Public-sector internal auditors are accountability actors whose written findings are increasingly exposed to proactive disclosure and freedom-of-information regimes. A 2020 survey of Canadian public-sector internal auditors found that perceived public disclosure requirements are widely seen as limiting the capacity to communicate audit findings, with stated concern associated with risk aversion, professional isolation, and a data-analytics orientation, and reduced among auditors who prefer an advisory role — but the evidence was cross-sectional and purely attitudinal. Whether disclosure requirements actually change audit behaviour and written reporting (gaming disclosure rules, omitting or softening adverse findings, shifting content to verbal channels, producing sanitized documents), what the causal structure of such chilling effects is, what disclosure design would mitigate it while preserving accountability, and whether the pattern holds beyond the surveyed Canadian population all remain empirically open. Later work by the same authors analysed the language of thousands of actual Canadian audit reports and showed that institutional reforms move reporting style, but did not test disclosure-induced chilling; no subsequent study has causally identified such effects or evaluated disclosure-design mitigations for them. The problem is to supply the missing behavioural, causal, and policy-design evidence.

Background

Internal audit is the function inside a government department or agency that independently reviews its operations, compliance, and performance and reports the findings to management and an audit committee; it is distinct from legislative auditors (in Canada, the Office of the Auditor General, a country's Supreme Audit Institution, which audits government on behalf of parliament). Internal auditors' value to accountability depends on candour: adverse findings must reach those who can act on them in written form.

Over the past two decades, governments have expanded transparency obligations — proactive disclosure, meaning the routine publication of documents such as audit reports without waiting for a request, and freedom-of-information (FOI) laws under which internal records can be demanded by the public. A substantial literature on the unintended consequences of transparency, anchored in work on blame avoidance in government and on the 'audit society', warns that officials who expect their written words to become public manage what they write: records may be sanitized, sensitive content moved to unrecorded verbal channels, or disclosure rules gamed. Whether this anticipated chilling effect — a deterrence of candid internal communication caused by the prospect of disclosure — actually occurs, and at what cost to accountability, is largely an open empirical question.

Liston-Heyes and Juillet (2020) provided the first profession-specific evidence. In a national online survey of Canadian public-sector internal auditors (459 responses from a frame of 1,500 Institute of Internal Auditors–registered auditors; 400 usable), they measured agreement, on a seven-point scale, with the statement that 'the public disclosure requirements of my organization limit my capacity to communicate findings effectively'; about 40% of respondents placed themselves on the concerned side of the scale. Ordered logistic regressions (a regression model for ordinal survey responses) showed that stated concern correlates positively with individual and organizational risk aversion, professional isolation, and a preference for a greater role for data analytics (automated analysis of large administrative datasets) in accountability processes, and negatively with a preference for an expanded advisory (consulting) role for the profession; professional certification and federal (as opposed to provincial or municipal) employment also predicted concern.

The design, however, was cross-sectional and attitudinal: it establishes correlations between self-reported attitudes and a self-reported concern, not behaviour. The authors therefore left a chain of questions open: whether the attitudes cause the concern (or vice versa, or both are produced by unobserved organizational factors); whether auditors who report concern actually change behaviour — gaming disclosure rules, omitting adverse findings from written reports, shifting content to verbal reporting, or producing sanitized documents; whether the pattern extends beyond the surveyed Canadian internal-audit population; what disclosure design would mitigate chilling while preserving accountability; and how technology adoption and the profession's shifting role shape these dynamics over time. Subsequent computational analyses of Canadian audit report language by the same authors have shown that reporting style responds to institutional incentives, which makes the behavioural questions tractable but leaves them unanswered for disclosure regimes specifically.

Problem Statement

The problem, left open by Liston-Heyes and Juillet (2020), is to establish — with evidence on observed behaviour rather than self-reported attitudes — whether and how public disclosure regimes chill the candour of public-sector internal audit, and what disclosure design would mitigate it while preserving accountability. This question encompasses four dimensions: (1) Behavioural reality: determine whether auditors and audit units whose work is subject to disclosure actually engage in the anticipated adaptive responses — gaming of disclosure rules, omission or softening of adverse findings in written reports, shifting content to verbal channels, production of sanitized documents — and how prevalent these responses are. (2) Causal structure: identify the direction and mechanism of the relationship between disclosure regimes, auditor attitudes (risk aversion, isolation, data-analytics orientation, advisory-role preference, certification), and reporting behaviour, separating disclosure effects from confounders such as organizational culture, formal independence structures, and concurrent reforms. (3) Mitigation: determine what disclosure design — the rules governing whether, when, and how internal audit records become public, and supporting measures — would mitigate chilling of internal-audit candour while preserving accountability. (4) Scope: determine whether these relationships hold beyond the surveyed Canadian internal-audit population — across jurisdictions, levels of government, and disclosure regimes. Whether public disclosure regimes actually chill internal-audit candour in observed behaviour, and what disclosure design would mitigate that chilling while preserving accountability, remains open; the existing cross-sectional, attitudinal evidence does not settle it.

References

  1. Liston-Heyes, C., & Juillet, L. (2020). Burdens of transparency: An analysis of public sector internal auditing. Public Administration, 98(3), 659–674. DOI: 10.1111/padm.12654. https://doi.org/10.1111/padm.12654
  2. Liston-Heyes, C., & Juillet, L. (2022). Institutional embeddedness and the language of accountability: Evidence from 20 years of Canadian public audit reports. Financial Accountability & Management, 38(4), 608–632. DOI: 10.1111/faam.12336. https://doi.org/10.1111/faam.12336
  3. Liston-Heyes, C., & Juillet, L. (2024). Does increasing auditors' independence lead to more forceful public auditing? A study of a Canadian internal audit reform. Public Administration, 102(3), 1045–1068. DOI: 10.1111/padm.12964. https://doi.org/10.1111/padm.12964
  4. Natan-Krup, D., & Mizrahi, S. (2024). Public Accountability and Bureaucratic Discretion: Why do Internal Auditors Stretch the Boundaries of Their Role? The American Review of Public Administration, 54(8), 699–716. DOI: 10.1177/02750740241265893. https://doi.org/10.1177/02750740241265893
  5. Hanskamp-Sebregts, M., Robben, P. B., Wollersheim, H., & Zegers, M. (2020). Transparency about internal audit results to reduce the supervisory burden: A qualitative study on the preconditions of sharing audit results. Health Policy, 124(2), 216–223. DOI: 10.1016/j.healthpol.2019.11.013. https://doi.org/10.1016/j.healthpol.2019.11.013
  6. Menke, W., Gomes, R., & Xavier, F. (2024). Impacts of AI-based anti-corruption audits on risk aversion in decision-making: a case study of the Brazilian ALICE tool. Global Public Policy and Governance, 4(3), 273–286. DOI: 10.1007/s43508-024-00098-1. https://doi.org/10.1007/s43508-024-00098-1
  7. Mikkelsen, K. S. (2026). How perceived red tape in freedom of information processing shapes public servants' trust in citizens: Evidence from Slovakia and Croatia. Government Information Quarterly, 43(2), 102135. DOI: 10.1016/j.giq.2026.102135. https://doi.org/10.1016/j.giq.2026.102135