Election Audits in America: Time for Innovation

When compared to innovation in voting methods and technology, audits have seen very little in the way of change.

The Omnibus Spending Bill of 2018 provides $380 million to be distributed to states to “enhance election technology and make security improvements.” With this clear intent combined with funding, there is a unique opportunity to ensure lasting change and innovation in election audit methodology.

Against the backdrop of concerns over the integrity of America’s elections, momentum has shifted from unauditable touchscreen voting systems to auditable paper ballots. New technology has made it possible to conduct, as Maryland did in 2016, statewide 100% audits in a fraction of the time of manual audits. However, when compared to innovation in voting methods and technology, audits have seen very little in the way of change.

Election Audits: A Historical Perspective

Election audits typically involve a review to determine the accuracy of the vote count (called a “results audit”), or to ascertain that proper procedures were followed (a “process audit”), or both.  Results audits are characterized by a comparison of two independently produced results (usually ballot counts and vote counts) that were derived from the same underlying data.  Results audits offer the greatest opportunity for innovation.

Eleven years ago, the Brennan Center for Justice issued a landmark report on the need for both results and process election audits.  Naturally, their recommendations are informed by the ways voters cast their ballots, and by the Center’s understanding of available technology to verify election results.  In 2007, most voters cast their ballots in precincts; half of U.S votes were cast on unauditable touchscreens.  Voters’ marks on optically scanned ballots were evaluated with 80-year-old mark-sense technology developed to process student tests.  At that time, the only means to verify the accuracy of voting equipment was through a limited, statistically weak and insignificant hand count of paper ballots. Verification of touchscreen cast ballots was too difficult or impossible.

Voting Methods Have Changed; Audit Methods Have Not 

Since the Brennan Center’s report, driven by the need to reduce costs and make voting more convenient and secure, voters have been given a wide range of methods to cast their ballots beyond precinct voting. Voters now have the option to cast ballots at early voting centers, through the mail, and on electronically delivered ballots. Furthermore, many expect that one day voters will be able to cast ballots securely on their smartphones.

Ironically, these innovations in voting methods have placed severe strains on conducting efficient post-election audits. The world envisioned in the Brennan Center’s report has been quickly overtaken by innovations in voting methods.  For example, when all ballots were cast in a precinct, a manual tally of those ballots (required by legislation) was relatively easy to conduct. However, when more and more ballots are cast in early voting centers as they do in Florida, or arrive through the mail as they do in Oregon, ballots from every precinct are comingled. Physically arranging the ballots by precinct to make the audit more efficient is a time-consuming, error-prone job and a security risk. Large counties have had to employ dozens of part-time staff simply to sort ballots for the audit.

Importantly, every time the ballots are handled by humans, the security risk increases. In short, the drive for operational efficiency and voter convenience has spawned innovations in voting methods that have outpaced manual audit methods.

Published May 01, 2018
Larry Moore - Founder
Larry Moore