What the data says about how pro-climate leaders can win down-ballot in North Carolina
By: Andrew Hong
Activists and organizers too often only use Presidential and US Senate election data to make key decisions on where to invest time and money in local and state races. While this approach is common in American politics, the practice can mislead organizations both on where true district pickup opportunities exist and the best strategy to flip said districts.
For instance, an affluent suburban district outside Charlotte may have swung for Biden in 2020, but may vote staunchly Republican in down-ballot races. Targeting that district over a State House district Biden lost but has flirted with Democrats down-ballot could lead to a missed opportunity and wasted resources. This is why we measured “ticket-splitting” across North Carolina to identify how partisanship changes up and down the ballot, and most importantly, where.
In another example, organizers often spend a lot of resources trying to activate new voters for overall turnout or persuade conservative voters to switch parties in a swing district. However in many close local races, there already exists the necessary number of ballots that voted for Democrats top-ticket, but simply left down-ballot contests untouched. Contacting those voters and providing education about down-ballot races and direct impact on their way of life can save immense resources and more efficiently increase down-ballot support from an existing pool of voters that already vote and already support progressive policies.
That is why Climate Cabinet is publishing this piece. Below we will show how “roll-off” rates across North Carolina can be the sole difference maker in state and local elections, and identifying ticket-splitting patterns can optimize organizing strategy to better target close local races to maximize climate action wins. These analyses can change the calculus of down-ballot political strategy.
TICKET SPLITTING
Ticket-splitting is the event when a voter supports two separate political parties and its candidates on their ballot. At Climate Cabinet, we are interested in measuring the drop (or increase) in pro-climate candidate support down the ballot in state and local positions that dictate overall climate policy. About 75-percent of the nation’s emissions goals will be met outside of the federal government, and Climate Cabinet Action exists to be the go-to resource in down-ballot climate politics.
We are looking at Democratic Party ticket-splitting rates here because Democrats largely support pro-climate legislation in American elections. Understanding states, districts, and voting precincts that have significant differences in down-ticket support helps organizers understand where climate action opportunities exist at state and local levels. This data often catches opportunities that are missed when one looks solely at Presidential and US Senate election results.
Likewise, measuring ticket-splitting rates can shed light on where local Democratic and pro-climate activists have been doing a stronger job than national Democrats at communicating the climate agenda to voters — and where national Democrats appeal more. Finally, we take a look at precinct-level ticket-splitting rates to start to understand demographic — not just geographical — trends in ticket-splitting rates, too.
North Carolinians vote equally pro-climate (Democratic) for Top-Ticket and Down-Ballot Races
When comparing every precinct’s results for the 2020 and 2022 U.S. Senate races to votes for local races (Congress, State Senate and House), North Carolinians voted for Democrats at both levels at roughly equal rates. There was little Democratic “ticket-splitting” where voters choose Democrats in statewide races and Republicans locally, generally. Democrats maintained a roughly equal level of support from top-ticket statewide contests to down-ballot local contests.
The 2022 difference between U.S. Senate and U.S. House results was 0.5% favoring U.S. House Democrats (House Democrats outperformed U.S. Senate Democratic candidate Cheri Beasley by 0.5% on average). That number was 0.03% for State House Democrats. Conversely, Beasley outperformed State Senate Democratic candidates by 1.4%. Overall, Democratic candidates maintained a remarkably stable level of support up and down the ballot.
Getting into the actual vote totals for Congress, 2022 was a good — not great — year for state and local Democratic candidates in North Carolina. US House Democrats won 47.7% of the vote compared to the 47.3% U.S. Senate Democratic Cheri Beasley earned that same year. Democrats earned more of the statewide congressional vote (sum of all U.S. House races across all districts) than Republicans in 2020, but were outperformed by Republicans in 2022. However the reason for that was because Republicans fielded a candidate in every district in 2022, but did not in 2020. When you throw out districts where Democrats ran unopposed in 2020, House Democratic candidates performed roughly the same in both 2020 and 2022.
In the State Legislature, 2022 was a solid year too. Democrats won 49.6% of contested State House statewide votes (41.4% of all races, including uncontested races with just one major-party candidate — Republicans contest more races than Democrats). Democrats also won 51.7% of contested State Senate races (39.8% of all races). It’s important to note that both statistics — with and without uncontested races — cannot fully reflect whether there is ticket-splitting favoring Democrats or Republicans down-ballot because two things are true about uncontested races: the party that doesn’t contest a race would’ve received more votes overall if they had fielded a candidate, and that parties tend to not contest races in districts that heavily favor the opposition party. So the “true” Democratic statewide total (if every district was contested) in State Senate and House races likely falls between 41.4% and 49.6% (likely closer to 49.6%) for State House, and between 39.8% and 51.7% (likely closer to the latter) for State Senate. As an aside, despite these small vote margins, Republicans enjoy a wide 71–49 House majority and 30–20 Senate majority as a result of Republican gerrymandering.
While average statewide support for Democrats remains consistent top-down the ballot, that is not the case for individual voting blocs. There were clear contrasting trends for ticket-splitting — and different kinds of ticket-splitting — depending on age and race.
Younger voters vote Democratic down-ballot (and older voters Democratic top-ballot)
Across all elections, there was a negative, but weaker, correlation between median age of a precinct and ticket-splitting rate for Democrats down-ballot. Put differently, older precincts split their ticket by voting Democratic for U.S. Senate and Republican for local races. Inversely, the youngest precincts split their ticket by voting Republican top-ticket and Democratic down-ballot. This could highlight a failure by National Democrats to attract younger voters in North Carolina, or a clear indicator of success among local Democrats to appeal to younger voters, or both. This relationship is existing but relatively weak, though, with the difference in Democratic drop-off being around just 1% between precincts with older and younger-than-average voters. Local organizers in North Carolina should keep investing in younger voters locally as they are likely to vote Democratic down-ballot.
