by Christopher Cooper
Normally, the weekly voter registration updates on the North Carolina State Board of Elections are met with a collective shrug from all but the most dedicated, data oriented, and dorky observers of the state’s politics.1
There have really only been three notable exceptions to that rule: September 2017 when Unaffiliated crossed Republican to become the second largest group of registered voters in the state, March 2022 when Unaffiliated became the largest number of registered voters in the state and yesterday when, for the first time in the state’s history, registered Republicans outnumbered registered Democrats in North Carolina.
In October I wrote a piece for The Assembly about this shift—what it means and why it matters. I encourage you to read it in the context of what just happened.2 Some primary things to keep in mind are:
The Democrats are dropping, The Republicans are stable, the Unaffiliated are rising. Democratic party registration is on the decline in North Carolina as it is throughout the South3—see the graph above and watch the blue line drop like my hopes for the Vols throughout this football season. But that should not imply that the Republicans are gaining. The Republicans have approximately the same share of the electorate they had in 1988. The big gainers over this time period are the Unaffiliated. The Republican party flex isn’t that they have risen, but rather that they have remained steady during a time of decreasing party registration.
The Unaffiliated4 are gaining because of generational replacement, not party switching. People switching from the Democratic to the Republican Party is an incredibly small share of this change (around 2%). The primary driver of this change is that old people are overwhelmingly registered with the Democratic Party5 and first time voters (young people and people from other states) are registering as Unaffiliated in huge numbers. When Democrats at the end of the life span….let’s just say exit the electorate (as well as this mortal plane)…they are replaced, overwhelmingly, not with Republicans, but with Unaffiliated voters.6
North Carolina’s semi-closed-primary system is a primary explanation for the rise in Unaffiliated voters. North Carolina used to have a closed primary system (Unaffiliated could not vote in a party primary). In 1988, the Republicans opened up their primary to Unaffiliated and 8 years later the Democrats followed suit (giving us the semi-closed primary system we have today). Those two policy decisions led directly to increases in Unaffiliated share of registered voters. If the Democratic and Republican parties closed their primaries, Unaffiliated registration would drop. I talk about this history much more in the piece for The Assembly.
North Carolina’s primary system is a major explanation for the rise in Unaffiliated, but not the only one. Young people are less enamored with institutions in general and political parties in particular than they used to be and some of the rise in Unaffiliated registration is due to these larger forces. There are also administrative explanations—people who leave the political party question of their voter registration application blank are coded as “Unaffiliated,” almost certainly leading to some of the rise in the Unaffiliated. And, voter registration information in North Carolina is public, so many people simply don’t want their next employer, date, or next-door-neighbor to know their party affiliation.
Unaffiliated is not the same as Independent. People who do not register with a party in North Carolina are referred to as Unaffiliated, not Independent. I get it, it’s easy to substitute one word for the other, but the distinction matters. Michael Bitzer, Whitney Ross Manzo, Susan Roberts and I found that Unaffiliated voters are—on average—are more moderate than members of the two major parties, but “on average” is doing a lot of work in that sentence. Most people do lean in one direction or another (it’s just that some lean left and some lean right). Those folks who identify with a party but are not registered with a party are best understood as “shadow partisans”—masking their ideology with an Unaffiliated registration. We argue that, on the whole, Unaffiliated voters are best understood as “unmoored voters.” They may prefer one partisan dock or the other, but with nothing tethering them to their party (like official party registration) they may find themselves adrift in certain situations. Said another way: most Unaffiliated voters are not independent voters, but most independent voters are Unaffiliated voters.
- The future candidate emergence problem is real. As I discuss in Chapter 8 of Anatomy of a Purple State, despite the rise in Unaffiliated voters, Unaffiliated candidates rarely win. This is setting up a real problem going forward. In a world where the majority of young people are opting out of parties altogether, yet for all intents and purposes, only members of the two major parties win, who is going to be left to run for office in the future?
Other Stuff that Caught My Eye this Week
The sage of North Carolina politics, Gerry Cohen, reminded me that people who pre-registered to vote before turning 18 and will turn 18 before the next election will enter the North Carolina voter registration database next Saturday. I’ll report back on the partisan characteristics of those newest North Carolina voters next week.
On Twitter, Stanford Political Scientist Andrew Hall predicted that “Claude Code [an AI tool] and its ilk are coming for the study of politics like a freight train. A single academic is going to be able to write thousands of empirical papers…per year…..” He then went on to prove his point by replicating and updating an influential paper about the effects of vote-by-mail. The entire process took him one hour. Claude Code scraped new data from the web, ran statistical models, produced the output and wrote an entire paper. It also placed all of the replication code on a publicly available repository so people can recheck the work. This is mind-blowing stuff and is far beyond what you might expect if your primary exposure to AI is playing around on Chat GPT.
I can’t possibly wrap my head around what this means or how we are going to adjust. There are lots of academics hashing this out on social media with their observations about how it will affect publication incentives, the load on reviewers, tenure expectations, etc. within social sciences, but I’m much more interested in—and concerned by—what this is going to mean for our understanding of politics and government (that is the point of Political Science, right?). I’ve got no solutions here and no answer, but I do think we need to keep that goal—knowledge and understanding—at the forefront of these discussions. My general sense is that information literacy, the ability to filter out of the information wheat from the chaff, and the ability to communicate those nuggets of truth that help us understand our world are going to take on increased importance in a world that is drunk on information.
Speaking of academic research, a team of scholars conducted a field experiment in North Carolina during the 2020 election and found that contacting ex-felons who have completed their sentence and reminding them to register to vote did increase voter registration rate. The specifics of the script didn’t however, matter. The paper is forthcoming in the Journal of Politics.
Applications are now open for the Spring class of the Institute of Public Leadership. If you’re looking for how to get involved in public leadership in North Carolina and want to be a part of a program with decades of success and members from multiple political perspectives and parties, look no further.
I identify with all 3 categories.
Huge shout-out to the good folks at The Assembly. If you’re interested in people, place and power in North Carolina, I recommend subscribing. Although their North Carolina politics coverage is excellent, it’s notable that many of their most-read stories from 2025 aren’t explicitly political. So, although you, like me, might come for the analyses of Tar Heel politics, you might stay for stories about religion, sports, and the joys of Dominoes thin-crust pizza.
The entire United States South used to be dominated by the Democratic Party (see V.O. Key’s classic Southern Politics in State and Nation for an explication of the problems of one-party rule). From Reconstruction through the early 1970s, the South was referred to as the “solid South” because of the near-complete hold the Democrats had over the region. Beginning in the 1950s, picking up speed with the southern strategy, picking up even more speed with Reagan’s elections in 1980 and 1984, and solidified through 2010’s Tea Party movement and Project RedMap, the South flipped from one-party Democratic control to two-party competition to one-party Republican control. Smart people disagree about how to characterize the region today, but no one disagrees that the Democrats are not the dominant party anymore.
“The Unaffiliated” also could have been the name of the people in Pluribus who did not join (If you don’t get this reference, it’s ok. Pluribus is a TV show that people love to talk about. I am one of those people).
For more, see Chapter 7 of Anatomy of a Purple State.
I’ve seen a lot of folks on social media imply (or state outright) that Republican registration exceeding Democratic registration is because scores of people are making an active choice to leave the Democratic Party for the Republican Party. This is not true.
