Skip links

A Failure to Launch? Kansas’ Republican Gubernatorial Contest and the History of Incumbent Governor Primary Performance

KEY POINTS FROM THIS ARTICLE

— As of Wednesday afternoon, Gov. Jeff Colyer (R-KS) trailed Kansas Secretary of State Kris Kobach (R) 40.6%-40.5% in Kansas’ gubernatorial primary. If Colyer’s deficit holds, it would mark the first primary loss for an incumbent governor in 2018.

— If the final outcome is similar to the current vote, Colyer’s defeat would make history: His present margin of defeat stands as the narrowest ever for an incumbent governor in a primary. Conversely, if Colyer wins based on the counting of outstanding votes and/or a recount, he could claim the record for narrowest primary win among incumbent governors.

— From the first statewide primaries in the late 19th century through Tuesday, incumbent governors have sought renomination in primaries 916 times. Out of those attempts, incumbents have won 830 times, a 91% renomination rate. But about one-third (300) of the 916 primaries in question featured no opposition for the incumbent seeking renomination. If we remove those races, incumbents in contested primaries won 530 out of 616 contests, an 86% renomination rate.

— Of the 616 contested primaries with incumbents, 541 featured elected incumbents and 75 unelected, successor incumbents. Overall, 87% of elected incumbent governors have won renomination in contested primaries while 78% of successor incumbents have done so.

“Tight as a tick” in the Sunflower State

The Kansas Republican primary for governor remains too close to call. As of Wednesday afternoon, Gov. Jeff Colyer (R-KS) trailed Kansas Secretary of State Kris Kobach (R) 40.6%-40.5% — a raw margin of just 191 votes — but thousands of provisional ballots still have to be counted, which could alter the outcome. However, if Colyer’s deficit holds, it would be notable because it would mark the first primary loss for an incumbent governor in 2018. Granted, Colyer is a “successor incumbent,” having moved from the lieutenant governorship to the governorship. Back in 2017, the Crystal Ball examined the electoral track records of successor incumbents in the post-World War II era and found that they performed worse than their regularly-elected counterparts in primary and general elections. Colyer is one of four successor incumbents who stepped up to a state governorship in the 2018 cycle and sought a full term. The other three — Govs. Kay Ivey (R-AL), Henry McMaster (R-SC), and Kim Reynolds (R-IA) — all won renomination.[1] Not coincidentally, Colyer had served the shortest time as governor prior to his primary up the quartet, only taking office on Jan. 31 after Gov. Sam Brownback (R) resigned to become President Donald Trump’s ambassador at-large for international religious freedom. If the final outcome is similar to the current vote, Colyer’s loss would make history: His present margin of defeat stands as the narrowest ever for an incumbent governor in a primary. Conversely, if Colyer wins based on the counting of outstanding votes and/or a recount,[2] he could claim the record for narrowest primary win among incumbent governors.

The Crystal Ball compiled data on every gubernatorial primary involving an incumbent in all 50 states up through this Tuesday. Our data set only includes primaries that occurred separately from the general election. That is, the data exclude Louisiana’s gubernatorial elections since 1975 because it is possible win outright in the first round of the Pelican State’s system by garnering a majority of the all-party vote. However, data from open primaries that nominated the candidate from each party with the most votes (Alaska and Washington have used such systems) and top-two primaries like those currently used in California and Washington are included. Unopposed primaries are also included in the overall count, including in cases where an incumbent did not appear on the ballot because of a state’s rules. For example, then-Gov. Jeb Bush (R-FL) did not appear on Florida’s 2002 primary ballot because the Sunshine State does not list unopposed candidates on its primary ballots (or its general election ballots). If a state used a convention-primary system where a candidate won enough support at the convention to avoid a primary, the incumbent is not included. An example of this would be Gov. Gary Herbert (R-UT) in his 2010 special election win and 2012 reelection, when he won sufficient backing at the state GOP convention to avoid a primary. It should be noted that many states once had prohibitions against incumbents running for renomination or traditions barring an immediate reelection bid (today, Virginia is the only state with a statutory prohibition against incumbents seeking a second consecutive term). Some states also moved from permitting consecutive two-year terms to a sole four-year term for a time. Factors such as these influenced the number of incumbents seeking renomination.

