What Might SB 326 Mean in Practice?

By Christopher Cooper

The "Election Integrity Act," SB 326 was filed in the NC Senate on March 18, 2021 by Senators Daniel, Newton, and Hise. The bill has a number of provisions, including (1) "prohibit[ing] the state board of elections and county boards of elections from accepting private monetary donations for certain purposes," (2) "appropriat[ing] funds to establish a program to identity and assist voters needing photo identification",  (3) "amend[ing] the date by which a voter must request an absentee ballot, and (4) amending the date by which a "mail-in absentee ballot must be received."

While the first two provisions are important and worthy of study, the third and fourth policy changes are ready-made for the type of empirical analysis that we try to provide on this blog. In this brief entry, I report results from a simple simulation of whose votes would not have been counted and who's would have been rejected in the past two General Elections if SB 326 had been the law.

To perform this analysis, I downloaded the ENRS file from the State Board for the 2016 and 2020 General Elections (note: the bill also applies to Primary elections, I just don't analyze them here), selected the accepted mail-in ballots, then compared the ballots requested at least 14 days before the election and returned by 5 PM on election day, to those that were accepted under that year's election law, but would have been rejected if the provisions of SB 326 had been in effect.

2016 General Election

If the provisions from HB 326 had been in place in the 2016 General Election in North Carolina, about 24,000 fewer mail-in ballots would have been accepted. As the table below suggests, compared to the ballots that would have still been accepted, the rejected ballots would have been more likely to be from African American voters, younger voters, and registered Unaffiliated voters.

2016 Mail Ballots Accepted & Rejected Had SB 326 Been in Place
	Accepted Under Proposed Guidelines	Rejected Under Proposed Guidelines
Total	167,687	23,914
Race
Black	9%	13%
White	84%	77%
Undesignated	3%	4%
Age
Avg age	56	46
Gender
Female	57%	57%
Male	41%	40%
Undesignated	2%	3%
Party Registration
Democrat	31%	32%
Republican	40%	35%
Unaffiliated	28%	32%
Ethnicity
Hispanic	1%	2%
Not Hispanic	79%	76%
Undesignated	20%	22%

Note: To Perform this analysis, I downloaded the 2016 General Election ENRS data from the NCSBE web site, selected only accepted mail ballots, and then compared the ballots requested at least 14 days before & returned by 5 pm on election day (left-hand column) to those that were accepted under the 2016 law, but would have been rejected if SB 326 had been law (right-hand column). 

 

2020 General Election 

If the provisions of HB 326 had been in place in the 2020 election in North Carolina, about 31,680 fewer ballots would have been accepted. The ballots that would have been rejected match the 2016 patterns in terms of age (rejected ballots would have come from younger voters). 

The partisanship trends are interesting and little different from 2016, however. In 2020, Republicans would have been overrepresented in the rejected category if SB 326 had been in place in 2020 (a difference likely due to the broader shifts in mail-in voting patterns in 2020). Similar to 2016, however, it is the Unaffiliated voters who would have been most disadvantaged by the proposed guidelines. 

The potential effects in terms of race, ethnicity, and gender in 2020 are a little harder to divine because of the rise of non-reporting of race, ethnicity, and gender (a disturbing trend that Michael Bitzer and I documented here). 

2020 Mail Ballots Accepted & Rejected Had SB 326 Been in Place
	Accepted Under Proposed Guidelines	Rejected Under Proposed Guidelines
Total	969,915	31,680
Race
Black	15%	15%
White	69%	60%
Undesignated	10%	19%
Age
Avg age	55	46
Gender
Female	53%	49%
Male	39%	37%
Undesignated	8%	14%
Party Registration
Democrat	45%	31%
Republican	21%	30%
Unaffiliated	34%	38%
Ethnicity
Hispanic	2%	3%
Not Hispanic	72%	63%
Undesignated	26%	34%

Note: To Perform this analysis, I downloaded the 2020 General Election ENRS data from the NCSBE web site, selected only accepted mail ballots, and then compared the ballots requested at least 14 days before & returned by 5 pm on election day (left-hand column) to those that were accepted under the 2020 law, but would have been rejected if SB 326 had been law (right-hand column).

Take-aways

The takeaways and lessons from this brief analysis will likely be in the eye of the beholder. Those in support of SB 326 will likely point to the relatively small number of ballots requested and cast outside of the proposed time period and the lack of overwhelming data that one race or party is consistently being disadvantaged by the proposal. 

Those in opposition to the bill will likely suggest that it is a solution in search of a problem. There is nothing that suggests that the past guidelines were producing an overwhelming burden on election boards. Further, opponents of SB 326 will likely  point out that the proposed guidelines would hurt younger voters who we are trying to get involved with the political process.

If this analysis suggests anything that both sides can get behind, it should be the problems of the lack of reporting of race and gender on voter registration. It should be frustrating to both Republicans and Democrats that politically important questions, such "is one race more disadvantaged by a policy proposal than another race?" are increasingly difficult to answer because of the lack of consistent data collection on race, gender, and ethnicity in North Carolina. This is a problem the NC State Board of Elections needs to work to resolve.

Two Caveats

Before I let this one go, I want to note three brief limitations of this analysis. First, I only analyze two elections--both are general elections and both are recent. Given the radical shifts in mail-in balloting in 2020, a more reliable approach would be to extend the analysis further back and include primary elections as well. I'll do that soon if I find the time (and if I do, I'll post the update here). 

 

The second limitation is that some of these ballots that are currently in the right-hand column are UOCOVA ballots, which are exempt from the new rules proposed under SB 326.

The third limitation can't be fixed with additional data or analysis. It is possible that even fewer people would be affected by SB 326 than implied here, as voters might know the deadlines and act accordingly.  There's simply no way to know for sure.

 

Errata

 

The original version of this post had incorrect totals in the right-hand column of the second table due to a coding error on my part. This version has the corrected data. Although the number of ballots that would have been rejected were listed incorrectly in the previous version of the post, the patterns and conclusions highlighted previously did not change. 

I also fixed three type-os in the previous draft (I had accidentally written HB 336 instead of 326) and added the caveat about UOCOVA ballots.

------

Chris Cooper is the Madison Distinguished Professor of Political Science and Public Affairs at Western Carolina University. He tweets at @chriscooperwcu