Connecticut’s audit law was signed into effect in 2007. SB 1051 (2015) subsequently authorized the use of electronic equipment in audits after January 2016 and SB 1202 established a working group to examine employing RLAs, including equipment and procedures necessary to implement them, and to oversee a post-election pilot program in 5-10 municipalities in 2021. In January 2022, RLA pilots were conducted in five towns and in February 2022, the working group recommended that Connecticut adopt comparison RLAs for federal elections, including all statewide races, at least one U.S. House of Representatives race, and at least 5% of state legislative races. (Risk-Limiting Audit Recommendations for Connecticut.)
The current audit statute requires 5% of voting districts to be audited. Both the number and type of office audited varies by election.
Unless otherwise specified, statutory references are to Conn. Gen. Stat. § 9-320f.
Voting Systems Used
All Connecticut election jurisdictions use hand marked paper ballots and optical scanners in polling places, with ballot marking devices available for accessibility. For the most up to date information please visit Verified Voting’s Verifier.
For an explanation on the types of voting equipment used, click here.
The audit statute requires the “manual or electronic tabulation of the paper ballots cast and counted by each voting tabulator subject to such audit.” The statute was amended in 2021 to consider “any central location used in a municipality for the counting of absentee ballots” as a voting district eligible for audit. Although the statute does not explicitly mention that provisional ballots must be audited, according to the office of the Connecticut Secretary of State, provisional ballots are included in the audit.
With the passage of SB 252 in 2016, the percentage of audited voting districts was reduced from 10% to not less than 5%, to be randomly selected by the Secretary of State.
The post-election audits are open to the public for observation and reporting. However, there is no mention of how much advance notice must be given. Also, the selection of the audit sample is open to the public. No mention is made in the statutes pertaining to audits regarding either officially appointed observers or the ability of observers to verify voter marks.
After the audit is conducted, the Secretary of State provides a report to The University of Connecticut, whose analysis describing any discrepancies identified is then filed with both the Secretary of State and the State Elections Enforcement Commission. The reports are open to public inspection and may be used as prima facie evidence of a discrepancy for any cause of action arising from the election If The University of Connecticut report indicates that a tabulator failed to record votes accurately, the Secretary of the State shall require that the tabulator be examined and recertified.
The random selection of offices for the audit are not required to be public or to be announced in advance.
Audit Counting Method
The audit is either a hand or electronic count. The statute originally specified that the audit will “consist of the manual tabulation of the paper ballots cast and counted by each voting machine” in the selected precincts. However, with the passage of SB 1051 in 2015, the use of electronic equipment for the purpose of conducting audits was authorized after January 1, 2016.
Type Of Audit Units
The audit uses precincts/districts as units and 5% of voting districts (precincts) are audited. Any central location used in a municipality for the counting of absentee ballots is treated as a voting district for the purposes of the audit.
Contests & Issues Audited
For elections in which the president or governor is on the ballot, the statute calls for auditing at least three contests, one of which must be selected by the Secretary of State in a random drawing. For municipal elections, three offices or 20% of the number of offices on the ballot, whichever is greater, are selected at random by the municipal clerk. For primary elections, the statute calls for auditing at least 20% of the offices on the ballot, including all offices required to be audited by federal law, plus one additional office, if there is one, selected in a random drawing by the municipal clerk. Ballot questions/issues are not audited.
If the manual count cannot be reconciled with the initial machine tabulation, the Secretary of State must investigate further to determine whether to decertify the voting tabulator or tabulators in question or to order the voting tabulator to be examined and recertified. The audit is to be expanded to a “discrepancy recanvass” if the difference between the manual or electronic counts and tabulator counts is more than 0.5%; cannot be resolved through other means (for instance, an accounting of improperly marked ballots); and would exceed the margin of victory if extrapolated to the entire contest (see subsections (f) and (o)).
Re-canvasses in Connecticut are not full hand recounts, but utilize a mix of both hand recounting and machine retabulation. Procedures for discrepancy re-canvasses can be found in Conn. Gen. Stat. §9-311, “Re-canvass in case of discrepancy.”
For recount laws, see Citizens for Election Integrity Minnesota website.
The audit must be completed before results are finalized, specifically between the 15th day after the election and two days before the canvass.
Binding On Official Outcomes
Statute does not provide for audit results to directly affect the reported vote totals. However, if the audit uncovers sufficiently large discrepancies (see “Addressing Discrepancies”), the Secretary of State is to initiate a discrepancy recanvass (recount) that can alter the official outcome.
Oversight & Conduct
The Secretary of State oversees the audits and is responsible for all audit decisions. The audit is conducted by the local registrars of voters. The random selection of the offices to be audited is conducted locally for municipal and primary elections, and otherwise handled by the Secretary of State. This leads to greater independence for general elections with statewide or federal offices on the ballot than for those with only local offices on the ballot.
The registrars of voters shall securely keep the tabulator keys and not permit any tabulator to be unlocked, for a period of 14 days from the election, unless otherwise ordered by a court or by the State Elections Enforcement Commission. (Conn. Gen. Stat. § 9-266.) All tabulators shall be collected immediately on the day after election or as soon thereafter as possible, and shall be secured and stored in a place or places directed by the registrars of voters. For more information see Conn. Gen. Stat. §§ 9-309-11.
Additional Targeted Samples
While the Secretary of State has authority to investigate discrepancies, the statutes do not explicitly mention the possibility of additional targeted samples as part of such investigation.
Conn. Gen. Stat. § 9-320f (audit provision)
RLA Recommendation Working Group: Risk-Limiting Audit Recommendations for Connecticut
Conn. Gen. Stat. § 9-266 (ballot protection)