Are the Polls Accurate? Here is the Unbiased Data.

[WelfareQueen]

All elections are won by one simple factor......turnout. It truly is that simple. When 40% to 50% of the electorate doesn't show up to vote in Presidential elections what is important is who does show up to vote.

Obama won in 2008 and 2012 because the Dems enjoyed a 7%-8% turnout advantage over the GOP. Romney got high voter turnout from traditional GOP voters...but Obama got more. In fact, minority voters turned out for Obama 30% higher than any other election in history.

Pollsters obviously got things right in 2008 and 2012. Most built in a 7%-8% Dem turnout advantage. However, in the 2014 mid-terms the pollster fucked up big time. Remember all of those Senate races that were "too close to call." Remember Kentucky.....North Carolina....Kansas....Colorado.....Georgia. The GOP won every close Senate race but one...in New Hampshire.

The reason....The pollsters on average had the GOP with a 1.5% turnout advantage, when in fact it was 4.5%. That 3% margin is the sole difference between winning and losing elections.

So that brings us to 2016. Here is data from Nate Silver (hardly a GOP supporter).

FiveThirtyEight's Pollster Ratings


You can see the built in bias from each polling organization. Notice most of the biggest pollsters and factoring in a slight Dem turnout advantage. Internal data I have seen is indicating the GOP should have between a +2% to 5% turnout advantage this election cycle. If so Trump probably wins the election. That is not currently reflected in virtually any polls. Please note...almost all the major polling organizations are now showing a slight Dem turnout advantage. I am very skeptical to say the least.

Here is another excellent article about polling from Nate Silver. I think this is very fair and accurate.

The State Of The Polls, 2016

I am copying Doc because he has been a political operative and I am curious what he thinks. Please add your comments if you are interested. Thanks.


theDoctorisIn

 

[Rexx Taylor]

well being we have about 40 million pissed off white guys right now,,they will be dying to vote!!

 

[theDoctorisIn]

Did you leave out a link? I'm curious to read any article by Silver, but I only see a link to the pollster ratings.

 

[WelfareQueen]

Did you leave out a link? I'm curious to read any article by Silver, but I only see a link to the pollster ratings.

Sorry...I just posted it. My mistake.

 

[theDoctorisIn]

As for the thesis of your post, you're 100% correct in saying that turnout is key - but it's not as simple as saying "+X %" advantage, turnout matters very differently in different places. Also, turnout is one of the few things that a campaign can directly influence via GOTV operations - and the Democrats have been getting very good at that in the last few years.

Also, the voter share that's truly controlled by their party (the Ds who always vote D, the Rs who always vote R) is shrinking by the day - the key is going to be which way the middle goes.

 

[theDoctorisIn]

Did you leave out a link? I'm curious to read any article by Silver, but I only see a link to the pollster ratings.

Sorry...I just posted it. My mistake.

Thanks for posting that, it's a great article. I've been a big fan of Silver for a while now.

I think it's an optimistic view of the state of polling, though. There's no question that polling is going through somewhat of a watershed situation right now - the fact that cell phones are replacing landlines for many people (myself included) particularly youngish people, who are also a rising voting demographic is a serious problem for pollsters, since polling cell phones is both more expensive and less productive than landlines. The prevalence of caller ID has also contributed to declining response rates as well. Some companies are trying online polling, but that's a whole different can of worms and there's little evidence that it has any statistical value.

 

[WelfareQueen]

As for the thesis of your post, you're 100% correct in saying that turnout is key - but it's not as simple as saying "+X %" advantage, turnout matters very differently in different places. Also, turnout is one of the few things that a campaign can directly influence via GOTV operations - and the Democrats have been getting very good at that in the last few years.

Also, the voter share that's truly controlled by their party (the Ds who always vote D, the Rs who always vote R) is shrinking by the day - the key is going to be which way the middle goes.

I agree with your comments, but would add one thing. Which side in this election has the most intensity (i.e. voters who may want to lean Dem or GOP but are not necessarily party affiliated).

Clearly Dem leaning voters in 2008 and 2012 had more energy after 8 years of Bush than voters who might lean toward the GOP. Clearly in the 2014 mid-terms after 6 years of Obama GOP leaning voters had more energy.

