Monday, April 6, 2020

Who Showed Up for North Carolina's Super Tuesday Primary Election?

It's been a month since North Carolina held its primary election on Super Tuesday, and the data is finally available for who showed up to cast ballots in the respective party primaries.

In total, the state saw a turnout rate of 31 percent, with over 2.1 million ballots being cast. This is slightly down from the overall 2016 March primary election, where 35 percent of registered voters showed up, casting 2.3 million ballots.

This year's Democratic primary ballots did exceed 2016's numbers: 1.3 million and 1.1 million four years ago. The drop was in Republican primary ballots, not surprising due to a lack of competitive primary at the presidential level.

Out of this year's primary elections in the state's 100 counties, 98 of them appear to have submitted their data for which voters showed up, with which party primary and vote method the voter used in casting a ballot.

The only counties missing so far are Duplin (with 8,894 votes cast) and Gates (with 2,024), or a potential 0.005 percent of the total 2.1 million votes cast. So, a very small amount that likely won't impact the overall figures that I analyze here.

In using the 'voter history' data file, I pulled out the March 2020 primary election data out for each registered voter, and then merged it with the April 4 voter registration data file (utilizing the county description and the voter's registration number to serve as the matching keys between the two data files). I isolated those voters (roughly 40 thousand) who had registered after the March primary, recognizing that I'll likely miss a few registrations due to processing of those who registered in the early voting period. Once all 100 counties have reported, I'll look more deeply into the numbers, but I'm not expecting significant shifts to occur.

First, to set some general data points about North Carolina's March primary: out of the 2.1 million ballots cast, 62 percent were in the Democratic primary, with 37 percent in the Republican primary, and the remaining one percent in all other party primaries (Constitution, Green, and Libertarian, along with 'unaffiliated' ballots).

In looking at the turnout rates for a variety of demographic characteristics of voters, starting with gender, there were some distinct patterns within the data.


Not surprisingly, female voters had a higher turnout rate than male voters; this turnout advantage has been well documented in the research on gender in voting.

In looking at turnout by racial categories of voters:



One-third of the white registered voters cast ballots in the primary election, with thirty percent of black/African American and 20 percent of all other races showing up.

Among party registrations, no real surprise with the turnouts among party affiliations.



As a reminder, registered unaffiliated voters can vote in either the Democratic or Republican primary, but not both. Data regarding which party primary unaffiliated voters selected shows that unaffiliated voters went Democratic (65 percent) than Republican (34 percent), and this isn't surprising: the Democratic Party's presidential contest garnered a lot of attention, while the most competitive 'top-of-the-ballot' race on the Republican side was the gubernatorial contest.

Among generational cohorts, it's not surprising that one group was a significant plurality of the electorate.


Not surprisingly, older voters (those over the age of 55 years) saw the highest turnout rate, while Millennials and Generation Z saw the lowest. Generation X saw a comparable turnout rate to the state's rate.

Finally, among the urban-suburban-rural regions of North Carolina:



There was only slight differences between the various regions, with voters within suburban areas inside urban counties having the highest rate, followed by central city voters, and then both surrounding suburban and rural county voters.

The next set of analyses looks at each of the major party primaries (Democratic and Republican) to see the differences among and within the various demographic characteristics, starting first with voter gender:



Women went two-to-one in selecting the Democratic Primary ballot, while men went more Republican in their vote casting.

In looking within both party's primaries at the electorates by gender:


The Democratic party primary was overwhelmingly more female, while the Republican party primary was evenly split.

By voter race, white voters went ever-so-slightly more Democratic (49.8 percent) than Republican (49.6 percent), but as is the usual case, black/African American voters went heavily to the Democratic primary (98 percent) and voters of color also went heavily Democratic (77 percent) over Republican (21 percent).

Among the Democratic primary electorate, whites were 57 percentage of the electorate, with black/African Americans being 33 percent and other voters of color being 10 percent. The Republican party primary was 95 percent white, one percent black/African American, and four percent other voters of color.

Among the generational cohorts, as defined by the Pew Research Center, the stark divide between younger and older voters is significant:


Three-quarters of voters under the age of 40 selected the Democratic Party's ballot over the Republican Party's ballot; this isn't surprising, because of the lack of a competitive 'top-of-the-ticket' primary contest on the GOP side. But if national dynamics regarding Millennial voters are at play in the Old North State, this Democratic advantage in party affiliation among young voters could be a stark signal going forward.

While Democratic primary voters were a majority among older voters, it is interesting to note that these voters had a higher than state percentage of Republican primary voters among those cohorts. Even without a truly contested presidential primary, older voters were more likely to select the GOP ballot.

Within the party primaries' electorates, this age difference is notable:


Within the Democratic Party's primary electorate, twenty-five percent of the voters were under the age of 40, with fifty percent over the age of 55 years. Among the Republican Party's electorate, only 14 percent were under 40 years old, with 61 percent over the age of 55. 

I'm going to dig a little deeper into this generational dynamic in a future post, most notably in looking at who showed up in 2016 and 2020 by generational cohorts. 

Finally, in looking at the 'regions' of North Carolina, based on urban, suburban, and rural dynamics:




The fact that over 80 percent of voters in central cities went to the Democratic ballot is pretty astonishing; more so that their urban-suburb voters also went Democratic as well. The only region to see a buck against the Democratic dominance was among surrounding suburban counties, which tend to be the most Republican areas of the state. 

In looking at each party's primary electorate and the regional composition:




A plurality of Democratic party ballots came from central cities, while a plurality of Republican party ballots came from surrounding suburban counties. 

Now, what does this all mean? I think there are some key take-aways from this data.

First, we can't use this information from the March primary to determine who will likely show up this fall. In 2016, 96 percent of those who cast ballots in the March primary voted in the fall general election, which hit 69 percent voter turnout overall. So it's likely that the March 2020 voters will be ones who show up to vote in the fall, but we can't estimate among the voters who didn't participate in the primary electorate what their turnout rate will likely be. 

Two important groups to watch this fall will be voters under 40 (Millennials and Generation Z) and those voters in the suburbs of urban counties. The numbers of voters under 40 going to the Democratic primary is impressive, but we'll need to see how those voters react to the fall campaign, especially with Joe Biden at the top of the ticket. 

In North Carolina, the 'suburban voters in urban counties' tend to the true battlegrounds of the state, if one looks at it from 2016's presidential election results:


If you want to watch the other regions, I would say: how much more Democratic do central city precincts go; conversely, do rural counties go even more Republican; and, is there any drop-off in surrounding suburban county precincts from the Republican dominance? 

Pending whether the COVID-19 virus re-emerges in the fall, expectations for voter turnout have been very high: perhaps into the 70 percent range in North Carolina. But as we have seen in 2020 so far, it's anybody's guess what may happen.