Elections and the Challenges of Technology

Our voting systems are a difficult marriage between ancient and modern tools. Keeping the proper balance is tricky but crucial.
August 24, 2017
Businessman writing a digital signature on a tablet.
Digital signatures can be problematic for election administrators. (Shutterstock)
By Phil Keisling  |  Contributor
Director of the Center for Public Service at Portland State University's Hatfield School of Government

Elections are arguably the most important administrative processes underlying our democracy. So when it comes to running America's elections with efficacy and integrity, when are business processes and technologies that are two millennia old the safest and most appropriate to use? And when do 21st-century tools make more sense?

This was one of many key questions explored at a two-day Election Sciences, Reform and Administration (ESRA) conference last month in Portland, Ore., that brought together more than 50 academic scholars and election administrators from across the country. Among them was Chris Walker, the elected county clerk of Jackson County, Ore. While many of the statutes and rules that govern elections have their origins in a "world of paper," Walker noted, "more and more, the administration of elections depends on the quality of electronic data transfers."

Oregon, for example, now allows and encourages citizens to go online to register to vote or to update an existing registration. This particular use of modern technology -- which almost 40 states, both red and blue, now use -- saves money, greatly reduces reliance on paper records and improves the accuracy of voting rolls. But it also carries with it a certain conundrum, which anyone who's recently used a credit card at a local business might attest to.

If you're like me, your touch-screen scrawl for these transactions may be far different than the way you sign paper documents, such as a voter-registration card. So even as election officials increasingly rely on software-enabled technology to reduce their reliance on paper, the existence (somewhere) of a physical signature is still considered a best practice.

But there's another issue beyond the contrast between a careful signature on a legal document and a credit-card scrawl -- an especially relevant fact for states and local governments that automatically mail ballots to every registered voter. (Colorado, Oregon and Washington state now use such a "vote at home" system; in last November's elections, 21 of Utah's 29 counties also took this approach). Before any ballot in these systems can be counted, the physical signature on the ballot's outer envelope must be matched to the official signature connected with that voter's registration record, typically a digitized version of another paper record.

While over 99 percent of such signatures are usually matched immediately, a residual number aren't. Fraud isn't the problem; rather, people's signatures can and do change over time, especially among voters at the youngest and oldest ends of the age spectrum. That leaves election officials with the task of contacting these voters to urge them to update their voter registration records, which many -- but not all -- do quickly enough to allow their ballot for a particular election to be validated and counted.

Other types of mismatches also pose a challenge. For example, I might be "Phillip A. Keisling" for purposes of certain government records, but "Phillip Keisling" or "Phil Keisling" for other purposes. And my address for DMV purposes might be different from the address I consider my "true home" for voter registration purposes -- if, for example, I'm studying or working out of state or stationed at a military base.

Technology now enables millions of such records to be matched against each other for purposes of updating voter-registration records and dealing with such common problems as duplicate registrations within and between states. But a "false positive" can lead to the cancelling of a valid registration. And in some cases, the less-than-careful use of such powerful technology has needlessly disenfranchised voters.

Partisans of all stripes are now debating election-administration issues with a ferocity not seen in decades, and in too many cases with far more ignorance than is healthy for a discussion of this importance. Too many conservatives obsess over fraudulent votes that simply don't exist; too many progressives are overly protective of long-outdated records that linger on official voting rolls, to the point that many jurisdictions have more registered voters than eligible citizens.

Fortunately, today we have far better tools than we did 50 years ago, when many existing election laws and practices were already a century or more old. Making our elections run even more effectively -- and be more trusted by our citizens -- isn't a matter of paper or software. Rather, it's a matter of figuring out how to best use each of these powerful tools in the optimal way. That's a goal that future ESRA conferences will focus on as we continue to bring scholars and practitioners together to share relevant new research and best practices in this vital arena.