Doc..in my mind this question will determine the election in 2016. Which side has more energy or intensity? Do you have any thoughts on this question?

 

[Sherry]

With that sort of methodology, accompanied by the negative press, I'm sure they expect it to have a psychological impact on swing voters in the polling booth...time will tell.

 

[WelfareQueen]

Please note.....per Nate Silver's data most big name pollsters currently have a slight Democrat bias. These include Pew, NBC, ABC, Fox, Reuters, Time, NYT Times...etc.

Gallup and Rasmussen are the only major pollsters I can see with a slight GOP built in bias.

 

[theDoctorisIn]

As for the thesis of your post, you're 100% correct in saying that turnout is key - but it's not as simple as saying "+X %" advantage, turnout matters very differently in different places. Also, turnout is one of the few things that a campaign can directly influence via GOTV operations - and the Democrats have been getting very good at that in the last few years.

Also, the voter share that's truly controlled by their party (the Ds who always vote D, the Rs who always vote R) is shrinking by the day - the key is going to be which way the middle goes.

I agree with your comments, but would add one thing. Which side in this election has the most intensity (i.e. voters who may want to lean Dem or GOP but are not necessarily party affiliated).

Clearly Dem leaning voters in 2008 and 2012 had more energy after 8 years of Bush than voters who might lean toward the GOP. Clearly in the 2014 mid-terms after 6 years of Obama GOP leaning voters had more energy.

Doc..in my mind this question will determine the election in 2016. Which side has more energy or intensity? Do you have any thoughts on this question?

I think it's an over-simplification. "Energy" and "intensity", party-wide, is too broad and undefined. Turnout is more complicated that feelings - the people who decide every election are the ones who aren't energetic or intense about either candidate - the ones who don't really care or follow politics. A good GOTV operation in the right place will make more a difference than the "intensity" of voters in, say, New York City.

 

[theDoctorisIn]

Please note.....per Nate Silver's data most big name pollsters currently have a slight Democrat bias. These include Pew, NBC, ABC, Fox, Reuters, Time, NYT Times...etc.

Gallup and Rasmussen are the only major pollsters I can see with a slight GOP built in bias.

Also Quinnipiac, Marist, Siena, LA Times, Ohio Poll, Fox/Opinion Dynamics, Public Policy Institute.

Those are all "major" pollsters, too.

 

[theDoctorisIn]

The truth of the matter is that there's no single bellwether for a Presidential election. Things like "energy" and "intensity" are nearly impossible to empirically measure, turnout depends on too many variables, other metrics like money, PAC support, endorsements, and so on are all just small parts of the bigger game.

 

[WelfareQueen]

Please note.....per Nate Silver's data most big name pollsters currently have a slight Democrat bias. These include Pew, NBC, ABC, Fox, Reuters, Time, NYT Times...etc.

Gallup and Rasmussen are the only major pollsters I can see with a slight GOP built in bias.

Also Quinnipiac, Marist, Siena, LA Times, Ohio Poll, Fox/Opinion Dynamics, Public Policy Institute.

Those are all "major" pollsters, too.

And interestingly.....Fox, CBS/NYT, ABC/Washington Post all currently have a .6% Democrat bias. When the spread is now 7 points between all polling organizations it tells me they are uncertain and the polls are not to be trusted very highly. The last three election cycles there was general consensus among pollsters and they were mostly correct.

The fact there does not appear to be strong consensus is troubling.

 

[sonic]

The effect of turnout is also relative to the closeness of the race. When push comes to shove, Trump is an exceptionally weak candidate who is faltering badly, and while a there is relatively small group who are very enthusiastic about him, the majority of Americans are not at all.
I like Nate's work a lot, and enjoy the 538 elections podcast - good stuff.
His polls-only forecast as of today has Trump with about a 16% chance of winning, with his more sophisticated Polls+ Forecast at a 24% chance.

 

[theDoctorisIn]

Please note.....per Nate Silver's data most big name pollsters currently have a slight Democrat bias. These include Pew, NBC, ABC, Fox, Reuters, Time, NYT Times...etc.

