System and Institutional Audit

The purpose is to examine whether a system or institution performs its bounded task in a traceable, reasoned, and corrigible manner. The audit concerns the relation among purpose, mandate, rules, roles, registration, classification, decision path, actual practice, effect, and possible correction.

The audit may be applied to public administration, education, health and welfare services, research, businesses, organisations, media, digital platforms, control arrangements, and other coordinated practices. It does not presuppose that the system is illegitimate, ineffective, or captured. Standardisation, division of labour, documentation, and hierarchy may be necessary; the question is what they do in the concrete matter and whether they continue to serve the purpose.

A precise audit requires system, institution, organisation, rule, procedure, practice, role, mandate, record, decision, output, and effect not to be used as though they were the same link. The distinctions below are operational: they identify what must be documented when a finding is formulated.

An institution may be lawfully established while its practice departs from its mandate. Conversely, a single deviation may result from error, uncertainty, or insufficient capacity without establishing a systematic failure of the whole institution. The audit must keep these levels distinct.

The mandate must bound the system, institution, process, or decision path being examined. It must identify the purposes and norms the arrangement is meant to serve, the time period and cases covered, and what lies outside the audit.

It is not sufficient to ask whether an institution is good, bad, bureaucratic, or corrupt. The mandate must make it possible to test concrete relations among requirements, registered grounds, actual handling, decision, effect, and response to correction.

The system should, as far as practicable, be traced from purpose and mandate through rules, roles, intake, classification, registration, assessment, decision, and action to actual effect, complaint, correction, and revision. A formally complete file does not by itself show that all relevant information was understood or that the effect corresponds to the purpose.

Traceability concerns both case and system. The audit should be able to show what information entered, who changed or assessed it, which rule or discretion was used, what decision followed, and how the system responded to counter-material and error.

• Purpose and mandate → hierarchy of norms → roles and delegation → intake and access.
• Intake → registration → categorisation → prioritisation → information gathering and examination.
• Grounds → professional or administrative assessment → recommendation → decision → implementation.
• Implementation → registered output → actual effect → deviation, complaint, or contradiction.
• Proposed correction → responsible decision → implementation → effect testing → revision.
• Document manual stages, automated rules, system integrations, transfers, and points where information may be lost.
• Record who had access to which material at which time and which changes are reviewable.
• Mark stages that are unavailable, unrecorded, or described only through the institution’s own account.

The questions must be adapted to the mandate and concrete process. They are designed to distinguish formal arrangement from actual operation and to make responsibility, discretion, information flow, and corrigibility visible.

System failure does not arise only through a defective rule. It may arise in the transition among rule, registration, role, capacity, technology, incentive, and actual action. The audit must therefore examine both individual stages and their interaction.

Formal compliance may be important and still insufficient. A system may follow its own procedure while producing an output that conflicts with superior rules, purpose, or the documented matter. Conversely, departure from an internal routine may be justified where a superior duty or concrete actuality requires it, provided the departure is authorised and reasoned.

System capture is not the same as bureaucracy, standardisation, error, delay, lawfully grounded confidentiality, disagreement, or a decision adverse to one party. F6 or F7 requires a bounded and documented relation in which the system’s own frame gains governing priority over the matter, purpose, or relevant grounds.

Capture may occur through category, register, indicator, procedure, automation, division of responsibility, or institutional self-description. What matters is not the name of the mechanism but whether it replaces relevant contact with what the system was meant to serve and resists relevant correction.

Findings must be linked to the part of the system examined: rule, data, classification, role, decision path, practice, effect, or correction mechanism. A mixture of strong and weak stages may require compound findings.

Confidence and sample must be recorded. A documented pattern in a bounded sample cannot automatically be assigned to all units or periods, while a single case may be decisive where it reveals a binding rule or a serious, reviewable failure.

Correction must address the documented mechanism and be selected at the lowest level capable of producing real effect. An error in one register field does not automatically require reconstruction of the entire institution; a binding rule producing systematic harm may require more than an individual reversal.

The correction plan must identify responsibility, authority, resources, deadline, transition, communication, protection of affected parties, possible retroactive effect, and how impact will be tested. Where several bodies share the system, responsibility for the whole must be made explicit.

• Correct inaccurate data, record, category, case number, access, or version and trace where the error has produced consequences.
• Separate mandatory rule from guidance and make discretion, exceptions, and reasons visible.
• Change forms or workflow so relevant individual information and counter-material can be recorded and assessed.
• Clarify roles, delegation, and a responsibility point capable of owning and correcting the matter across units.
• Change indicators and incentives where they drive conduct conflicting with purpose, rights, or documented effect.
• Provide parties meaningful access to grounds, contradiction, complaint, and correction within lawful limits.
• Temporarily suspend or restrict automated or manual practice where documented risk is serious and reversible action is necessary.
• Test the change against case, pattern, and unintended effects; define rollback where correction worsens the situation.

System and Institutional Audit is not by itself a legal compliance audit, financial audit, security audit, accreditation, statutory inspection, workplace investigation, clinical assessment, data-protection impact assessment, or inquiry with coercive public powers. Where such questions are decisive, the proper mandate and expertise must be used.

The document must not be used to totalise an institution as good, evil, captured, or deceptive; to assign hidden motives to individuals; to bypass confidentiality, privacy, or legal process; or to make the auditor’s model of the system immune from counter-material.

The audit must be documented in the common Workbench or an equivalent traceable system. The record should include mandate, system map, hierarchy of norms, roles, data sources, case sample, process chain, findings, contradiction, correction, and review plan.

Review must examine whether correction actually improved correspondence, due process, service quality, responsibility, and effect. It must also test whether the change merely moved the problem to another form, register, level, contractor, or informal practice.