BREAKING: MIDNIGHT 'VOTE FRAUD' FILES JUST LEAKED IS GAVIN NEWSOM'S CAREER OFFICIALLY OVER?

California Rejected 148,427 Mail Ballots in Its June Primary—But the Biggest Problem Was Deadlines, Not Proven Fraud

California designed its election laws to make it difficult for an otherwise valid mail ballot to be thrown away.

Every registered voter receives one.

Ballots mailed on time can arrive after Election Day.

Voters with certain signature problems are given an opportunity to correct them.

And counties can continue processing eligible ballots well after television networks have begun announcing winners elsewhere in the country.

Yet nearly 150,000 mail ballots still failed to count in California’s June 2 primary.

The latest statewide figures make the problem slightly larger than first reported.

An earlier California Secretary of State dataset cited by The Associated Press listed 148,241 rejected vote-by-mail ballots, representing 1.73% of mail ballots returned.

The Secretary of State subsequently updated the report.

As of August 6, the official statewide total had risen to 148,427 rejected ballots out of 8,581,393 returned.

Another 8,432,966 were accepted, meaning 98.27% of returned mail ballots passed the state’s acceptance process.

So there are two numbers worth keeping in view.

Nearly 150,000 ballots were rejected.

More than 8.4 million were accepted.

Both are true.

The first deserves scrutiny because each rejected ballot represents a vote that ultimately did not enter the tally.

The second matters because the rejection data do not show an election system in which enormous percentages of mail ballots were failing.

The statewide rejection rate remained below 2%.

What stands out most is why ballots were rejected.

The largest category by far was what the Secretary of State labels:

“Ballot was not received on time.”

The latest report puts that number at 93,666.

Another 43,976 ballots were rejected for nonmatching signatures.

8,335 lacked a voter signature.

743 were rejected because the voter had already voted.

Smaller categories included ballots missing from envelopes, multiple ballots returned in one envelope, missing required identification and miscellaneous county-specific issues.

Those numbers immediately raise an obvious question.

How can more than 93,000 mail ballots be too late in a state that already allows mailed ballots to arrive a full week after Election Day?

The answer is more complicated than saying 93,666 Californians simply voted late.

California law distinguishes between when a ballot is cast and when election officials receive it.

Under Elections Code Section 3020, a mailed vote-by-mail ballot can generally be considered timely if election officials receive it no later than seven days after Election Day and the postal evidence indicates it was mailed on or before Election Day.

That distinction is critical.

California does not provide seven extra days to decide whom to vote for.

A ballot arriving during the grace period still has to satisfy the state’s rules showing that it was timely cast.

The law also accounts for some postal imperfections.

If an envelope has no postmark, an undated postmark or an illegible postmark, other information—including the date on the signed ballot envelope under the conditions specified by state law—can potentially establish timeliness.

The statute also allows other information from the Postal Service or a bona fide private carrier to indicate that a ballot was mailed on time.

That means the Secretary of State’s “not received on time” category should not casually be translated into:

93,666 people voted after Election Day.

The state report does not establish that.

Nor does it provide a subcategory showing exactly how many ballots physically arrived more than seven days after the election versus how many arrived within the grace period but carried postal information that failed the timeliness requirement.

Election experts interviewed by AP believe postmarks are a major part of the explanation.

Kim Alexander, president of the nonpartisan California Voter Foundation, said the late-ballot problem appeared to be driven primarily by postmarks that were too late rather than ballots simply arriving beyond the seven-day receipt period.

AP also reported that many voters waited until the final days of the election during a crowded gubernatorial primary.

That brings the U.S. Postal Service into the story.

And here, another distinction matters.

It is not quite accurate to say USPS suddenly adopted an entirely new postmarking procedure in 2026.

The Postal Service says its underlying postmarking practice has not changed.

Machine postmarks have long generally been applied at originating processing facilities rather than necessarily at the first mailbox or local facility where a piece of mail entered the postal network.

What has changed is the postal transportation network.

USPS acknowledged that adjustments to its transportation operations mean some mailpieces may no longer reach the originating processing facility on the same day they are deposited or collected.

As a result, the machine-applied postmark can carry a date later than the date on which the Postal Service first obtained possession of the envelope.

That sounds like a technical postal detail.

In an election system where legal validity can depend on a postmark date, it can become much more significant.

Imagine a ballot placed into the postal system on Election Day.

If it reaches an automated processing facility the next day and receives a next-day postmark, the date printed by the machine does not necessarily prove the voter waited until the day after the election to mail it.

USPS itself says a processing-facility postmark date does not necessarily reflect the first date on which it possessed the mailpiece.

That is why California election officials were already warning about the problem before the June primary.