As referenced above, Colyer could go down as the narrowest loser or winner among all incumbents who sought renomination in a primary. Table 1 presents the 20 closest primary losses for incumbent governors in U.S. history, and Table 2 presents the 20 closest primary wins for incumbents.

Table 1: 20 closest primary defeats for incumbent governors in U.S. history

Notes: In the “party” column, “D” and “R” signify Democrat and Republican, respectively. In the “type” column, “P” and “R” signify primary and primary runoff, respectively.

Sources: Archived election results; CQ Press Guide to U.S. Elections, vol. II (6th ed.); Newspapers.com

Footnotes: 1) Results for the Kansas Republican primary as of the afternoon of Wednesday, Aug. 8, and the race remains uncalled; 2) unelected “successor incumbents”; 3) special primary for a special election for governor; 4) Joseph died about one month after winning the Republican primary, so the state party had to replace him on the general election ballot; 5) technically “acting governors” but sought full terms while serving in this role.

As Table 1 shows, a close primary loss for an incumbent sometimes augured poorly for the incumbent party. Besides Colyer’s possible defeat, eight of the other 19 cases saw the incumbent party lose in the general election. Many Democrats believe that Kobach’s divisiveness — he has gained notoriety and infamy for his evidence-less claims about voter fraud, among other things — and the stark unpopularity of the recently departed Brownback could give Democrats their best shot at winning the Sunflower State’s governorship since former Gov. Kathleen Sebelius (D) won reelection in 2006. Kansas is sometimes referred to as a three-party state comprised of conservative Republicans, moderate Republicans, and Democrats. Some GOP insiders fear Kobach will turn off moderate Republicans, to the detriment of the party in both the gubernatorial election and in other important races on the ballot, such as U.S. House contests in KS-2 and KS-3. Should Kobach retain his slender primary lead, he will face state Sen. Laura Kelly (D) and independent Greg Orman in November. In the Crystal Ball’s ratings, the general election race Leans Republican, though Kobach’s far-right views could make him more vulnerable than Colyer might be in a general election. Orman, who ran as the de facto Democrat in Kansas’ 2014 Senate contest, seems more likely to win over votes that Kelly might otherwise attract, which would probably ease the path to victory for the eventual Republican nominee.

Table 2: 20 closest primary victories for incumbent governors in U.S. history

Notes: In the “party” column, “D” and “R” signify Democrat and Republican, respectively. In the “type” column, “P” and “R” signify primary and primary runoff, respectively. “OP” signifies open primary whereby primaries are open to all voters and candidates from all parties run together, but the candidate from each party with the most votes advances to the general election; the primary data for Alaska’s 1978 open primary include just the votes for Republicans.

Sources: Archived election results; CQ Press Guide to U.S. Elections, vol. II (6th ed.); Newspapers.com

Footnotes: 1) Unelected “successor incumbents”; 2) Hickel ran as a write-in candidate in the general election and finished second to Hammond 39%-26%; 3) Langer ran as an independent candidate in the general election and defeated Welford 36%-35%.

Should Colyer win after the outstanding votes are counted and/or after a recount, he would likely win by less than one-tenth of a percentage point (0.10 points), which would hand him the title for the closest-ever primary win for an incumbent governor. The present record holder is Gov. Jay Hammond (R-AK), who bested Wally Hickel (R) by 0.12 points among the GOP candidates in Alaska’s 1978 open primary. Hickel, a former governor who left office to serve in President Richard Nixon’s cabinet, then ran as a write-in candidate in the general election and finished behind Hammond 39%-26%. The parties of this narrow group of primary winners had slightly more success in the general election than the parties of the primary losers in Table 1, with the incumbent party losing six of the 19 completed general elections. This makes sense, intuitively, because the incumbents reached the general election in each case and incumbents tend to be stronger candidates for their parties than non-incumbents. Gov. Bruce Rauner (R-IL) managed a slender primary renomination back in March and sits at the bottom of this list; the Crystal Ball views him as a major underdog in November, rating the race as Likely Democratic.

