If adopted, county clerks would hand-check selected election races, while election workers would count all ballots even when the poll list shows fewer voters.
County clerks would check selected statewide races by hand, using random samples of paper ballots before election results are certified.
When ballots outnumber people listed as voting, election workers would count every ballot and record the number and type of extra ballots. A large mismatch would require a state investigation.
The Elections Commission would have to allow a qualifying comparison of reported and hand-counted ballot batches when assessing voting-system error rates.
County clerks would manually audit selected election contests using random samples of paper ballots. Audits cover the statewide candidate or referendum contest with the most votes and, if available, one additional contested statewide race randomly chosen by the Elections Commission at a public meeting. Races decided by 0.25 percent or less are excluded. The audit uses statistical methods to limit the risk of certifying an incorrect preliminary result. If an audit leads to a full manual count, that count replaces electronic-system totals for the official result. The bill analysis says audits would begin with elections held in 2027. Each audit must finish before the county canvass is completed and results are certified.
Election inspectors and absentee ballot canvassers would count all ballots and document any excess over the poll list. This applies when ballots outnumber the electors recorded on the poll list, including ballots handled by absentee ballot canvassers. Workers must factor each documented discrepancy into the risk-limiting audit. The Elections Commission must investigate if it exceeds 0.5 percent of ballots cast in the audit unit. The process would apply to selected contests beginning January 1, 2026, and to all elections beginning April 1, 2027.
The Elections Commission would have to allow a qualifying batch-level comparison for voting-system error-rate assessments. This concerns the error-rate assessment of voting systems. The comparison must meet commission standards and compare each unit's reported count with a manual count. Differences guide whether more units are checked, until the risk limit is met or the election receives a full manual count.
No hearings or votes scheduled.