Sunday, March 1, 2020

NC's Democratic Primary Early Voters: Diversification in Many Ways

With Saturday's final day of early voting, it's time to take a deep dive into the data to see who showed up to cast absentee ballots before the "Super Tuesday" March 3rd primary election in the Old North State.

North Carolinians have three main options in casting ballots: absentee by mail, absentee onestop (which is no-excuse early voting in-person), and Election Day voting (including transfers and provisional ballots).

In this year's preliminary total, close to 800,000 ballots were cast in all party primaries through February 29 (over 500,000 of those ballots are in the Democratic primary), and while there will be some "settling" by counties and remaining processing, the following descriptive analysis should give us a sense of who showed up. We'll have to wait until Tuesday (March 3) night, shortly after the polls close at 7:30 PM, to get a sense as to how these early voters decided.

The major contest of this primary is on the Democratic side, especially with a presidential nomination battle that has been entertaining and confusing at the same time. As I'm writing this blog piece Sunday evening, Pete Buttigieg is the latest candidate to suspend his campaign, following Tom Steyer's announcement last night after a disappointing, multi-million dollar effort in the Palmetto State. 

This analysis will compare between this year's early vote in the Democratic primary to the early vote in the 2016 Democratic primary, focusing on absentee voting and various demographic factors. 

First, the Democratic Party's primary electorate is diversifying based on voter party registration. With North Carolina's two major parties having 'semi-closed' primaries, only registered Democrats and registered unaffiliated voters can participate in the Democratic primary. In 2016's Democratic early voting electorate, 81 percent of the absentee ballots came from registered Democrats, with 19 percent from unaffiliated voters. But that ratio changed considerably this year. 


Registered Democrats dropped to 73 percent of the early ballots cast, with 27 percent coming from registered unaffiliated voters. This is most likely due to the growing number of younger voters, under the age of 40, who are registering unaffiliated in significant numbers. In the February 29 registered voter data file, Millennials are 41 percent registered unaffiliated, 34 percent registered Democrat, and 24 percent registered Republican, while the youngest voters--those of Generation Z--are 47 percent unaffiliated, 29 percent Democratic, and only 22 percent Republican. This continued flight from party registration doesn't indicate political independence, but rather 'choice' by younger voters, and will likely have a stronger influence moving into future primary elections. 

North Carolina's Democratic party is a majority-minority political party, with voters of color having a crucial impact in primary elections. As of February 29, the party's racial breakdown among Democratic registered voters was 47 percent black/African American, 43 percent white, and 11 percent voters of all other races. In this year's early voting on the Democratic side, voters of color constituted nearly the same percentage as in 2016, but there was an important shift going on within that group.

White voters were 55 percent of the early vote electorate in the 2016 contest between Clinton and Sanders, and they were 58 percent of 2020's early vote in the Democratic primary. Voters of color dropped the three points, but the difference among those voters is critical to look at. 


Black/African American voters were only a third of the early ballots this year, while voters of all other races were nearly 10 percent of the early voting electorate. Black voters could have been waiting for neighboring South Carolina to show it some direction, and it will be important to watch who shows up Tuesday and whether black voters make up the difference. In 2016, the total Democratic primary electorate (both early and Election Day ballots) was 57 percent white, 36 percent black/African American, and 7 percent voters of all other races. 

Another key divide that may come into stark relief this year is by generational cohorts. Among Democratic absentee voters, young voters actually increased their percentage of the early vote electorate from 2016's percentages:


Millennials and Generation Z (basically any voter under the age of 40) were 21 percent of this year's Democratic primary early vote, compared to 19 percent four years ago (and with just Millennials eligible). Generation X, the bridge generation between Millennials and Boomers, increased two points, while Boomers increased one over their percentage of the Democratic primary electorate in 2016. The increases came at the expense of the oldest generations, the Greatest and Silent, dropping from 17 percent down to 13 percent. 

In further comparing the generation cohorts, the selection of which party primary by each cohort gives an interesting picture as well. The 2020 primary election, lacking a competitive presidential contest, may be a significant reason for the great gulf among young voters, but the 2016 comparison gives a hint that young voters are preferring one party over the other. 


