Thanks to the Office of the Auditor-General for posting this blog on our last Leaders Integrity Forum.
At the final Transparency International New Zealand Public Sector Leaders Integrity Forum in July, senior public sector leaders gathered to discuss “groupthink” and how it can contribute to integrity failures in organisations. In this blog post, Melanie Webb (Assistant Auditor-General, Legal, Knowledge, and Practice) reflects on the discussion.
The forum was chaired by Karen Chang, Director of the Serious Fraud Office. We heard from Professor Paul 't Hart of Utrecht University, who has spent three decades studying how governments talk themselves into disasters, and from Dr Warren Young, formerly of the Independent Police Conduct Authority, on what this looks like in New Zealand.
In the early 1970s, the psychologist Irving Janis read a school essay his daughter had written about the 1961 Bay of Pigs invasion of Cuba. The invasion went ahead three months into President Kennedy’s first term and was defeated within days. What caught Janis's attention wasn't that the plan was bad. A bad plan is, as we know, not unusual. It was that the room had contained people who could see the plan was bad, and it went ahead anyway.
Janis called what happened in that room “groupthink”: the way a close-knit group’s pull towards agreement can override its members’ willingness to test what is in front of them.
Three ways a doubt disappears
Reading the accounts of people involved in the decision-making, Janis found three different things going on.
Some people had doubts and never voiced them. We all recognise this one. It is what we have built our integrity systems around: speak-up channels, disclosure protections, and assurances that raising something won't cost you.
Some people voiced their doubts, were brushed aside, and let it drop. Janis was struck by how little resistance this took. The dissenters didn't fight and lose. They mentioned their concern, felt the temperature of the room, and the discussion simply moved on. Nobody had to try particularly hard to shut them down, because they shut themselves down. Professor 't Hart noted that the pressure here is often gentle to the point of invisibility. It might take only a throwaway comment dressed up as friendly banter to move the conversation on to the next point.
Some people never had the doubt in the first place. This is the interesting one, because there was no concern to raise, no decision not to speak, and nothing for a speak-up process to catch.
Professor Paul 't Hart made a point that sat underneath everything else discussed in the forum: a group of clever people can be collectively… not clever. What a group knows is only ever what its members are willing to put on the table, and that is shaped by what each member thinks the other members are thinking. And in the case of the Bay of Pigs, people thought everyone else was thinking, that the President wanted to overthrow Castro and the plan to do so was a sure thing.
What it looks like from the inside
Dr Young brought this closer to home, describing what investigators found in a recent IPCA inquiry. Asked why they hadn't questioned what they were being told to do, people said it hadn't been their job to question it. When pressed, some went further: it hadn't been their job to think about whether it was being done properly.
This wasn’t about wilful inaction. They weren't saying they had noticed something was wrong and decided to do nothing. Instead, they were describing a boundary, real or perceived, around their role. This was something they were expected to do, not something they were expected to question. The inquiry found no bad faith. No corruption. Nobody knowingly protecting anybody. What Dr Young described instead was tunnel vision; people who didn't perceive that what was in front of them was wrong. These are Janis’s third group, sixty years later.
An organisation can have excellent speak-up processes and still miss this. Those processes help when someone has a concern. They cannot help when nobody thinks there is anything to be concerned about. And harder questions sit further upstream. What have we taught people to notice? What have we taught them to disregard? What have people learned to understand as “somebody else's job”? And what has become so familiar, accepted, or settled that nobody thinks to question it?
The tell
So how would you know? A group where doubts have stopped forming doesn't necessarily look like a group in trouble. It can look confident, cohesive, well aligned, and can appear to be performing well.
Professor 't Hart highlighted a warning sign in how a group responds to challenge. Groups where doubts have stopped forming often start to characterise the people who disagree with them. They might describe a critic as being unable to understand their operating environment. They might say that the media are looking for a story, or that a reviewer has applied a framework that doesn't fit their context.
Any of those things might be true. The tell isn't the content, it's the reflex; the instinct to explain the critic rather than consider the criticism.
