Split Ticket

Editor’s Note: In December 2024, the Split Ticket WAR model received a major methodological upgrade that resulted in WAR score changes. The findings remain directionally the same, but the updated WAR scores are found here.

Our 2020 House wins-above-replacement (WAR) model showed that spending still matters in American politics. Among other things, the new 2022 edition proves that candidate quality, or the lack thereof, can fundamentally impact competitive races.

Controversial contenders paid a bigger penalty across the board this cycle than they did in 2020. In many cases, as 2022’s statewide elections showed, Republicans tended to nominate weaker candidates than Democrats did — perhaps a byproduct of former President Trump’s continuing influence in GOP primaries.

We’ve already examined candidate quality’s impact on senatorial, gubernatorial, and Secretary of State races. Now we’re turning our attention to the recent House of Representative elections to analyze candidate effects in more detail. Namely, how strong was each candidate relative to his or her opposition? How much did the result in any given district differ from the performance that a generic slate of nominees would have been expected to produce?

Methodology

Our wins-above-replacement metric provides a quantifiable “score” for each district that displays whether the Republican or Democrat performed better relative to data-based expectations. During the design process, our goal was to account for candidate-independent factors intrinsic to every district and the national environment.

We started by controlling for each seat’s racial composition. This was essential given both the strong correlation race has with partisanship and differing electoral patterns observed among distinct demographic groups nationwide. Next, we controlled for district partisanships in the 2016 and 2020 presidential elections, yielding a good measure of overall lean and expected down-ballot lag.

We then controlled for the midterm shift that districts’ states experienced. As we have previously covered, individual states experienced different environments this year. New York and Florida had massive red waves, whereas Michigan and Colorado swung leftward despite President Biden’s low midterm approval ratings. With this in mind, a Colorado Democrat outrunning Biden seems a lot less impressive than a Democrat in New York City or Florida doing so. Any model quantifying candidate strength should deal with this accordingly.

Lastly, we controlled for a couple of election-specific factors, such as incumbency and money spent by candidates and outside committees. While money’s impact has been declining of late, it still matters, as our 2018 and 2020 models showed.

By accounting for all of these factors, we crafted a measure of candidate strength that answers a simple question: how big was the candidate quality differential on a per-seat basis?

Authors’ Note: We urge appropriate caution when using this metric. Similar to baseball’s wins-above-replacement metric, our model is not granular enough to split candidates by a quarter of a point. It is impossible, for example, to use this metric to determine whether Pat Ryan, whose WAR in NY-18 was D+2.5, was a stronger candidate than Susan Wild, whose WAR in PA-07 was D+2.2. That said, we can use our tool to conclude that both were probably stronger candidates than Eric Sorensen, whose WAR was D+0.7 in IL-17.

ANALYSIS

Democrats Overperformed in Battleground Districts

One key question many observers had in the wake of the 2022 election cycle was whether swing seat Democrats overperformed more than swing seat Republicans. The evidence we have indicates that was the case.

We classify districts as battlegrounds if they voted within 10 points of the nation in 2020, or have an opposite party incumbent despite lying outside of this partisanship range. (In TX-34 we factored in both incumbents). Of the 82 House districts in the battleground category, Democrats overperformed in 46 compared to the Republicans’ 34 ー a stark reinforcement of the candidate quality points discussed above.

The evidence suggests Republicans’ 222-213 House majority is also slightly smaller than it could have and should have been with better nominees, giving Democrats an easier path to winning back the House in 2024 even with expected redistricting in North Carolina and Ohio. In other words, the GOP has a candidate problem where it counts the most: swing districts. Even if Republicans had had a better national environment on Nov. 8, candidate effects likely still would have limited their majority.

Democrats Probably Won More House Seats Than They Should Have

By numerically quantifying underperformance, the WAR model allows us to estimate which House races would have had different outcomes if “generic pairs” of Republican and Democratic candidates had been nominated. For example, which Democratic-won seats would have flipped Republican with a more standardized set of nominees? Our conclusions suggest that along with their disproportionate battleground nominee strength, Democrats won one more seat in 2022 than they would have with generic nominee pairs.

