Evidence Ledger · Protocol candidate · reviewed 2026-09-04

Using behavioral science to improve compliance in government records management: Evidence from a large field experiment

The reviewed source can support a conservative practical protocol.

Brali decision: Protocol candidate. This page summarizes a reviewed evidence boundary; it does not reproduce the source and it does not turn one paper into a universal prescription.

What the source supports

In the studied Canadian government setting, a multi-component package combining a recurring calendar commitment, timely records-management guidance and behavioral framing reduced network-drive storage relative to status quo control, with effects persisting over subsequent months. This supports testing a bounded Brali protocol that reserves a recurring maintenance block and uses one organization-approved records action per session when retention and archival rules are already known.

What it does not establish

Ten minutes every week is the universally optimal records-maintenance dose.
A calendar reminder alone produced the observed effect.
Personal-benefit framing is superior to institutional or injunctive-norm framing.
The intervention proves employees deleted only records that should have been deleted.
Deleting old, duplicated or large files indiscriminately is safe.
The intervention has been shown to improve retrieval speed, reduce security incidents, increase productivity or improve information-access compliance directly.
The reported monetary savings generalize to other organizations.
The intervention is already validated for private-sector organizations or personal knowledge-management systems.
Generative AI or automation should automatically delete or archive records on the basis of this study.

Important limitations

The primary outcome was storage volume, an objective administrative measure but only a proxy for records-policy compliance; individual file decisions were not observed.
The intervention bundled pre-commitment, recurring reminders, weekly practical guidance and motivational framing, so the causal contribution of each component cannot be isolated.
The study took place in one Canadian provincial government, whose accountability structures, retention requirements, infrastructure and employment context may differ from private, nonprofit or personal settings.
Deidentified data prevented subgroup analysis by employee age, gender, tenure or occupation.
The unbalanced panel experienced attrition over time, although reported robustness checks did not materially change the findings.
The study did not test alternative maintenance frequencies or durations, so the 10-minute weekly schedule should not be treated as an optimized dose.
The study did not directly measure retrieval quality, data-breach risk, employee productivity or downstream information-access outcomes.
The treatment worked in an environment with mandatory records-management training and expert-approved retention guidance; applying cleanup actions without equivalent policy knowledge could create harm.

Source context

Using behavioral science to improve compliance in government records management: Evidence from a large field experiment

Source type: primary-study
Design: Pre-registered randomized field experiment using repeated administrative observations of employee network-drive storage from January through December 2021. Employees were randomized in June to an injunctive-norm intervention, a personal-incentive intervention, or status quo control; the control group received the injunctive-norm intervention later in October, providing an additional within-study replication pattern. The treatment was a bundle rather than a factorial test of individual components.
Population: 22,774 employees across 35 agencies in one Canadian provincial government. Employees used centrally managed network drives, allowing monthly storage measurement. Data were deidentified, so treatment-effect heterogeneity by age, gender, tenure or occupation could not be examined.
Exposure / intervention: An initial behaviorally framed email and calendar invitation. Accepting the invitation created a recurring 10-minute weekly records-management appointment for two months. Each reminder included one expert-approved practical tip, such as emptying the recycling bin, finding large duplicate files or moving sensitive files to secure government archives. Initial messaging emphasized either personal benefits or institutional/injunctive norms.
Outcomes: Monthly network-drive storage volume per employee; Persistence of storage reduction after intervention; Relative performance of personal-benefit versus injunctive-norm framing

Citation: Hopkins VC, Kormos C. Government Information Quarterly. 2026;43(2):102106.

DOI: 10.1016/j.giq.2026.102106

What this changes in Brali

Historical review target

records maintenance block

This identifier was used as an evidence-review target but was never published as a canonical Brali protocol. It is retained as provenance and deliberately does not link to a public route.

Editorial note: Protocol-design direction: reserve a short recurring records-maintenance appointment; at each session execute one approved, low-ambiguity action such as clearing disposable material, resolving duplicates or moving sensitive records to the correct archive. Treat 10 minutes weekly for two months as the tested implementation, not a universal prescription. Where retention rules are unclear, stop and resolve policy rather than deleting. Any future AI assistance should surface candidates and preserve human accountability for retention or deletion decisions.

Use the boundary, not just the headline

When an AI agent or a person retrieves this decision, preserve both the supported claim and the unsupported or overstated claims. Dropping the boundary would turn a reviewed source into a stronger claim than Brali actually maintains.

All Evidence Decisions → · Research & trends → · Machine-readable review →