On May 26, Secretary of State Shirley Weber’s office advised that some mail might not receive a postmark until the following day and encouraged early returns because of those postal circumstances.

The concern was therefore not invented after 93,000 ballots were rejected.

Officials knew in advance that the relationship between the date a voter handed over an envelope and the date ultimately appearing in a machine postmark could create problems.

The June data suggest that question now deserves closer examination.

But that still does not prove USPS caused all—or even most—of the 93,666 deadline rejections.

Some voters undoubtedly may have mailed ballots too late.

Some envelopes may have arrived outside California’s seven-day statutory window.

Some may have carried postmarks or other information showing they were not timely.

And postal-processing delays may have affected another portion.

The statewide rejection report does not break those possibilities apart.

That uncertainty is one of the most important facts in the story.

The numbers identify a problem.

They do not yet fully identify its cause.

The comparison with earlier primaries makes that problem more noticeable.

AP reported that California rejected 108,982 mail ballots in the 2024 primary, or 1.56% of mail ballots.

In the 2022 primary, 105,818 were rejected, at a similar rate.

The 2026 total is therefore substantially higher in raw numbers, while the rejection rate increased more modestly to 1.73%.

Lateness was also the largest rejection category in the earlier data reviewed by AP.

About 76,000 ballots were rejected for lateness in the 2024 primary and roughly 70,000 in 2022.

The latest 2026 statewide report now places that category at 93,666.

So saying deadline problems are the largest single reason for rejected mail ballots is well supported.

Saying they necessarily account for most of the entire increase requires more caution because other categories, particularly signature problems, also contribute substantially to the statewide total.

The signature numbers deserve their own examination.

California’s updated report lists 43,976 nonmatching signatures and 8,335 envelopes with no voter signature.

Those are not minor categories.

Together they represent more than 52,000 ballots.

California provides a process under which voters whose mail-ballot signatures are missing or fail verification can submit a signature statement to attempt to cure the problem.

For the June primary, the Secretary of State listed June 24 as the applicable statewide deadline for those signature corrections.

That means an initial signature problem does not automatically equal a permanently rejected ballot.

The figures appearing in the statewide rejection report represent ballots that ultimately remained unaccepted in the data submitted by county election officials.

Signature verification is also one reason California can take longer than some states to finish election processing.

Speed and validation are different objectives.

A state can produce very rapid unofficial results by requiring virtually everything to be in election offices by Election Day.

California instead permits certain ballots to arrive afterward and gives some voters additional opportunities to resolve signature defects.

That increases the amount of legitimate post-election administrative work.

And that distinction became nationally significant only weeks after California’s primary.

On June 29, the U.S. Supreme Court decided Watson v. Republican National Committee.

The case involved Mississippi, not California.

Mississippi law allowed certain absentee ballots postmarked by Election Day to arrive as many as five business days afterward.

Republican challengers argued that federal statutes establishing a national Election Day meant ballots also had to be physically received by election officials on Election Day.

The Supreme Court rejected that argument 5–4.

Justice Amy Coney Barrett’s majority opinion concluded that the federal Election Day statutes establish when the voter’s choice must be made but do not themselves impose a nationwide Election Day deadline for state officials to receive the ballot.

Mississippi could therefore continue counting qualifying ballots that were timely cast but arrived later.

The ruling matters directly to the political debate around California.

It means the mere fact that election workers are still receiving or counting qualifying ballots after Election Day does not, by itself, make the practice unlawful under the federal Election Day statutes addressed in Watson.

The Court did not order every state to adopt a grace period.

It left room for states and Congress to set the applicable rules.

California currently chooses seven days for qualifying mailed ballots.

Mississippi uses five business days for the absentee ballots covered by its statute.

Other states use different systems.

And that exposes one of the central misunderstandings surrounding slow election counts.

Counting after Election Day is not the same thing as voting after Election Day.

A ballot can be legally cast by the deadline and then spend days moving through the mail, verification and canvassing process.

Likewise, election workers can continue examining and tabulating ballots after Election Day without allowing voters to continue changing their choices.

Those are administrative processes following the casting of the vote.

That does not mean every late-arriving ballot is valid.

California rejected 93,666 ballots under its timeliness category precisely because they did not satisfy the state's final deadline rules as applied by county officials.

It means only that the existence of a post-election count is not evidence that illegal votes are being added.

That distinction became particularly important after President Trump attacked California’s lengthy June count.

Trump alleged widespread wrongdoing while results continued to change as counties processed mail ballots.

The Justice Department’s Los Angeles office subsequently announced “multiple election fraud investigations” and sent a federal prosecutor to observe operations at Los Angeles County’s main ballot-processing center.

