Every cycle, there’s always a state that seems to be truly poised to surprise. Despite a clear partisan lean, all the conventional fundamentals – polling, fundraising, candidate quality, and more – lead both parties and election analysts to take it seriously. And yet, in most cases, these surprise states turn out to be nothing more than a mirage, retaining the status quo and making everyone question why they thought this cycle could be different.
To be sure, upsets in previously safe states do happen from time to time; Maryland’s 2014 gubernatorial race and Virginia’s 2021 gubernatorial race come to mind. But in many cases, potential upsets turn out to be nothing but smoke.
Hawaii Senate (2012)
One of the strangest races to emerge on the national scene in recent years, Hawaii’s Senate race in 2012 is a clear example of overestimating candidate quality and trusting polls over fundamentals. Hawaii has long been one of the most Democratic states in the country, with Republicans rarely if ever truly competing in federal races. After a closer-than-expected 10-point win in 2004, Democratic dominance returned in 2008 as Barack Obama carried it with 71.9% of the vote to John McCain’s 26.6%.
Despite this, 2012 initially seemed like it could be an exception; former two-term Governor Linda Lingle, a Republican, opted to run for the open Senate seat being vacated by long-serving junior Senator Daniel Akaka. Lingle actually outspent Democratic nominee Mazie Hirono, and what few polls were conducted often indicated a race within high single-digits to low double-digits. By all accounts, with the exception of the state’s baseline lean, this race met what most would regard as the traditional definitions of competitiveness. With the exception of the Cook Political Report, which rated the race as Leans Democratic, all outlets gave this race a Likely Democratic rating.
While Lingle outperformed Mitt Romney by an impressive margin of nearly 17 percentage points, she still lost 62.6%-37.4% – a margin that, in hindsight, seems entirely predictable given Hawaii’s almost unflinching Democratic lean.
Kansas Senate (2014)
For a time, it seemed like Kansas was poised to elect its first non-Republican to the Senate since the 1930s. The state itself was mired in a budget deficit widely attributed to the inept policies of Governor Sam Brownback. Incumbent Pat Roberts was also wildly unpopular, nearly losing a primary to a gadfly right-wing candidate and facing constant challenges to his residency. These two factors led the Kansas Democratic Party to make an unusual decision – they opted to drop their candidate from the ballot, leaving Independent Greg Orman as the only opponent for Roberts.
Orman – a self-proclaimed moderate who voted for Barack Obama in 2008 and Mitt Romney in 2012 – suggested he would caucus with whichever party held a majority in the Senate. While publicly neutral, federal Democrats quietly aided Orman behind the scenes. From October on, Roberts led in only six of 23 polls, and for a time the race was rated as Leans Independent by Sabato’s Crystal Ball. Ultimately, only the Crystal Ball (which moved the race to Leans Republican in their final ratings) correctly predicted the outcome; Roberts won by a decisive 10-point margin, well outside the Tossup threshold most outlets rated it as.
So what happened? Well, it’s undeniable that national Republicans went in hard to protect Roberts. But at the same time, the idea that Kansans would vote for someone just because they had an (I) next to their name instead of a (D) turned out to be wrong. Moreover, Orman’s refusal to lay out clear policy positions or to commit to caucusing with a particular party made it easy to tie him with national Democrats. Despite what the polls suggested, when push came to shove, Kansas acted predictably to its partisan lean.
South Dakota Senate (2014)
After longtime Democratic Senator Tim Johnson opted to retire, South Dakota was seen as one of the easiest pickup opportunities in the country. This turned out to be reality, as Republican Mike Rounds led every poll and won by a margin of nearly 21 percentage points. However, for much of the race, prognosticators expected it to be more interesting.
Rounds was universally regarded as a poor fundraiser and a weak candidate, but Democrat Rick Weiland failed to garner much excitement. However, the presence of former Republican Senator Larry Pressler as a liberal-leaning independent was seen as a complicating factor that could split the vote. By October, Democrats had convinced themselves that the race was winnable despite Rounds’s double-digit polling leads and began spending in hopes of pulling an upset. Election analysts moved the race in response; most moved it to Likely Republican, but the Cook Political Report moved it to Leans Republican.
