Can partisan polls be manufactured to create illusory momentum in campaigns?

[bluestateguy]

Let's say I am a candidate for president or senator in a high profile election campaign. Things are not going so well, and we feel like the election is starting to slip away. The media coverage is mostly negative and the polls stubbornly show us down 3-5%.

Now let's say we hatch a scheme to create momentum for the campaign that may not actually be there in reality. So we go to a pollster who is affiliated with our party and say "I'm going to need you to go to the bakery for me. Bake me up a poll showing me ahead by 3-5 points. Do whatever you have to do to get me a poll showing me slightly ahead."

The pollster does as he is asked, gets paid, and the campaign then blasts their new poll showing that they are ahead. The hope then is that the media will run with the new polling numbers and change the course of their coverage with a new narrative of a campaign that is "on fire" and gaining traction. The hope will be that the manufactured momentum will generate real momentum.

I wonder how often this happens in campaigns.

[cthulu2016]

1. What really happens is similar, but not quite as bad

Campaigns do private polling and then release the results only if they are favorable

[Reasonable_Argument]

2. Often but not like you think

Most polling agencies have a political bent but they do it through how they ask the questions. Example:
Q1. Do you support the government keeping people from going bankrupt due to medical expenses?
Most people would say yes to this making it look like the president's policies are popular but they could also phrase it like this.
Q2. Do you support the government getting more involved in the health care system?
Most people wouldn't support this and would answer no so it would be spun to make the president's positions sound more unpopular.

This is the status quo in polling today.

[RKP5637]

5. Yep! Well said!!! We see this all the time. And often, somewhat

the same, the results of polls are promoted, but the actual questions are not reported hence masking, of course, the prose used in the weighted question.

[slampoet]

10. Great post. +1

[moondust]

3. I've always suspected

that you could simply poll heavier in certain area codes or zip codes that would be more likely to yield the results you want. But I have no evidence to back it up.

[unc70]

4. Lots of ways to slant polls

For example, including the option "haven't enough information" will typically decrease the favorable responses. Using any "not" word will strongly influence how people answer.

[el_bryanto]

6. Pretty often I'd guessed. I've had a few push polls

and of course this technique goes back to Nixon if not before.

I miss the good old days of real democracy when the winner would be the one who bought his supporters the most beer.

Bryant

[hughee99]

7. I'm sure this happens quite a bit,

the pollsters challenge isn't to make up numbers, it's to create a series of specific poll questions that will produce a desired result.

[uponit7771]

8. Not really, but the M$M usually takes polls that are under sampled and un even weighted and taughts

...them as true

[BumRushDaShow]

9. Smart canidates pay for their own "internal polls"

(i.e., they hire companies who specialize in this) without all the media hype and purportedly designed to give a truer sense of what is going on.... And this is often what you see them reacting to. The results of these are never made public, although you'll occasionally hear a pundit reference a candidate's "internal polling which showed such and such is happening" (usually described in a general sense if the advisor or surrogate happens to share that info).