“No Evidence” Is Not the Same as “No Issue”

Election arguments become less useful when different claims are compressed into the same phrase. “An incident occurred,” “a process failed,” “a pattern exists,” and “the result was changed” are separate propositions. Each requires evidence that matches its scope.

The point is not to assume election systems are flawless or to assume every allegation is meaningful. It is to identify exactly what the record establishes before moving to a larger conclusion.

Start by defining the claim

When someone says there is “no evidence,” ask: evidence of what? A specific allegation, a documented violation, a broader pattern, or an effect on a particular result?

Apply the same precision to claims that misconduct has been proved. Evidence about one incident does not establish every allegation about a different contest, jurisdiction or election. The scope of the conclusion has to stay connected to the scope of the record.

A ballot drop box in Denver, October 2020. Contextual election imagery; this photograph is not connected to the North Carolina case discussed in the article. Photo: Jami430. Source · CC BY-SA 4.0. Image unmodified.

Use a claim-scope ladder

A useful way to evaluate an election claim is to move through the questions in order rather than jumping from an allegation to an outcome.

• Allegation: What exactly is being claimed, by whom, about which contest and period?

• Documented incident: What primary record establishes that a specific event or violation occurred?

• Process issue: Does the record identify an administrative, procedural or security problem that needs correction?

• Scale: How many ballots, voters, records or jurisdictions are actually implicated by the evidence?

• Outcome effect: Is there evidence capable of showing that the certified result was affected, or does the record stop short of that conclusion?

Evidence can be strong at one rung and weak or absent at another. A documented incident can justify investigation or reform without establishing a statewide pattern or a changed outcome.

Read the actual finding

North Carolina's State Board of Elections ordered a new election in the 2018 Ninth Congressional District contest after examining evidence of misconduct and irregularities. Its March 13, 2019 order is useful because it shows what a formal finding looks like: a specific proceeding, a defined evidentiary record and a remedy tied to that contest.

That case is a worked example, not a national incidence rate. It cannot establish how often similar misconduct occurs elsewhere, and it should not be used as a substitute for the record in another election.

Audits, recounts and certification answer different questions

The U.S. Election Assistance Commission separates post-election auditing, recounts, canvassing and certification because they are different parts of election administration. A recount retabulates votes under the rules of a jurisdiction. An audit tests specified aspects of the election process or tabulation. Canvassing and certification reconcile and formalize results.

The EAC's 2024 policy survey reports that nearly all responding states conducted logic-and-accuracy testing and that two-thirds reported some form of post-election tabulation audit. Those figures describe administrative practices. They do not establish a national fraud rate.

Likewise, the 2024 Election Administration and Voting Survey covers election administration across all 50 states, the District of Columbia and U.S. territories and reports more than 158 million counted ballots. That dataset is useful for understanding administration and scale, but it should not be repurposed as evidence that misconduct is either ubiquitous or nonexistent.

A process problem is not automatically an outcome claim

An election can have an administrative problem worth fixing even when the available evidence does not show that the final result changed. The reverse distinction matters too: saying a problem did not alter an outcome does not make the underlying process issue imaginary.

That is why safeguards should be debated at the level they address. If a proposal concerns chain of custody, voter-list maintenance, ballot reconciliation, signature procedures, machine testing or post-election audits, evaluate the evidence for that specific problem and the tradeoffs of the proposed response.

What a Well-Supported Election Claim Should Show

• the exact claim and jurisdiction;

• the primary record behind the allegation or finding;

• the number of ballots, voters, records or locations actually implicated;

• the administrative or judicial disposition;

• whether the record establishes an incident, a process problem, a broader pattern, an outcome effect, or some combination of those categories;

• and what remains unresolved.

That standard allows documented problems to be examined without turning every irregularity into proof of a stolen election, and it allows unsupported claims to be rejected without pretending election administration can never fail.