By Michael Bitzer
With four in ten of the 7.8 million North Carolina registered voters having selected ‘unaffiliated’ in their party registration, pundits continue to describe these voters as ‘political independents’—with the idea that they are up for grabs between the two major parties (which themselves are at 30 percent each).
A recent podcast that explores the political U.S. South highlighted whether North Carolina was ready to ‘flip to the Democrats’ with this year’s U.S. Senate election—with Mitch Landrieu observing that the state’s unaffiliated voters are “otherwise known as independents—in the middle…”
And one of the state’s leading partisan strategist—Democrat Morgan Jackson—noted that unaffiliateds “don’t want to affiliate with either political party,” while Republican strategist Paul Shumaker observed that the unaffiliated voters are concerned about what impacts them directly in deciding their vote—and by extension, are not partisan loyalists.
Being registered unaffiliated tells us how someone is registered. It does not, by itself, tell us how that person thinks, votes or identifies politically.
So are these unaffiliated voters really independents and therefore up for grabs in North Carolina? The answer, at least from the evidence presented here, is not necessarily—and perhaps not even in the same way from Manteo to Murphy.
This has been a serious research question to better understand North Carolina’s politics. In 2022, several colleagues (Drs. Christopher Cooper of Western Carolina University, Whitney Ross Manzo of Meredith College, and Susan Roberts of Davidson College) and I published a study of the state’s unaffiliated voters as the ‘unmoored voter’: those who are not just simply shadow partisans but, on average, are distinct from the two major parties.
In our analysis, over half of the unaffiliated registered voters who participated in the 2012, 2016, and 2020 presidential primaries could be termed “shadow partisans,” who consistently voted in one party’s primary all three years. And those shadow partisans split evenly between being ‘Democratic’ and ‘Republican’ in their consistency.
The other portion of unaffiliated primary voters—47 percent—we determined were “floaters”: those who shifted from one party's primary one year to another based on the electoral environment or the choice of candidates.
In our statistical analysis, we generally found that the unaffiliated North Carolina voter occupied the ‘go-between’ space regarding attitudes towards politics: they were more conservative than Democrats but more liberal than Republicans, and were generally in between the two parties when it came to satisfaction about the president at the time (2021 under Joe Biden) and the country. They also had distinctive policy attitudes on things like free community college and raising taxes on high income earners (again, between the two parties), while being the group that was most supportive of a third party over the two major party identifiers.
This research offers an important warning: treating all unaffiliated voters as a single pool of persuadable independents can be misleading.
What About The Unaffiliated’s Voting Patterns?
Another analysis of NC’s unaffiliated voters is the 2024 election and the percentage of partisan (or not) registered voters in a precinct’s electorate, compared to Trump’s percentage of the precinct vote share. This scatterplot analysis takes each precinct’s registered party’s percentage of voters casting ballots and compares it to Trump’s two-party precinct vote share.
Taking the two major parties, one would think as the percentage of registered Republicans increases in the precinct’s electorate, so too would Trump’s vote percentage—and indeed, that’s exactly what happens, as shown by the black line-of-fit.
 |
| Scatterplot created by Julius.ai |
The black line is the single statewide weighted fit. Its R-squared value is 0.913, indicating an exceptionally strong relationship (91 percent) between registered Republican share and Trump’s precinct vote share. And it’s practical vote implication is important as well: for every 1-point increase in a precinct’s registered Republican share, Trump’s two-party vote share was expected to increase by about 1.28 points.
But as many N.C. political observers know, North Carolina politics isn’t just shaped by partisan dimensions, but also geographic/regional aspects. The above graph delineates precincts within one of four regions: central cities in an urban county, the outside the central cities but inside the urban county (urban suburbs), the surrounding suburban counties to an urban county, and finally the rural county precincts—in different colors.*
And when you break each of these four regions into their own separate lines, the Republican increase is similar across the regions.
 |
| Scatterplot created by Julius.ai |
The relationship is strongest in Urban Suburb precincts (R-squared=0.961) and Central City precincts (R-squared=0.956), followed by Surrounding Suburban County precincts (R-squared=0.883). It remains strong but comparatively lower in Rural County precincts (R-squared=0.773).
Now, take the Democratic precinct percentage in the electorate, and naturally one would expect the line to go down (meaning, as the percentage of registered Democrats in a precinct’s electorate increases, Trump’s vote share decreases).
 |
| Scatterplot by Julius.ai |
And indeed, it does, not just state-wide but also within the four regions.
 |
| Scatterplot created by Julius.ai |
For every 1-point increase in a precinct’s registered Democratic share, Trump’s two-party vote share was expected to decrease by about 1.12 points, based on a vote-total-weighted statewide fit. And while the R-squared values were slightly lower than the Republican values, the Democratic percentage in a precinct’s electorate was strongly related to Trump’s vote percentage.
Now, the Unaffiliateds—what happens to Trump’s two-party vote share when the Unaffiliated percentage in a precinct’s electorate increase?
 |
| Scatterplot created by Julius.ai |
The overall relationship is weakly negative, with a weighted R-squared value of only 0.020. Compared to partisan registration that is highly predictive of presidential voting, unaffiliated registration, by itself, is not.
The estimated slope is -0.422, meaning that across all precincts, a 1-point increase in unaffiliated registration share corresponds to an estimated 0.42-point decrease in Trump’s two-party vote share. However, the low R-squared means this single statewide line summarizes relatively little of the precinct-level variation.
