A few weeks before an audit by the supervisory authority, management sat down with everyone who would face the auditors and drilled into them not to say anything “wrong”. Nobody was asked to lie. “Wrong” was whatever is true and harmful anyway. The rest of the meeting assigned roles and settled the agreed lines for the sensitive topics. This was theatre, and elaborate theatre at that: weeks of preparation so that an authority would see a state of affairs that does not exist.
Another audit, same counterpart: the regulator. This time the audited organisation laid its known weaknesses openly on the table and asked the auditors to write findings about them. The auditors, used to probing and digging to get at results of that kind, were visibly surprised. The reason was pragmatic: these problems had resisted internal resolution for a long time. What had been stuck in the line for months suddenly moved under the pressure of an official finding. This audit was not passed in the sense of a spotless record. It was used.
I have experienced both scenes myself, on the audited side of the table. The same instrument, two attitudes. “Audit theatre” is a popular accusation, but those who raise it usually describe only the first scene and blame the tool for it. That is a convenient diagnosis, and it is wrong.
The Critique Has a Point
The critique of auditing has substance. Michael Power described audits as “rituals of verification” back in 1997: they produce a credible image of order, not a measurement of safety. What audits structurally cannot deliver, and what therefore has to stand beside them, is something I laid out here in June (“What Happens Between Audits”). None of that has changed.
But “audits have limits” does not add up to “audits are theatre”. Theatre is not a property of the instrument. It is a way of using it. And it arises fairly precisely where an organisation treats the audit as an exam to be passed rather than an outside view to be used.
How the Theatre Arises
The exam mindset is not a character flaw; it is rational. When a failed audit triggers questions, reports and justifications up the line, and in the extreme case puts the licence to operate at risk, a sensible organisation prepares for the exam the way one prepares for an exam: it learns the answers to the expected questions. The documentation is groomed for the reading auditor, the interviews are rehearsed, the site tour follows the polished route. Every single step makes sense. The sum is a staging that hides exactly what an audit could see.
Two features of the inflated audit make this worse. The first is the ambition of full coverage: whoever wants to “cover” thirty processes in four days checks at checklist speed, and checklist speed produces confirmation, not surprise. The second is the scope: the people questioned are the ones doing the work. The decisions that shaped their working conditions (budgets, staffing levels, deadline pressure) lie outside the audit’s perimeter. An audit that questions only the people executing the work and spares the decision-making level will reliably find deviations among those with the least say over their causes.
Theatre is not a property of the audit. It is the way an organisation uses it when it wants to pass rather than to see.
What a Good Audit Does Differently
A good audit is, first of all, narrow. It gives up the ambition to capture everything and takes on a few things properly instead: three processes in depth rather than thirty at the surface. An honest sample can surprise; full coverage can only conclude.
It observes instead of merely asking. An interview answer can be rehearsed; an observation cannot. Watching the actual work reveals the gap between work-as-imagined and work-as-done. Erik Hollnagel has shown that precisely this gap is the interesting finding: where practice deviates from paper, practice usually has a reason.
And it audits the decision-making level too. The most uncomfortable question of a good audit is not addressed to the specialist whose handling deviates from the procedure, but to the level whose planning made that deviation necessary. A single finding that lands there moves more than ten at the front line.
I saw this work most clearly at an internal audit with a mixed team: one person from the internal audit function, two externals from partner organisations. Auditors who knew the same work from their own operations came not as an inspection body but as an imported outside perspective. What they brought were not verdicts, they were inputs: comparisons, ideas, blind spots, and they measurably moved the audited unit forward. The best findings of that audit were gifts.
The Finding as Leverage
Back to the second scene from the beginning. A finding creates an obligation that internal papers cannot create. The same weakness, documented in meeting minutes for months, acquires a deadline, a budget and an escalation line the moment it becomes an official finding. One can object that this too is staging, just with the sign reversed: again the auditor is shown a prepared play. The objection has a point. But it overlooks the direction in which truth flows. The theatre withholds what is true from the audit and distorts its picture. Ordering findings supplies it with truths it would hardly have found on its own, and sharpens its picture. If that is misuse, it is the kind every supervisory authority should wish for. An organisation that has understood this stops fearing audits and starts commissioning them. It puts the topics where it needs pressure on the auditor’s table and lets the audit deliver the obligation it cannot create internally. The tool has a handle. Nobody forces you to grab it by the blade.
Even the good audit remains a snapshot, and what happens between audits still needs the ongoing learning practice the June essay was about. But the difference between theatre and tool is not decided in the audit report. It is decided in the attitude with which an organisation meets the exam, weeks before anyone writes a finding. One organisation trains not to say anything “wrong”, and passes. The other says what is true and gets something for it.
Sources
- Michael Power – The Audit Society: Rituals of Verification, Oxford University Press 1997
- Erik Hollnagel – Safety-I and Safety-II: The Past and Future of Safety Management, Ashgate 2014