Thursday, October 21, 2010

The Horner Effect, Part 3

As the Independence Party candidate in the Minnesota Governor's race, Tom Horner faces an uphill battle to get elected. Because of this the question that everyone ends up asking is; who will he take more votes from, Mark Dayton or Tom Emmer? History suggests that the success of Independence party candidates in Minnesota comes more at the expense of Democrats than Republicans, but the dynamics this year are different. Tom Horner is himself a former Republican and the GOP candidate, Tom Emmer, hails from the fringes of Republican extremism. Given those factors, the conventional wisdom has become that if Tom Horner hurts anyone it was going to be Tom Emmer.

This is the final chapter of a three part series on what effect we can expect Tom Horner to have on the Minnesota Gubernatorial election. In this installment I'm going to look at the polling and who it shows Tom Horner is having more of an adverse effect on. Part one of the series was a look at past Minnesota elections and part two was a look at how the 2010 election is different from those past Minnesota elections.

With that then, here is Part 3...

The Horner Effect in action

Let's get right to it, here is a chart of the partisan support for Tom Horner from the cross-tabs of the public polls which release such information.

Image Hosted by

Except for in the most recent MPR/Humphrey poll Horner's getting only slightly more self identified Republicans than Democrats. In fact the only two polls that show any significant gap in support for Horner between partisans are both MPR/Humphrey polls. Just looking at this supports the theory that Tom Horner will pull more Republicans than Democrats, if only slightly, but there are other factors to consider as well.

In the StarTribune poll they expicitly asked voters who they would vote for if Tom Horner was not in the race, those results:

Image Hosted by

In this scenario Mark Dayton's lead expands, with almost half of Horner voters moving to his column while Tom Emmer is getting less than a third of those voters. Tom Horner's partisan share in the StarTribune poll is 9D/11R/38I, so if we assume that all the D's and R's go home than Mark Dayton is winning the remaining independents 38% to 21%. This tendency for independents to go for Dayton more than Emmer is supported by the favorable/unfavorable numbers in the same poll.

Image Hosted by

This would seem to be the reason why independents are moving to Dayton more than Emmer when Horner's name is removed from the poll. So while Tom Horner may be getting slightly more GOP support than DFL support, he's apparently also keeping a lot of Dayton leaning independents in his column. To much shouldn't be drawn from just one poll of course, but this is the only poll so far that has explicitly asked for voters preferences without Tom Horner in the race, so it will have do.

What happens on election day

The conventional wisdom around third party candidates is that come election day their support collapses as some of their supporters decide to vote for someone with a shot of winning. Let's take a look at some recent Minnesota history to see if that theory holds up. Here are graphs of the polls released in the last month of the election for 2002 & 2006 Gov and 2008 Senate (I'm using 2006 Gov instead of 1998 because their wasn't enough polling in 1998 for any kind of analysis):

Image Hosted by

Final Polling Average: 23%
Election Day Total: 16%
Difference: -7

Image Hosted by

Final Polling Average: 7%
Election Day Total: 6%
Difference: -1

Image Hosted by

Final Polling Average: 16%
Election Day Total: 15%
Difference: -1

In Tim Penny's case, as you can see from the graph, his support eroded rather quickly leading up to the election. If we instead just use the last four polls in the average he ends up with 17%, for a difference of -1, just like the other two races. In recent Minnesota elections than, the third party collapse theory hasn't held up, instead what we see is an across the board 1 point decline between what the candidate was polling and what they received on election day.

In fact Rasumussen pointed this out in the polling memo of their most recent poll of the Governor's race:

In the Minnesota governor’s race, Rasmussen Reports has made a decision not to use our traditional leaners model. Normally, that model shows support falling off for a third-party candidate. However, in Minnesota, third-party candidates often defy that trend, and a look at the initial preference data suggests that may be happening this year.

Based on this I think it's hard to argue for any kind of Horner collapse and it's safe to assume that Tom Horner's final vote share won't be more than a couple points less than his final polling average, which right now looks like this:

Image Hosted by

Current Polling Average: 15%

It certainly appears as though any momentum he may have had is gone and he's settling into the Dean Barkley zone of 15%. It's possible that this year is not like past years in Minnesota and Horner's support will collapse to a greater degree on election day than it has for past candidates, but I wouldn't bet on it. Minnesota independents have shown in the past that they are perfectly happy sticking with their guy to the bitter end.

The Wrap up

Can Tom Horner actually win? At this point that seems doubtful, rather than continuing to climb into the twenties when he was at 18% in the polls, he instead has fallen back into the mid-teens. At this point he would need something dramatic to happen to have a shot. So what is The Horner Effect, what has been the point of this whole three part series?

Tom Horner appears to be masking what would otherwise be a more substantial Dayton lead under two candidate conditions. It's unlikely that he ends up costing Mark Dayton the race because he's getting the support of enough Republicans for the independents not to matter as much, but without Horner it's not inconceivable that Mark Dayton would be staked to a larger lead in the polls than he currently has, say 8-10 points rather than 4-5.

There is a flip side though, unlike past IP candidates it appears that Tom Horner is soaking up enough Republican votes that Tom Emmer is hitting a lower ceiling than past GOP candidates have. We won't know how any of this plays out until election day of course, but the GOP this year is in a position they're not used to being in, having members of their party defecting to the IP candidate. So while Tom Horner may be having a somewhat adverse effect on Mark Dayton, what he has really done is make this race very difficult for Tom Emmer, who is having to fight to keep his base together.

In a certain sense this entire line of thought is irrelevant though. The presence of a third party candidate in major statewide races in Minnesota doesn't appear likely to end anytime soon, so asking the question of who would have benefited from that candidate not being in the race becomes an exercise in after the fact finger pointing.

It's hard to argue that the biggest factor at play here is not luck. Roger Moe got unlucky that a former Democratic congressman ran as the third-party candidate, and while the dynamics of that race were volatile and many feel Moe was a less than good candidate, any shot he might have had of winning was probably lost the moment Tim Penny entered the race. In retrospect the same might be said of the 2010 election, Tom Emmer was going to have a tough time getting elected in Minnesota because of who Tom Emmer is, but when you add Tom Horner to the mix, his road becomes that much tougher.

No comments:

Post a Comment