Outside challenge might be the only way a group like this encounters a question it has stopped asking itself. If the first response is to dismiss the person raising it, the organisation has lost an opportunity to introduce doubt back into the discussion.
An answer from an unlikely place
Asked how organisations with strong hierarchies manage this, Professor 't Hart pointed to research on high reliability organisations. In fields like air traffic and electricity distribution, mistakes can be extremely expensive and sometimes catastrophic. These organisations have learned to make raising doubts and asking questions part of standard processes.
Aviation offers a familiar example. Crew resource management developed after investigations into accidents in which experienced flight crews had failed to act on information that was available to them. Part of the response was to establish that pilots were expected to question a captain's decision when they had concerns about what was happening, even though the captain remained responsible for the flight. The industry developed agreed ways to raise and escalate concerns. Questioning became part of the process rather than something left entirely to individual judgement.
The analogy has obvious limits, not least that a public organisation is not an Airbus A380 (even if most of us have experienced the particular joy of 'building the plane while flying it'). Many of our decisions play out over years rather than minutes. Many of our problems are not clear-cut, and there might never be a moment afterwards that definitively settles whether a particular concern was warranted. But the underlying idea carries across. Asking questions and raising doubts can become part of the ordinary way an organisation works, rather than depending on an individual finding the confidence to speak up at exactly the right moment.
Same people, different room
Eighteen months after the Bay of Pigs, President Kennedy and many of the same advisors worked through the Cuban Missile Crisis. This time, the process was very different.
Kennedy tasked his advisors with identifying as many different courses of action as possible. Someone was given the specific role of arguing the opposing case. The group was split so that rival options were developed and tested against each other rather than converging early. And importantly, Kennedy stayed away from some of the sessions, so that discussions weren’t shaped by his views.
At the Bay of Pigs, the room had converged on what it assumed the President wanted. This time, he took that assumption out of the room.
Sometimes the most useful thing a leader can do is create some distance. If the room already knows what the most senior person thinks before anyone else has spoken, a later invitation to disagree may have less effect than we would like. That might mean leaving a team to work through an issue without you, seeking separate advice from people likely to see it differently, and being willing to take seriously a recommendation you had no hand in shaping.
The point is to create conditions in which people can work out what they think before they have to respond to what the most senior person thinks.
Working backwards
Janis’s three groups need different things, and those things need to work together.
People who have doubts but keep them to themselves need mechanisms and protections. This group need safe ways to speak up, and most organisations have invested in those.
People who raise a doubt and let it drop or are shut down need to see and experience a culture where challenge is treated as useful rather than disruptive.
The third group, where the doubt never forms, requires clear role and process expectations. People need to understand that questioning how things are done is part of their job. That means building questioning into way the work is done, as aviation has. It also means running discussions the way Kennedy did during the Cuban Missile Crisis.
One way to test how your organisation is doing on that front is to take the last genuinely awkward challenge it received and work backwards. Who raised it? What happened when they did? Was the substance of the challenge discussed, or did the conversation turn to the person raising it? Did anyone change their mind? And if the challenge came from outside the organisation, how was it received? You could do the same with a decision that later turned out to be wrong. Not to find who was at fault, but to ask what questions were asked at the time, what assumptions went untested, and whether anyone had the opportunity to see the problem differently.
The people involved in the Bay of Pigs decision were not short of intelligence. What changed eighteen months later was how they approached the decision. Groupthink is not something an organisation decides to do. It is what is left when nobody has made arrangements for the awkward question to be asked.
A farewell to the Leaders Integrity Forum
This was the final Public Sector Leaders Integrity Forum in its current form. Over the last 10 years, these sessions have brought people together from across the public sector to talk candidly about integrity in a setting that made those conversations possible.
The Office of the Auditor General thanked Transparency International New Zealand, and in particular Anne Gilbert, for creating and sustaining the space, and to all the speakers who have shared their time, experience, and wisdom.
John Ryan (TINZ Patron and former Auditor-General) and Sir Peter Boshier (former Ombudsman and current TINZ Director) are working together to craft a new version of the forum - occurring less often perhaps but with more impact. There will be more to hear about this in 2027.
This article has been reproduced with minor changes from the OAG website.