But when broken out by each of the four regions, a fascinating pattern emerges.
 |
| Scatterplot created by Julius.ai |
The regional relationships differs substantially, and thus when combined, gives us the very weak and slightly downward line.
In Central City precincts, higher unaffiliated-registration shares were associated with higher Trump vote shares. The weighted slope was +0.53, meaning that a one-point increase in unaffiliated registration corresponded to about a 0.53-point increase in Trump’s two-party vote share. Even so, the relationship was modest, with unaffiliated share explaining about 9% of the variation in Trump vote share.
A similar, but weaker, positive relationship appeared in Rural County precincts. There, a one-point increase in unaffiliated registration was associated with about a 0.46-point increase in Trump vote share. However, the model’s explanatory power was substantially low, with an R-squared values of 0.037 (meaning, barely 4% of the Trump vote share can be explained by this one factor). In practical terms, unaffiliated registration in rural areas appears to lean somewhat more Trump-friendly, but it is far from a dominant predictor.
The pattern reverses in suburban settings. In Urban Suburb precincts, a one-point increase in unaffiliated registration was associated with an estimated 0.84-point decrease in Trump vote share. The relationship was modest but more pronounced than in the urban-core or rural models, with an R-squared value of 0.116.
The strongest negative regional relationship appeared in Surrounding Suburban County precincts, where a one-point increase in unaffiliated registration was associated with about a 0.89-point decrease in Trump vote share. The model explained roughly 10.6% of variation in Trump support.
This regional contrast explains why the single statewide unaffiliated-registration line is not especially informative on its own. Unaffiliated registration does not appear to represent one consistent electoral bloc across North Carolina. In central-city and rural precincts, a larger unaffiliated share was associated with somewhat stronger Trump performance. But in urban-suburban and surrounding suburban-county precincts, the relationship ran in the opposite direction: precincts with more unaffiliated voters tended to give Trump a lower vote share.
The divergence appears to suggest that unaffiliated N.C. voters could be shaped heavily by regional context. In the two suburban North Carolina regions, unaffiliated registration may capture voters who are less aligned with the Republican Party and less supportive of Trump. In central-city and rural areas, by contrast, unaffiliated registration may include more voters who are culturally or electorally receptive to Trump despite not registering as Republicans.
A caution to throw into the mix: first, this scatterplot analysis is a precinct-level relationship, not individual-level evidence about how specifically unaffiliated voters voted. Secondly, due to the regional R-squared values ranging only from 0.037 to 0.116, this may only tell us that unaffiliated registration is informative, but substantially less powerful than Republican or Democratic registration in explaining Trump’s precinct vote share.
North Carolina’s Unaffiliated Are Not One Thing
So, are North Carolina’s unaffiliated voters really independents, sitting in the political middle and waiting to be won over?
The evidence, as is the case in many areas of North Carolina (and American) politics, suggests a more complicated answer.
North Carolina’s unaffiliated voters are not one homogeneous electoral bloc. Some are “shadow partisans,” consistently behaving like Republicans or Democrats despite their registration. Others are genuine floaters who move between candidates, parties and elections. And 2024’s precinct-level evidence suggests that even the relationship between unaffiliated registration and Trump’s performance varies considerably depending on where those voters live.
That may be the most important finding here. The word “unaffiliated” tells us something about how a North Carolinian is registered. It tells us considerably less about how that person thinks politically or how they will vote.
And geography appears to matter.
In suburban areas, higher levels of unaffiliated registration were associated with lower Trump vote shares. In central-city and rural precincts, the relationship moved in the opposite direction. None of these relationships is nearly as powerful as the connection between partisan registration and presidential voting—and the relatively low R-squared values for the unaffiliated models are an important reminder not to overstate what the data can tell us.
But they do challenge the idea that North Carolina’s unaffiliated voters can simply be placed in a statewide “middle.” Yes, they may be a bridge between the two parties, but there’s complications crossing the canyon.
For campaigns, that distinction matters. A candidate looking at the state’s 40 percent unaffiliated registration and seeing a massive pool of independent voters up for grabs may be looking at the wrong number. The more useful question is not How many unaffiliated voters are there? but rather Who are these unaffiliated voters, where are they, and how have they behaved politically?
North Carolina’s growing unaffiliated electorate may indeed be an important source of electoral volatility. And in a state like North Carolina, that has an intense base of partisan support on both sides, little movements among the remaining voters can have big consequences. But volatility is not the same thing as ideological moderation, and registration status is not the same thing as political independence.
The “unaffiliated” label may tell us that a voter has chosen not to join either major party, for a variety of reasons. Increasingly, the challenge for political analysts is understanding what that choice actually means.
-------------------
* The classifications have examples such as for central cities, Charlotte in Mecklenburg County; Raleigh in Wake County; urban suburbs are Huntersville in Mecklenburg; Wake Forest in Wake County; surrounding suburban counties are Gaston and Union counties to Mecklenburg; Johnston to Wake; all of these classifications are based on the U.S. Office of Management & Budget’s bulletin.
--------------------
Dr. Michael Bitzer is a professor of politics and history and director of the Center for N.C. Politics & Public Service at Catawba College. This post was originally published at the Center's Substack account on August 17, 2026.
