This August, the U.S. election system was cast into doubt. Donald Trump suggested that it might be rigged – presumably to help Hillary Clinton win. Russians allegedly hacked voter registration systems in Arizona and Illinois, although, according to the FBI, they didn’t succeed in tampering with voter rolls.
Voters and their experience at the polls
Part of the reason that U.S. democracy is so stable — and that citizens accept election results instead of, say, rioting and setting the Capitol on fire — is that most American voters are confident that their ballots are counted accurately in our elections. That’s what we’ve found in our research over the past decade. And that confidence can be improved or harmed by how state and local election officials manage elections — whether their favored candidates win or lose.
We’ve found that when voters have problems voting — if, for example, they find the ballot confusing, poll workers unhelpful, long waits in line or uncertainties about whether their absentee ballots were received or counted by the election office — they probably will be less confident that their vote will be counted correctly. Our research suggests that a bad experience at the polls can reduce voter confidence by nearly 10 percent. And absentee voters are less confident than in-person voters.
Improving voter confidence in a N.M. county
Over the past decade, we have systematically observed elections in Bernalillo County, N.M., using a methodology described in our book “Evaluating Elections.” At the same time, we have surveyed those voters on how confident they felt about the election’s integrity. From our observations, we’ve recommended changes in how voting was conducted. Because County Clerk Maggie Toulouse Oliver put into place many of our recommendations, the surveys have shown us whether — and how — improving election administration improves voter confidence.
What have we recommended? Over the past eight years, Bernalillo County has professionalized poll worker training, with specific training for each job, and with each poll worker assigned to a single job during voting. The county moved voting to larger and more accessible locations, with enough parking, and more effectively moved voters through voting. Voters can use a “voter line” app that points them to the closest polling location with the shortest line; a faster experience is typically a better one for the voter.
Let’s look more closely at a couple of other important examples — such as the hot-button issue of voter ID. Voters are particularly sensitive to how they are asked to prove their identity at the polls. Whatever the state’s policy might be, officials need to implement it consistently in every voting location and with every voter.
In New Mexico, a voter can either show a physical ID to the poll worker or simply state her name, address and birth year. In 2008, our election observations concluded that workers in about one-fourth of Bernalillo County’s polling places were doing it wrong. By 2010, that was true for one-third of all polling places. We reported our findings to the county, which revamped poll worker training to emphasize how to authenticate voters correctly. In 2012 and in 2014, nearly all polling locations were identifying voters correctly.
Then there’s the sanctity of the secret ballot, so critical to the U.S. idea of democracy. Research by Ryan Claassen and fellow researchers (which you can find here as a pdf) shows that being confident that one’s ballot is private raises voters’ confidence in the election. In 2008, our election observers also noticed that, as they walked from the voting station to the machine that tabulated votes, citizens weren’t able to keep their paper ballots from being visible. We suggested adding a privacy sleeve in which voters could shield the ballot from view while carrying it to increase voter privacy and feelings of efficacy; Bernalillo County introduced privacy sleeves countywide in 2014, and sure enough, voter surveys revealed that voter confidence increased.
Here’s how we measured voter confidence
Since 2006, we have asked voters in Bernalillo County, “How confident are you that your vote was counted as intended?” Respondents can answer: very confident, somewhat confident, not too confident or not at all confident.
That’s how we found that better administrative practices increased voter confidence. Figure 1 shows the percentage of voters who were very confident their vote was counted correctly. Voter confidence has increased from 39 percent in 2006 to 57 percent in 2014, despite a mix of Democrat (2006, 2008, 2012) and Republican (2010, 2014) winners across elections.
So what should election officials do this November?
Take these simple, evidence-based steps:
- Train poll workers to do the job consistently and effectively. How voters and poll workers interact has a profound influence on voter confidence. And poll workers who understand their important role perform their jobs better.
- Improve transparency. Election offices can do a number of things. For instance, they should make sure their website is user-friendly and easy to navigate. Or they could install webcams so that citizens can observe vote counting and other processes in the central election office.
- Report timely, accurate data at the precinct level, disaggregated by voting mode (i.e., in-person or absentee). Election “geeks” need accurate and disaggregated data for their analyses. That’s standard for identifying possible voting problems.
- Audit election practices throughout the election. Post-election ballot audits can offer evidence that election results have integrity. Respond to problems promptly and actively. Track complaints and problems on social media and other tools, and respond quickly.
Michael Alvarez is a professor of political science at Caltech and co-director of the Caltech/MIT Voting Technology Project. Lonna Rae Atkeson is a professor of political science at the University of New Mexico, where she directs the Center for the Study of Voting, Elections and Democracy. Thad E. Hall, PhD, is a subject matter expert at Fors Marsh Group, an applied research company that helps organizations and federal agencies make research-backed decisions.