Election polling and predictions

--or--

Post-election brain dump

I'm not going to spend a lot of time on any one topic but instead just get a bunch of stuff out that's twirling around in my head. Each of these topics is probably deserving of it's own post, but that will come later. Right now I have to get this stuff out of my head so I can move on with my life.

Governor Projection

The results are still trickling in, but I now have enough information to put up a preliminary projection.

This is based on partial results from the counties, so take it with a grain of salt, but right now I project the final numbers to be:

Mark Dayton (D): 45
Tom Emmer (R): 41

See this post for background info.

Outcome Prediction System

I'm going to try something tonight, that is to predict the final outcome, as returns come in, based on vote totals in some key counties.

Sound fun?

I'll provide a little background on what I'm doing in this post and than start an actual Outcome Estimator post when returns start coming in.

I will be doing this at MN Progressive Project so that I don't have to keep two separate threads going.

Final GPI; Dayton with a 5 point lead

Here is the final GPI.



Additionally Nate Silver puts Mark Dayton's odds of winning at 86% and his weighted average is virtually identical to the one above.

Mark Dayton has lead the entire month of October by about five points, but this race will all come down to turnout. The DFL has the advantage in that regard, we just need to capitalize on it.

That said, there is still a 14% chance that everyone is wrong. One of the data points that makes up that 14% is of course the 2006 Minnesota Governor's race. In that contest Mike Hatch had a three point lead in the simple polling average 45.5-42.5 and lost 46-47, a four point swing for Tim Pawlenty.

To avoid that fate again we need to GOTV like mad. So everyone who doesn't have anything to do today, GOTV.

Predicting turnout on Tuesday

I wrote a post before the primary election on predicting turnout. That was quite a bit more involved since there was the issue of the new summer date to consider. Predicting turnout this Tuesday should be a little more straight forward since there are not really any unknown variables to consider, unless you think there will be a GOP surge.

I'm not discounting that possibility, I just don't think that it will have nearly the same effect in Minnesota where we already turnout at a high rate.

Races to watch on election day; MN Senate edition

I want to start with a bit of clarification, I omitted from the House edition post, the methodology behind the three metrics I'm using and why I'm only using those three metrics. To be clear, these were not chosen at random, but rather have had the most predictive power in past Minnesota elections that I've analysed.

A few notes about those metrics.

When I refer to the amount of money raised by a candidate, I talking about the amount raised in individual contributions this year, not total amount raised or cash on hand or anything else. This metric gives us some idea of how much support a candidate has in the absence of polling.

This is by no means a perfect measure, but it has decent predictive power in that incumbents who raise significantly more than their opponent rarely lose, and those who raise significantly less, are in danger of losing.

The other metrics I'm using are; 2006 vote shares and hPVI.

With that out of the way, let's get on with the Minnesota Senate edition of "Races to watch."

One more quick thing about that PPP poll

In my rush to get the GPI update up last night before heading off to a series of Halloween parties, I overlooked something interesting in the crosstabs. That is, who people that have already voted, voted for. Those numbers are:

Public Policy Polling (10/30, those who have voted, no trend lines):

Mark Dayton (D): 55
Tom Emmer (R): 36
Tom Horner (I): 6

It's not a huge part of the sample, only 9%, but still a good indicator for Mark Dayton.

That is all.

GPI update; PPP 's first MN-Gov poll shows a small Dayton lead

Public Policy Polling (10/30, likely voters, no trend lines):

Mark Dayton (D): 43
Tom Emmer (R): 40
Tom Horner (I): 15
Undecided: 3
(MoE: ±2.2%)

A few things about this poll that should be noted. One, PPP interviewed 2,058 likely voters for a tiny 2.2% margin of error. This is the largest sample any pollster has yet taken of the Minnesota Governor's race and it should make us that much more confident in the results.

New CD8 poll confirms tight race

SurveyUSA (10/28, likely voters, no trend lines):

Rep. Jim Oberstar (D): 47
Chip Cravaack (R): 46
George Burton (CP): 3
Undecided: 4
(MoE: ±3.9%)

We finally get a public poll of congressional district eight and it confirms the internal poll that the Chip Cravaack campaign released a few weeks ago. It's clear that this race is tight and the internals of this poll have to be troubling for Jim Oberstar.

The partisan composition of the poll is 41% Dem, 28% GOP and 28% Indy, Cravaack is keeping it close by winning independents 56-32, which doesn't bode well for Oberstar's chances of winning the remaining undecided vote.

One caveat, of course, is what I talked about earlier today in regards the the SUSA poll of the Governor's race, automated polls this year have skewed in the GOPs direction by an average of two points, according to Nate Silver. Survey USA in particular has skewed right by four points on average.

Regardless of that, Jim Oberstar is in a close race, something he's not used to and he will need to make sure that democratic voters in CD8 turnout for him on election day. And it appears that is exactly what he is doing.

GPI update; SurveyUSA goes cellphone style, GPI stabilizes

SurveyUSA (10/28, likely voters, 10/14 in parentheses):

Mark Dayton (D): 39 (42)
Tom Emmer (R): 38 (37)
Tom Horner (I): 13 (14)
Undecided: 6 (4)
(MoE: ±4%)

This is the first poll in a while thats shown Mark Dayton below 40% and while his numbers have slipped since the last SUSA poll, that could just be float within the margin of error rather than a decline.

Emmer camp disputes MPR poll, releases internal

Public Opinion Stratagies (10/28, likely voters, no trend lines):

Mark Dayton (D): 40
Tom Emmer (R): 40
Tom Horner (I): 13
Undecided: -
(MoE: ±4.38%)

In response to the release of the MPR/Humphrey Institute poll showing Tom Emmer down by 12 points to Mark Dayton, the Emmer campaign released an internal poll by POS conducted on October 24th and 25th that shows the race tied.

