In an August 30 report, NBC’s Meet the Press claimed that Democrats had an advantage in voter registrations in North Carolina compared to 2016:
And North Carolina has added 1.3 million more voters to its rolls since 2016. Democrats had an edge of more than 56,000 in new registrations, while 583,000 others registered as unaffiliated.
(Note: there were 6,918,150 voter registrations on election day, 2016. If you followed NBC’s report unquestionably, you could be forgiven for believing that there are around 8.2 voter registrations in North Carolina today due to those “1.3 million more voters.” In fact, North Carolina has 7,117,610 voter registrations as of September 12. That 1.1 million registration difference alone should be a warning that NBC is taking you “lies, damned lies, and statistics” territory.)
NBC got this bit of information from TargetSmart, a “Democratic political data and data services firm.” Here is what they say about how voter registration has changed since
Since 2016, more than 1.3 million people have registered to vote in North Carolina. Of those, 393,409 voters have registered as Democrats while 336,966 have registered as Republicans, giving Democrats a 56,443 vote advantage.
Here is the problem: TargetSmart only reports half the equation for finding current voter registration numbers (Current voters = 2016 voter registration list + new registrations added to the list). As I noted previously, there are not 393,409 more Democrats now than in 2016, there are not 336,966 more Republicans than in 2016, and there is not a 56,443 Democratic vote advantage compared to 2016.
So, what did NBC’s Meet the Press and TargetSmart fail to share? They did not point out that, while 1.3 voter registrations were added to voter rolls, another 1.1 registrations were removed (for example, when a voter dies or leaves the county where they were registered). They should have included registration removals in their measurements to find the net change in voter registrations (Current voters = 2016 voter registration list + new registrations added to the list – old registrations removed from the list).
To use a painful analogy for many Tar Heel fans, what NBC said is like saying how great Carolina did against Clemson last year because they scored 20 points while ignoring that Clemson scored 21.
The Civitas Voter Registration Changes page, using data from the NC State Board of Elections, provides the net change in voter registrations for any time frame over the past twelve years. For example, by setting the “Changes from” box to November 8, 2016 (election day), and the “to” box to September 12 (the latest available date), we find that there has been a net gain of 54,364 Republicans and a net loss of 180,552. In other words, the 336,966 new Democratic registrations since the 2016 election were more than offset by the 517,518 Democratic registrations that left the voter rolls over the same period.
The net change also includes about 577,000 registrations that were removed from voting rolls in January of 2019 and another 229,000 removed over the course of several months after the 2016 elections. Those removals are people who move away or death was not reported to the county board of elections or who otherwise did not vote in four federal election cycles (about eight years) and could not be reached by the county board of elections. By definition, the people attached to those registrations did not vote in 2016. So, while measuring net voter registration change is a more accurate measure of changing voter registration than just measuring new voter registrations, it also includes registrations of people who would not have voted even if they were still on the voter rolls.
So, a measurement of change in the relative voting strength, rather than registration, of the two major parties would measure the net change in voter registration other than during periods of list maintenance. For the 2020 election, that would mean from February 4, 2017 (after the 2017 list maintenance) to December 22, 2018, and from Feb 2, 2019 (after the 2019 list maintenance) to today. Those measures indicate a gain of roughly 145,388 (53,628 + 91,760) active Democratic registrations and 238,644 (80,256 + 158,388) active Republican registrations, giving Republicans a net advantage of about 93,256 more active registered voters compared to 2016. Unless January of every odd-numbered year is somehow a period of unexplainable Democratic gains, this number that number slightly undermeasures the Republican advantage in the net change of active voters over the past four years.
Either of the measures presented here provides a more accurate picture of what is happening with voter registration in North Carolina than what NBC presented.