The table above also indicates that the GOP had a bigger candidate quality problem than Democrats did. Of the 15 House seats that our findings suggest would have voted differently had generic nominees been on the ballot, a majority (9) ended up in the Democratic column.

A look at the campaigns waged in some of the districts mentioned shows exactly why this was the case. In OH-09 and AK-AL, both Trump-won districts, Democrats Marcy Kaptur and Mary Peltola won comfortably thanks both to low-quality opponents and their compelling personal brands.

Kaptur has served in Congress since 1983 and enjoys strong union connections in a working class, traditionally-Democratic seat. Her challenger, J.R. Majewski, beat two credible state legislators in the GOP primary despite never having held elected office. National Republicans later cut Majewski off from their spending apparatus after allegations surfaced that he misconstrued his military service record.

Peltola, who won a stunning upset in an August special election to replace the late House Dean Don Young, managed to establish a reliable brand much faster, exemplified by her “fish, family, freedom” slogan. Former Governor Sarah Palin, whom Peltola beat in two instant-runoffs, suffered from high unfavorability ratings and a string of controversies dating back to her 2008 Republican Vice Presidential nomination.

Ranked-choice voting undoubtedly enabled Peltola’s initial election in August, but the scale of her recent victory for a full term suggests that she likely would have won under the conventional system too. Pre-election polls showed her beating both Begich and Palin in head-to-heads.

Strong Democratic incumbents like Jared Golden (ME-02) and Matt Cartwright (PA-08) benefitted from the same dynamic in Republican-trending, Trump-won seats. Challengers Bruce Poliquin and Jim Bognet were more conventional candidates, but the GOP would have needed some extraordinary recruits to flip both districts.

Republicans Bo Hines and Joe Kent also cost the GOP two open seats: NC-13, and WA-03. The first one, based primarily in the suburbs Raleigh, marginally backed Biden in 2020 and comfortably elected experienced state legislator Wiley Nickel this cycle. Despite his relative youth and inexperience, Hines probably suffered most among swing voters because of their strong ties to Trump.

The second district, a historically-inelastic Trump +4 seat in southwestern Washington, wouldn’t have been that competitive without the former President’s influence. Moderate Republican Jaime Herrera-Beutler lost the jungle primary to far-right challenger Joe Kent and Democrat Marie Gluesenkamp Perez after voting to impeach Trump following January 6th. Gluesenkamp Perez eventually pulled off an upset in a race that Kent had been favored to win despite his controversial bid.

Henry Cuellar, a long-time Democratic incumbent and top overperformer, is not on the above list because TX-28, though Republican-trending, still broke for Biden by 7 points. In other words, a generic Democrat would have been expected to hold onto the seat. Only Cuellar could have won by 13, though, a sign that his Politiquero pedigree continues to hold weight in South Texas.

Zapata County (Trump +5), for instance, supported Cuellar by 46 while maintaining near presidential turnout. Strengthened by his Catholic background and social conservatism, Cuellar’s regional pull overshadowed an FBI investigation into accusations of illicit dealings with Azerbaijan and delivered him a cakewalk victory against otherwise-credible Republican Cassy Garcia.

Similar candidate quality dynamics also helped victorious Republicans, though to a lesser extent. All six of the seats Democrats would have been expected to win under generic conditions are within the states of New York, California, and Oregon ー each of which lurched significantly rightward between 2020 and 2022.

In the Golden State, Republicans Mike Garcia (CA-27) and David Valadao (CA-22) maximized crossover appeal to win reelection in double-digit Biden seats made bluer in redistricting. Democrats simply did not get the challengers that they needed to match such formidable opposition. Meanwhile, in the east, a similarly strong incumbent was moderate Republican Brian Fitzpatrick, who retained his Biden +5 PA-01 at a canter, despite our model suggesting that his seat should have flipped in what was an exceptionally tough cycle for Pennsylvania Republicans. Even if the national environment is bluer in 2024 (a presidential cycle), national Republicans worried about their House majority may take solace in the strength of Garcia, Fitzpatrick and Valadao.

