What’s Happening in South Dakota?

South Dakota is normally a red state. And by red, this means very red. Donald Trump won the state by 30 points in 2016 and by 26 points in 2020. In past eras of the New Deal and of the 1980s Farm Crisis, Democrats were more competitive, but today, South Dakota’s heavily white and rural demographic easily carries the GOP to victory.

With these factors in mind, incumbent Republican Gov. Kristi Noem should normally be a lock for re-election. A strong conservative in a Republican state does not typically lose as an incumbent. But, for reasons Split Ticket will explain below, the site rating for this year’s gubernatorial race will move from Safe Republican to Likely Republican.

The first big reason to acknowledge a more competitive race is that candidate quality matters more in state level races than federal races. Split Ticket has produced numerous wins-above-replacement models to show that while top-line partisanship is increasing in overall salience, it is still not itself 100% determinant of electoral outcomes. Candidate quality deltas matter, and arguably explain “underperformances”. For example, Massachusetts Sen. Elizabeth Warren became the first incumbent Democrat at the federal level, to lose one of the state’s 14 counties to a Republican, since 1996.

This effect is only magnified in a smaller electorate like the South Dakota voting pool, where persuasion of each individual voter has yields a higher percentage of the electorate than in a bigger voting pool. More relevant than the Elizabeth Warren example, is that Kristi Noem is the only Republican to allow a genuinely competitive South Dakota governor’s race since 1986. She won by under 4 points, against Democratic State Sen. Billie Sutton, who was a good candidate in his own right.

For reference, the last time a Democrat won the governorship was in 1974, when Kristi Noem was three years old. A map is pictured below.

What this map shows is also the second key factor in South Dakota’s relative closeness: the old Democratic coalition of wheat farmers in the northeast of the state. As the state is situated at the confluence of many larger cultural regions, the coalitions involved pitting politically swingy and culturally Midwestern wheat farmers versus politically red and culturally Western ranchers. More recently, this coalition is exemplified in the 2002 Commissioner of School and Public Lands election, shown below. Eastern South Dakota is blue, reflective of the partisanship of the farming areas at the time, while western South Dakota votes much like neighboring Wyoming. These dynamics allowed South Dakota Democrats like Tom Daschle and Tim Johnson to compete on equal footing with the GOP from the 1980s until the mid-2010s.

The 2018 election featured a semi-throwback to this coalition. Rural areas which were solidly pro-Trump in 2016 reverted massively back to the Democrats in 2018, though not by Bryce Healy margins that would likely have won Sutton the governorship. But this election is proof that with a good enough Democrat and a bad enough Republican, a race can be a lot closer than a 26-point GOP presidential edge would suggest. Noem won the open gubernatorial seat by only 3 points against Billie Sutton, the Democratic minority leader of the state Senate.

To better understand the elasticity of South Dakota, a table of the 2018 statewide results is listed below.

Even with the gubernatorial contest taking up the lion’s share of attention in South Dakota, there was a fair amount of variability between the Democratic vote share, ranging from the 35% in the Secretary of State race to the 45% in the Attorney General race. This is proof that a halfway-decent Democratic candidate can run a race against a below-average Republican that merits a rating more competitive than Safe Republican. Democrats have a solid recruit in State House Minority Leader Jamie Smith, the leader of the lower chamber’s eight Democratic members. While not from a rural area like Billie Sutton, Smith represents a recruit with institutional support and with potential to earn the votes of disaffected Republicans who dislike Noem.

Moreover, Noem has also been dogged by a recent scandal in which she was accused of improperly using her position to get her daughter an appraiser’s license. All of these aforementioned factors combine to illustrate that fundamentals plausibly point to a closer-than-expected gubernatorial contest.

The other big aspect to consider is the set of polling indicators Split Ticket has received. Two polls have become public news via social media. The first is a Noem internal, which had the race at a 45-42 Noem lead, and Split Ticket has verified its authenticity with senior South Dakota figures. The second was a poll from SDSU, which also had Noem up 45-41 over Smith. In addition to this, we are aware of two other polls, both of which show a race that is far closer than expected. The most recent of these was a poll shared with Split Ticket by a senior South Dakota figure in which Noem was at 46%, Smith at 40%, and libertarian Tracey Quint at 6%, with an MOE of +/- 3%.

All of this is to say that Noem is highly unpopular in South Dakota at the moment, especially relative to partisanship. South Dakota insiders have mentioned that polling finds her underwater with both women and independent voters, and she runs behind incumbent Republican senator John Thune by massive margins in every poll conducted, indicating that the closeness of the race may not simply be an artifact of bad polling.

It should still be noted that a Noem victory is highly likely in this Trump +26 state, and that fundraising and fundamentals both strongly favor the incumbent for re-election. A Smith upset is extremely unlikely. But there exists a tiny lane under which this is now plausible, based on the evidence we do have, and it may be just enough to change the rating of this race.

Based on a combination of historical precedent, fundamentals, and candidate quality metrics, Split Ticket sees enough justification that the South Dakota governor’s race is no longer Safe Republican. While a Republican win is still expected, it is not a surefire deal, and Noem is virtually certain to underperform Trump’s 26-point margin. This race is now Likely Republican. Noem should win at a canter, but she may underperform virtually everyone in the process.

Ed. note: This article was updated on Oct. 11 at 1:30 PM EST to link to the full, released SDSU poll with the final numbers.

I’m a software engineer and a computer scientist (UC Berkeley class of 2019 BA, class of 2020 MS) who has an interest in machine learning, politics, and electoral data. I’m a partner at Split Ticket, handle our Senate races, and make many kinds of electoral models.

Leave a Reply

%d bloggers like this: