Utah signature auditors provide data and red herrings; Utah media inaccurately report findings

On October 15 2024, the Utah Legislative Audit Group reported their findings on verifying the signatures of Governor Spencer Cox's petition to run for re-election in the Utah Republican primary after losing his bid to be the nominee at the GOP state convention.

Although Utah legislative auditors provided a table of statistical sampling to legislators, they elided over their own data that found Spencer Cox likely did not gather enough signatures to be on the Republican primary ballot. The statistical methods used and reported by the committee show that Cox likely was nearly 100 signatures deficient, yet they refrained from directly articulating the shortfall for Governor Cox, instead offering assurances that he qualified to be on the ballot because he "did everything he could" or that the timing for his submission was unfortunate.

However, Utah Code 20A-9-400-411 does not empower auditors to referee ballot qualification based on their subjective sympathies for the level of effort provided by the candidates or their timing for submitting signatures. Instead, Utah code permits auditors to use statistical methods to determine and verify signature validity and percentages. Nowhere in Utah code does it instruct auditors to provide opinions on when exceptions to signature thresholds can be forgiven, based on subjective or novel criteria. Utah code is clear on meeting the signature threshold. It emphasizes the importance of certifying the petition so that "each name is that of a registered voter who is qualified to sign the signature packet." The auditors' numbers show this did not happen for candidate Cox, as reflected in their own table of data.

The auditor's table shows a sample of one thousand signatures that they reviewed. In that sample, they found incorrectly validated signatures at a rate of 2.4% and incorrectly invalidated signatures at 1.9%. In other words, 2.4% of the signatures were invalid, but 1.9% of the signatures had previously been incorrectly invalidated by a county clerk, so those signatures were added back in. Projecting both of those percentages (which is allowed by code) to all 28,000 signatures gathered results in 572 sum total signatures that should have been invalidated from Cox's petition. Because Cox submitted only 492 additional signatures that could have been counted if invalidated signatures were encountered, he likely submitted 80 few too signatures in the original submission, and approximately 93 signatures deficient if counting his additional signatures. Because statistical calculations based on sampling include a margin of error, the auditors included numbers for a possible "low" sample error and a "high" sample error alongside the numbers they actually examined in their sample.

Source: Utah Office of the Legislative Auditor General, "A Performance Audit of the Signature Verification Process"

After the report was presented, Utah media widely misrepresented the findings and often completely omitted the results of the intent of the audit– to determine signature verification and verification rates. None appeared to report the obvious result that based upon both Utah code for signature limits and the statistical analysis allowed by code, Cox did not qualify for the primary ballot after losing his bid to be the nominee at the convention. Instead Utah media repeated the legally irrelevant factors presented in the auditor's opinions which they are not authorized to referee. Utah media then erroneously announced findings that he qualified for the ballot based on these extra-legal factors.