and/or sample size is all wrong, you can get any result you want...I'm sure I could come up with a poll showing Clinton leading in the deepest of red states like Oklahoma, Idaho, Wyoming, etc if I oversample Democrats or use a statistical model that might have worked in the 1964 election but isn't relevant for 2016...
6
Sample size not as big deal. The difference between 500 sample and 1000 sample is not twice as good.
But yes the method of determining likely voter is important. But also important is when the electorate just doesn't really know what they intend to do and thus the polling reflects that additional uncertainty.
But yes the method of determining likely voter is important. But also important is when the electorate just doesn't really know what they intend to do and thus the polling reflects that additional uncertainty.
1
the dang thing is scary.
polls wrote bernie off in the begining.
polls show hillary extremely distrusted.
polls show drumpf up and down.
predicting this race is going to be verry difficult, and we all
may end up in deep dark yogurt.
polls wrote bernie off in the begining.
polls show hillary extremely distrusted.
polls show drumpf up and down.
predicting this race is going to be verry difficult, and we all
may end up in deep dark yogurt.
5
Worried about the polls? This lifelong Dem hasn't been worried one iota. It's Hillary's to lose. For all the bashing of Hillary from almost every corner of the opposition universe she has absorbed it all and prevailed. She gets up, dusts herself off and marches on.
Trump is a joke, a jerk, a complete buffoon. When you have the most prominent Republicans, former nominees and presidents, shun this convention leaving Trump with having to fill half his speakers roster with members of his family you are looking at disaster. There are delegates who don't want to comment publicly or be photographed due to the embarrassment. They know they have a loser and are doing their best to stay away from this stink. The GOP seems determined to keep their reputation as the Stupid Party. Nominating Trump guarantees it.
DD
Manhattan
Trump is a joke, a jerk, a complete buffoon. When you have the most prominent Republicans, former nominees and presidents, shun this convention leaving Trump with having to fill half his speakers roster with members of his family you are looking at disaster. There are delegates who don't want to comment publicly or be photographed due to the embarrassment. They know they have a loser and are doing their best to stay away from this stink. The GOP seems determined to keep their reputation as the Stupid Party. Nominating Trump guarantees it.
DD
Manhattan
169
Isn't it possible the GOP wants this? With all of the attention on the national race with a forgone outcome, they are free to invest everything in keeping Congress red. Complain and obstruct, but achieve nothing. Sounds like their dream plan.
6
@Dennis
You say "It's Hilary's to lose" and you say you're not worried. I hope your confidence is justified. But I think that if it's Hilary's to lose, there's good reason to worry.
You say "It's Hilary's to lose" and you say you're not worried. I hope your confidence is justified. But I think that if it's Hilary's to lose, there's good reason to worry.
2
Dear Robert T:
I'm not sure if the GOP does much thinking at all. If they did they would have done something to prevent this Trump train wreck. They're stuck with the guy and trying their darnedest to extricate themselves from the aftermath.
DD
NYC
I'm not sure if the GOP does much thinking at all. If they did they would have done something to prevent this Trump train wreck. They're stuck with the guy and trying their darnedest to extricate themselves from the aftermath.
DD
NYC
2
The "Polls" include 4 candidates. Your article included 2. Subtraction?
R's & D's do not like their choices.
Independents like the 2 top rackets less.
Gary Johnson has been trending up.
You further the myth that there is a binary choice - that is false reporting.
And, you know that exposure means influence.
Have Integrity. Analyze, Report, but do not introduce bias. Gary is a real option.
R's & D's do not like their choices.
Independents like the 2 top rackets less.
Gary Johnson has been trending up.
You further the myth that there is a binary choice - that is false reporting.
And, you know that exposure means influence.
Have Integrity. Analyze, Report, but do not introduce bias. Gary is a real option.
