Elections office says hand-count audit of SD 9 matched machine totals; staff to provide reconciliation details

The Tarrant County Elections Administrator told the Commissioners Court on March 10 that a hand-count audit of the SD 9 runoff validated the electronic tally, and that a higher apparent rate of mail-ballot "rejections" reported publicly was a misunderstanding of reconciliation reports.

Elections staff described how the ballot board — an independent, party-appointed body that conducts statutory hand-count audits — performed the SD 9 audit. The administrator said the county expanded the usual sampling for that audit and asked the secretary of state to allow a complete hand check because the race provided an opportunity to investigate image anomalies reported during the counting process. "It was a 100% match," the official reported when asked how the hand count compared with machine totals.

The office also addressed concerns about absentee and vote-by-mail reconciliations. Staff said the large numbers some members of the public cited combined ballots that were mailed but not returned with those that were rejected; the actual rejection rate for SD 9 was about 2 percent, within the historical range for comparable elections (the office provided averages of roughly 1–2.6 percent on sampled prior contests). The administrator said the secretary of state's online curing tool remained unavailable in the state system and the office was providing alternative curing routes (in-person curing or mail). "The online curing function is not working in the new software they released," the administrator said, and the secretary of state has been notified.

Commissioner Simmons asked whether the full-county hand count was necessary and whether the office had budget authority to incur the extra costs of an expanded audit. Elections staff said the statutory, randomized hand count must occur, and that the county had elected to extend the sample to all ballots to test whether the image anomalies represented a systemic problem; they said the work remained within the elections budget but that final invoices were still being reconciled.

The elections office agreed to provide commissioners certified reconciliation data showing mail ballots issued, returned and counted, returned but rejected, not returned, and surrendered at the polls; staff said most of those numbers are already posted on the department's website and promised to attach data sources and calculations to reconciliation entries so the court and public can verify them.

The administrator concluded by telling the court that, based on the audit results and current data, the office did not recommend routine, countywide full-hand counts for future elections. "The post-election hand count of all ballots in November — no, ma'am," the official said in response to a commissioner asking whether such an approach was needed for general elections.

Commissioners requested a follow-up with final accounting of the costs of the expanded audit and certified reconciliation spreadsheets for SD 9 and the recent primary.