In 2016, Generation X, Boomers, and the Greatest/Silent generations all selected the Republican primary, albeit by rather close margins. The Millennial generation, however, went decidedly in 2016 for the Democratic primary ballot; this year, the lack of a Republican presidential contest most likely had a significant impact on this year's 80 percent of Generation Z and Millennials picking the Democratic ballot, but to see that level of disparity among the generations is something to watch for in future election years. 

Another important factor when it comes to Democratic party politics is the urban/surrounding suburban/rural county division. Among Democratic registered voters (as of February 29), sixty percent of Democrats were in urban counties, while 18 percent were in surrounding suburban counties and 22 percent in rural counties. However, the early voting primary electorates tended to be more urban in nature. 


2020's early votes in the Democratic primary were more urban than in 2016, with rural county votes dropping two percentage points from four years ago. This 'urbanization' of the Democratic Party has been trending for some time, and also is reflected in the increase in voter turnout, to be discussed later. 

In regards to the gender divide, the Democratic party typically sees an advantage among female voters: 41 percent of North Carolina registered female voters are Democratic, 30 percent unaffiliated, and 28 percent Republican, according to the latest figures from the North Carolina State Board of Elections. Within the Democratic party registration, women are 58 percent to 39 percent men. 

In 2016's Democratic primary early votes, women made up 60 percent of the vote; in 2020, it slipped to 57 percent, but there was a higher than normal percentage of voters who did not indicate a gender on their voter registration record. 


Finally, the fact that this year's early vote for the Democratic primary, and for the primary election overall, was ahead of 2016's numbers needs some mention and discussion. The state saw the early vote turnout at 11 percent of all registered voters (as of February 29), and had a 33 percent increase in Democratic ballots being cast early over 2016's Democratic primary election. 

While a third of the 100 counties in the state saw a decline in Democratic primary early vote turnout, 27 counties saw increases above the state's increase of 33 percent in the Democratic primary over 2016:

CountiesIncrease over 2016's Democratic Early Vote Numbers
Wake79%
Union78%
Iredell74%
Yadkin73%
Brunswick70%
New Hanover63%
Northampton63%
Mecklenburg59%
Forsyth56%
Moore56%
Alamance55%
Henderson51%
Guilford47%
Johnston45%
Bertie44%
Stokes41%
Currituck40%
Lincoln39%
Anson38%
Durham36%
Chatham35%
Harnett34%
Gates34%
Transylvania34%
Pamilco34%
Cabarrus34%
Buncombe33%

Many of these counties that saw increases above the state average are urban counties, such as Wake (home to Raleigh), Mecklenburg (home to Charlotte), New Hanover (Wilimington), along with surrounding suburban counties. 

Many have asked me, what will Tuesday bring? I don't believe in 'predictions' as a political scientist, other than what the data and trends may show up. But in 2016's Democratic primary, over one-third of the ballots came before Election Day, with two-thirds being cast on Election Day. Here are the results for each vote method broken down:


If Democrats copy the 2016 primary electorate, then with a little over half-a-million ballots coming in early, it could be expected that another million votes come in on Election Day, if the one-third/two-third division holds from four years ago. But we'll just have to wait and see how Tuesday evening play outs. 

Two other notes from the 2016 Democratic primary election, for both voter race and generational cohorts on Election Day. 

In 2016, we saw more black voters use early voting than on Election Day, with 43 percent of all that year's ballots (1.1 million cast in the Democratic primary) ultimately from voters of color. 


Next, while Boomers were a plurality of early votes (46 percent) and Millennials being 18 percent, on Election Day, Boomers slipped to 41 percent and Millennials were one in five Election Day voters, making up that same percentage (20) when all the ballots were cast. 


We will see what happens come Tuesday, and then later this month, get the data on who voted on Election Day to see what trends were similar or different from 2016. But this we know: 2020's Democratic presidential primary contest has established itself as a very different type of campaign from 2016's contest, in many diverse ways.