Gallup and Rasmussen are the only major pollsters I can see with a slight GOP built in bias.

Also Quinnipiac, Marist, Siena, LA Times, Ohio Poll, Fox/Opinion Dynamics, Public Policy Institute.

Those are all "major" pollsters, too.

And interestingly.....Fox, CBS/NYT, ABC/Washington Post all currently have a .6% Democrat bias. When the spread is now 7 points between all polling organizations it tells me they are uncertain and the polls are not to be trusted very highly. The last three election cycles there was general consensus among pollsters and they were mostly correct.

The fact there does not appear to be strong consensus is troubling.

There is consensus, though. A point or two in either direction doesn't make the polls conflict.

 

[WelfareQueen]

The effect of turnout is also relative to the closeness of the race. When push comes to shove, Trump is an exceptionally weak candidate who is faltering badly, and while a there is relatively small group who are very enthusiastic about him, the majority of Americans are not at all.
I like Nate's work a lot, and enjoy the 538 elections podcast - good stuff.

I think right now no one has a real good idea about turnout, but I think the GOP will end up having around a 2%-3% turnout advantage. Generally the party that has been in power 8 years has less enthusiasm. That certainly held true in 2008. But if I am wrong I will freely admit it.

 

[NYcarbineer]

All elections are won by one simple factor......turnout. It truly is that simple. When 40% to 50% of the electorate doesn't show up to vote in Presidential elections what is important is who does show up to vote.

Obama won in 2008 and 2012 because the Dems enjoyed a 7%-8% turnout advantage over the GOP. Romney got high voter turnout from traditional GOP voters...but Obama got more. In fact, minority voters turned out for Obama 30% higher than any other election in history.

Pollsters obviously got things right in 2008 and 2012. Most built in a 7%-8% Dem turnout advantage. However, in the 2014 mid-terms the pollster fucked up big time. Remember all of those Senate races that were "too close to call." Remember Kentucky.....North Carolina....Kansas....Colorado.....Georgia. The GOP won every close Senate race but one...in New Hampshire.

The reason....The pollsters on average had the GOP with a 1.5% turnout advantage, when in fact it was 4.5%. That 3% margin is the sole difference between winning and losing elections.

So that brings us to 2016. Here is data from Nate Silver (hardly a GOP supporter).

FiveThirtyEight's Pollster Ratings


You can see the built in bias from each polling organization. Notice most of the biggest pollsters and factoring in a slight Dem turnout advantage. Internal data I have seen is indicating the GOP should have between a +2% to 5% turnout advantage this election cycle. If so Trump probably wins the election. That is not currently reflected in virtually any polls. Please note...almost all the major polling organizations are now showing a slight Dem turnout advantage. I am very skeptical to say the least.

Here is another excellent article about polling from Nate Silver. I think this is very fair and accurate.

The State Of The Polls, 2016

I am copying Doc because he has been a political operative and I am curious what he thinks. Please add your comments if you are interested. Thanks.


theDoctorisIn

Are you aware that every time the RWnuts like you claim the polls are biased because a Democrat is winning it turns out that you were full of shit?

 

[WelfareQueen]

Please note.....per Nate Silver's data most big name pollsters currently have a slight Democrat bias. These include Pew, NBC, ABC, Fox, Reuters, Time, NYT Times...etc.

Gallup and Rasmussen are the only major pollsters I can see with a slight GOP built in bias.

Also Quinnipiac, Marist, Siena, LA Times, Ohio Poll, Fox/Opinion Dynamics, Public Policy Institute.

Those are all "major" pollsters, too.

And interestingly.....Fox, CBS/NYT, ABC/Washington Post all currently have a .6% Democrat bias. When the spread is now 7 points between all polling organizations it tells me they are uncertain and the polls are not to be trusted very highly. The last three election cycles there was general consensus among pollsters and they were mostly correct.

The fact there does not appear to be strong consensus is troubling.

There is consensus, though. A point or two in either direction doesn't make the polls conflict.

