Imagining a World without Polls By Jonah Goldberg

[mystery-ak]

http://www.nationalreview.com/node/426669/print

 Imagining a World without Polls
By Jonah Goldberg — November 6, 2015

What if the polls just stopped working?

Admittedly, this needs work as a plot device for a Stephen King novel. But for politics, it might be pretty awesome.

This week, businessman Matt Bevin won a stunning upset in the Kentucky governor’s race. It was only the second time in more than four decades that a Republican took the governor’s mansion in the Bluegrass State. Bevin’s margin of victory: nine percentage points.

Bevin’s win was big political news for a lot of reasons. Kentucky’s state health-care exchange, Kynect, was supposed to be the shining success story of Obamacare. Bevin vowed to dismantle it, a fatal mistake according to many inside-the-Beltway types.

The results in Kentucky — along with state-senate elections in Virginia — also demonstrated that however successful Barack Obama has been as a president, he’s been terrible for the Democratic party. On his watch, Democrats have lost more than 900 seats in state legislatures, 12 governorships, 69 congressional seats, and 13 Senate seats. The GOP, according to the Washington Post’s Chris Cillizza, has full or partial control of 76 percent of state legislatures.

continued

[Bigun]

Ahem!  As I have been saying polls are pretty much worthless and getting more so every day because of lifestyle changes among the polling sample base.

[aligncare]

"What if the polls just stopped working?"

Well then, poll dancers hit the floor.

[Oceander]

"What if the polls just stopped working?"

Well then, poll dancers hit the floor.

[Oceander]

Imagine there's no heaven
It's easy if you try
No hell below us
Above us only sky
Imagine all the people living for today
Imagine there's no countries
It isn't hard to do
Nothing to kill or die for
And no religion too
Imagine all the people living life in peace

*  *   *

[Right_in_Virginia]

Ahem!  As I have been saying polls are pretty much worthless and getting more so every day because of lifestyle changes among the polling sample base.

I don't think polls are worthless.  They are often designed for a desired result and those results are used to set the agenda and media dialogue.

[Bigun]

I don't think polls are worthless.  They are often designed for a desired result and those results are used to set the agenda and media dialogue.

I that regard you are entirely correct!   

[Right_in_Virginia]

I that regard you are entirely correct! 

 

[truth_seeker]

Is there a post election analysis for Kentucky? Why were the latest polls so far off?

If done correctly, there is nothing wrong with the "science" of polling, which is merely statistical sampling.

 

[Bigun]

Is there a post election analysis for Kentucky? Why were the latest polls so far off?

If done correctly, there is nothing wrong with the "science" of polling, which is merely statistical sampling.

I have never argued otherwise.  What I do argue with is accepting any and all public consumption polling at face value.  The same people who know how to do perfectly valid statistical sampling know how to skew the sample in any fashion they might wish to skew it and routinely do so.

 

[truth_seeker]

I have never argued otherwise.  What I do argue with is accepting any and all public consumption polling at face value.  The same people who know how to do perfectly valid statistical sampling know how to skew the sample in any fashion they might wish to skew it and routinely do so.

Any thoughts on why the polls were wrong, for Kentucky?

[Bigun]

Any thoughts on why the polls were wrong, for Kentucky?

Yes. Primarily faulty assumptions and the rapidly increasing difficultly in obtaining a valid sample.

[truth_seeker]

Yes. Primarily faulty assumptions and the rapidly increasing difficultly in obtaining a valid sample.

I am interested in two largely separate subjects;

--Dishonest polls intended to influence the outcome, and

--Difficulty to take a statistically representative sample in "fair" polls

The first makes no attempt to be accurate, but the latter does.

I am interested in the shortcomings of polling, with regard to methods and the latter.

[Fishrrman]

I don't talk to pollsters.

Mainly, because the phone isn't answered at my house.

But even if I did have to talk to one of them, I'd lie!

[PzLdr]

I'd rather imagine one without Liberals. Most polls don't deliberately lie all the time.

[Bigun]

I am interested in two largely separate subjects;

--Dishonest polls intended to influence the outcome, and

--Difficulty to take a statistically representative sample in "fair" polls

The first makes no attempt to be accurate, but the latter does.

I am interested in the shortcomings of polling, with regard to methods and the latter.

What you and I would call "dishonest" polls are the ones which start with an end result  and then design the poll to get exactly that result.  It's all about making the necessary assumptions, knowing how to frame questions, who to call, and what time to call ect. These are done ALL the time for all kinds of reasons.

"Honest" polls  are those which actually seek to measure the opinions of certian spectrum of the population about a specific subject (general public, voters, registered voters, republican primary voters, etc.) . 

Those polls require accurated data about the spectrum of opinions they seek to measure and the ability to contact a broad spectrum of the folks within that spectrum on a timely basis. Both of those things are becoming increasingly difficult to get/do for a variety of reasons including the increasing mobility of the general population (data becomes outdated much more quickly now than in the past), greatly reduced numbers of people who have actual land line telephones, and the increasing unwillingness of those wh do have them to participate in the poll. to name a few.

