Not quite as skewed: CNN's latest poll is still propaganda
David Sirota just sent me the results of the latest CNN New Hampshire poll. It shows Bernie leads in the two issues most important to the people of New Hampshire - Single Payer Healthcare and climate change.
Here is the poll about which issues resonate most with Democrat voters in New Hampshire:
Here is where Bernie rates in those issues:
Healthcare
Climate change
There's even more good news. Bernie is leading overall in New Hampshire:
Plenty of good news right? You may be asking yourself, but BBM, you said it was skewed. Surely CNN wouldn't skew it in Bernie's favor would they?
No. They would not. However, they can't skew it against Bernie either. Everyone knows Bernie is going to win New Hampshire. To say he isn't goes against everyone's common sense and is what we'd call Explicit Propaganda. Propaganda only works when it is implicit.
The implicit propaganda here is that Bernie is winning, but not by much. However, one only has to look at their own numbers to realize how the polling has been massaged in order to skew the results so that Bernie is only slightly ahead.
Here is CNN's age distribution of their poll:
Note that the age group 18-34 represents 27% of those polled. 35-49 are 23%. The largest group is 50-64 at 30% and finally, the elderly, 65+ at 20%.
But this is the actual age distribution in New Hampshire, per datausa:
Now, we have to do some maths here. First, to get the actual % of each age category, we have to deduct those under age 18. That's 19% of the total, leaving each category as a % of the remaining 81%. Then we need to combine ages 18-24 and 25-34 to match CNN's category of 18-34. The remaining age categories don't perfectly match but we can estimate closely.
The actual percentage of each age group in New Hampshire is:
18-34: 31%
35-49: 22%
50-64: 27%
65+:19% 20%
Now, for visual ease, here is CNN's age distribution again:
18-34: 27%
35-49: 23%
50-64: 30%
65+: 20%
So as you can see, it is slightly skewed against young people aged 18-34. The age group that most identifies with Bernie Sanders. But we are only talking 4% here. It isn't nearly as skewed as most polls that have been touted for the past year.
However, it is safe to say that Bernie picks up 3 of that 4% and Biden/Warren lose it. The other 1% is probably negligble, but let's give it to Warren, just to show how much that affects the results. A small 4% difference.
Before:
Bernie - 21%
Warren - 18%
Biden - 15%
Adjusted for actual age distribution of New Hampshire voters:
Bernie - 24%
Warren - 17%
Biden - 13%
Two things stick out: Sanders goes from a slight lead - in a state it's already considered a given he will win - to a commanding lead. The other thing that sticks out is just how badly Biden is lagging.
Just a little fudging in their polling procedures and CNN has boosted the perception that Biden and Warren are hot on Bernie's tail, when the reality is, Bernie is holding a very strong lead.
Comments
However....
First, CNN *is* skewing the reporting. When they have a poll showing Warren at 27 and Biden at 25 they have no problem saying "Warren ahead". When Bernie is st 21 and Warren is at 18 they say "virtual tie for first."
Anyway...so I deal with polls all the time. I remember well the first time I was outraged by apparent skewing by a pollster. He had some really high 50+ number. When the election came it turned out he was pretty close to the actual turnout.
Historically older people vote, younger people vote less often. The trick here is that Bernie is strong with the group that might or might not turn out. I think they will but it's clear that CNN is using a typical voting model.
Battle Of, your article needs to get around. Young'uns need to get out there and vote! Maybe being told they won't vote will motivate them!
What morons think Biden is good for healthcare?
Good for insurance companies.
I've seen lots of changes. What doesn't change is people. Same old hairless apes.
NH is a semi-closed primary
According to Ballotpedia:
So an unaffiliated voter (looks like 57% of the electorate from the table above) can turn up at a Democratic polling place, register as a Democrat, and vote that day? Or does such a voter have to register x days before the primary? I just spent 10 minutes trying to look it up with no success.
How the hell do pollsters take this into account either way? How can they predict how many voters will do this?
Given that Bernie will undoubtedly pick up a goodly number and percentage votes from this pool, the polls seem necessarily skewed against him if they don;t have a good handle on what seems mighty hard to predict, no?
That's how it works in Illinois
In the past one had to skip an election to change parties but now you can flip every time. Maybe a court ruling.
No pre-registration required.
I've seen lots of changes. What doesn't change is people. Same old hairless apes.
Then there's my conspiracy theory
That they're buying data from Facebook or Google and selectively polling the people who will give them the answers they want, or they can manipulate with their questions. (I suspect that's what they did with registration shenanigans in the 2016 primary) Or they're just making it up. They have enough of a history to pull believable sounding numbers out of thin air.
On to Biden since 1973
@doh1304 And you're probably the
Immigration went from 15% to 4%
I wonder why the massive change?
I've seen lots of changes. What doesn't change is people. Same old hairless apes.
Immigration is probably very personal
In baseball history the greatest racists (Cap Anson for example when John McGraw wanted to play Eddie Grant) were northerners. Maybe they just randomly polled people who did not hire or work beside immigrants, or alternately they may have disproportionately polled people who worked in occupations that have suffered severe wage erosion due to the exploitation of immigrant labor.
I live in a condo complex with a diverse community. If I had not driven a taxi for 30 years I would have a very ignorant but tolerant opinion towards immigrants. Instead I have suffered thousands of dollars of losses due to taxi companies use of racism as a weapon against their drivers.
Immigration is a class issue for most Americans, whether they know it or not. Therefore opinions on immigration may vary widely depending on cultural (economic) circumstances.
On to Biden since 1973
That would throw the whole poll into question
if there was as 10% variance between samples.
I've seen lots of changes. What doesn't change is people. Same old hairless apes.
It would, wouldn't it?
We know about rigging by under representing young people or over representing some other demographic, so now maybe they're over representing Hillary voters or Biden supporters based on their Google history or Facebook profiles.
On to Biden since 1973
could be!
I've seen lots of changes. What doesn't change is people. Same old hairless apes.