The earliest statewide gubernatorial primary featuring an incumbent took place in 1898, when Gov. William Ellerbe (D-SC) survived a primary runoff to win renomination in the Palmetto State two years after he won South Carolina’s first statewide primary (as well as the first primary runoff). Because the Democratic primary was tantamount to election in South Carolina (and most of the South) at that time, Ellerbe’s primary win also assured him of a general election victory in both 1896 and 1898. Ellerbe died in 1899, and Lt. Gov. Miles McSweeney (D) succeeded him as South Carolina’s governor. In 1900, McSweeney became the first successor incumbent to seek renomination in a primary, making it through a runoff en route to winning a full term of his own. The first elected incumbents to lose a gubernatorial primary both fell in 1908: In Georgia, Gov. Hoke Smith (D) lost to Joseph Brown (D) by five percentage points (52.6%-47.4%),[3] and in Washington’s first statewide primary, Gov. Albert Mead (R) lost to Samuel Cosgrove (R) by four points (28.3%-24.2%) in a crowded field.[4] In 1919, Gov. William Runyon (R-NJ) became the first successor incumbent to lose a primary while seeking a full term. Runyon narrowly lost a September primary to Newton Bugbee (R) by about four points while serving as acting governor of the Garden State. Runyon held that post because Gov. Walter Edge (R) had resigned the governorship to take his seat in the U.S. Senate (Edge won a Senate race in November 1918 and waited until May 1919 to take his place in Congress’ upper chamber).

From those early primaries through Tuesday’s nomination contests, incumbent governors have sought renomination 916 times in primaries (these data include the Kansas race based on its current result). Out of 916 attempts, incumbents have won 830 times, a 91% renomination rate. Of course, incumbents are often unopposed in primaries. About one-third (300) of the 916 primaries in question featured no opposition for the incumbent seeking renomination. If we remove those races, incumbents in contested primaries won 530 out of 616 contests, an 86% renomination rate. Table 3 lays out the data for incumbent governors and primaries based on their elected status and if they had intraparty opposition.

Table 3: Incumbent governors and primary performance

Sources: Archived election results; CQ Press Guide to U.S. Elections, vol. II (6th ed.); Newspapers.com

The vast majority of these primaries featured elected incumbents. Of the 616 contested primaries with incumbents of some type, 541 had elected incumbents and 75 had successor incumbents. Unsurprisingly, based on both intuition — unelected gubernatorial incumbents would naturally be weaker, having not previously won the office they hold — and the Crystal Ball’s previous research for the post-World War II period, successor incumbent governors have performed worse in primaries than elected ones. Whereas 87% of elected incumbent governors have won renomination in contested primaries, 79% of successor incumbents have done so. While the sample of successor governors in contested primaries is small — just 75 — the percentage who won primaries in that time coincidentally about matches the 75% success rate for such incumbents in 2018, where three of four have won (assuming Colyer does indeed lose).

In 2018, seven of seven elected incumbent governors have sought and won renomination in primaries. On Saturday, Hawaii’s primary features perhaps the most likely opportunity for an elected incumbent governor to lose a primary in 2018. Gov. David Ige (D-HI), who himself crushed incumbent Gov. Neil Abercrombie (D-HI) in Hawaii’s 2014 Democratic primary, is in a closely-fought contest with Rep. Colleen Hanabusa (D, HI-1). Hanabusa herself came up just short in a 2014 primary challenge to Sen. Brian Schatz (D-HI), an appointed incumbent. Now she is taking on Ige, who earlier in the cycle appeared to be well on his way to defeat. A March Mason-Dixon poll found Hanabusa up 47%-27% as Ige struggled with the fallout from controversy regarding a false missile alert in January that his administration failed to correct quickly in part because Ige forgot his Twitter password. Morning Consult has consistently found Ige with one of the highest disapproval ratings of any governor (49% disapproval in the first quarter of 2018, 46% in the second quarter). But Ige’s position in the primary appears to have recovered to some extent — Mason-Dixon went back into the field at the start of July and found Ige ahead 44%-40%, while a late July poll for the Civil Beat found Ige up 43%-34%. Should Ige lose, that would make elected incumbent governors seven for eight (88%) in primaries in 2018. Assuming the seven incumbents seeking renomination in states with primaries after Hawaii’s successfully win renomination — most are heavy favorites or unopposed — an Ige loss would mean elected incumbents would go 14 for 15 (93%) in 2018 primaries.

