2016 Postmortem
Related: About this forumBrand Spanking New CNN Iowa Poll -HRC 54% SBS% 36% MOM 4%
The gender gap between Clinton and Sanders has narrowed in the last month, with both men and women breaking significantly in the former secretary of state's favor (51% Clinton to 40% Sanders among men, 57% Clinton to 34% Sanders among women; last month it was 48% Clinton to 44% Sanders among men and 62% Clinton to 31% Sanders among women).
Clinton's edge on handling economic issues is smaller in Iowa than it is nationally, 47% in Iowa say they think Clinton would do the best job handling the economy while 42% say they trust Sanders most, while our recent national poll found Clinton holding a 58% to 31% advantage on the same issue. Likewise, Iowa's likely caucusgoers are more likely to think that Sanders would do the most to help the middle class (50%) than to say Clinton would (43%), a question that split national Democrats almost evenly (47% Clinton to 44% Sanders).
http://www.cnn.com/2015/12/07/politics/donald-trump-ted-cruz-iowa-poll/index.html
randys1
(16,286 posts)do this well.
DemocratSinceBirth
(99,714 posts)EOM
randys1
(16,286 posts)JaneyVee
(19,877 posts)Homogeneous states with low delegate count. Be honest, 6 months ago you thought Bernie would be shellacking Hillary.
randys1
(16,286 posts)I did think he thought he HAD to do something to get the conversation going before the poor become 89% of the population, or whatever.
BlueCheese
(2,522 posts)I haven't decided which way to vote myself (and my state votes so late it doesn't really matter). But Sanders has added an important voice to the primary debate, and his issues deserve to be heard.
George II
(67,782 posts)MrWendel
(1,881 posts)George II
(67,782 posts)riversedge
(70,321 posts)FloridaBlues
(4,008 posts)NurseJackie
(42,862 posts)Agschmid
(28,749 posts)oasis
(49,415 posts)DemocratSinceBirth
(99,714 posts)EOM
oasis
(49,415 posts)DemocratSinceBirth
(99,714 posts)George II
(67,782 posts)DemocratSinceBirth
(99,714 posts)msrizzo
(796 posts)Didnt you.😄
Dem2
(8,168 posts)53.6-36.3
Edit: NM, I hit refresh and it changed, though my statement still stands I guess:
Hillary Clinton 54.7%
Bernie Sanders 37.4%
Martin O'Malley 3.7%
HerbChestnut
(3,649 posts)That this poll reflects the last CNN Iowa poll almost exactly. Then it was Hillary 55, Bernie 37, Martin 3. So if anything it looks like things have stalled a bit in Iowa for the time being.
workinclasszero
(28,270 posts)56 days until the BS stops and reality steps into the picture around here
Attorney in Texas
(3,373 posts)Zequest
(18 posts)What went wrong in this period between these two polls? Nobody got know him?
Attorney in Texas
(3,373 posts)Zequest
(18 posts)When one visits HuffPo, this is what we see
As you can see, your numbes (HRC 49 vs. Sanders' 40.8) are not the numbers given by HuffPo.
Attorney in Texas
(3,373 posts)JaneyVee
(19,877 posts)Alfresco
(1,698 posts)YUP!
JI7
(89,276 posts)Winning both would do a lot towards helping in later states.
Iliyah
(25,111 posts)Go Hillary!
Alfresco
(1,698 posts)Persondem
(1,936 posts)Cha
(297,734 posts)DCBob
(24,689 posts)Looking good for Madame President!
riversedge
(70,321 posts)the sound of that
riversedge
(70,321 posts)OilemFirchen
(7,143 posts)FloridaBlues
(4,008 posts)SidDithers
(44,228 posts)Sid
NurseJackie
(42,862 posts)mcar
(42,382 posts)madokie
(51,076 posts)we've seen this show before. I still remember what the view was like, Actually
RichVRichV
(885 posts)[IMG][/IMG]
First thing to notice is that the under 50 crowd has N/A listed for their results. So you'd think we can't figure out their totals, but it's actually blatently easy. If you look at the top of the poll, the Democratic likely caucus goers is 442 and the total has a margin of error of +/-4.5%. The formula 0.95/sqrt(442) works out for this. Using that we can see how many are in the 50 and older crowd. With +/-5.5% it comes out to ~298. 298/442 = 67.4% of the total.
