← Election Auditing in the US

Select any state to see a summary of its 2024 audit, with links to the law and the published results.

Depth or breadth: audits built to confirm a winner checked few races; audits checking every race used ballot images or distributed precinct/machine sampling 56 audit programs plus the 4 states with no audit, 5 November 2024 General Election, by how the sample is sized and how many races each county checks. Risk-limiting audit Tiered hand count Fixed-rate hand count Full computer re-count Sample scanner re-count Records review No audit People read paper ballots No person reads a paper ballot (computer or scanner re-count, or records only) Breadth: how many races each county checks → None One Two or three Four or more Every race, ≥ 1 precinct/machine per county Risk-limiting sized to confirm the winner Grows when a race is close, or covers every ballot Fixed share, grows if a discrepancy is found Fixed share No sizing rule, or no audit Depth: how the sample size is set → AZ1 AZ2 MD1 MD2 RI1 RI2 CT NM WA1 WA2 CO OH NJ PA GA NV VT IN NE OK MI1 MI2 MA FL1 FL2 NC WI MN DE1 DE2 WV KS VA1 VA2 SC1 SC2 OR SD MT WY DC AK UT MS AR NH CA HI ID IA TX IL KY MO NY TN AL LA ME ND recounts from ballot images Audits checking four or more races sample only a few precincts or scanners per county, so most local races are never checked. NY is shown by its rule; its 2024 practice is unconfirmed. TN and WA1 use a hand count only in some cases. A state’s two programs are numbered (MD1, MD2). Select any state for details. Rules as in force for the 5 November 2024 General Election. Source: author’s compilation from state statutes, rules and audit records.