CNN's Harry Enten's analysis of the new Monmouth Poll

[Skya Rhen]

Snip...

If you average Monmouth's poll with the three other high-quality national polls taken this month from CNN/SSRS, Fox News and Quinnipiac University, you get a clear Biden lead: Biden at 28%, Warren 19% and Sanders 14%..

Snip...

When you break it down further, you see that Monmouth did not poll at a vastly later date than others or among a different set of voters. Monmouth conducted its poll on August 16-20 among voters who say they are Democrats and independents who say they lean toward the Democratic Party.

Interestingly enough, CNN did its own poll of the race August 15-18 among this same defined group. Unlike Monmouth, CNN put Biden up with 29% to Sanders' 15% and Warren's 14%.

The CNN poll looks a lot more like other polls of the race taken this month when it comes to Biden. Fox News was in the field from August 11-13 (within a week of the Monmouth and CNN polls) and showed Biden at 31%, Warren 20% and Sanders 10%. Unlike CNN and Monmouth, Fox didn't preassign voters to the Democratic primary based on party identification. Instead, they asked all voters which primary they'd vote in.

You'll note that in all of these other polls, Biden's doing at least 10 points better than he was doing in the Monmouth poll. It seems unlikely to me that Biden lost so much ground within the two-day period of CNN ending its own poll.



https://edition.cnn.com/2019/08/26/politics/monmouth-poll-analysis/index.html

[Princetonian]

1. Hmmm... what were they smoking at Monmouth that day?

[Iliyah]

6. What ever it is I want some . . . .

LOL

[Princetonian]

34. Maybe it was the brownies...

[Thekaspervote]

55. Oh too too funnieeeee!!

[Princetonian]

79. Hey, Thekaspervote!

[Thekaspervote]

2. I'll take a 538 analysis of the polls any day. Even if that were to show my fav was not in the lead

Thx for posting Skya Rhen!!

[melman]

9. "I'll take a 538 analysis of the polls any day"

Good news. Nate Silver is a big fan of Monmouth..





https://fivethirtyeight.com/features/which-pollsters-to-trust-in-2018/

[Thekaspervote]

22. So... his pal Harry didn't really see it as a big shift... no one is nixing the poll.

I just prefer to listen to the experts. I’m sure Harry Enten didn’t have to talk his “take” on things over with Nate Silver b4 putting his statement out there

[George II]

25. Yes, I prefer to listen to the experts. And not cherry picked experts.....

....and Enten is one of the best.

[Gothmog]

45. From Nate Silver re this particular poll

Link to tweet

Link to tweet

[melman]

50. Oh

So he changed his mind on Monmouth now that he doesn't like their results?


I AM SO SHOCKED.

[Gothmog]

51. Look at the sample size and margin of error in this poll

Link to tweet

[Thekaspervote]

53. FYI: Nate Silver is a self described libertarian and statistical geek

Why would anyone think that he’s playing favorites? if he were to “skew” the numbers to suit himself or his views, 538 would die the death of a million cuts.

[melman]

54. "Why would anyone think that he's playing favorites?"

Maybe from reading the bullshit he posts on twitter?

[Thekaspervote]

56. Well, there you have it. If it's on twitter

[Cha]

69. Nate Silver Tweets it as he sees it.

[George II]

86. I'll take Nate Silver's assessment of this being an outlier any day.

[NYMinute]

3. People who flunk statistics get excited by meaningless polls

Monmouth poll published today is meaningless and flawed to the core.

I wonder if David Sirota paid some buddy to do a sub-sample and publish the results. That is something that Lee Atwater, Karl Rove and Roger Stone did routinely.

[Princetonian]

8. Interesting bit about the sub-sample sleight of hand. Good to know.

Has Monmouth published their sampling methodology yet?

[dgauss]

18. This Monmouth poll looks like Susan Sarandon's handiwork.

No doubt in cahoots with David Sirota. The only question is, what role did Jill Stein play in all this?

