Understanding Student Voting Rates

​In January 2013, eAIR featured a story about the National Study of Learning, Voting, and Engagement (NSLVE). eAIR spoke with Nancy Thomas, NSLVE Project Director, to get an update on this study on college student voting rates.

circle.JPG            Tufts.jpg         
eAIR:  We’d like to know how things are going with NSLVE, but before that, can you remind readers of what the study is and what you are trying to achieve?
In a nutshell, we want to learn about patterns in college student political engagement, and our flagship initiative is a study of college student voting rates. (Voting isn’t the best indicator of political engagement, but it’s an important and objective one.) To measure student registration and voting, we “married” enrollment records with publicly available voting records. We worked with de-identified student-level data so that we could also correlate voting with specific data such as age, class level, graduate and undergraduate level, field of study, and, when possible, by gender, race, and ethnicity. We had two NSLVE goals: first, to provide a service, to offer campuses a free, easy, and nonthreatening way to obtain their students’ aggregate voting rates, and second, to build a national database for future research.

eAIR: You launched NSLVE in January 2013. How has it gone? How many campuses signed up to participate?

I’ll start with the easy question. We had hoped to engage 125 campuses in the study, and we are delighted that over 260 signed up by August 2013. We paused our recruiting to work on the data that we had already collected, which proved to be easier said than done. The good news: our database now consists of nearly 2.3 million student-level records, representing about 11% of all college and university students nationally. It’s not big enough, but it’s a terrific start. The 260 institutions received individualized campus reports from us, the first of this kind ever.

eAIR:  Is there any “bad” news? 

There’s no “bad news” per se, but it took us a lot longer than expected to understand variations in the ways campuses report their enrollment information, explore the source of problems with some voting records, and troubleshoot things that didn’t quite look right for individual campuses. We discovered inconsistencies in, for example, how campuses report their full- and part-time students, their degree-seeking and non-degree-seeking students, and U.S. citizens. Some campuses serve high numbers of nonresident aliens (international students, mostly) and resident aliens, both documented and undocumented. We have no direct way of controlling for resident aliens, so it’s possible that students deemed “eligible” to vote are actually non-citizens who are ineligible to vote, resulting in a lower voting rate than, for example, peer institutions that do not serve many non-citizens. On the voting data side, our biggest challenge was in voting method – many election officials simply don’t collect how people vote. We needed to sort out these things.

eAIR: We realize you haven’t completed your analysis of the participating campuses, but is there anything interesting you’d like to share?

Absolutely. For the campuses in our study, voting rates ranged from 71% to 25%, although some institutions with low voting rates may have disproportionately high numbers of students who are ineligible to vote. Students at urban institutions seem to vote at lower rates than students in suburban or rural institutions. Campuses with student bodies comprised of 90% in-state students have an average voting rate 6.5 percentage points higher than campuses with 25% in-state students. The most significant predictor of an institution’s voting rate appears to be the average age of students. For example, campuses with an average age of 27 have voting rates about 8% higher than campuses with students whose average age is 21.

We’re also confirming insights suggested by other research, such as the fact that voting rates are higher in highly contested states like Ohio, and in places with interesting ballot initiatives like Colorado and Washington; however, despite these consistencies, there are a couple of campuses with surprisingly low rates at which we might want to take a closer look. We’ve put together a complicated model for predicting voting rates, and we plan to hone in on outlier campuses – those with unpredicted high or low rates. We’ll be examining all of these preliminary observations moving forward. 

eAIR: What does “taking a closer look” mean in this context?

We have funding for follow-up case studies, which will include site visits, interviews, and qualitative inquiry on campuses with potentially interesting stories. Our preliminary data are useful, but don’t tell the entire story by any stretch of the imagination.

We don’t want a campus to view its voting rate as an indicator of institutional quality until we have a deeper understanding of the campus environment. For example, we know of a campus where the local electoral officials moved the voting booths 10 miles from campus. That’s a tough barrier to overcome, so perhaps a lower voting rate there has nothing to do with the quality of the student learning experience. Students at another campus complained that election officials were hostile, asked for additional identification, and questioned campus addresses. These kinds of non-statutory barriers to voting might prompt action on the part of students, making them even more determined to vote. Or these barriers could be serious deterrents. We want to explore these questions further.

eAIR: What else will you be considering?

We know that education attainment matters – that people with college experience vote at higher rates than their non-college peers – but we don’t understand why that is. We want to take a closer look at campus climates and potential trigger points for political engagement. Why is one campus on fire politically, while another is disengaged? Where do discussions of controversial issues happen? What is the context for voting on campus, in the community, and in the state? 

eAIR:  Do you want more institutions to join the study?

We do. We need to add geographic and institutional diversity to our pool of campuses and fill out our comparison groups. Right now, if campuses sign up, they can still receive registration and voting rates for Fall 2012. And the 2014 elections are around the corner.

eAIR:  Are you aware of any particular reasons campuses might not participate?

We heard some pretty silly reasons for not participating. One campus refused because it was seeking Carnegie Classification for the “engaged campus” and the administration was afraid that a low voting rate would hurt the school’s chances of being awarded that designation. All results are confidential; we send them only to one designated individual per campus. No institution is required to share its rates for any reason, and certainly not for Carnegie classification or accreditation purposes.

We also heard from a few campuses that this was something they simply didn’t want to know because they assumed it would be bad. Ridiculous.

Some campuses erroneously assumed they would have to do a lot of work, particularly a survey. No. No survey. Promise.

Many campuses were worried about FERPA and student privacy rights, especially places without in-house attorneys. In those cases, an administrator would express an interest and send an inquiry to an off-campus legal office, never to be seen again. This is unfortunate. We dotted our “I’s” and crossed our “T’s” on the privacy concerns. We posted an extensive legal analysis on our website in the form of an FAQ.

The only other thing I will say about FERPA is that 260 campuses signed up, and many not only have legal counsel, they have access to large legal staffs. FERPA should not be a barrier to participating in NSLVE.

eAIR: How can institutions sign up, and is there a deadline?

Institutions can still sign up to receive Fall 2012 registration and voting rates. Signing up is easy. Simply download an authorization form, have a senior administrator (president, provost, vice president, IR director, or registrar) sign it, and send it to us. The instructions are on the form. That’s it.

We match enrollment and voting records when we have a large enough number of campuses to justify the expense, and our plan right now is to do another run in April. That means that campuses should sign up now.

eAIR:  Are there ways institutions can avoid some of the problems you experienced with the quality of the student data? 

Yes. Campuses can be more consistent about what they send to the National Student Clearinghouse. Specifically, please send records for all students and fill out the multiple data elements for each student. To avoid misleadingly low voting rates, be sure to include information on nonresident alien status and whether or not a person has a Social Security number. The more your campus sends in to the Clearinghouse, the more accurate your report.