Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Elections topic
No spam. Unsubscribe anytime.
Legislative auditors: signature-verification process finds some errors tied to subjectivity; lawmakers and candidates press for more transparency and rule fixes
Summary
Auditors told a Utah Legislature interim committee that signature verification for nomination petitions detected significant invalid signatures but showed measurable error rates driven by subjective judgment; the lieutenant governor’s office and county clerks described rule, audit and transparency changes they will implement.
Get email alerts on the Elections topic
No spam. Unsubscribe anytime.
Auditors from the Office of the Legislative Auditor General told a Utah Legislature interim committee they tested signature verification on statewide nomination packets and found the verification process generally caught significant invalid signatures but produced measurable error rates driven largely by subjectivity in signature comparison.
Jesse Martinson, audit manager for the legislative auditors, told the committee the audit examined candidate nomination packets submitted for three statewide campaigns — including packets for then-Lieutenant Governor Spencer Cox, Senator John Curtis and Derek Brown — and tested a random sample of 1,000 signatures for each candidate. “It appears to us that the existing signature verification process is able, to reasonably identify significant instances of invalid signatures,” Martinson said, summarizing the office’s conclusions about the verification process while noting variability in error rates across the samples.
The audit’s sample found two types of errors: signatures that were validated but should have been rejected, and signatures that were invalidated but should have been accepted. For Governor Cox’s sample the auditors reported an incorrectly validated error rate of 2.4% and an incorrectly invalidated rate of 1.9%; when extrapolated to the full population of reviewed signatures those sample rates would correspond to an estimated several hundred incorrectly validated signatures and a smaller number incorrectly invalidated. Martinson told lawmakers the differences in error rates are attributable to the inherently subjective nature of visual signature comparison and to sampling variability.
Recommendations and legislative response
The auditors recommended the lieutenant governor’s office calculate an empirical error rate for the verification process and factor that rate into the number of signatures required of candidates so there is a buffer to account for expected verification error. The audit also noted the terms used in guidance — such as “substantially similar” and “reasonably consistent” — are not clearly differentiated or defined in statute, administrative rule or the LG’s manual.
Ryan Kelly, director of elections for the lieutenant governor’s office, told the committee the audit “was very helpful” and described several responses already under way. Kelly said the lieutenant governor’s office and the legislature advanced changes in SB 164 that will require the reviewing officer to verify additional signatures — up to 10% beyond the threshold — and create a 1% audit of verified signatures. He also said the LG’s office is revising administrative rule language to clarify verification standards and training materials and that the office will make nomination packets viewable to candidates and party chairs prior to certification.
Davis County Clerk Brian McKenzie, whose office performed verification for the statewide races under contract to the LG’s office, told the committee that clerks have long required verification training and that claims some temporary workers lacked training were incorrect. McKenzie acknowledged errors identified by the audit and said the county and the LG’s office are implementing chain-of-custody and transparency changes.
Public comment raised additional concerns
Two public commenters who ran or supported candidates in the 2024 statewide cycle pressed officials to make additional changes or to expand audits. Michael Clara and candidate Natalie Clausen described difficulty obtaining access to nomination packets, alleged chain-of-custody irregularities and, in Clausen’s case, broader concerns about election-related documents such as cast-vote records and mail-envelope handling.
Clara told the committee he believes some “spare” or backfill signatures used to meet a candidate threshold were not fully traceable in public records and asked the committee to investigate missing cover sheets and packet records. Clausen asked the committee to examine whether privately withheld names in petitions are being treated consistently with statute and asked the committee to review allegations of improper inducement of a write-in candidate and whether investigative agencies responded adequately.
Officials’ next steps
The lieutenant governor’s office told the committee it is preparing revised administrative rules to: clarify verification standards and terminology; require an audit sample and calculation of an empirical error rate to inform how many extra signatures a candidate must provide; and make nomination packets viewable to candidates and party chairs in unredacted form until certification. The Davis County Clerk said he will improve chain-of-custody documentation for packets and that his office is prepared to implement the additional verification and audit steps.
Lawmakers signaled they will review proposed rule language in an upcoming committee meeting and asked election officials to return with clearer documentation of chain-of-custody processes and training completion for temporary verification staff.
Ending
Committee members said they will monitor the implementation of the administrative-rule changes and the effect of the SB 164 provisions on error rates and transparency. Several members urged a wider review of related election materials, including cast-vote records and county-level canvas materials, where legally permitted, so the public can better understand how election outcomes were produced.