With the polls close in Hawaii, it seems unlikely that Ige could lose by the sort of margin that he won with in the 2014 gubernatorial primary. Ige defeated Abercrombie by 35.9 percentage points, the worst defeat for an incumbent ever. Table 4 displays data for the 20 largest primary defeats for incumbent governors.

Table 4: 20 worst primary losses for incumbent governors in U.S. history

Notes: In the “party” column, “D” and “R” signify Democrat and Republican, respectively. In the “type” column, “P” and “R” signify primary and primary runoff, respectively. “OP” signifies open primary whereby primaries are open to all voters and all candidates from all parties run together, but the candidate from each party with the most votes advances to the general election; the primary data for Washington’s 1940 open primary include just the votes for Democrats.

Sources: Archived election results; CQ Press Guide to U.S. Elections, vol. II (6th ed.); Newspapers.com

Footnotes: 1) Failed to make a primary runoff; 2) unelected “successor incumbent.”

Back in 2014, the Crystal Ball examined how Abercrombie made history for the entirely wrong reason. The other two recent members of this list of 20 are Gov. Frank Murkowski (R-AK), who lost badly to future vice presidential nominee Sarah Palin (R) in Alaska’s 2006 GOP primary, and Gov. Jim Gibbons (R-NV), who fell to current Gov. Brian Sandoval (R-NV) in the Silver State’s 2010 Republican primary. Note that Gov. Preston Smith’s (D-TX) 8.7% in the 1972 Democratic primary is the smallest share of the vote ever won by an incumbent governor in a primary or primary runoff, a figure made even more remarkable by the fact that he was unopposed two years earlier for the Texas Democratic nomination.

Conclusion

The Kansas primary result may not be finalized for some time, especially if there is a recount. Nonetheless, Colyer’s current 0.1-point margin of defeat would rank as the closest loss for an incumbent governor in the history of primary elections. Should Colyer lose, the result would be somewhat unsurprising given the greater vulnerability of unelected, successor incumbents in primaries compared to elected incumbents. With a potentially close and competitive general election ahead in the Sunflower State, the narrow GOP primary result could delay efforts to unify the party, and a Kobach candidacy might deflate turnout among Republican moderates and/or shift some of their votes into the Democratic or independent column. Kobach might also further juice turnout among Democrats, though they may be as enthused and motivated as they can be in the Age of Trump. Then again, Orman’s presence in the race could help Kobach triumph over splintered opposition.

Footnotes

1. Ivey handily won her primary with 56.1% of the vote while Reynolds had no primary opposition. McMaster had to battle through a primary and primary runoff, but he defeated John Warren (R) 53.6%-46.4% in South Carolina’s GOP runoff.

2. Kobach’s office would manage a recount, and there appears to be no law forcing Kobach to recuse himself from overseeing a recount in a race involving himself.

3. In the 1910 Democratic primary, Smith avenged his 1908 defeat by defeating Brown by two points (see Table 1). Before 1917, Democratic Party primaries in Georgia did not officially determine the nominee. Instead, the primary allotted delegates at the convention by county unit vote, which then officially nominated a candidate. Even after implementing direct primaries, however, Georgia still used the county unit system to determine outcomes, so primary winners and losers were not decided by the overall popular vote until 1962.

4. Washington used a ranked-vote primary system from 1907 to 1917. In the 1908 primary, no GOP candidate won a majority, so ranked votes decided the Republican nomination. Henry McBride won 32.9% of first-choice votes compared to Mead’s 31.7% and Cosgrove’s 25.0%. But when second-choice votes were added to each candidate’s total, Cosgrove received 28.3% of all first- and second-choice ballots compared to Mead’s 24.2% and McBride’s 23.4%.