6
Neither Gary Johnson nor Jill Stein has any chance of being elected president. A vote for Stein is a vote for Trump, and I don't know what a vote for Johnson is, other than a wasted vote. For those in the real world concerned about this nation and a Trump presidency, this election is indeed a binary choice, the starkest choice in 100 years.
6
No, just another Nader or Perot.
3
Be careful America, the motivation of the disenfranchised can surprise. The vast majority of polls said Brexit would fail.
50
But they were all close.
3
Exactly.
1
Thanks for this. I stopped paying attention to polls because they were so widely divergent.
Wishful thinking?
1
As the late Harold Wilson once said, "A week is a long time in politics." This is especially true this year, when shocking events follow each other daily. Although many judge Donald Trump not fit to be president, terrorist attacks, the murders of police, and persistent economic worries may goad a sizeable fraction into voting for him.
Although they won't, Times readers, including myself, should probably ignore the polls until after Labor Day.
Although they won't, Times readers, including myself, should probably ignore the polls until after Labor Day.
8
The baseball analogy seems very apropos given how the media treats every election like a major sports contest, irrespective of the profound consequences upon all citizens (fans) in America.
7
I'm glad this article was written. These days polls seem to be used to sensationalize the news rather than give rational perspective on what is going on in the world. In a world where social media, and media in general seem to leap at any bit of information as if it is complete and objective, it is wise to step back and have perspective on what is going on. I wish this kind of reporting was more the norm rather than the exception in our confusing times. Frankly, we need that sort of view rather than our current diet of sensationalism.
47
It is incredibly distressing to me to see that a man as unfit and as crude and despicable as Donald Trump has any real chance of winning the Presidency of the United States. His very candidacy insults the Office.
284
So we should elect someone who is corrupt and is a serial liar? Many people have no idea who Hillary Clinton is. Read a few of anti-Clinton books. If just 10% of what was written in those books were true most people would look at Trump in a different light. Clinton has a long, documented history of ethical lapses that should concern every intelligent voter. When they see her time and time again refuse to answer even the simplest question about important matters, most intelligent people would begin to wonder about her willingness to tell the truth about anything. I laugh through most of her interviews. Her husband was a master at deflection. She just looks like a pathetic third grader trying to lie her way out of taking responsibility for spilling a glass of milk.
4
If you are seeking a truth benchmark of 10% from authors like Glen Beck and Ann Coulter as the basis for your Trump vote - you could end up voting for Hillary! BTW... She's a liar and they only have to hit the truth one out of 10 times? If truthiness matters (thanks Mr Colbert) Hillary rates far better than Trump. Right?
3
Edward Warren - I think you need to look at Hillary's motives and the outcomes of her actions. Unlike Trump, Hillary doesn't live to hurt other people.
7
Polling has been given a rough ride of late. In pretty rapid succession it has failed to accurately predict Cameron's Conservative majority in the UK, Trudeau's Liberal majority in Canada, the Brexit vote result among other surprises here and abroad. It would appear that one would be wise to take any poll or amalgamation of polls in the Clinton-Trump contest with a rather large grain of salt.
40
Brexit polls showed leaving with the advantage in the final week.
3
As pointed out "poll shopping" is very dangerous. It lead the Republicans, including their guru of raw numbers, Karl Rove, to make such catastrophically bad inferences from the polls in 2012, to assume Romney would win easily, when by the end of September of that years, careful state-by-state analysis, attempted to neutralize personal bias, indicated that Obama would easily top 300 electoral votes and the Democrats would hold the Senate, but not retake the House.
And, when looking at a "poll of polls" it's very important to throw out the outliers, as they are usually wrong. A rare exception with this year's Democratic primary in Michigan.
This election will turn on turn-out--which party marshals its members and hangers-on best. Thus Trump's otherwise-inane selection of Pence was a sound attempt to rouse Cruz voters and GOP donors. Yet the way he back-handed both Christie and Gingrich may come back to haunt him, though both would have been, from the outside looking in, bad picks.