[andy58-in-nh]

What you and I would call "dishonest" polls are the ones which start with an end result  and then design the poll to get exactly that result.  It's all about making the necessary assumptions, knowing how to frame questions, who to call, and what time to call ect. These are done ALL the time for all kinds of reasons.

"Honest" polls  are those which actually seek to measure the opinions of certian spectrum of the population about a specific subject (general public, voters, registered voters, republican primary voters, etc.) . 

Those polls require accurated data about the spectrum of opinions they seek to measure and the ability to contact a broad spectrum of the folks within that spectrum on a timely basis. Both of those things are becoming increasingly difficult to get/do for a variety of reasons including the increasing mobility of the general population (data becomes outdated much more quickly now than in the past), greatly reduced numbers of people who have actual land line telephones, and the increasing unwillingness of those wh do have them to participate in the poll. to name a few.

A common problem with political polls is that they necessarily depend on assumptions based upon previous election cycles, especially those immediately preceding. Even as facts, conditions and circumstances change on the ground, standardized methodologies assume stasis, and thus introduce error into their own measurements. 

[Bigun]

A common problem with political polls is that they necessarily depend on assumptions based upon previous election cycles, especially those immediately preceding. Even as facts, conditions and circumstances change on the ground, standardized methodologies assume stasis, and thus introduce error into their own measurements.

Exactly right! That is precicely what I meant when I simply said  "bad assumptions".

[truth_seeker]

So if the polls are no good, is Santorum really in the lead in Iowa and not Trump? He won Iowa last time.

No, we accept Trump is leading, because of polls. Then we turn around and say we don't believe polls.

I am interested in how Bevin won in Kentucky. Why polls were wrong, what the GOP di to get better turnout.

And to learn if Bevin's campaign made effective use of electronic media.

[ABX]

Ahem!  As I have been saying polls are pretty much worthless and getting more so every day because of lifestyle changes among the polling sample base.

It would be interesting to see how differently people look at candidates without group/tribal mentality. How much of what is popular to each of us is because it is perceived popular to those like us?  There are certain candidates who have run before but didn't get an iota of support because the media pretty much ignored them, then later they run but they are given attention so it creates a popularity cycle.

[Oceander]

So if the polls are no good, is Santorum really in the lead in Iowa and not Trump? He won Iowa last time.

No, we accept Trump is leading, because of polls. Then we turn around and say we don't believe polls.

I am interested in how Bevin won in Kentucky. Why polls were wrong, what the GOP di to get better turnout.

And to learn if Bevin's campaign made effective use of electronic media.

Good point.  The first part might be confirmation bias.

[alicewonders]

So if the polls are no good, is Santorum really in the lead in Iowa and not Trump? He won Iowa last time.

No, we accept Trump is leading, because of polls. Then we turn around and say we don't believe polls.

I am interested in how Bevin won in Kentucky. Why polls were wrong, what the GOP di to get better turnout.

And to learn if Bevin's campaign made effective use of electronic media.

I live in Kentucky.  For weeks, months even - I got an occasional email or flyer from Bevin's campaign and heard an occasional commercial on the air.  I got a little worried because I was hearing anti-Bevin commercials which were really over the top lies - like Bevin wanted to drug-test older people on Medicare.  They were revolting commercials and they seemed to air at every commercial break on the radio.  The ones I saw on TV were equally prevalent and dishonest. 

As it got closer to the election - a few weeks before - polls were saying Bevin was five points behind Conway.  I was in despair and really had resigned myself we were going to get another crappy Democrat governor (we've gotten used to that here).  A day or two before the election, I heard Bevin on Beck's show and Hannity's show and it seemed he was playing down his chances of winning - giving the statistics that we've only had - like two Republican governors in 75 years.  Depressing!

But, Bevin said the polls were tightening and they were pretty much tied - he was begging people to get out and vote.  When I voted - I was the only voter at the polling location.  I hadn't seen that in a while! 

I kept checking the election returns and saw Bevin was slightly ahead - but I figured Louisville or Lexington hadn't come in yet.  When they called it for Bevin - you could have knocked me over!

The only explanation I can think of (other than cheating by grossly manipulating the polls) is that Conway's commercials were so insulting and on so very often all through every single day - that people just got sick to death of hearing them.  I know I was.

Couple that with the state going bankrupt covering the thousands of new people on Medicaid because of Obamacare - and that October 31st was the last day our power plant was able to get coal, 800 people laid off at our local steel plant.......our state is like 70% Democrat - old time union job Democrats.  I think the Democrats just didn't have the heart to get out and vote.

 

 

[Bigun]

http://www.c-span.org/video/?400380-1/discussion-2016-presidential-election