Polling & November 2
Polls Polls Polls… what to believe.
I’ve been noticing that the generic ballot polls with Gallup and Rasmussen tend to swing in opposite directions–when Rasmussen has a high R/D gap, Gallup is close, and when Rasmussen is close, suddenly Gallup has a huge gap.
I don’t know why this happens.
Generic ballot polling is very tricky anyway. Even with Rasmussen’s weekly sample of 3500 people, that’s an average of only 8 people per district. If one person changes his/her mind, the percentage can jump by 12.5% for that district. Generic polls are national, not by district.
Rasmussen released its Generic Ballot poll today showing the lowest advantage for the GOP all year. It read: GOP’s generic ballot lead down to 3% in Rasmussen among “likely voters.” Then, Gallup comes out and says among likely voters: Republicans lead by 13%. Then there was this: Among voters ‘not affiliated’ with either major party, Republicans hold an 18-point lead.
The context of these polls is Rasmussen had their numbers at 43-43% when the Republicans won in 1994 and in the final pre-election Gallup generic ballot for 1994, they showed a 46-46 tie. And, Rasmussen’s national polls are usually spot on… but their state polls are always iffy. Problem, this is a national poll.
Take a look at the Gallup turnout models – this should make the liberals want to bury their heads in the sand:
Scenario GOP Dem
Registered Voters 46% 43%
High turnout likely 53% 40%
Low turnout likely 56% 38%
As I stated, Gallup is going the opposite direction than Rasmussen. If Gallup is right, this will be an utter Democrat wipe-out of historic proportions. Take a look at the likely voter turnout models Gallup just reported: +13 GOP or +18 GOP, take your pick.
What can we make of all this? Well… if Gallup’s LV numbers hold up, it will be a nightmare for Democrat’s come November! Or, it could be it is a bad sample, or it could be the race has significantly tightened. The same tightening happened in ‘94. Stories and polls came out saying that the races were tightening and so on and so on….
It makes you wonder. Maybe all the Dem advertising is having an effect? Maybe the GOP did peak too soon? Maybe voters are just now paying attention? GOP candidates running crappy campaigns? Take a breath.
Based on statistical modeling of the historical relationship between the national vote and seats, any situation in which the Democrats have less than about 47% of the actual two-party national vote for Congress (i.e., 53% voting for the Republicans and 47% for the Democrats among those voting for one of the two parties) would strongly predict that Republicans would win enough seats to gain control of the U.S. House of Representatives. If there is a widely disproportionate skew in turnout toward Republican voters and their national vote lead ends up being in the double digits, the Republican gains would be very substantial.
In 13 midterm elections between 1950 and 2006 for which Gallup estimated the preferences of likely voters, the correlation between the Republican margin among likely voters in the final pre-election poll and the actual Republican margin in the national popular vote is an impressive 94%. Moreover, the average difference between the predicted Republican vote margin and the actual Republican vote margin in these 13 elections is only 2 percentage points.
Based on some regression analysis conducted by Larry Sabato, we can estimate the number of seats Republicans would be expected to win in the 2010 midterm election depending on the results of Gallup’s final likely voter poll. In order to win a majority of seats in the House, Republicans will need a lead of 2-3 percentage points in Gallup’s final likely voter poll. Hence, the predicted Republican Seats in 112th House would be:
Gallup chart: Source: Gallup Poll and data compiled by Larry Sabato
That said, Rasmussen is more accurate according to comparing their pre-polls in 2008 to the actuals: The following list ranks the 23 organizations by the accuracy of their final, national pre-election polls (as reported on pollster.com).
1. Rasmussen (11/1-3)**
1. Pew (10/29-11/1)**
2. YouGov/Polimetrix (10/18-11/1)
3. Harris Interactive (10/20-27)
4. GWU (Lake/Tarrance) (11/2-3)*
5. Diageo/Hotline (10/31-11/2)*
5. ARG (10/25-27)*
6. CNN (10/30-11/1)
6. Ipsos/McClatchy (10/30-11/1)
7. DailyKos.com (D)/Research 2000 (11/1-3)
8. AP/Yahoo/KN (10/17-27)
9. Democracy Corps (D) (10/30-11/2)
10. FOX (11/1-2)
11. Economist/YouGov (10/25-27)
12. IBD/TIPP (11/1-3)
13. NBC/WSJ (11/1-2)
14. ABC/Post (10/30-11/2)
15. Marist College (11/3)
16. CBS (10/31-11/2)
17. Gallup (10/31-11/2)
18. Reuters/ C-SPAN/ Zogby (10/31-11/3)
19. CBS/Times (10/25-29)
20. Newsweek (10/22-23)
The relative polling accuracy of Rasmussen.
But, Rasmussen’s numbers have been petty volatile recently. In fact, Rass’s poll has oscillated rather regularly for the last 8 months, (due to algorithm changes) although this is the biggest dip so far. Just last week, Rasmussen had 0bozo approve/disapprove at +1 one day (September 28th), yet the very next day (Sept 29th), 0bozo’s approve/disapprove was back to -6, a swing of -7!
I saw this happen in the Washington Post and Rasmussen polls in the VA governors race last year also. McDonnell had a double-digit lead over Deeds after the convention and primaries. This closed to about 4% then went back up to 9% the week before the election. McDonnell won by 23%.
