A review of publicly available voting data from the Texas Secretary of State’s office shows repeated changes to voter records during the 2026 primary election’s early voting period, an issue that election observers say could just as easily surface in states like Nevada, which rely on similarly structured digital voting systems.
The analysis examined daily snapshots of early voting data for both Democratic and Republican primaries, tracking records from the start of early voting through Election Day on March 3. Each snapshot serves as a complete record of ballots cast, including voter identification, method of voting, and timing.
But comparisons between those snapshots tell a less stable story.
Vote totals tied to specific days—such as February 17—shifted multiple times across successive reports in both party primaries. Under normal expectations, once a ballot is cast and recorded, historical totals for that day should remain fixed.
At the individual voter level, the data showed patterns that raise additional questions. Tens of thousands of voter IDs appeared in one dataset, disappeared in another, and later reappeared. The pattern occurred in 37,039 voter IDs in the Democratic primary and 87,336 in the Republican primary.
There were also instances where additional ballots appeared associated with voter IDs that had already submitted a ballot. The analysis identified 35,169 such cases in the Democratic dataset and 35,772 in the Republican dataset.
In other cases, voter records that initially showed a ballot cast were missing from the final dataset at the end of early voting. Those deletions totaled 10,558 in the Democratic primary and 66,146 in the Republican primary.
Overall, the Democratic dataset, containing 1.4 million records, showed 80,903 changes to voter histories over time. The Republican dataset, with about 1.26 million records, reflected 137,693 such changes, including both additions and removals.
A smaller number of records also indicated multiple ballots associated with a single voter ID. In the final Democratic dataset, 54 voter IDs were tied to more than one ballot, while 1,549 appeared in the Republican data.
Although the findings do not determine the cause of the discrepancies, they illustrate how digital election records can change after their initial publication in ways not easily explained.
The concern is not only in Texas.
Nevada utilizes a centralized voter registration database and reports election data through a layered system that aggregates information from counties before being reflected in statewide totals. Like Texas, Nevada also processes large volumes of early and mail-in ballots, often updating records over multiple days as ballots are verified, counted, or corrected.
Election officials in Nevada have previously stated that updates to voter files can occur for routine reasons, including ballot curing, duplication checks, and late-arriving data uploads from counties. However, the Texas data illustrates how those same processes, if not clearly documented and communicated, can appear as unexplained additions, deletions, or shifts in vote history.
One notable issue raised in the Texas review involved fluctuating candidate vote totals in at least one Democratic congressional primary, where publicly reported numbers decreased at certain points during the count. While such changes can result from corrections, election systems are generally expected to move in one direction, upward, as ballots are added.
Nevada’s reporting system has similar expectations, but is not immune to revisions. During past elections, the state has also updated totals as batches of ballots were processed or errors were corrected, leading to real-time confusion among observers.
Texas primary elections are administered by political parties, with the Secretary of State providing underlying voter data. In Nevada, county officials run elections under state oversight. But both systems rely heavily on digital recordkeeping that is periodically updated.
Election officials in Texas have not publicly explained the specific reasons for the changes identified in the data. Likewise, without detailed audit trails or public-facing explanations tied to each update, it can be difficult to determine whether such shifts stem from routine administrative processes or something more significant.
What the Texas example makes clear is not necessarily what happened, but what is possible.
In any state where voter records are updated dynamically, like Nevada, the same patterns, appearing, disappearing, and changing vote records, can emerge if systems allow for post-entry modifications without clear, transparent tracking.
The Old Blog in the Desert 
Leave a comment