The problem is that in Iowa in 2012 the over 50 crowd only counted for 54% of the voting totals. CNN's own exit polling from 2012 verifies this. As you can see by comparing the totals column to the 50 and older column, Bernie pickes up substantially with the sub 50 crowd. Hillary does better the older the crowd gets.
They also list margin of errors breakdowns for the 50-64 crowd (+/-8.0%, ~140 people, 31.6% polled) and 65+ crowd (+/-7.5%, ~160 people, 36.1% polled). That puts the 50-64 crowd 12.8% higher sampling then they should be, and the 65+ crowd (Hillary's best group) at 38% higher sampling than they should be vs 2012 exit polling. All this means that the under 50 crowd (Where Bernie does best) is being undersampled by 41% (~144 ppeople, 32.6% here. Should be 46% according to 2012 exit polling)
This means that either CNN thinks the youth turnout is going to drop like a rock from 2012, or they're blowing smoke up our asses with the sampling.
Hillary currently has a lead in Iowa. It's not as high as the poll results say it is.
stevenleser
(32,886 posts)CNN and other polling agencies aren't tailoring their samples to a general election and they shouldn't. They are tailoring it to what they expect caucus turnout will be like and I am guessing they have some info that tells them what 2008 caucus turnout was like that helps them do that.
RichVRichV
(885 posts)Since there was no Democratic presidential primary in 2012, let's go back to Iowa 2008 Democratic primary.
17-29 --- 22%
30-44 --- 18%
45-64 --- 38%
65+ ----- 22%
Now we can't convert perfectly because the 45-64 age group doesn't line up with the 50-64 age group. But if we assume equal distribution across the group then that puts us at
under 50 - 49.5%
over 50 - 50.5%
Thank you, you just made my argument stronger.
stevenleser
(32,886 posts)And Iowa has caucuses not primaries.
RichVRichV
(885 posts)If we assume the 45-49 vs 50-64 numbers come out with the same difference as the 30-44 vs 45-64 age groups do then it points to a 44% under 50 and 56% over 50. Since the variance probably doesn't fall at either end of the extreme, it's not far off from my original 46% vs 54% in the 2012 general. My original math stands. The sampling on the poll is wrong.
And this info is from CNN's entrance polls for the Iowa caucus in 2008. It's accurate and a valid comparison.
DemocratSinceBirth
(99,714 posts)EOM
RichVRichV
(885 posts)numbers for Iowa. The demographic breakdowns were pretty much equal. This information isn't hidden, go look it up.
ABC Iowa caucus exit poll
CNN Iowa caucus entrance poll
Washington Post Iowa caucus entrance poll
NYTimes Iowa caucus entrance poll
The turnout numbers aren't what's important, it's the demographic breakdowns of the caucus that matter to the accuracy of the polls. The polls can only be considered accurate if they correctly identify the likely demographic breakdowns for the caucus (especially when the vote skews so heavily along those demographics as it's doing in these polls). Right now this and most other polls are not matching the 2008 Iowa caucus demographics. That means either they expect a substantial drop in sub 50 voting from 2008 caucus or they are using improper sampling.
Alfresco
(1,698 posts)ismnotwasm
(42,014 posts)coyote
(1,561 posts)The 18-50 year group is totally missing or has an extremely low sampling group. It looks like a shit poll to me.
18-34 group is N/A
35-49 group is N/A
http://edition.cnn.com/2015/12/07/politics/poll-results-ted-cruz-donald-trump-iowa/index.html