Communications of the ACM
The Risks of Election Believability (or Lack Thereof)
Credit: Patricia McKnight / USA Today Network
Despite or perhaps because of COVID-19 health concerns, a record 155 million ballots were cast for President in the 2020 general election, with both the winner and runner-up each individually getting more votes than any candidate in U.S. history. Yet, according to post-election polling,11 only two in three voters felt confident the election was free and fair. Even the voter-verified paper ballot9 for direct-recording electronic voting machines (which enables hand-counting via an audit or 100% tally) does not in and of itself create sufficient credibility in the election results.
Trustworthiness in elections is inherently a total-system problem (as considered more generally2). Every part of the overall process (for example, voter registration, ballot layouts, casting and counting, audits, and recounts) provides potential points of compromise. Problems may result from human errors, intentional manipulations (such as ballot tampering, creative disinformation, and insider fraud), imbalanced redistricting (for example, local and state gerrymandering), Electoral College issues, unlimited funding and targeted advertising (for example, Citizens United, Cambridge Analytica), delays at the polls due to malfunctioning equipment, and even the effects of environmental conditions (such as weather, power outages, and pandemics). Education is also a vital consideration, especially in the ability of voters to cope with changing conditions, new instructions, politically biased reporting, and a host of other incidental or even disingenuous factors.
No entries found