An election audit is a systematic examination of electoral records, procedures, systems or controls designed to determine whether specified requirements were followed and whether the evidence supports the accuracy, reliability or integrity of an electoral process. The exact meaning varies because different jurisdictions use the term audit for different activities. An audit can examine financial transactions, voting technology, registration records, ballot accounting, procedural compliance, results data or security controls.
It can occur before an election, during an electoral process or after results have been produced. An election audit should have a defined scope, criteria, methodology and evidence base. It is therefore different from a general review or informal inspection. The central idea is that an independent or appropriately separated examination tests whether the process or records meet specified requirements and identifies discrepancies, weaknesses or exceptions requiring attention.
Audits can be procedural, financial, technological or statistical. A procedural audit may examine whether election officials followed required steps for ballot handling, reconciliation, storage or counting. A financial audit may examine election expenditure and financial records where the relevant authority has that responsibility. A technology audit can examine system controls, access permissions, software management, logging, security practices and testing arrangements.
A statistical audit can use statistical evidence to test whether reported results are consistent with the underlying ballots or records, where the electoral system and available data permit such testing. These forms of audit answer different questions, so the term audit should not be treated as automatically meaning a complete recheck of every vote. A narrowly scoped audit may provide strong evidence about one aspect of an election while saying little about unrelated aspects.
One important distinction is between an audit and a recount. A recount directly repeats or rechecks the counting of ballots under the applicable rules, whereas an audit generally examines evidence about whether electoral procedures, records or outcomes meet specified standards. An audit can sometimes involve checking ballots or comparing records, but its purpose and methodology may be different from a recount.
A recount is usually tied to a particular contest and number of votes, while an audit may examine controls across many polling locations or electoral processes. In England, recount procedures allow votes to be recounted when appropriate, while verification checks ballot paper numbers against ballot paper accounts before counting. The two mechanisms can therefore complement one another without being interchangeable.
Election audits depend heavily on reliable records. Relevant evidence can include ballot accounts, poll books, chain of custody records, reconciliation documents, electronic logs, financial returns, system configurations, testing records and official result statements. The auditor needs to establish what evidence is authoritative, how it was created, whether it was protected against alteration and how samples or records were selected. Audit independence is also important.
If the people responsible for a process are the only people determining whether their own procedures complied with the rules, the audit may provide less independent assurance. Depending on the legal system, audits can be performed by internal audit teams, independent auditors, electoral authorities, government audit institutions, specialist technical organisations or other authorised bodies.
A useful election audit should identify criteria against which evidence is assessed. These criteria can come from legislation, regulations, official procedures, technical standards, contractual requirements or documented controls. The auditor then examines evidence and records exceptions where actual practice differs from the required standard. The existence of an exception does not automatically mean that votes were changed or that an election result was inaccurate.
Some findings may be administrative weaknesses with no demonstrated electoral effect, while others may require corrective action or further investigation. A good audit therefore distinguishes the existence of a control failure from evidence that the failure caused an electoral outcome to be wrong. This distinction prevents both exaggeration and underestimation of audit findings.
Audits can contribute to election security and electoral integrity by identifying weaknesses before they become serious incidents. Pre election audits or assessments can examine whether systems are configured correctly, whether procedures are documented and whether contingency arrangements exist. Post election audits can examine whether records reconcile, whether procedures were followed and whether discrepancies require investigation. Auditing can also provide institutional learning because repeated findings may identify weaknesses in staff training, documentation, technology or management controls.
The value of an audit consequently extends beyond proving that one election was conducted correctly. It can create recommendations for improving future elections. However, recommendations need to be separated from factual audit findings because an auditor may identify a control weakness without having authority to determine electoral policy.
Audits also have limitations. A sample based audit cannot necessarily establish the condition of every individual ballot or record, although appropriate statistical methods can allow carefully defined conclusions about larger populations. A technology audit may establish that certain security controls exist without proving that no attack ever occurred. A procedural audit can identify noncompliance without establishing that the final result changed. Audit evidence can also be affected by incomplete records, poor documentation or limitations in the audit methodology.
Consequently, audit reports should state their scope, methodology, evidence and limitations clearly. The absence of an audit finding should not automatically be interpreted as proof that no irregularity existed; it means that the audit did not identify an issue within the scope and methodology applied.
Election audits are ultimately assurance mechanisms within a wider electoral accountability system. Their usefulness depends on independence, expertise, access to evidence, appropriate sampling or testing, transparent methodology and meaningful follow up. They should be distinguished from observation, which focuses on watching and documenting electoral proceedings, and from recounts, which repeat or recheck vote counting. They should also be distinguished from investigations, which may seek to establish whether a specific allegation or suspected wrongdoing occurred.
An audit can identify evidence that leads to an investigation, just as an investigation can reveal weaknesses requiring an audit. Used together with verification, observation, recounts, complaints and judicial remedies, audits can provide a structured means of testing whether electoral processes operated according to defined requirements.