There is much more than a 1 point spread between the polls. The range is from Dem +2.5 to GOP +4.5. Granted....those are the outliers....but a seven point spread is a huge difference.

 

[WelfareQueen]

All elections are won by one simple factor......turnout. It truly is that simple. When 40% to 50% of the electorate doesn't show up to vote in Presidential elections what is important is who does show up to vote.

Obama won in 2008 and 2012 because the Dems enjoyed a 7%-8% turnout advantage over the GOP. Romney got high voter turnout from traditional GOP voters...but Obama got more. In fact, minority voters turned out for Obama 30% higher than any other election in history.

Pollsters obviously got things right in 2008 and 2012. Most built in a 7%-8% Dem turnout advantage. However, in the 2014 mid-terms the pollster fucked up big time. Remember all of those Senate races that were "too close to call." Remember Kentucky.....North Carolina....Kansas....Colorado.....Georgia. The GOP won every close Senate race but one...in New Hampshire.

The reason....The pollsters on average had the GOP with a 1.5% turnout advantage, when in fact it was 4.5%. That 3% margin is the sole difference between winning and losing elections.

So that brings us to 2016. Here is data from Nate Silver (hardly a GOP supporter).

FiveThirtyEight's Pollster Ratings


You can see the built in bias from each polling organization. Notice most of the biggest pollsters and factoring in a slight Dem turnout advantage. Internal data I have seen is indicating the GOP should have between a +2% to 5% turnout advantage this election cycle. If so Trump probably wins the election. That is not currently reflected in virtually any polls. Please note...almost all the major polling organizations are now showing a slight Dem turnout advantage. I am very skeptical to say the least.

Here is another excellent article about polling from Nate Silver. I think this is very fair and accurate.

The State Of The Polls, 2016

I am copying Doc because he has been a political operative and I am curious what he thinks. Please add your comments if you are interested. Thanks.


theDoctorisIn

Are you aware that every time the RWnuts like you claim the polls are biased because a Democrat is winning it turns out that you were full of shit?

Settle down Beavis.....I am not saying the polls are inaccurate and the quote of bias is from Nate Silver....not me. Read the information before you comment.

 

[NYcarbineer]

All elections are won by one simple factor......turnout. It truly is that simple. When 40% to 50% of the electorate doesn't show up to vote in Presidential elections what is important is who does show up to vote.

Obama won in 2008 and 2012 because the Dems enjoyed a 7%-8% turnout advantage over the GOP. Romney got high voter turnout from traditional GOP voters...but Obama got more. In fact, minority voters turned out for Obama 30% higher than any other election in history.

Pollsters obviously got things right in 2008 and 2012. Most built in a 7%-8% Dem turnout advantage. However, in the 2014 mid-terms the pollster fucked up big time. Remember all of those Senate races that were "too close to call." Remember Kentucky.....North Carolina....Kansas....Colorado.....Georgia. The GOP won every close Senate race but one...in New Hampshire.

The reason....The pollsters on average had the GOP with a 1.5% turnout advantage, when in fact it was 4.5%. That 3% margin is the sole difference between winning and losing elections.

So that brings us to 2016. Here is data from Nate Silver (hardly a GOP supporter).

FiveThirtyEight's Pollster Ratings


You can see the built in bias from each polling organization. Notice most of the biggest pollsters and factoring in a slight Dem turnout advantage. Internal data I have seen is indicating the GOP should have between a +2% to 5% turnout advantage this election cycle. If so Trump probably wins the election. That is not currently reflected in virtually any polls. Please note...almost all the major polling organizations are now showing a slight Dem turnout advantage. I am very skeptical to say the least.

Here is another excellent article about polling from Nate Silver. I think this is very fair and accurate.

The State Of The Polls, 2016

I am copying Doc because he has been a political operative and I am curious what he thinks. Please add your comments if you are interested. Thanks.


theDoctorisIn

Are you aware that every time the RWnuts like you claim the polls are biased because a Democrat is winning it turns out that you were full of shit?

Settle down Beavis.....I am not saying the polls are inaccurate and the quote of bias is from Nate Silver....not me. Read the information before you comment.

You're not biased? That's the best lie I've ever heard on USMB.