An academic model spots irregularities in three counties. What might we find if we looked at more?
When people talk about election integrity, we often think in extremes—stolen votes, dead voters, hacked machines. But in a country of 3,000 counties and 170 million registered voters, most trouble doesn’t shout. It whispers. And sometimes, the math hears it first.
We found the link to this report on the Election Truth Alliance website.
A new statistical report from Dr. Walter Mebane, a political scientist and statistician at the University of Michigan, quietly released on May 6, 2025. It analyzed the 2024 presidential vote in just three Pennsylvania counties: Allegheny, Erie, and Philadelphia.
🧠 Sidenote: What is eforensics?
Eforensics is a statistical method developed to detect potential electoral fraud using only publicly available vote counts. It doesn’t prove wrongdoing—but it can flag anomalies that deserve further scrutiny. It’s used worldwide by election monitors and academics.
Using this method, the report asked a simple question: Do the precinct-level vote patterns in these counties look normal? The answer; not entirely.
In its most detailed model, Mebane’s analysis flagged 186 precincts with what the model calls “incremental fraud”—patterns that deviate from expected turnout and vote splits just enough to suggest something may have been altered. These irregularities added up to nearly 29,000 votes. That’s about 24% of the statewide margin between Trump and Harris—not enough to change the result, but certainly not insignificant.
The paper doesn’t claim intentional wrongdoing. It doesn’t speculate about who might have benefited. And it doesn’t try to draw legal conclusions. What it does say is that in these three counties, the numbers are strange enough to stand out—and if anomalies exist here, they may exist elsewhere too.
⚠️ Sidenote: Why Archive This?
Reports like this have a way of disappearing. We’ve seen it with federal sources—like when the CDC quietly removed health guidance related to myocarditis, menstrual changes, or vaccine risks. Once documents become politically inconvenient, they’re often taken down or edited. That’s why BitsRail exists—to preserve the evidence before it vanishes.
That’s why we’ve archived the full report here on BitsRail, at
👉 https://bitsrail.com/serve_document.php?id=26
This is one area in a large country that has enough statistical anomalies to warrant a wider search into election data. And that’s really the point. Election integrity doesn’t mean finding a single smoking gun. It means paying attention when data—quietly and mathematically—signals something out of place.
This article isn’t just a summary. It’s a call to action.
We believe the American public deserves a full, independent accounting—not just of these counties, but of any jurisdiction where vote patterns seem too strange to ignore. The longer we wait, the harder it becomes to access the original evidence. That’s why we’re preserving it—and why we need readers like you to keep asking questions.
Photo credit JillWellington
