The Hot Seat: Oklahoma’s 5th District

Welcome to The Hot Seat, a new blog where I break down 2018 US House predictions from a House model I built that varies from all the other election handicappers out there, and then I attempt to explain why it thinks that way. First up is Oklahoma’s 5th congressional district, currently held by two-term Republican representative and Deadliest Warrior frequent presence Steve Russell, that my model believes is much more competitive than it initially appears.

Crystal Ball: Safe R

Cook Political Report: Safe R

Inside Elections: Safe R

My Model: Tossup

District Overview:

OK-5 is smack dab in the middle of the state and consists of most of Oklahoma county and then the whole of Pottawatomie and Seminole counties. However, the clear majority of people live in Oklahoma county, so the story of OK-5 is really the story of Oklahoma county, with the other two being important in the margins.

The chart above shows that, when the three counties are combined, what percentage of votes belonged to Oklahoma county. Now OK-5 doesn’t have the complete county as a part of it, so realistically only about 86 or 87% of the vote in this race will be from there. Based on past vote totals, about 9% will come from Pottawatomie and the last 4% will be Seminole. The graph also shows that while the county’s share has been increasing a bit, the interesting tidbit is that the drop from an on-year election to a midterm has steadily closed, and turnout here will be not necessarily be an advantage to the more Republican rural counties in an off-year.

The Fundamentals:

Election handicapping, as an industry, relies a lot on how a seat looked or acted in the past, but Oklahoma recently had a massive shake-up. Third parties finally started getting on the ballot statewide in 2014, which they largely haven’t been able to do since Oklahoma slapped the toughest hurdles to qualify in the nation after Nader’s presidential runs. In the 2014 gubernatorial run, two Independents (Kimberly Willis & Richard Prawdzienski) received a combined 3.2% statewide. They only got 2.8% in Oklahoma county as the margin from the 2010 gubernatorial race shifted 9.7% towards the Dems, a clear swing. In 2016, only Gary Johnson made the presidential ballot, and captured 5.75% statewide, an impressive performance for a state that had almost no past elections to build off. Romney got 66.8% statewide in 2012 with no 3rd parties, and Trump only dropped to 65.3%, so Johnson’s run appeared to have cut into the Democrat’s margin heavily. This isn’t the case in Oklahoma county however, where Johnson got a higher 7.14%, which cut entirely into Trump’s lead going from Romney’s 58.3–41.7 win in the county to Trump’s 51.7–41.2 win over Clinton there. This isn’t hugely surprising given that Marco Rubio actually won this district in the primary, and Trump came in third in the county. The chart below plots the GOP top of the ticket’s candidate performance in Oklahoma county since 2008, the dark red is the raw vote and the split to the pink is the two-party vote when minor parties started running. You can see here a trend downwards recently that heavily accelerates in the recent cycles, and not just because of 3rd parties.

Next up is the top of the ticket baseline itself for the model. While it is not used heavily here, it is worth it to look at Trump’s current standing in the district, or the best approximation. In 2016, an R+1.1 year, Trump captured only 53.2% in this district, down from Romney’s 59.2% in 2012. In a good Dem year, which looks to be about D+6 or 7, it is not unreasonable to assume that Trump’s favorables are under 50% here, though his unfavorables may not reach that point, with a lot of undecideds or tepid feelings. The shift here of down 6 points is also stark, considering he only fell about a point statewide and the model penalizes the GOP for that. Mary Fallin in 2014 only got 52.3% and with the hard Trump shift in OK-2, this is for sure, the most left-leaning district in the state. Through forecasting, we can guess that this seat will most likely lean 9 points to the left of the statewide total. This means that statewide, the gubernatorial candidate will need to outrun Mary Fallin’s 2014 percentage of 55.81% from a good GOP year to establish down ballot coattails. This seat does run a couple of more percentage points towards Republican for Congress and with the incumbency advantage is likely to outrun the top of the ticket’s percentage by about 4 points given past district behavior and demographics that in a regression, basically break even. This is because the district is a bit younger, has more African Americans in it and has a higher gender ratio to women than most districts. It is also a tad more educated but this is all canceled out because about half the district is Evangelical, with some additional Methodists, and religion could play a huge crucial part here in the final few percentage points if this seat gets close

Downballot, it doesn’t look too different in the congressional races and the model gives a boost for avoiding a drop, but it may be a bit artificial on competitiveness. Russell did the worst for a GOP congressman since redistricting by a little bit but still received 57.1% of the vote, dropping 3 points from 2014, and that was an open seat. However, a 3rd party Libertarian, Zachary Knight, took 6.1% of the vote. This was up from the last couple of elections, where it was only about 4%, but it does not appear as if he is running again. Third parties have been trending hard Dem since the election, and if there is no minor party candidate on the ballot, it could boost the Dems.

The Personal Touch:

It would be irresponsible to at least not give the context and what goes on in this seat, so this section will be reserved for the details. Things like who’s running, some background to the state or district environment, and any intrusion by the national parties into this race.

We’ll start off with the widespread dissatisfaction with of Republicans in the state, which currently control all levers of power. A Morning Consult poll in January 2018 put Governor Mary Fallin’s approval at a dismal 28–61 approval/disapproval split. A Sooner Poll survey pegs it at 35.3–58.1 for her favorable/unfavorable numbers. The GOP-controlled legislature is even lower at 31.2–57. The discontent with the current rule is reflected in the special elections held after the 2016 elections, where we have eight examples to look at in the chart below. These were all Republican-held seats before the elections, though 4 have since flipped.

