Close-Up: Independent Candidates in Sparsely Populated Prairie States Are Ruining Our Probabilities

Polls in the Great Plains delivered another defiant middle finger to our conventional wisdom about the Senate races this year. SurveyUSA released its latest survey of the state, and while Democratic no-name Rick Weiland remains at his sub-30% support, the major change is that this poll is the first one to put independent and former Republican Sen. Larry Pressler in a clear second place behind former Gov. Mike Rounds.

There's a bit of a paradox here, though. When we talk about third party candidates we inevitably wonder: is the candidate spoiling votes for the Democrat or the Republican? In South Dakota, before ever looking at the polls, we would guess that Pressler was spoiling votes for Rounds, the Republican, since Pressler represented South Dakota in the Senate as a Republican for 18 years before losing the seat to the man who currently holds it, retiring Sen. Tim Johnson. The trends seem to confirm this--Pressler's support has generally grown since the summer, while Rounds still remains ahead but seems to have trouble breaking the low 40s in the polls compared to the summer, where his best performance was 59% in the first iteration of the YouGov panel survey. (Note, however, that that survey did not include Pressler as an option.)

Larry Pressler, then (1979) and now (2013).

Why South Dakota is like Kansas

This trend, however, must be qualified. The SurveyUSA poll also asked questions about what would happen first if Weiland were not running and second if Pressler were not running. When Pressler was removed as an option, about 3 in 5 of his supporters throw in with Weiland, bringing the race to a 47-47 tie between Weiland and Rounds (something we've never seen in the polls). And when Weiland was removed as an option, about 3 in 4 of his supporters move to Pressler's column, catapulting the former senator into a 15-point lead over Rounds. What appears to be happening, especially looking at the candidates' favorability ratings, is that Pressler (net favorable by about 8 points on average) is taking away votes from Weiland (net favorable by about 3 points on average), dividing what could be a more unified opposition to Rounds (net unfavorable by about 4 points on average), who receives at best tepid support from his own party. Additionally, throughout the year, the relatively inexperienced Democrat seemed horribly unequipped, both financially and politically, to defeat the Republican, with less money in a pretty red state. Sound like Kansas to you?

Why South Dakota is not like Kansas

If it does, I apologize in advance for bursting your bubble. First off, South Dakota is not as monolithically red as Kansas is. It has historically had no problem electing Democrats to Congress. For 8 years between 1997 (when Sen. Johnson took office) and 2005 (when Senate Minority Leader Tom Daschle was defeated by less than 5,000 votes), the Mount Rushmore State was represented in the Senate by two Democrats. From 2004 to 2011, its sole congressional district was held by a Democrat, Stephanie Herseth, who won landslide elections in 2006 and 2008 and lost by just two percentage points in 2010. (Incidentally, this means that for seven months between June 2004 and January 2005, South Dakota's congressional delegation was entirely Democratic.)

Second, Larry Pressler is a different breed of independent than Greg Orman is in Kansas. While Orman was unknown to Kansas voters until about August, Pressler had been a senator from 1979 to 1997--which may explain why in the SurveyUSA poll he's actually the most popular candidate among Generation X'ers (ages 50-64), who began voting during Pressler's tenure. In other words, he was well known even before he launched his Senate campaign in 2013. However, there's a catch--what Pressler has in name recognition he lacks in campaign funds, having only reported about $100,000 to the Federal Election Commission so far. Orman's last report (which is almost certainly an underestimation, given how slow the FEC is to publish their reports) puts him at almost $700,000.

Third, Pressler has made no statement on which party he will caucus with should he be elected. While Orman has historically flirted with both parties (he ran for the Democratic Senate nomination six years ago), Pressler is definitively a career Republican (is three terms in the Senate good enough?). That's no guarantee that he will caucus with Republicans if he wins, but it's a pretty good guide, although it is true that he dissociated from the Republican Party in 2013 when he could conceivably have won the nomination. There are several possible reasons he did so:

1. He thought his views were no longer in step with the Republican Party: "I don’t think I’ve moved. I think the party has moved." For example, although he voted for 1996's Defense of Marriage Act, he now has come out in favor of same-sex marriage equality.
2. He wanted to play kingmaker--possibly getting perks for deciding which party gains the majority. Of course, Greg Orman's potential position as kingmaker complicates this.
3. He doesn't care about winning; he just wants to go back to the thrill of the chase.
4. He didn't think he could win the Republican nomination. (See #1.)

We can safely toss out #3: he definitively stated that he "intends to win." #1 and #4 look like the most reasonable choices--most of his positions on social issues (gun control, abortion rights, marriage equality), his support of tax increases to close the deficit, and his endorsement of President Obama in 2008 and 2012 don't do him any favors among the conservative crowd. While it does seem like his positions associate him with mainline Democrats than mainline Republicans, the fact is that he has never been a Democrat and has associated with the GOP for a good 40 years now. As a result, our model treats his choice of caucus as a toss-up--50-50 in determining Senate control (much like Orman's choice in the event of a 50-49 plurality for Republicans). 

Finally, the truth is that we have no idea whether Pressler is actually receiving the 28% support our polling average gives him. This is best expressed in terms of the standard deviation of the polls we use to compute the average. The smaller the standard deviation, the more certain we are that a candidate's actual level of support is pretty close to the number he gets in the average. In some races, like Michigan's gubernatorial, the standard deviations for the candidates are a little less than 2--so we're pretty sure the candidates are actually receiving the levels of support the polls show. Another way to think of it is that lower standard deviations mean the polls have more consistently shown similar results. Weiland and Rounds's standard deviations are at a reasonable 2.07 and 2.68 respectively. But Pressler's standard deviation is at a whopping 7.20, largely because the SurveyUSA poll gives him 32% while the YouGov poll, which is also in our average, gives him just 12%. So we don't know how close to 28% Pressler's "real" level of support is.

How the race is messing with our simulations

The title of this post is misleading; in the next post I'll elaborate on what this whole South Dakota shake-up does for the election simulations, but for now suffice it to say that an expanded version of the simulator has been created to take into account a third party candidate who has a good shot at winning the election (so far no election has seen this, so we've been able to get away with just simulating elections between Democrats and Republicans). Our three-way simulator gives Rounds an 89% chance of victory, with Pressler winning a little over 10% of the time and Weiland defeating both candidates less than 1% of the time. Despite Pressler's gains, this race is still LIKELY REPUBLICAN--although it isn't nearly as clear as others.