Did Your Mailed Ballot Count: The Unrecognized Unreliability of Voting By Mail

Voting By Mail (VBM) was developed to support absentee voters. It was originally intended to handle canonical absentee voters who now fall under the Uniformed and Overseas Citizens Voting Act (UOCAVA) and those with legally acceptable reasons for being unable to appear at the polls on Election Day. Its use slowly expanded to more casual justifications, such as those with planned Election Day travel. More recently, there has been a trend of further expansion to on-demand VBM in many states. As a result, the percentage of VBM ballots has skyrocketed, with little research regarding its impacts on security, privacy, reliability, and accuracy on U. S. elections. In virtually every close election, the outcome must await tabulation of VBM ballots. Yet, VBM may be the least reliable voting approach in wide spread use today. Vote By Mail fraud is recognized by some as possibly the single greatest security vulnerability in U. S. elections. The lack of in-person, at-the-polls accountability makes absentee ballots the tool of choice for those inclined to commit fraud," the Florida Department of Law Enforcement concluded in 1998, after a mayoral election in Miami was thrown out when officials learned that "vote brokers" had signed hundreds of phony absentee ballots.1 Conversely, others recognize theoretical weaknesses in VBM, but generally dismiss its practical impact [1, 2]. Others continue to promote VBM expansion [3, 4]. In this paper, we identify inherent, widespread vulnerability in VBM systems and illustrate their practical impact with numerous examples. We show specifically why VBM systems are not auditable and demonstrate how their unreliability can negatively impact real elections.