Tangentially, I'm guessing that if Son-in-Law Kushner convinces Trump NOT to promise Christie the AG slot, Christie may resign from the campaign. (Christie put Kushner's crooked dad in prison).
And, when looking at a "poll of polls" it's very important to throw out the outliers, as they are usually wrong. A rare exception with this year's Democratic primary in Michigan.
This election will turn on turn-out--which party marshals its members and hangers-on best. Thus Trump's otherwise-inane selection of Pence was a sound attempt to rouse Cruz voters and GOP donors. Yet the way he back-handed both Christie and Gingrich may come back to haunt him, though both would have been, from the outside looking in, bad picks.
Tangentially, I'm guessing that if Son-in-Law Kushner convinces Trump NOT to promise Christie the AG slot, Christie may resign from the campaign. (Christie put Kushner's crooked dad in prison).
11
The media wants horse races thus they give lots of attention to polls that further this narrative so that inevitably means talking up Trump who sticking with racing parlance is seen as the outsider. Clinton almost certainly doesn't mind since they serve to dispel any sense of complacency amongst Democrats or anyone who views the prospect of Trump with horror. The other fact to bear in mind is that these polls aren't worth much until late August. Cohn's analysis right now is probably not far off the mark. Clinton has a lead of =/- 4% and has had for a couple of months. Obama won in 2012 by about 3.5%.
77
Unless there is an overwhelming majority for a given candidate, races that are more problematic cannot be reflected in polls in which only a thousand persons are canvassed. For a poll to reflect the population,it must put together a bewildering number of elements:
Percentage of men vs women
Percentage of young vs old
Percentage of rural vs urban
Percentage of potential Black voters
Percentage of Latinos and other ethnic populations
Percentage of land lines vs. cell phones
Percentage of registered likely voters
Percentage of new potential voters
So a poll must assess what weight to give each of these elements to arrive at a right conclusion. Anything else will skew the results and invalidate the poll. They say that polls are just a snapshot of a moment, it may be true but it is a blurred one.
Percentage of men vs women
Percentage of young vs old
Percentage of rural vs urban
Percentage of potential Black voters
Percentage of Latinos and other ethnic populations
Percentage of land lines vs. cell phones
Percentage of registered likely voters
Percentage of new potential voters
So a poll must assess what weight to give each of these elements to arrive at a right conclusion. Anything else will skew the results and invalidate the poll. They say that polls are just a snapshot of a moment, it may be true but it is a blurred one.
12
Or you use statistical science to randomly poll people, accepting that the margin of error reflects the risk that your sample size wasn't large enough to likely have selected a representative sample. No pollster deliberately tries to get the right percentages of each of the classifications you listed other than considering the landline versus cellphone question. To do so would make the sampling non-random.
2
Sample size is seldom the issue.
Although, dubious, there are pollsters who try to "adjust" for demographics.
The biggest question is operational defining who is a likely voter.
How likely to vote do you have to be to be a likely voter? Is a a would be voter who claims to be 85% likely to vote but has missed voting I the last 4 elections really a likely voter? Is the disheartened voter who self describes as only 40% likely to vote but who has never missed an election in 40 years, really not a likely voter?
Good statistical science is required, but also required as W E Deming once noted is an understanding of the underlying substantive process, in this case the psychology and science of who votes and who doesn't and why?
Although, dubious, there are pollsters who try to "adjust" for demographics.
The biggest question is operational defining who is a likely voter.
How likely to vote do you have to be to be a likely voter? Is a a would be voter who claims to be 85% likely to vote but has missed voting I the last 4 elections really a likely voter? Is the disheartened voter who self describes as only 40% likely to vote but who has never missed an election in 40 years, really not a likely voter?
Good statistical science is required, but also required as W E Deming once noted is an understanding of the underlying substantive process, in this case the psychology and science of who votes and who doesn't and why?
Add to that - nuanced questions.
