North Carolina’s deep and lasting political split is often characterized as a bimodal phenomenon; a 50/50 split between liberals and conservatives, Democrats and Republicans, or urban areas versus rural. Indeed, this has long been the conventional political wisdom in our state. Yet what would you say if we told you that this characterization isn’t quite right?
Recently, we completed a deep analysis of Carolina Forward’s growing trove of polling data, including the cross-tabulations of different subcategories of respondents. And we found something eye-opening: clear and broad voter consensus across many high-profile issues. On issue after issue, we found a big majority of voters pretty much aligned on the major questions facing North Carolina, with one exception: conservatives.
The big split in public opinion is simply not liberals versus conservatives. Instead, it’s conservatives against everyone else. Poll after poll shows that North Carolina’s conservatives are increasingly way outside the mainstream of public opinion, with moderates and liberals – who together make up a big majority of voters – much more closely aligned.
Countless words have been written about how deeply hyperpartisan gerrymandering has distorted North Carolina’s politics for the last decade. This analysis adds a new quantifiable measure to that tally of evidence. Simply put, North Carolina’s politics are being more and more skewed to the extreme right by aggressively partisan gerrymandering. How skewed? Let’s see.
How Do Voters Describe Themselves?
In Carolina Forward’s polling, we allow respondents to describe themselves on a 5-step spectrum of Very Liberal, Somewhat Liberal, Moderate, Somewhat Conservative or Very Conservative. This is standard practice for the industry, and because we draw a representative sample of North Carolina voters, the breakdowns of respondents across these categories vary only a little. Thus, our weightings of partisanship and ideology are nearly identical to that of other polling organizations, like both Elon University and even the far-right John Locke Foundation. You can read more about Carolina Forward’s poll sampling and methodology here.
Moderates are a big plurality of North Carolina’s electorate. These percentages move a few points one way or another from poll to poll (see the margin of error), but fundamentally they remain very stable. North Carolina is simply a very moderate state.
Yet this is not the pattern that emerges in issue polling. As we will see, on issue after issue, the majority of liberals and moderates are in broad agreement. A minority of conservatives, however, still get their way.
By using some simple math to collapse Liberal and Conservative categories respectively by their sample weightings, we can arrive at combined Liberal/Moderate/Conservative issue breakdowns that bear out this phenomenon more clearly.
Consensus or Controversy?
It surprises many people to learn that many issues that are portrayed as controversial in the political media are actually not that controversial at all among voters. The problem is that some voters’ views count for more than others.
Take automatic voter registration (“AVR”), for example. AVR is simply not a very controversial issue among most North Carolina voters. It commands +16 net approval overall, and a whopping +39 net approval among moderates. Liberals and moderates are plainly in agreement.
Yet conservatives alone are strongly opposed:
Or there is the issue of qualified immunity for police officers. Qualified immunity is massively unpopular. Ending it is something like a policy unicorn: it is an issue that all racial groups, as well as urban, suburban and rural voters all agree on. (See our in-depth post on this topic.)
The only subcategory of voters who want to keep qualified immunity? You guessed it:
Take the issue of housing. North Carolina, like most of the country, is in the middle of an acute housing crisis. Big majorities of voters from different kinds of areas agree that state leaders should act on the issue, even if it means overriding some local control. Yet once again, conservatives disagree:
On many issues, like housing, the guiding logic for conservatives seems to be a basic and unmovable opposition to any action by the state at all. Take as another example the “For the People Act,” the foundational voting rights legislation currently pending in the U.S. Senate. The Act enjoys majority support among North Carolina voters, and is not even particularly controversial among liberals and moderates. Yet, as we have seen, the same minority of voters enjoy a veto:
We even see this pattern in the most supposedly “controversial” issue of all: abortion. It turns out that the right to choose is not controversial at all among most voters, even while conservatives remain steadfastly opposed:
The aberration of conservative opinion even extends beyond the purely issue space. Alone among North Carolina’s voters, who overwhelmingly trust teachers by a towering 24 points, conservatives alone do not:
In the Trump era, we have seen the American right-wing movement steadily become more paranoid, suspicious and distrusting of anything outside its own echo chamber. We see clear evidence of the same here. Self-described conservatives have demonstrably moved well outside of the mainstream of a society they increasingly look at with wariness and contempt.
By gerrymandering our state legislature’s district lines to be overwhelmingly safe for Republicans, the locus of our state’s political power has shifted from general elections to party primaries – specifically, Republican primaries. Republican primary voters are (in)famous for their extremist far-right views, to which aspiring candidates must pander to be successful in their party’s primary. (Those voters are also almost exclusively white.) Since those candidates are all but guaranteed to win in rigged districts, this has steadily moved our state’s politics into the far-right extreme, and significantly removed from the median voter’s preferences.
North Carolina deserves better elections, and a better class of representatives as a result.