Black voters vote Democratic down-ballot
Another consistent progressive voting group that historically supports Democrats is Black voters. This electorate is by far the largest minority racial bloc in North Carolina and is pivotal to progressive success in the state. Black voters, especially in rural areas, tend to support local Democratic candidates more than statewide Democrats, especially comparing between U.S. Senate and Congressional candidates. This relationship is also weak — much like with young voters — , with the difference between Democratic drop-off being around 2% on average for precincts with more and less Black voters than the average precinct.
In 2020, Congressional and State Legislative Democratic candidates did noticeably better than U.S. Senate Democratic candidate Cal Cunningham among majority-Black precincts, particularly rural Black precincts. While Congressional Democrats outperformed Cunningham by 0.7% in majority-white precincts, Congressional Democrats outperformed Cunningham by 2.6% in majority-Black precincts — nearly four times as much. This persisted in 2022, but to a smaller extent. This coincided with a noticeable drop in Democratic support for president between 2016 and 2020 among the same rural Black precincts in North Carolina. The key takeaway is clear: as (national) Democrats improve in the white suburbs, they are losing support in working class rural Black (and other nonwhite) communities especially in statewide federal races.
Overall, the recent trend lines in North Carolina suggest that Black and younger voters ticket-split with Republicans top-ticket and Democrats down-ballot; conversely older and non-Black voters vote Democratic top-ticket and Republican down-ballot. These statistics are interesting beyond just how North Carolinians ticket-split: it may be an indicator for future changes in the electorate’s partisanship trends. We see this in rural Black communities where some voters support Republicans in federal races while still voting Democratic locally, and simultaneously voting increasingly Republican in presidential elections.
While ticket-splitting rates are relatively low statewide (<2% change in support across all levels of elections and years), there are specific key regions that have substantive ticket-splitting in either direction. Specifically, white affluent suburbs split their tickets for Democrats in statewide contests and Republicans down-ballot, where rural Black regions ticket-split with Republicans top-ticket and Democrats down-ballot. Understanding where those areas are and which ways they split their ticket provides crucial information to organizers on where to invest in that simply looking at presidential results can misguide.
ROLLOFF
“Roll-off” rates are the percentage of voters who vote for top-ticket, statewide races but do not vote in down-ballot local contests. Also called an “undervote”, this is an important metric to measure because it can shed light on where activists can target outreach to voters that support pro-climate, Democratic candidates nationally, but are skipping key down-ballot races entirely. This can also optimize where climate organizations try to increase down-ballot awareness to activate pro-climate voters who already vote, to vote in local and state races.
Overall, there was an average roll-off rate of less than 1% in 2022 for state house, senate, and Congressional contests in North Carolina. This means less than 1% of voters who voted for U.S. Senate in 2022 did not vote in local races — when throwing out uncontested races. This translates to roughly 30,000 voters who voted for US Senate, but not state races.
There was greater roll-off in 2020, where nearly 2% of voters who voted for US Senate left state house and senate contests blank. This suggests there are more voters who only vote for top-ticket contests in presidential elections than midterm elections. This roll-off rate translates to roughly 100,000 voters who voted for US Senate, but not local races in 2020. This rolloff rate was enough to potentially flip State Senate District 9, where Republicans held on by just 1,150 votes or less than 1%. Turning out voters who only voted top-ticket can change the outcome of elections in places like State Senate District 9.
There was little correlation between race and rolloff rates, suggesting White, Black, and all other races vote in local races at similar rates in North Carolina.
There was also a very weak association between age and rolloff rates. In both 2020 and 2022, precincts with older voters voted down-ballot more often than younger precincts. But the association between age and roll-off is much weaker than with drop-off (ticket-splitting). This means that encouraging voters in young areas (who tend to support progressive policies) to vote down-ballot will only very marginally boost local pro-climate candidates. This also shows that even though less voters in young areas vote down-ballot, those same precincts still support local Democratic candidates more than national Democrats — an encouraging sign for local Democrats.
Overall, roll-off rates were still highest the more local a contest became (State House races had less votes than Congressional races than statewide races). Local candidates across the entire state are well-served by educating pro-climate voters to complete their whole ballot, especially during presidential years.
CONCLUSION
Overall, we see that looking at ticket-splitting and roll-off rates at a precinct level can paint a fuller picture of voting in North Carolina. The persistent roll-off rates, especially in presidential years, can cost pro-climate local candidates key districts; and mistargeting local races from tunnel-visioning ourselves to just top-ticket election data can waste resources meant to win pro-climate majorities. The data shows that young and Black voters are key to progressive success in North Carolina, and even more so to state-level progressive success as young, Black voters are an even greater share of the Democratic base in local races. Looking at surface-level, top-ticket data overshadows the fact that younger, Black voters are key to down-ballot success. And that non-nuanced approach can cost pro-climate activists crucial seats in North Carolina state government, as well as a future pro-climate generation of Gen Z, Black voters.
METHODOLOGY
Ticket-splitting (or drop-off) rates were calculated as the difference in Democratic vote share percentage between the US Senate contest and either US House, State Senate, or State House contest for each precinct. Precincts with uncontested races were thrown out of aggregated analyses (ex: statewide totals and regression analyses), which are defined as contests without either a Democrat or Republican candidate.
Roll-off (or undervoting) rates were calculated as the percentage of voters who voted in the US Senate contest but did not vote in either US House, State Senate, or State House contest for each precinct. Precincts with uncontested races were thrown out of aggregated analyses (ex: statewide totals and regression analyses), which are defined as contests without either a Democrat or Republican candidate.
Data sources: Redistricting Data Hub election results, L2 Registered Voter Data