I'll let the infamous Cullen Sheehan take it from here:

All polls are snapshots in time. The new MPR/Humphrey Institute results are an unfortunately predictable snapshot in absurdity. The HHH poll's track record rivals only television weather forecasters for accuracy. Four years ago this same poll predicted that Mike Hatch would beat Tim Pawlenty by six percentage points. Minnesota should be as confident in this poll's prediction as Governor Hatch was during the last gubernatorial campaign.

GPI update; Emmer below 30 again

MPR/Humphrey (10/28, likely voters, 9/29 in parentheses):

Mark Dayton (D): 41 (38)
Tom Emmer (R): 29 (27)
Tom Horner (I): 13 (16)
Undecided: 20 (20)
(MoE: ±5.5%)

MPR released it's last poll of the Governor's race this morning and they don't find a whole lot has changed since last time, other than Horner losing a few points to the other two candidates, which other polls had already shown.

If the undecided number seems high, it is, but The Humphrey Institute lumps undecided into the same category as "don't know," "refused," and "voting for someone else." It would be nice to see what the undecided number is all on it's lonesome.

New CD1 poll; Tim Walz at 50%

SurveyUSA (10/27, likely voters, 10/15 in parentheses):

Rep. Tim Walz (D): 50 (47)
Randy Demmer (R): 41 (42)
Steven Wilson (I): 4 (4)
Undecided: 4 (5)
(MoE: ±4.1%)

This could just be float within the margin of error, rather than a Tim Walz surge, but either way there just aren't enough undecided voters left for Randy Demmer to pull off the upset. These are exactly the kinds of seats that Democrats need to win if they want to hang onto control of the house, but the rest of the Midwest might not be so friendly.

Races to watch on election day: MN House edition

This is the first post in a series about what races will be the most interesting to follow on election day. If you are looking for early indications about who's going to have a good night, these are some races you'll want to pay attention to.

A few notes about the metrics I'm using.

When I refer to the amount of money raised by a candidate, I talking about the amount raised in individual contributions this year, not total amount raised. This metric give us an idea of how much support a candidate has in the absence of polling.

This is by no means a perfect measure, but it has decent predictive power in that incumbents who raise significantly more than their opponent rarely lose, and those who raise significantly less, are in danger of losing.

The other metrics I'm using are; 2008 vote shares and hPVI.

With that out of the way, let's get on with the Minnesota House edition of "Races to watch."

GPI update; another day, another poll

St. Cloud St. U (10/25, likely voters, no trend lines):

Mark Dayton (D): 40
Tom Emmer (R): 30
Tom Horner (I): 19
Undecided: 5
(MoE: ±5%)

While the initial reaction to this poll might be to think that Tom Horner is back in the thick of it, there are reasons to be skeptical.

GPI update; the Horner slide confirmed

StarTribune (PSRI) (10/24, likely voters, 9/26 in parentheses):

Mark Dayton (D): 41 (39)
Tom Emmer (R): 34 (30)
Tom Horner (I): 13 (18)
Undecided: 12 (12)
(MoE: ±3.9%)

This is now the third pollster to show Tom Horner sliding from his high-water mark of 18% down into the low teens, so it's clear that he has lost support, the question now becomes weather he can hold onto what he's got left or if the slide continues.

GPI update; Horner's slide continues

Rasmussen (10/22, likely voters, 10/6 in parentheses):

Mark Dayton (D): 44 (40)
Tom Emmer (R): 41 (38)
Tom Horner (I): 10 (15)
Undecided: 4 (5)
(MoE: ±4%)

This is the third poll in a row to show Mark Dayton with a 2-4 point lead over Tom Emmer and Tom Horner losing support since the last poll by the same pollster, although by a much more significant margin than we've previously seen. Despite what I wrote yesterday, there is now reason to believe that Tom Horner's support is in fact eroding, similar to what happened to Tim Penny in 2002, although I would certainly like to see another poll showing him this low before jumping to any conclusions.

GPI update; SurveyUSA is back in the field

SurveyUSA (10/14, likely voters, 9/15 in parentheses):

Mark Dayton (D): 42 (38)
Tom Emmer (R): 37 (36)
Tom Horner (I): 14 (18)
Undecided: 4 (4)
(MoE: ±3.7%)

This is confirmation of where the current average has the race, although there are a few things about the poll itself worth pointing out.

GPI update; Ras now says no to leaners

Rasmussen (10/6, likely voters, 9/22 in parentheses):

Mark Dayton (D): 40 (34)
Tom Emmer (R): 38 (36)
Tom Horner (I): 15 (18)
Undecided: 5 (2)
(MoE: ±4%)

A few things change with this update of the GPI, the first of which is something truly radical, standardizing the way I format new poll results. I put a lot of thought into what the most effective way to do it would be and after careful deliberation I decided to crib Swing State Project's poll formatting style, so if it looks familiar you know who to blame.

GPI update; what happened to a tight race?

MPR/Humphrey released a new poll of the governor's race today and it's good news for the blue team. Not only do they show Mark Dayton with an 11 point edge, 38-27, the internals paint an even better picture.

• Minnesota voters have awoken from their summer slumber. More than 8 out of 10 Minnesotans are interested in the November elections, a substantial increase that is being propelled by the energizing of formerly turned-off Democrats.

•Horner is crippling Emmer's campaign, soaking up one-fifth of Republican voters while taking only half as much from Dayton.

• Dayton is tapping voter frustration and appears better positioned among voters who have not yet declared their support for a candidate.