Douglas W. Jones, CC0, via Wikimedia Commons Born Isopod and voters of the United States of America, CC BY 4.0, via Wikimedia Commons

Ballots Under Review: Court Case Disputes 2024 Election Results

When the numbers didn’t add up, a judge told the state to look again.

by Met Middleson

June 11, 2025

It began with a statistical red flag: thousands of ballots in Rockland County’s 2024 election recorded a vote for president but no vote for Senate. That pattern proved statistically far from random, with undervote rates exceeding 30 percent in some Trump strong districts compared to normal levels elsewhere. Now a judge has ordered a hands-on review of those ballots, not just in the Senate race but including the presidential selections themselves.

VERIFIED DISCREPANCIES

The 2024 general election certified Donald Trump as the winner in Rockland County, New York, with 83,543 votes, a dramatic 11-point swing from 2020. Yet more than 15,000 ballots recorded no vote at all for the U.S. Senate race.

That total, 15,366 undervotes in the Senate race, amounted to just over ten percent of all ballots cast in Rockland County. Thousands of voters selected a presidential candidate but left one of the most prominent races on the same ballot line blank. According to SMART Elections, the drop-off pattern was unusually concentrated in certain districts and statistically inconsistent with expected voter behavior. Their analysis suggests that Senate undervotes were more common in areas where Donald Trump performed strongly, raising questions about whether machine misreads selectively failed to capture down-ballot votes.

By contrast, drop-off rates in historically Democratic districts were lower and more consistent with past elections. The largest anomalies clustered in Republican-leaning precincts, suggesting the possibility that down-ballot selections were misread or misrecorded even as presidential choices were properly tallied.

Further evidence came from Senate candidate Diane Sare. In two districts voters submitted affidavits swearing they voted for her, yet the official tally listed fewer Sare votes than sworn statements. In ED 39 nine affidavits were filed while the certified count showed five Sare votes. In ED 62 five voters testified, but the count showed three.

These are not abstract claims; they are sworn statements matched to real ballots. Together they raise the prospect that the tabulation process failed to register certain selections beyond the Senate race, potentially affecting other contests on the same ballot.

SYSTEMIC PATTERNS AND DROP-OFF LOGIC

The ballots in question were all processed through Dominion ImageCast scanners. These machines read selections across the entire ballot, from president to local judgeships. But in Rockland County’s 2024 general election, the machines consistently reported complete presidential selections followed by missing entries for U.S. Senate and beyond.

That pattern appeared at scale. Of the more than 15,000 ballots that skipped the Senate race, most had a valid presidential vote. In precinct after precinct, voters selected Trump or in fewer cases Harris, but appeared to skip every other contest. Statistically, that level of drop-off is rare. Even in low-profile races, some residual vote activity typically remains. Here, large blocks of ballots show a clean top-line vote followed by silence.

Election integrity advocates point out that high undervote rates can signal either voter intent or machine behavior. In this case, the consistent undervoting aligned more with geographic clusters than with demographic expectations. Districts with higher Trump support were more likely to show extreme drop-off between the presidential and Senate races, while districts with tighter margins showed more typical vote patterns.

This matters because all races on the ballot are recorded from the same scan. A machine error that misreads one contest could affect the others. While there is no confirmed evidence that presidential votes were lost, the lawsuit argues that the failure to verify results from any part of the ballot raises concerns about the accuracy of the whole.

LEGAL RESPONSE AND DISCOVERY STATUS

New York Supreme Court Justice Rachel Tanguay ruled that the case presented by SMART Elections and its co-plaintiffs could proceed to discovery. Her order was based on specific allegations backed by voter affidavits, certified result comparisons, and documented undervote statistics.

The plaintiffs are not alleging widespread fraud. Their claim centers on the argument that Rockland County’s Board of Elections certified results that failed to reflect the true intent of voters. They point to machine misreads, undervote anomalies, and the Board’s refusal to allow a recount or full ballot inspection despite formal requests and testimony from affected voters.

Under New York law, parties to an election challenge can request discovery when they show reasonable grounds for concern. In this case, the court found the evidence credible enough to require the Board of Elections to release machine logs, scanner records, ballot images, and related documentation.

The legal action seeks no reversal of election winners. It asks for a full hand count of ballots in the affected districts and a public accounting of any discrepancies between machine tabulations and the physical vote. If differences are confirmed, the plaintiffs say the county’s certification process may need procedural reform, particularly in cases where voter testimony conflicts with machine output.

As of this writing, subpoenas have been issued, deadlines are pending, and the court has not placed limits on which parts of the ballot may be reviewed. That includes the presidential race.

ELECTION PROCEDURES UNDER SCRUTINY

For Rockland County officials, the 2024 election was certified without incident. Machines reported totals. Tallies were transmitted. No recount was triggered, and no audit was performed. But the court’s decision to allow discovery has turned a routine certification into a case study in how election procedures respond, or fail to respond, when numbers are contested by the public.

New York does not require an automatic hand recount unless the margin between candidates is razor thin. It also does not mandate that undervotes or blank contests be investigated unless formally challenged. That leaves gaps in oversight, especially when undercounts are not spread evenly but instead cluster in certain precincts or among specific races on the ballot.

In this case, the plaintiffs allege that voters were disenfranchised not by fraud or suppression, but by a system that refused to double-check itself. They argue that sworn affidavits from voters should have triggered a review, and that the Board of Elections’ refusal to provide ballot images or scanner logs limited their ability to prove the discrepancy.

The court has not ruled on whether those claims are accurate. But by granting discovery, it opened the door for those records to be examined and for the procedures that blocked a recount to be tested in full public view.

WHAT COMES NEXT

The court’s discovery order puts Rockland County’s 2024 ballots, scanners, and certification procedures under formal review. The Board of Elections is now required to provide machine logs, ballot images, and other records that could confirm or contradict the plaintiffs’ claims.

If those records show that votes were miscounted, skipped, or dropped from the certified results, it may not change the outcome of any race. But it would challenge the assumption that certified results always reflect voter intent, and could lead to legal or legislative pressure for stronger recount protocols in future elections.

For now, the case continues. No findings of fraud or machine malfunction have been made. But one fact is no longer in dispute. A New York judge found the evidence credible enough to look again.