Wouldn’t touch it with a ten point poll: Polling uncertainty and the prospect of whisker-thin elections

• In Event: Elections, Turnout, and Ballot-Box Policy Making

Abstract

Recent conspiracy theories in the United States posit that electoral fraud occurred during presidential elections where margins of victory were incredibly small. The implicit suggestion is that extreme ex post election closeness is an indicator of fraud. This study proposes a theoretic model for why extreme closeness (e.g. margins of victory within 1%) would actually be an unlikely outcome of systemic electoral fraud based on uncertainties from pre-election polling. It then tests this theory against a data set of more than 1000 elections graded as fraudulent or likely clean. Even after removing fraudulent elections designed to be blowouts, contrary to the conspiracy theories fraudulent elections were less likely to be within a margin of victory of  .5% or 1%. This finding can be a tool in testing whether future elections are fraudulent. While it does not usurp the need for vigilance, it may tone down heated arguments during prolonged vote counting periods, such as Georgia in 2020.

Author

• Joe Henry Uchill, RAND School of Public Policy
  Presenting Author