[questionseverything]

35. lol

but is she headquartered in Vermont or new Hampshire?

nm..states are interchangeable now

[bluewater]

4. Biden's AVERAGE support has fallen since he announced on April 25th.

538 says to average results from different pollsters to minimize the effects of any systematic errors they have.

Monmouth University Poll is an A+ rated poll with an 80% accuracy according to 538, so it can't be ignored or dismissed.

Over the next few months we will see how the polls change. I don't expect the online tracking polls that survey the same pool of voters over and over (Like MorningConsult and HarrisX, for example) to show much change yet, but it will be interesting to see how the highly rated LIVE polls go.

Right now, looking at the RealClearPolitics Tracker of average poll results, Biden has LOST SUPPORT since he announced his candidacy for the 2020 nomination.

On April 25th, when he announced, Biden was at 29.3% AVERAGE support.

Today, Biden is at 27.2% AVERAGE support.

Anyone can see the rise and fall of Biden's support at the RCP Polls Tracker here:

https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html

[Princetonian]

5. It's an outlier, nonetheless. Happens to the best of polls with an 5.7% MOE.

[bluewater]

10. The polling average on RCP Polls is what it is

https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html



Biden's AVERAGE poll support is now lover than when he announced on April 25th.

The polls are what the polls are.

[BannonsLiver]

13. Omigerd!111

[bluewater]

20. lol

Hey, but the polling average is what the polling average is.

[Princetonian]

26. RCP poll averages: Joe 27.3%, Bernie 16.7%, Liz 16.2%

I'll take it.

[Thekaspervote]

16. I ask the same thing almost everyday day

[NYMinute]

62. Polling or trolling?

[bluewater]

11. RCP Polls shows Biden's AVERAGE of all Polls support has dropped since he announced.

https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html



That is not based on any single poll.

[Princetonian]

28. Yes, by 2.1%. His spread is +10.5 and is ahead of Liz by +11.1

I'll take it.

[bluewater]

36. Warren has GAINED 9.7% while Biden lost that 2.1%, net 11.8% gain for Warren.

Everyone can see the actual trend for themselves here:

https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html

Biden LOSING support over the last 4 months is why many people consider him a fragile front runner.



It will be interesting to see what the polls do over the next few months.

[Thekaspervote]

14. You are right!! Mr. Biden's number have gone down. In political speak it's called an initial bump

Did you read what Harry Enten from 538 said about the Monmouth poll today? If you did then you know mr. Biden’s numbers remain pretty much the same. No big shift

[George II]

33. Here are the averages on April 25 and today:

Biden 29.3, 27.2 -2.1
Sanders 23.0, 16.7 -6.3
Warren 6.5, 16.2 +9.4
Harris 8.3, 7.5 -0.8

Why no reports of Sanders "plunging"? He's down 3X Biden's drop, yet Biden in "plunging"!

Essentially Sanders and Warren have been trading support, Biden and Harris have been stable.

[questionseverything]

38. warren up by over 9 would be considered better than stable

[bluewater]

39. Warren is up +9.7% during the 4 months Biden lost -2.1%

[George II]

41. I meant to say Harris, sorry. Warren and Sanders a trading support back and forth.

[Princetonian]

37. Yes, a 2.1% drop. Joe's spread is +10.5, is ahead of Liz by 11.1%

I'll take it.

[Thekaspervote]

46. Oh no... a 2.1% drop....the sky is falling!! Snark

Hi Princetonian

[Princetonian]

64. Hi, Thekaspervote!

By the way, the 2.1% drop referred to the weekly variation probably due to the Monmouth poll which Nate Silver said was an outlier because its sample size was so small (238). Another poster noticed Monmouth has overrepresented Under 50 voters. The MOE was 5.7% which is huge.

Polls don't mean much now but are a source of fun now.

