The last Pack Poll in 2013 found that N.C. State students support requiring voters to show government issued identification in order to cast ballots in person. Students’ support for voter ID follows recent, controversial changes to N.C. election laws—one of which requires everyone in 2016 and afterwards to show government issued proof of identity before voting.
One possible reason for supporting the new law is the perception that ID laws prevent fraud. A poll of adult North Carolinians conducted last December showed that 54% believe voter identification laws will help protect the state against fraudulent impersonation in the polls (“Civitas Poll: NC Opposes Voter ID Suit, ACA”).
To investigate this possibility, the Pack Poll measured students’ perceptions of voter impersonation fraud. On average, students grossly overestimate the percentage of voters caught impersonating other voters. Students guessed that 6.7% of all voters impersonated someone else to vote, although the median estimate was a more sanguine 3%. In reality, since 2000 less than 1% of any kind of fraud cases involved voter impersonation (News21), and individual cases of election fraud are exceedingly rare.
The Pack Poll finds that support for voter ID laws increases as perceptions of fraud increase. Only 51% support ID laws if they believed voter impersonation occurred less than 1% of the time, but jumped to 75% and 81%, respectively, among those who over-estimated this kind of fraud (between 1% and 5%) or significantly over-estimated it (6% or more).
We also conducted an experiment to study how providing students with background information about the ID law affects support for it. At random, students read one of three versions of the question. The first version asked, as many other polls have done, “Do you favor or oppose requiring people to present a government issued photo ID every time they attempt to vote in person?” When asked this way, 80% of students said they supported voter ID requirements.
However, a second version added the following information before asking the same question: “Current law in North Carolina requires first time voters to verify their identity, and then sign an oath affirming their identity in future elections. Utility bills and student ID were acceptable forms of ID. A recently passed law will require all voters in 2016 to verify their identity by, with a few exceptions, showing government issued photo identification. Student ID and utility bills will no longer be valid forms of ID.” Given this information, support for voter ID dropped to 65%.
A third version also provided this same background information and also included commonly heard arguments for and against voter ID laws. Support for voter ID in this version was unchanged compared to the second version.
Almost all Republicans on campus support voter ID, regardless of the question wording. Overall, 91% of Republicans, 72% of Independents, and 45% of Democrats said they supported voter ID. Perhaps one reason for this partisan gap is that, on average, Republicans thought more fraud took place. Republicans believed that over 8% of people were caught in a single election impersonating someone else to vote, while Democrats’ said this percentage was smaller (5%).
Among Democrats who were correct in their perception of voter impersonation (less than 1%), just 30% supported voter ID. However, a slim majority of Democrats who overestimated or grossly overestimated voter impersonation supported voter ID. Less fluctuation occurs among Republicans’ opinions—they were unaffected by perceptions of fraud. Among Republicans that were accurate about voter fraud, 88% still supported voter ID, while about 91% supported voter ID who thought fraud as more common.
Democrats were also more affected by the question wording experiment than Republicans. Just 37% of Democrats opposed voter ID without providing context to the question, but over 60% opposed voter ID in both version of the question that added information. Meanwhile, 97% of Republicans supported voter ID in the version without context, and their support never dropped lower than 84% in the other two versions.
It’s difficult to analyze the reasons why people truly support for new voting ID laws. For example, a person who supports ID laws without thinking too much about it might then say fraud is more prevalent, because that would make sense and justify the preference. Or, a person might get exposed to an argument about fraud being prevalent, believe it, and then support ID laws said to prevent it. In a survey like ours, it is nearly impossible to know which came first, the opinion about voter ID or information about fraud/voter ID that leads to opinions about it. For Republicans, though, information doesn’t significantly affect their levels of support for voter ID laws; Democrats’ levels of support, conversely, are highly dependent on the information they are exposed to.
Political scientists have examined the trends in states that have recently proposed voting legislation. In data collected from 2006–2011, proposals for tighter restrictions on voting were considerably more likely to occur in states where non-white voter turnout had been increasing. Additionally, every proposal that required photo ID was passed in a Republican majority legislature (Bentele). These findings reinforce the opinion data we collected showing a deep partisan divide.
- Bentele, Keith G., and Erin O’Brien. “Scholars Strategy Network.” Scholars Strategy Network RSS. The Society Pages, 2 Dec. 2013. Web.
- “Civitas Poll: NC Opposes Voter ID Suit, ACA.” Civitas Institute. N.p., 16 Dec. 2013. Web. 20 Jan. 2014.
- News21. http://votingrights.news21.com/interactive/election-fraud-database/
The Pack Poll is a representative survey of NCSU 1,032 undergraduates. The survey was conducted over the internet using email to sample students between Nov 7-12, 2013. The response rate is 23%, and the margin of sampling error is +/-2.97%.