Ultimately, Pressler finished a distant third, winning only 17% of the vote. Democrat Rick Weiland didn’t even clear 30% of the vote, while Rounds won with 50.5%. While that’s not an impressive showing in terms of raw vote percentage, a 21-point win is still a 21-point win.
All the Democratic Stretch States (2020)
2020 perhaps had the largest number of mirage states in recent memory. In a half-dozen heavily Republican states (Alaska, Kansas, Kentucky, Mississippi, Montana, and South Carolina), Democrats were deemed to have at least credible chances of winning. However, all of them fell short by double-digit margins:
- Alaska: Democratic-aligned Independent Al Gross kept polls within striking range, leading most outlets to rate the race as Leans Republican; Elections Daily rated it as Likely Republican. Incumbent Republican Dan Sullivan won by a 12.7% margin, outperforming both Donald Trump and Don Young.
- Kansas: Polls consistently indicated a close race, with ex-Republican Barbara Bollier appearing to be in striking range of flipping Kansas for the first time since 1932. With the exception of Elections Daily, which gave this race a Likely Republican rating, all other outlets deemed this a Leans Republican or Tilt Republican race. Republican Roger Marshall ended up winning by a healthy margin of 11.4%, about three points behind Trump.
- Kentucky: Amy McGrath excited national Democrats, raising boatloads of money against Mitch McConnell. Polling showed a race in the high-single digits to low-double digits range. Credible outlets like the Cook Political Report and Sabato’s Crystal Ball took the race seriously enough to rate it as Likely Republican; McConnell ended up winning by 19.5%, five points behind Trump.
- Mississippi: This inelastic, heavily Republican state gave perhaps the most impressive Democratic over-performance. A handful of outlets, including Sabato’s Crystal Ball, felt Mike Espy – who ran a spirited campaign in an earlier special election – made this race worth watching. While Cindy Hyde-Smith did pull out a win by 10 percentage points here, this was far behind Trump’s 16.5% margin of victory.
- Montana: Popular Governor Steve Bullock led in a number of polls against incumbent Senator Steve Daines, leading Elections Daily and Inside Elections to rate it as Leans Republican; others were more skeptical, with Sabato’s Crystal Ball deeming it to be a Likely Republican race. While Bullock did overperform, he still lost by 10 percentage points.
- South Carolina: In what inexplicably became the second-most expensive race of the cycle, Democrat Jaime Harrison raised a metric ton of cash to take on Lindsey Graham. Polls showed a close race, but Elections Daily was alone in deeming this state a Likely Republican race; the Cook Political Report went so far as to deem it a pure Tossup. Despite $276.9 million being spent, the final result was nearly identical to the Presidential race, with Graham prevailing by 10.2%.
What did all of these races have in common? Polls and fundraising. Even in implausible, inelastic states like South Carolina, it would have taken a major polling error to suggest something wasn’t actually happening on the ground. While in hindsight it’s easy to say that these races were always overhyped, more or all of the traditional indicators of a competitive race were met during the campaign.
Can Mirages be Predicted?
I can difficult to determine which races are truly competitive and which ones will turn out to be mirages. In this cycle, for example, there’s been increasing discussion about Oklahoma, a state that is beyond Safe Republican at the federal level. Many traditional standards of competitiveness are clear: Stitt’s approval rating is, at best, slightly above water, the RGA is spending, and polls have shown a competitive race. On top of this, a traditionally resonant issue in rural communities – schooling – has taken a focus, as has Stitt’s tumultuous relationship with Oklahoma’s Five Tribes. Going off these standard metrics, a rating of Likely or even Leans Republican could be justified.
On the other hand, it’s worth examining these elements with a cautious lens:
- Stitt’s approval rating is nowhere near as bad as Matt Bevin’s was when he lost the 2019 Kentucky gubernatorial race.
- Polling shows implausibly high approval ratings for Joe Biden in a state where he only won 33% of the vote, as well as implausibly high margins for federal Democratic candidates, who tend to perform even worse than Biden did.
- The salience of rural schools as a motivating factor to get an inelastic demographic to switch sides is unclear.
When looking through this angle, some cracks in Oklahoma’s potentially competitive status became a bit more clear. Similar logic can be applied to New York, a similarly uncompetitive state where a handful of recent polls have indicated a closer-than-expected race. While I wouldn’t discount either of these races off outright, I do think they are states that could well be looked back on as mirages.