The backlash here is evident, as the average shift from the presidential margins of the last two cycles was around 30%, and except for one, beat the margins by over 25%. A lot of this is Dem enthusiasm from low turnout elections, but these numbers are shockingly high. In addition to natural anti-Trump mobilizing, the Oklahoma teachers’ unions are planning to go on strike, which shows deeper organizational power in the state in the run-up to the elections.

The Republicans: Steve Russell is still well-liked and will be running again for office. At the governor level, there are six candidates currently running, but the major two frontrunners are popular Oklahoma City mayor Mick Cornett and the current Lieutenant Governor Todd Lamb. Both have a shot and it’s wide open with a lot of undecideds, as the recent Sooner Poll shows:

According to this poll, about half of the Republican voters in the 5th district support Cornett, but his lower name ID has led to a significant vote bust outside his area. Lamb is well-known and does decent in the rural parts, but his ties to the current administration remain a hinderance. A poll from the previous August had Lamb up on Cornett 32.1–29.2, and both have seen drops, Richardson has grown a tad since then.

Update (7/26): This race has gone into a runoff between former OKC mayor Mick Cornett and businessman Kevin Stitt, who pulled an 11th hour comeback, primarily by dumping a ton of cash into his home media market of Tulsa. Stitt is weaker in the Oklahoma City portion and likely would be in the general, more so than Cornett.

A very recent Sooner poll shows a huge general election split here, with Cornett beating Democratic candidate Drew Edmundson 43–41, while Edmundson leads Stitt 46–37 in the 5th district.

Update (8/30): After the runoff, Kevin Stitt, whose base is in Tulsa, won the runoff by preventing Cornett from dominating in the Oklahoma City market, and is now the nominee, expecting to do weaker in the classic Republican areas. Also, Kendra Horn won a dominating victory in the Democratic party primary, Horn would go on to win by about 50 points, and enter the general with a strong campaign infrastructure and some cash raised. As of now, despite what could be renewed attention from a 25% chance to win in the 538 model, OK-5 has remained under the radar from either party, at least publicly for now.

The Democrats: The Democratic field to take on Russell is wide open, and virtually no money has come into this race, because it’s seen as unwinnable. The four candidates consist of three virtual unknowns: Eddie Porter, Elysabeth Britt, and Kendra Horn, who appears to have raised the most cash on hand among those that filed with six figures. The last one is Tom Guild, a perennial candidate and former college professor who has run for the 5th district before, first losing in the general to now-Senator James Lankford in 2012, then losing this district’s primary twice the next two cycles to the same guy. By virtue of name recognition alone, he may be the frontrunner, but if he wins, it will likelier be due to external factors and the fact that people know who he is. Another very credible challenger is Kendra Horn, whose cash on hand may be able to boost her name ID and favorability and capture the nomination.

In the gubernatorial race, the frontrunner is a solid Democratic recruit and a former Oklahoma Attorney General from 1995 to 2011, who ran and barely lost the nomination in the 2010 election. National groups are eyeing this as a spot to take a sleeper governor’s mansion, so some money may come in that can trickle down ballot.

Update (7/26): After raising the most cash out of the candidates, Horn narrowly missed the cutoff in the crowded field at 43.8% to Tom Guild’s 17.9% in second place, and is likely to be the assumed nominee for now. There is also some tensions as video footage captured Guild removing a Horn sign in the final hours of the campaign. As of the last quarter filings, Horn had raised $417,000 but spent most of it in the primary. Guild had raised much less at $76,000, and only has over $17,000 left on hand. Russell has spent $252,000 so far and has only $474,000 on hand, much less than most incumbents. He has also raised primarily from PACS, as opposed to the Dems.

The Verdict:

The model currently rates this as about a Tossup/Tilts Dem race, with Russell getting a little under 50%, but he could eke out a plurality win if a third-party candidate files in this district, which according to Ballotpedia, they have not done yet. However, I am betting against the model (which has not worked in a special election yet) and writing my personal dissent on this one as a Tossup/Tilts R. The main reasons for this being the possibility of a plurality win and the big question here regarding the most important factor in the gubernatorial race statewide: Is Mick Cornett able to win the primary, bank on local goodwill and distinguish himself from Fallin and the frustrations at the Republicans controlling the state legislature, or does he get swamped by external factors? My prediction is that if Cornett can outperform statewide and in Oklahoma county a bit, beating the model’s district benchmark, this seat can be saved. If Todd Lamb is the nominee and is tied to Fallin, this seat becomes a lot more in danger of a flip, with no local reservoir of goodwill to tap into. I will solidify a prediction after the primary happens, and we’ll see if there are last minute third party candidacies. The takeaway here is that this is not a safe seat, and should be getting a lot more attention, which appears to have finally caught at least the eye of at least someone following a special, and should be on the list of races to keep tabs on

A huge shout out to the following people for helping me and using their data:

Miles Coleman (@JMilesColeman)

Everybody at Daily Kos for their pres-by-district breakdown, especially David Jarman (@DavidLJarman)

Dave Leip’s uselectionatlas

Photo Credit: Hammon’s American Politics Guide: Total Politics Bill

Analyzing Elections From Upcoming Battlegrounds to Historical Results