[Celerity]

60. no, he is down over 14 points from his peak, and trending lower, if Sanders was not in the race he

be in serious trouble with Warren over the long term trend (much of the Sanders support that said they would go to Biden was pure name recognition for both Biden and Sanders.) He still maybe in trouble IF she starts to make inroads with A-A's and older voters and the other candidates' supporters break to her.

https://www.democraticunderground.com/1287254977#post57

[Princetonian]

66. Not relevant. I was observing the weekly change.

[Celerity]

67. long term trend is more relevant than a 1 week snapshot nt

[Princetonian]

68. What matters to me: Joe's double digit poll average advantage is holding steady at RCP.

I expected things to tighten even further. No worries.

Polls are fun but meaningless at this point. December is when they will matter.

Warren needs AA voters not a polling popularity contest. She has little support there. Ditto with Bernie.

[Celerity]

76. Toss Bernie's overall A-A support levels onto Warren, and Biden is really in trouble in regards to

her. Bernie has the 2nd highest A-A support levels overall when the totality of the polls is measured, only tailing Biden on a cumulative average. Harris was biting into both Biden and Bernie, but has fallen back fairly across the board, and I expect Warren to make inroads in that vector and also amongst older voters if she actually is started to be seen by those groups as more viable for the general than she is at the present.

Like you said, it is still relatively early in the cycle. I am not swayed by just one poll, so do not think I am anywhere near calling the race even yet. Biden is still the clear front runner, but cracks are being slowly exposed. Biden is polling in that Monmouth poll at 6% with voters under 50, and in even Biden-friendly polls, he still lags badly in that group.

3 recent polls:

https://www.monmouth.edu/polling-institute/reports/monmouthpoll_US_082619/

Biden has suffered an across the board decline in his support since June. He lost ground with white Democrats (from 32% to 18%) and voters of color (from 33% to 19%), among voters without a college degree (from 35% to 18%) and college graduates (from 28% to 20%), with both men (from 38% to 24%) and women (from 29% to 16%), and among voters under 50 years old (from 21% to 6%) as well as voters aged 50 and over (from 42% to 33%). Most of Biden’s lost support in these groups shifted almost equally toward Sanders and Warren.

https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/u4gcv1suy6/econTabReport.pdf







https://www.foxnews.com/politics/fox-news-poll-8-15






I find it amazing that the number one target for negative posts by Biden supporter is Sanders. Without him in the race, Warren would more than likely at least on a glide-path to possibly overtaking Biden ATM. IF she starts to clearly pull ahead of Sanders by a fairly wide margin and sustains it for months, and he loses badly in Iowa and NH, etc, and yet still refuses to drop out and endorse her, the Biden folks owe him a lifetime supply of whatever Bernie's favourite food is, as he will be the blocker of her path at that point. On the flip side, I see no plan B for the centrist to centre-left liberal route at all, which truly worries me, as IF (a big if granted) Biden himself goes wobbly that leaves us (our Party) sitting with its 2 main candidates being drawn from the far left, which is asking for electoral trouble IMHO. I do not like 'all your eggs in one basket' type of thinking.

I want to wait to see the next several weeks polls en toto before I can be more comfortable in judging the extent to which this Monmouth poll was either (A) an outlier or (B) more reflective of an actual significant change in the dynamics. Anyone who is claiming to definitively label it, now, with certainty, either way is simply projecting their own wishes onto it IMHO.

[Celerity]

57. Biden peaked at 41.4% (May 9-11), so is down 14 points from then

[Princetonian]

65. We are talking apples and oranges. I used the link by the poster I replied to to see the recent #

[Celerity]

74. The poster I replied to went all the bay back to April. I simply showed Biden's overall peak in that

the time period of then until now. That was the sole purpose of my post to them.

