Our Actual Vote analyses as Case Studies
Every case study so far has asked the same counterfactual: if independent evidence had existed, what would it have changed? This one asks nothing hypothetical. Our own published Actual Vote analyses are case studies in their own right — of the method already doing exactly what the previous cases only imagine: producing independent, public evidence about whether vote reporting was faithful, and publishing it for anyone to check.
I. What our analyses are
For each audited election, volunteers record precinct poll tapes at the close of polls. America Counts transcribes them and compares the tape totals against the officially reported results, contest by contest — then publishes the full analysis, with every recording linked, every comparison shown, and the analysis code released so anyone can re-run it. The recordings are captured independently, on election night, outside any government chain of custody. That last point is the whole game: the evidence exists whether or not the official record is later corrected, lost, or disputed.
In other words, each published analysis is itself a case study — of the methodology working, on a real election, in public.
II. The track record
Across 13 county-level election audits in Florida, Georgia, and North Carolina, volunteers have independently audited 1,037,636 individual votes. Across 7,034 tape-to-official comparisons, every apparent discrepancy that ever arose was explained by a mechanism the county itself identified — and none remain unresolved.
III. Why a clean match is powerful evidence
It’s tempting to read “no unexplained discrepancies” as an anticlimax. It isn’t. Look at what the comparison actually does: it takes two records produced completely independently — the poll tapes our volunteers photograph at the precinct, and the totals the government publishes through its own reporting chain — and checks whether they agree. When two independent sources that are supposed to match do match, across thousands of values, that agreement is strong, positive evidence that the reporting layer was faithful. (The only thing that would weaken it is a hidden tautology — two figures that were never actually independent — which is exactly why we are careful that the tape and the official result come from genuinely separate paths.) This is the evidence the eighteen cases above never had.
A word on how the analysis actually unfolds, because it matters. The first pass is messy. Much of the pipeline is automated, but turning videos of paper tapes into a clean dataset still takes real hand work, and the raw comparison surfaces plenty of values that don’t line up at first. Those are not discrepancies — not yet. Most are ordinary artifacts: a transcription slip on our end, or a piece of benign, valid reporting logic we simply hadn’t accounted for — write-in handling, provisional ballots, a missing tape, an undocumented aggregation rule we can back out from the data we already have. Working those out is routine, and it’s the real meaning of “we don’t assume a mismatch means error or fraud.” What survives that process — and what the published analysis is about — is the verified sense in which the tapes and the official results agree, together with any discrepancy we cannot explain. (Our 2022 Wake County analysis is a good worked example: the apparent mismatches traced to ordinary canvass mechanisms the county itself confirmed.)
IV. A transparency finding along the way
In our Georgia analysis, a large share of the precinct values our volunteers transcribed had no officially published counterpart to compare against. Finding nothing wrong is valuable; discovering that there was nothing to check against in the first place is its own kind of finding — a transparency gap that independent auditing surfaces simply by trying to do the comparison.
V. What you can do with these
Unlike the historical cases, you don’t have to take our word for it or imagine the counterfactual. The recordings are public, the comparisons are published, and the code is open. You can watch a poll tape, read the official number, and check the match yourself — or use an analysis as one more independent input when you decide how much confidence the official results have earned. Read the published analyses →
VI. How this realizes the benefits
This is where the benefits described elsewhere on the site stop being promises:
- Transparency — the evidence is public and precinct-level.
- Defending against bad-faith accusations — a clean, independent record is the strongest possible answer to a false claim of fraud, from any direction.
- Deterrence — a known, growing public corpus of precinct tapes changes the incentives for anyone tempted to manipulate reporting, and for anyone tempted to invent a manipulation that never happened.
The other eighteen cases make the argument that vote reporting is vulnerable and under-verified. These analyses are the answer in the same collection: not “Actual Vote would have helped,” but “Actual Vote did its job — and here is the evidence.”