Wednesday, July 10, 2019

Previewing North Carolina for 2020's Election

With an upcoming interview for Spectrum News' "Capital Tonight" that previews the 2020 election in North Carolina, I thought it might be worthwhile to share some findings of the Old North State's electoral dynamics in the past several presidential election years, as we are slated to another "competitive battleground/lean GOP" state for next year's campaign.

As a reminder, since 2008, the state has witness some of the closest presidential elections in the nation among the states, following the 13 percentage-point victories for George W. Bush in 2000 and 2004:

Since 2008, the number of voters casting ballots has gone up about 200,000 each election (+194,583 in 2012 from 2008's total, and +236,192 in 2016 over 2012's total). Voter turnout has typically been 68 to 69 percent of the registered voter pool, equating to around 4.5 million ballots cast.

To put things into a comparative perspective, North Carolina tends to be about 3.5 percent more Republican since 2008 than the nation, using the Republican presidential candidates' performance nationally to the state's GOP performance.

Thus, it's a fair assessment to say that North Carolina is a "center/lean-right" kind of a state.

The following data is from the Cooperative Congressional Election Study, which, along with the American National Election Study, is a significant survey of Americans in election years, and provides some data, trends, and patterns to consider as we move into the 2020 election.

First, the CCES data on party identification among voters. One of the first set of analysis that I typically do is look at party identification, by strength of identification for partisan and "leanings" by self-described independents. Nationally, as a comparison, the partisan identification by strength shows this breakdown of the 2008 to 2016 presidential elections:

What these three graphs indicate are:

  1. partisan loyalty is predominate among those who "strongly" and "not very strongly" identify with one party or the other;
  2. independents who "lean" towards one party are just as, if not more, loyal to the party that they lean to (typically more loyal than those "not very strong" identifying partisans); and, 
  3. while those who identify as partisan (from strong to independent-leaning) may fluctuate from election to election, those who identify as "pure" independents are typically no more than 12 percent of the electorate and are willing to split their votes among the two parties, or, in the case of 2016, go third-party in their vote choices.

In North Carolina, for the same analysis, we see a comparable pattern, though some differences are evident:

Again, like at the national level, North Carolina's strong partisans tend to be strong party loyalists when it comes to their presidential vote choice: 94 to 98 percent of them will vote for their party's presidential candidate.

However, among "not very strong" partisans, we see some greater chances of party defections, most notably among Democrats: in 2008, 31 percent of not-very-strong Democrats crossed over and voted for Republican John McCain, but that party defection rate has decreased since. Perhaps this is due to those identifiers becoming more loyal in their attachment (hitting 81 percent for Clinton), or perhaps the classic southern moderate-to-conservative Democrat (who tends to be an older voter) may be seeing a generational replacement.

Another point of departure from national trends are North Carolina's independent voters, who tend to lean more Republican than nationally. This may be the foundation of the state being a "center/lean-right" state.

Taking these party identifications and collapsing them into Democratic (strong to independent-leaning), Independent (no leaners), and Republican (again, strong to independent-leaning) categories, we see another set of general patterns about the Old North State.

First, this graph demonstrates the electorate's party composition:

Notice how in 2008's election, when Obama won the state by 0.3 percentage points, Democratic identification was 52 percent, while Republicans were 40 percent. However, in 2012 and 2016, Democratic identification slipped, dropping eight points by 2016 compared to 2008, while Republican identification increased a total of three points, leading to Trump winning the state by less than 4 percentage points (helped by a significant advantage among NC independent voters).

In looking at this collapsed party identification and presidential vote choice for each of the past three presidential elections in North Carolina:

Notice that in 2008, Obama was able to keep pure independents close with McCain, 48-43 percent, while he dropped 15 percentage points in 2012, down to 33 percent among pure independents against Romney. Interestingly, Clinton only lost one percentage point among the pure independents in 2016 compared to Obama's performance four years earlier, but Trump fell eight percentage points compared to Romney in 2012, while independents defected to "someone else" in the presidential race by an increase of ten percentage points over 2012's seven percent. This "defection" by independent voters from Trump to "someone else" may be the "never Trump" voter, but that is a hypothetical guess.

Also key to point out: Democratic identifiers (including leaners) were tied with Republican identifiers in 2016, a stark comparison to 2008's electorate. But in all three past presidential elections, partisan identifiers, including leaners, went 90 percent or greater for their party's presidential candidate, a signal of clear partisan loyalty among identifiers (for more evidence of NC's partisan loyalty in voting, see the section "Ticket Splitters" in this post).

Of course, trying to guess what North Carolina's electorate will look like in November of 2020 is a Herculean predictive effort, something I'm not willing to do at this point in time. But these patterns--of voters who identify as partisan being the significant majority of voters, that partisan voters are party loyalists when it comes to presidential voting, and that North Carolina "independents" (who are fewer than 15 percent of the electorate) tend to lean Republican--are probably going to hold in next year's election.

Unless something dramatic and fundamental happens in the state--which, with generational replacement occurring, could indeed happen in 2020--we'll all just have to wait and see what happens in the Old North State come next November.