At the time, First Assistant U.S. Attorney Bill Essayli said California had serious structural vulnerabilities but provided no public specifics about the investigations.

Los Angeles County said the federal prosecutor was given an overview and walkthrough of its publicly observable processing operation.

One of the most notable reactions came from Republican gubernatorial candidate Steve Hilton, whom Trump had backed.

Hilton advocated major changes to California’s mail-voting system, including tighter receipt rules.

But he also said his own campaign had been monitoring the count and had seen nothing illegal that would warrant legal action.

That distinction has become even more relevant with time.

As of August 14—more than two months after the Los Angeles federal prosecutor announced the election-fraud probes—no case arising from those broad June-primary allegations had been filed, according to the latest reporting on the investigations.

That does not mean an investigation can never produce charges.

Investigations can take time.

It does mean an announced investigation should not be described as proof that fraud occurred.

A prosecutor saying he is looking for evidence and a prosecutor filing a case supported by evidence are two different stages.

As of mid-August, the second has not followed from the broad June announcement.

And the rejection statistics themselves do not demonstrate widespread fraud.

A ballot rejected because it arrived too late is not inherently fraudulent.

A ballot rejected because a signature could not be matched is not inherently fraudulent.

A ballot returned without a signature is defective, but defect is not the same thing as an intentional attempt to cast an illegal vote.

Even the 743 ballots coded “voter already voted” should not automatically be translated into 743 proven criminal double-voting cases.

The Secretary of State’s table records an administrative rejection reason.

It does not make findings about the voter’s intent or establish a criminal offense.

In one sense, that category demonstrates a safeguard doing what it is supposed to do.

Once an eligible ballot from a voter has been accepted, another submitted ballot associated with the same voter is not simply added to the count.

The duplicate is rejected.

The statewide table records the result of that control, not a criminal conviction.

Broader research likewise does not support treating California’s 148,427 rejected ballots as evidence of a massive mail-voting fraud problem.

A 2025 Brookings Institution analysis examined documented mail-voting fraud cases across the 2016, 2018, 2020 and 2022 general elections and estimated an average rate of roughly four cases for every 10 million mail votes.

Brookings acknowledged limitations in the available fraud databases, including that the database it used was not exhaustive, but concluded that even substantially more expansive assumptions still produced an extremely low incidence rate.

That research should not be stretched beyond what it proves either.

It does not establish that fraud is literally impossible.

It does not tell us the disposition of every questionable ballot in California’s June primary.

It addresses the broader empirical claim that mail voting routinely produces widespread documented fraud.

The available evidence does not support that claim.

California’s June numbers point toward a different concern:

eligible voters losing ballots because of administrative deadlines, postal timing or unresolved signature problems.

That is an election-integrity issue too.

Election integrity is not only about stopping ineligible votes.

It is also about making sure valid votes are not discarded unnecessarily.

The challenge is doing both at the same time.

The county numbers show that experience varied substantially across California.

The latest Secretary of State report places Tulare County at the highest rejection rate, 3.52%.

The updated figures now put Merced County at 3.38%, followed by Alpine County at 3.36% and San Luis Obispo County at 3.35%.

That is another correction to the earlier version of the story.

The preliminary data AP used had Merced and Alpine both at 3.36%.

The August 6 supplemental report changed some county totals slightly, including Merced’s.

Los Angeles County, despite being at the center of Trump’s post-primary criticism and the Justice Department’s attention, had a rejection rate of 1.34% in the latest statewide report—below the statewide 1.73% rate.

Its huge population meant it still accounted for 24,806 rejected mail ballots in raw numbers.

Those county differences could be important for understanding what happened.

If postmark timing were the only factor, researchers would want to know why rejection rates were considerably higher in some regions than others.

Differences in voter behavior, postal transportation, county processing, demographics and signature verification could all potentially matter.

The statewide table alone cannot determine the relative contribution of each.

There is another limitation in the data that should prevent partisan conclusions.

The Secretary of State’s rejection report does not break rejected ballots down by party registration.

AP specifically noted that absence.

So the figure cannot support a claim that 148,427 Democratic ballots were rejected.

It cannot support a claim that 148,427 Republican ballots were rejected.

And without additional voter-level analysis, it cannot establish that the rejection pattern systematically benefited one candidate or party.

Those claims would require evidence beyond the statewide rejection table.

What the table can show is where election administrators should look.

More than 63% of the rejected mail ballots fall into the state's timeliness category.

Nearly 44,000 had signatures that ultimately were not accepted as matching.

Thousands more were unsigned.

Those are concrete failure points.

The postal issue may be the most urgent because the rule depends partly on a timestamp the voter does not entirely control after surrendering the envelope.

USPS says most machine postmarks are based on when the mailpiece reaches and undergoes its first automated processing operation.