Let’s also not forget that Rasmussen underestimated the Scott Brown’s surge by at least 3 or 4 points. Do that in this case (also think about the MoE) and you have 48 – 39 or 49 – 38. And consider this: On the same day, it’s also hard to imagine that 10% of those polled by Rasmussen changed their minds about Obamacare (around 60% down to 50%) in just two weeks.
There is no poll out there that I trust. George Gallup a wonderful Christian man, and a pioneer in polling, is long gone. Am I supposed to believe that Rasmussen is some type of a Republican so his word is gold, sorry I do not buy it? There are so many variables, as the election closes there will be many polls saying that there is a shift back to Democrats, I predicted this, and its was not hard to predict with a liberal media.
I am not condemning Rasmussen because the numbers are not to my liking! I do not believe ANY poll even if it seems that the numbers are to my liking. I trust Rasmussen pretty well over others though. I think this is something of a statistical outlier. It’s bound to happen from time to time.
Fox News poll is pure garbage, done by a liberal pollster out of one of the New England States, Opinion Dynamics. Sure they may actually give you an honest poll when it is not crunch time, but when it counts they can not keep there liberal politics out of it. The evidence is strong in the Fox case. Their accuracy is in the middle somewhere.
NBC News/Wall Street Journal poll is also garbage. Chuck Todd actually has some input as to the wording of the questions. Even if you had a honest pollster, what about the people making the calls, are they non partisan robots? Nobody is non partisan, whether it’s the phony non-partisan White House Budget Office, Gallup, Fox or whatever.
Democrats typically have a 3% to 4% advantage in the generic ballot, based on voter registration. A tie in the generic ballot means big gains for the GOP, and a lead of 3 or 4 points fully supports the idea that the GOP will win 40 or more seats in the House and seven or eight seats in the Senate.
I’ve never seen a poll that over-sampled Republicans before. If Republicans are tracking equally as well within their party as Democrats, and we would have to assume they are, and they hold an 18% point lead among “unaffiliated” voters, what does that say about the sample weighting of Dem’s to Pubs in Rasmussen’s poll? Among voters not affiliated with either major party, Republicans hold an 18-point lead. Just wanted to say it again.
Right now everything is in flux because the media is scaring the voters of the new conservative candidates. Angles’s a kook, O’Donnell’s a witch, Miller’s an extremist, and Rubio is anti-Hispanic… blah blah blah. Voters are in a fog of confusion. The more they see and hear and comprehend of the conservative candidates vs. the socialist-progressive incumbents with their fascist boots on their economic trachea, the big picture will come into focus. And, wait till more bad economic numbers come in.
I’ve heard rumors that the U3 unemployment rate will rise to 9.7% – although they may be setting up an ‘unexpected surprise.’ Why look at that… unemployment stayed at 9.5% or fell to 9.3%, only to be adjusted upward after Nov 2. NEVER in recent American history has “Progressivism” ever dominated on its merits, and it won’t this election, either.
If you believe Gallup, the Dem’s chances haven’t improved at all. I don’t see genuine “middle of the roaders” and “independents” being swayed by Democrat overtures. This November is NOTHING but an electoral mandate against Obozo and his gang. Don’t have doubts or be depressed at all. I assumed that the earlier polls showing a twelve or nine-point advantage for the GOP were outliers, because numbers of that kind would mean a pickup of one hundred seats in the House (or more) and that is unlikely to occur.
Everyone should take a few deep breaths and realize polls fluctuate. The closer we get to election the mood of the electorate will start to move back up for pro-America capitalistic policies and anti-Obama opposition. Weekend polling always heavily favors (D) and the pollsters tend to always over-sample RATS.
And, for the second month in a row, the number of Americans who identify themselves as Democrats has fallen to a record low. My guess is that the remaining independents break roughly as the decided have broken, on about a 4:3 basis toward the GOP.
The same happened in 1994 if you look at Gallup. Republicans had big lead then it diminished to even early October then went back up 6 points in the last week of the election. I think this probably will happen again. The break against the incumbent usually happens about a week or so before Election Day.
And, few independent voters are going to vote for a “D” this time around. If someone holds a seat, the challenger will win if he/she is within 3-4% in the polls. Any Republican will win any empty seat within 3%. The ‘Enthusiasm Gap’ will account for 4% or 5% for the GOP. Turnout is key.
Jay Cost at Real Clear Politics has a good analysis of where the race stands as of Oct 4, comparing several turn out scenarios, and focusing on the critical independent vote. The major problem for the Democratic party? It is averaging less than one out of three independent/unaffiliated voters.
Importantly, all of these polls indicate that party voters have generally sorted themselves out – Republican voters have already broken to Republicans, Democrats to Democrats. Most of the undecided voters at this point are independents. Polls always tighten as election day nears in the districts.
However, the momentum is what counts, and the Democrats have none. The key is the likely voter turnout models – it should make liberals curl into the fetal position: +18 or +13 GOP. Unbelievable! …or believable?
Polls are nice. Votes are nicer. Winning elections are wicked nice.
Worst Polling = Adults.
Next Worst = Registered Voters (Gallup).
Most Accurate = Likely Voters (Rasmussen)