Our most surprising finding is perhaps NY-17, which DCCC Chair Sean Patrick Maloney narrowly lost to challenger Mike Lawler. This Biden +10 seat, primarily based in Westchester County, divides New York City and Upstate New York. Republicans invested heavily in the 17th to take down Maloney, a historically strong incumbent with leadership status.

In the wake of his loss, the ex-congressman blamed Governor Kathy Hochul’s underperformance for his defeat. While this certainly contributed to his loss, our model suggests that a generic Democratic incumbent still should have beaten a standard Republican; the WAR score for NY-17 was R+2.7, and the seat was decided by just 0.6%. It is difficult to subjectively determine how much of the outcome depended on Lawler’s strengths or Maloney’s weaknesses, but there are indications that the incumbent’s own choices may have cost him reelection.

Maloney’s biggest faux pas was deciding to run in NY-17, which overlapped with just 25% of his original seat. His move pushed out fellow congressman Mondaire Jones, fueling bad blood among new constituents and lessening the benefits incumbents normally accrue from district ties. Unwise in any situation, taking on new turf proved especially dangerous this cycle given New York’s palpable rightward shift.

The DCCC Chair’s past victories established his reputation as a strong incumbent, an accolade confirmed by his D+6.4 wins-above-replacement score from 2020. Switching districts ultimately had a negative impact on his crossover potential. Had Maloney stayed in NY-18 (Biden +8) he probably would have held on, a possibility confirmed by Pat Ryan’s victory in the same seat in November.

Extreme Candidates Likely Paid An Electoral Penalty

Our district findings do indeed suggest that “extremist” candidates often struggle to consolidate support, especially in swing districts where persuadable voters can mean the difference between close races and blowouts. Four of the worst performing Democratic and Republican incumbents of 2022 could be classified as “fringe” members: Cori Bush (MO-01), Ilhan Omar (MN-05), Majorie Taylor Greene (GA-14) and Lauren Boebert (CO-03).

Bush and Omar, representing the St. Louis and Minneapolis-St. Paul metros, respectively, are among the House’s most progressive members — referred to as “the squad”. Greene serves rural northwestern Georgia and Boebert Colorado’s Western Slope.

Of these four representatives, three also underperformed according to the 2020 WAR model (with Cori Bush being the lone exception). This is hardly surprising, considering Split Ticket’s prior discovery that extreme candidates tend to pay electoral penalties, an ideological quandary we plan on analyzing in a followup article.

Omar, for example, ran behind Biden more than any other incumbent Democrat in 2020 (-17.3), nearly lost her last primary, and underperformed by 9.1 points in 2022. Unlike swing district underperformers, though, Omar will have no issues winning future reelections in her Biden +64 seat as long as she secures the Democratic nomination — a glaring uncertainty at best.

Boebert, meanwhile, proved that a mixture of poor candidate quality and a hostile political environment can endanger competitive seats — even when they have a decided lean toward one party. CO-03, her Trump +8 district, did vote for Governor Jared Polis, but it is hard to argue that a mainstream incumbent like Scott Tipton, whom Boebert beat in the 2020 primary, wouldn’t have won by more than 546 votes. In other words, her win was extremely underwhelming by any standard, even after accounting for the awful night that Colorado Republicans had.

Conclusion

If there was ever an environment or a cycle to remind us that candidate quality was still important, then 2022 was probably it. Across every major office and chamber, one thing was clear: candidate quality mattered, and it mattered enough to change the results of several races.

In hindsight, this is not a surprise, even with the acceleration of polarization. Voters are not robots, and the electorate responds negatively against candidates who are viewed as out of step, disconnected, or extreme, just as they still respond positively to good candidates who work for all constituents.

It is still possible to reach voters on the other side and in the middle while still turning out your own base. It is what many strong candidates do, and it is how several elections were won in this cycle. Our model is just one way of quantifying a candidate’s strength, but it is, to our knowledge, currently the only such metric out there. We hope that you find it informative in learning about how candidates performed and how and where elections were won and lost.

—

We hope you enjoyed this piece. If you did, consider subscribing to get our articles delivered straight to your inbox — it’s free!