Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments February 10, 2020 erstellt
Beschreibung:
Post-election audits should provide the public with evidence that election results are fair and accurate. However, New York state's current audit has no guarantees that erroneously reported election results will be caught and corrected at a high rate. When outcome-changing tabulation errors are confined to a single machine, the New York audit will fail 97% of the time. The story is not much better when a handful of machines are affected. On the other hand, a risk-limiting audit with a 5% risk limit is statistically guaranteed to catch at least 95% of wrong outcomes. We simulate audits of precinct level election returns from the 2016 presidential race to illustrate the differences. We recommend replacing the current New York audit with a risk-limiting audit