I have no doubt that there is "poll shopping" by organizations/individuals looking, not for accuracy but for confirmation. Given the performance of the media, the pols themselves, "respected" elites, professional societies, governmental organizations, why should we assume that polling organizations are paragons of virtue? They too want to make a buck. They too are "giving the customer what they want".
Then too, maybe, like in baseball, some are simply not as good at it as others.
Then too, maybe, like in baseball, some are simply not as good at it as others.
5
Actually what I'd like to see is more reporting on Senate and House polls. The Republicans have shown that they are a hugely disproportionate blocking presence in any but an overwhelmingly Democratic congress. (And let us not forget what they have done in the way too politicized court system.)
146
Concur. If HRC were to win, a continued Republican-run Congress would be looking from its first day in session not just to block her program, but for concocted reasons to impeach her. To run out the clock until election day 2020. Only the election of a Democratic Senate this year would be able to stifle such efforts arising from a Republican House. A consistent Clinton lead in the polls in the next few months should free up serious time, money, and people within the next few months to help take back the Senate from the Right.
7
With more than twice as many Republican Senate seats up for re-election this year (reverse of 2014) and with Trump at the top of the ticket, there's a very good chance that President Clinton will be working with a Democratic Senate -- at least for her first two years. Evan Bayh jumping into Indiana's U.S. Senate race is a very strong indication that a Democratic Senate majority is on the horizon. He is not one to take big risks, and has never lost an election.
4
Thanks for the analysis of poll results. The column provides an experienced perspective that I do not get from individual poll results.
30
"It’s a lot like baseball. Even great baseball players go 0 for 4 in a game — or have rough stretches for weeks on end. On the other end might be a few multihit nights with extra-base hits, or a spectacular few weeks."
That is a terrible analogy. One baseball game is an experiment of n=4 at bats (four samples to determine how well the hitter is performing). Any poll, no matter how small, typically has hundreds or thousands of samples (respondents.) There may be sampling bias or any other number of methodological flaws at work, but its not the same as asking four people for their opinions.
That is a terrible analogy. One baseball game is an experiment of n=4 at bats (four samples to determine how well the hitter is performing). Any poll, no matter how small, typically has hundreds or thousands of samples (respondents.) There may be sampling bias or any other number of methodological flaws at work, but its not the same as asking four people for their opinions.
6
Really, it's a very good analogy. Four at-bats in the course of a single game tells little about a batters ability, nor does a dozen at-bats over a three game series. But six-hundred at-bats over the course of a season shows who are the best hitters. Performance in thousands of at-bats over the course of a long career are what separates even good players from Hall-of-Famers.
However, baseball like politics is much more complex. Neither can be reduced to a simple formula or analogy.
However, baseball like politics is much more complex. Neither can be reduced to a simple formula or analogy.
12
It would seem you missed the point of the analogy. He is comparing the CANDIDATE to a baseball player, not the poll. The candidate has hot and cold streaks, and these are reflected in the time-slice of the poll. But, like a quality player, the quality of the candidate is reflected by the overall results, not any particular sample.
6
What is important in evaluating polls is the margin of error, usually given as + or - some value. Most people don't know what this means. It's simply this: the margin of error gives a range of values in which there is a 95% chance the true value will fall. Stated another way, there is only a 5% chance the true value will fall outside this range. There can also be a more precise estimate such as 1% level of confidence in which there is a 99% chance that the true value will fall in this range. In large part the level of confidence in a poll is determined by the number in the sample; greater numbers give higher levels of confidence; also, how variable the sample is. If the first twenty randomly drawn subjects all answer the same way, you probably don't need any more as long as those subjects were selected randomly. Generally, people equate two values that fall within the margin of error as the same - the difference is statistically insignificant.
Finally if the margin of error is very low, you really don't need a large number of subjects. If you sample 10 subjects and they all respond the same way, that's probably adequate.
Finally if the margin of error is very low, you really don't need a large number of subjects. If you sample 10 subjects and they all respond the same way, that's probably adequate.
1