Institutional governance and assurance whitepaper

Institutional Review Index Whitepaper

The Institutional Review Index (IRI) is Terra Vita Hub’s reviewer-assurance and meta-governance layer: a transparent, evidence-linked mechanism for reviewer behaviour, reviewer consistency, bias and divergence detection, programme calibration, and cross-programme comparability.

Where this document belongs

This whitepaper sits with the public Institutional Review Index, the Institutional Assurance Layer, Governance Architecture, and the Governance Spine & Assurance Annexes. It completes the public explanation of the three assurance layers: structure, proof, and reviewer oversight.

Governance Spine — controlled structureIAL — proof and reconstructionIRI — reviewer assurance and calibration

Authority boundary

The IRI supports oversight, calibration, and reviewer-process assurance. It does not replace human reviewers, statutory decision-makers, procurement bodies, auditors, approved MRV methodologies, sovereign authority, or financial decision-makers.

Governance stack position

LayerQuestion answeredInstitutional function
Governance SpineIs the operating structure controlled?Defines evidence intake, reviewer routing, workflow controls, approval posture, audit lineage, and export readiness.
Institutional Assurance LayerCan the proof be reconstructed?Links evidence, reviewer attribution, audit trails, controls, and assurance artefacts into a verifiable proof environment.
Institutional Review IndexCan the reviewer process itself be trusted?Evaluates reviewer behaviour, consistency, calibration, bias, divergence, and cross-programme comparability.

Core IRI components

Reviewer Performance IndexEvidence alignment, timeliness, completeness, control adherence, and decision-quality signals.
Reviewer Consistency ScoreStable criteria application, drift detection, evidence interpretation, and comparable reasoning.
Bias & Divergence DetectionUnexplained outliers, thematic or geographic skew, peer divergence, and anomaly detection for human review.
Programme Calibration LayerCross-programme baselines and comparability without erasing local programme rules or sovereign context.
Institutional Oversight DashboardRole-governed dashboard outputs, reviewer assurance signals, calibration alerts, and audit-ready exports.

Scoring and safeguard posture

The whitepaper defines IRI scoring as evidence-linked and explainable. Scores and alerts must be reconstructable from source evidence, reviewer action, rationale, controls, programme context, thresholds, and audit records. Bias, divergence, and drift signals are oversight prompts; they are not automatic disciplinary actions or final decisions.

No black-box reviewer scoringEvery signal requires an explanation trail and source-record linkage.
Human review gateSensitive findings require authorised human interpretation before operational consequences.
Privacy-aware disclosureReviewer-level detail is role-governed; committee exports can use aggregate or authorised views.

Institutional use cases

The IRI is relevant where review decisions must remain trusted across institutions, programmes, countries, reviewers, evidence types, or funding cycles. It supports ministries and national authorities, DFIs and development banks, climate funds and donors, auditors and fiduciary reviewers, programme management offices, and portfolio governance teams.

Minimum acceptance criteria

CriterionReviewer test
ReconstructabilityCan every IRI signal be traced to evidence, decision, rationale, reviewer, criteria, threshold, and timestamp?
ExplainabilityCan an authorised reviewer understand why a score, flag, or alert exists?
Authority boundaryDoes the IRI avoid making approvals, release decisions, statutory determinations, or MRV certifications?
Role governanceAre reviewer-level details visible only to authorised users?
Sovereign-safe exportDo exports respect jurisdiction, residency, programme, and disclosure rules?
Human oversightAre bias, divergence, and drift alerts routed to human review before operational consequence?

Whitepaper preview

The PDF includes the full whitepaper, diagrams, output register, minimum data schema, committee review questions, and glossary.

Companion public materials