Now that you bring up your points and thoughts, here is my reply to you specifically:

Recent numbers over a short 1 week or so snapshot show far less than long term trends. Try to play arbitrage spreads in FOREX based off week-over-week-only interval interval analysis. You will soon be bankrupted. Use leverage and you may go into margin-call/liquidation territory in a few hours, even if your overall sense of direction for the spreads was correct.

[BootinUp]

12. Enten, ptoouui!! / sarcasm

[DrFunkenstein]

15. Biden has surged in the polls since Obama's former VP announced

I haven't seen crowds so worked up in all my time. Any suggestion that Biden had hit his ceiling after riding on Obama's coattails is completely fanciful. The other campaigns are in full meltdown as their numbers slowly dwindle away.

It's safe to say this race was OVER before it began.

[Watchfoxheadexplodes]

17. Someone wake me up when Joe is not

The frontrunner

[Thekaspervote]

19. Funnieeeee!!

[still_one]

21. I have said this before and I will say it again, National Polls are not a good indicator since

states determine the president l

[Thekaspervote]

23. And I sure wish we had more current polls for the states!!

It does really give a much better picture

[still_one]

31. Agreed. Things are still a long way off. I try to avoid OPs that go out of their way to snipe at

Last edited Mon Aug 26, 2019, 05:06 PM - Edit history (1)

other candidates. I am not referring to discussing differences on issues, but creating posts to belittle and bully

They play this like it is some kind of a game, and they do a disservice to the candidate they support

[Thekaspervote]

47. Oh gee a voice of sanity!! Thank you

I too go out of my way to not snipe at other candidates, because as you said it is a great disservice, not only to the said candidate but to the Dems as a whole.

I haven’t seen such road rage since I had to drive Illinois I-55 to work daily!!

[still_one]

58. Wow, I didn't realize Illinois I-55 was tough

[Thekaspervote]

61. Around O'hare at rush hour.. ughhh

[George II]

40. Here are the first four states' results:

Iowa, Biden 26.0, Warren 18.0 - Biden gets 59% of the delegates, Warren gets 41%, the others zero
New Hampshire, Biden 21.0, Sanders 19.3 - Biden gets 52% of the delegates, Sanders gets 48%, the others zero
Nevada, Biden 30.0, Warren 15.3, Sanders 15.3 - Biden gets 50% of the delegates, Warren gets 25%, Sanders gets 25%
South Carolina, Biden 38.0, Sanders 14.3, Warren 12.7, Harris 12.0 - Biden gets 100% of the delegates, the others zero

All of this negating minor differences due to variations in some districts.

Simple math (Iowa, New Hampshire, Nevada, South Carolina):

Biden - 24 + 12 + 26 + 54 = 116
Warren - 17 + 0 + 14 + 0 = 31
Sanders - 0 + 12 + 14 + 0 = 26
Harris - 0 + 0 + 0 + 0 = 0

Biden comes out of the first four states with 67% of the delegates, Warren with 18%, Sanders with 15% and Harris with 0%.

Could be all but over even going into Super Tuesday, where Biden will clean up!

[Thekaspervote]

48. Thx for posting some state polls...

[still_one]

81. I hear you George. Still a ways off though, and I think if the nominee is determined

before the last primary is held, the earliest it would be is after Super Tuesday

but we will know in a little more than six months

[Cha]

80. Mahalo, still_one!

[DallasNE]

24. More Bickering Over Polls

More taking the eye off the ball and playing right into Trump's hand.

In last weeks news we learned that the government has adjusted down by 501,000 the number of jobs created in 2018 AND the CBO adjusted up their estimate on the size of the budget deficit over the next 10 years by $800 billion.

If that wasn't bad enough the State Department issued a report saying ISIS is having a strong comeback in Syria and Iraq and they are getting established in Afghanistan, exactly what the experts predicted would happen.

And this doesn't even touch on the clown show in France where Trump once again made a complete fool of himself while Melania swooned over Canadian PM Trudeau.

These items are just within the past week.

So why does the poll of the day get the only billing? The distractions are driving the hard news off the air waves and that needs to stop.