Transportation adjustments can mean that date is later than the day the Postal Service first received the mail.

California law has mechanisms for handling missing, illegible or undated postmarks and other postal evidence.

But the June rejection spike suggests officials need better information about how many ballots actually encountered a mismatch between deposit date and machine-postmark date.

Without that analysis, two very different scenarios remain mixed together.

One is a voter who waits until after Election Day and sends a ballot that is legally too late.

Rejecting that ballot simply enforces the deadline.

The other is a voter who transfers a ballot to the Postal Service by the legal deadline but receives a later processing postmark because of postal transportation.

That raises a much harder question about whether the election system is accurately measuring when the vote was cast.

Those scenarios should not be treated as interchangeable.

Nor should California’s seven-day window itself be blamed for the rejection increase without evidence.

The grace period actually allows more timely cast ballots to survive postal transit.

Eliminating it would not solve delayed delivery.

It would simply change which ballots fail the deadline.

The Supreme Court’s Watson decision confirms that states have room under the federal statutes at issue to use such receipt grace periods.

That does not settle the policy argument.

Critics can still argue California's deadline is too permissive and prolongs uncertainty.

Supporters can argue it protects voters from ordinary mail delays.

The legal question and policy question are separate.

After Watson, federal law does not impose the Election Day receipt interpretation the challengers sought.

California can still debate whether seven days is the right state policy.

The same separation should apply to California’s slow count.

There are legitimate criticisms to make.

Waiting weeks for close races to become final frustrates voters and creates fertile ground for misinformation.

Election administrators have an interest in improving speed without sacrificing verification.

California can examine whether its technology, staffing, signature procedures and postal coordination can produce faster results.

But slow counting is not itself evidence of fraudulent counting.

The June primary demonstrated that clearly.

As more mail ballots were processed, some candidate margins changed.

Trump cited those movements as suspicious.

Yet even the Republican gubernatorial candidate he supported said his observers had not seen illegality warranting legal action.

That is how mail-heavy elections can behave.

Election-night totals are incomplete.

Later batches can have different partisan characteristics from earlier batches.

Adding lawful ballots changes the totals because the count was not finished—not because votes were manufactured after the fact.

There is an irony in the current controversy.

California is often criticized for taking too long because it creates multiple mechanisms intended to preserve potentially valid ballots.

Yet the June data show critics can simultaneously argue that the system throws away too many ballots.

Those concerns pull in opposite directions.

Shorten all deadlines and results may become faster, but more ballots caught in postal transit could be rejected.

Extend opportunities to cure problems and more ballots may survive, but certification takes longer.

There is no election-administration design that maximizes speed, access and verification without tradeoffs.

California has chosen to place comparatively more weight on giving potentially valid mail ballots time to complete the process.

Even so, 148,427 did not make it through.

That is the number state officials should investigate.

But it should be investigated for what it actually represents.

Not as 148,427 fraudulent votes.

Not as 148,427 votes cast after Election Day.

And not as proof that California’s entire election system failed.

It is a collection of ballots rejected for multiple administrative reasons, dominated by one category involving timeliness.

The most useful next question is therefore not:

“Who cheated?”

It is:

“Why did 93,666 returned mail ballots fail California’s timeliness rules, and how many of those failures came from voter delay versus postal processing?”

The current statewide report cannot answer that.

Nor can it tell us whether the recent postal-network changes disproportionately affected specific counties.

Those questions require more granular evidence.

If California officials find that most ballots were simply mailed after the legal deadline, then voter behavior is the primary problem.

If they find substantial numbers were handed to USPS on time but received later postmarks because of transportation and processing, the policy problem looks very different.

And if particular counties show unusual patterns unrelated to USPS, those patterns deserve separate investigation.

That is what evidence-based election scrutiny should look like.

The Justice Department can investigate actual allegations.

California can investigate administrative failures.

USPS can explain the effect of its transportation system.

Researchers can compare rejection patterns.

And political candidates can argue over whether the seven-day rule should remain.

What they should not do is pretend those separate issues have already produced the same conclusion.

As of August 16, the latest official California data show 148,427 rejected mail ballots, a 1.73% rejection rate, with 93,666 falling into the timeliness category.

The Supreme Court has confirmed that states may legally count qualifying ballots received after Election Day under the federal statutes challenged in Watson.

The Postal Service has acknowledged that a processing postmark can be later than the date it first possessed a piece of mail.

And despite the Justice Department announcing multiple California election-fraud investigations after the primary, the latest reporting says those broad investigations had produced no charges by mid-August.

Those facts leave California with a real problem.

Just not necessarily the one its loudest critics are describing.

Nearly 150,000 ballots failed to count.