[DrFunkenstein]

29. Probably Because We've Had Depressing/Infuriating News For Two Years

And it is more fun to think about a time when Trump is no longer protected from criminal prosecution.

[still_one]

82. I think you know the answer to that. "braging rights", or worse, just trying to intimidate the

competition

Until the candidate field narrows down, along with other reasons, I don't think the polls mean very much at this stage. Especially the national polls

[stonecutter357]

27. K&R

[redstatebluegirl]

30. I am going under the assumption that we have a lot of work to do.

I volunteered over the weekend, and will continue to help whenever I can. We have to outwork them and not look at any polls. Run like we are losing all the time. I love Joe, he is my choice. I still feel strongly he can beat the orange wonder. I will support the nominee, hoping it is not the two I dislike, but I will hold my nose and vote to keep this evil being out of the white house again.

[Princetonian]

32. +1000

[Thekaspervote]

49. Me too!!

[Cha]

73. Mahalo, redstatebluegirl!

[redstatebluegirl]

75. Hello and hugs Cha!

[Cha]

85. Aloha to you, redstatebluegirl!

[karynnj]

42. The lost ground would have to be even more severe than the differences suggest

An estimate of the results for the three dates polled by both - the 16th and 17th and 18th - would be estimates for the same unknown actual. It is that "actual" that is said to be within the confidence interval (usually) 95% of the time. Clearly the CNN estimate could be on the high end and the Monmouth one on the low end for the sum of the three days. To my knowledge neither pollster has talked of the difference between the results on different days. (Remember polls where something big really happened and they show major differences in the two segments of the poll.)

It would be very difficult to shift a 4 day estimate and a 5 day estimate as much as the difference of these two polls by assuming the 15th was higher for Biden and the 19th and 20th lower.

At this point, unless more polls show this shift, the likelihood is that the Monmouth poll is an outlier.

[Skya Rhen]

44. **** This ****

[Thekaspervote]

59. Agree!

[doxyluv13]

52. Wow, CNN is desperate to pump up Biden

This is a rationalization, not anything like a valid statistical analysis. The difference of a couple days can make a big difference, especially when the candidate is regularly doing things that seem doddering or out of touch. I could see Biden's numbers going up it he did something to put his gaffs behind him, or to recognize them as such. But he hasn't. The Monmouth poll could be an outlier, but CNN's poll showing Biden's lead increasing when there has only been bad news for him is more likely to proven wrong.

[Cha]

71. No they're not. That's just your opinion.

Maybe people are making up their own minds about Joe Biden and have a Much Different Opinion of him than you do.

[Demsrule86]

77. No they are not...

[5starlib]

63. What was the racial breakdown of the poll?

I'd be really curious about that.

[Skya Rhen]

78. They lumped the Hispanics, AA, Asian and Other together: Sanders 22%, Biden 19%, and Warren 14%

Seems questionable...

[Scurrilous]

70. Boink.

[Vegas Roller]

72. K&R nt

[highplainsdem]

83. Tweet from Obama/HRC pollster John Anzalone:

https://www.democraticunderground.com/1287255551

[highplainsdem]

84. Another pollster (Republican this time) on why this Monmouth poll is useless:

Thread starting with this tweet:

Link to tweet

This monmouth poll is "based on 298 registered voters who identify as Democrats or lean toward the Democratic Party".

Guys, just chuck it into the recycling bin. it's useless. <300 people to represent a national electorate?



this survey *theoretically* has a plus or minus 5.7% margin of error. but that's just not how actual modern political polling works. With declining response rates, you need more sample to hit quotas on various demographic categories. you just can't do that with n=298



which is to say nothing of the fact that "identifying as democrats or leaning towards the democratic party" doesn't make you a primary voter

[MineralMan]

87. Further, other polls taken after the Monmouth poll

do not show any such drop for Biden.

[Cha]

88. KICK!