The Lie the Organisation Tells

Part 6 of 8 in a series on lies and deception at work.

Why do organisations tell lies their own staff can see straight through? From Challenger to the Post Office, the people doing the work usually know first.

In 1986 the physicist Richard Feynman looked into how likely NASA thought it was that a space shuttle would be lost on any one flight. The estimates he found ranged from about 1 in 100 to 1 in 100,000. The working engineers gave the first figure. Management gave the second.

Feynman spelled out what the second number meant. You could launch a shuttle every day for 300 years and expect to lose one. He was asking because Challenger had broken apart 73 seconds after launch that January, on the programme’s 25th flight, killing all seven people on board. In his report he asked what lay behind management’s “fantastic faith in the machinery”.

Nobody had to stand up and lie for that second number to exist. It was built up over years, out of optimistic assumptions and flights that had got away with it, by people who mostly believed it. The people doing the work knew better. The people furthest from the work believed it most.

That gap is what this essay is about. The earlier essays in this series were about individual people and their lies. This one is about the lie the organisation tells. It often has no single author, and it’s usually obvious to everyone except the people telling it.

One place I worked had a sales process. It was meant to be the very heart of the company’s values, and everyone was supposed to follow it. Day to day, most people didn’t. Management reported that everyone was doing it right and the process was being followed. The truth was the opposite. It came out when mystery shoppers visited the store and head office sent back feedback that the process wasn’t being followed. Anyone on the shop floor could have told them that.

A lie with no author

When we picture a lie, we picture a liar: one person who knows the truth and says something else. Institutional lies rarely work like that.

In May 2024 Sir Brian Langstaff published the final report of the Infected Blood Inquiry. It ran to 2,527 pages on how more than 30,000 people in the UK were infected with HIV and hepatitis through NHS treatment, and how more than 3,000 of them died. He asked whether there had been a cover-up. His answer was yes, but not a small group plotting together. It was something “more subtle, more pervasive and more chilling”. Some documents were deliberately destroyed. The truth was hidden for decades.

One phrase runs through his findings: lines to take. In government, a line to take is an agreed form of words for answering a question. It’s written once, then repeated by ministers, officials and press officers who may never have checked it. For years the line was that people infected through contaminated blood products had received the best treatment available at the time. Margaret Thatcher used it in 1989 when turning down calls for compensation. The Inquiry said the line was wrong and that using it was unacceptable.

This is how most institutional lies are built. The people who know the truth don’t say the words. The people who say the words don’t know the truth. Everyone in the chain can feel honest, because each of them is only passing on what they were given.

Diane Vaughan, the sociologist who wrote the definitive study of the Challenger decision, called one version of this structural secrecy. An organisation divides what it knows between departments, layers and specialisms until nobody holds the whole picture.

It was fine last time

Vaughan’s book, The Challenger Launch Decision, came out in 1996. What she found was more ordinary than rule-breaking, and more worrying.

The seals in the shuttle’s booster rockets, the O-rings, had been damaged on earlier flights. It was damage that shouldn’t have happened at all. Each time it was noticed, discussed and then accepted, because the shuttle had come back safely. The next bit of damage was judged against the last one rather than against the original standard.

Vaughan called this the normalisation of deviance: an unacceptable practice gradually becomes acceptable because nothing bad has happened yet. Feynman noticed the same drift. Safety criteria, he wrote, grew gradually less strict, and having flown a risk before without failure was taken as an argument for flying it again.

The night before the launch, engineers at Morton Thiokol, the company that made the boosters, recommended against launching in the forecast cold. Their managers went into a private discussion. According to the evidence later given to the presidential commission, Jerry Mason, a senior vice-president, asked Robert Lund, the head of engineering, to take off his engineering hat and put on his management hat. The recommendation was reversed.

Vaughan concluded that nobody had broken the rules. The decision came out of the organisation working normally. That’s the uncomfortable part. You don’t need villains for an institution to end up saying something untrue. You need a system that rewards good news, and enough time.

Workplaces have their own O-rings.

The helpful workaround

Normalisation has a friendly face, and it’s often your best colleague.

In a study published in 2003, Anita Tucker and Amy Edmondson followed nurses in nine American hospitals and recorded the problems that got in the way of their work: missing supplies, wrong information, broken equipment. The nurses almost always fixed the problem in front of them and carried on. They found the missing item, borrowed from another ward, made it work. Hardly anyone reported the problem so that it could be fixed for good.

Patients were looked after, so in one sense this was good work. But the same problems came back shift after shift, and managers never saw them, because the nurses had absorbed them. The researchers described an illusory equilibrium. Things looked stable from above because the people below were quietly paying for it. The nurses held up as ideal were often the ones best at working round problems without complaint.

Nobody in those hospitals was lying. Yet the picture at the top was false, and it was false because of the effort of the people at the bottom. Every organisation runs on some of this. The danger comes when the gap between the picture and the reality grows wide, and the people who could close it have learned not to try.

Why the top hears least

In the fourth essay I looked at why individuals keep bad news from their bosses. The same thing happens across a whole organisation, and it has its own causes.

In 2000 Elizabeth Morrison and Frances Milliken argued that silence at work is often collective. It isn’t simply lots of people separately deciding to keep quiet. It’s a climate, and they traced it to beliefs managers often hold without saying so, such as that staff are mostly out for themselves and that management knows best. They also pointed to research showing that criticism from below is judged less accurate and less legitimate than criticism from above. Later they summed it up in one image: a chief executive with no clothes, whose nakedness everyone can see, being complimented on his outfit.

In 2014 Nathanael Fast, Ethan Burris and Caroline Bartel found a sharper version. Managers who doubted their own ability to do the job asked their teams for fewer ideas and judged people who spoke up more harshly. In an experiment, making managers feel less capable was enough to make them less willing to hear suggestions. The researchers put it down to ego defensiveness. The managers who most needed bad news were the least able to take it.

The cleanest figures can come from the most frightened team

The fourth essay left a question open: why do some teams feel safe enough to deliver bad news when others don’t? The best answer comes from a result that surprised the researcher.

In the mid-1990s Amy Edmondson, then a doctoral student at Harvard, set out to show that well-led hospital teams made fewer drug errors. She surveyed nurses on eight hospital units about their managers and colleagues, while nurse investigators tracked the errors. The best-led units, with the best working relationships, had higher error rates, not lower.

A second researcher, who hadn’t been told her findings, went and observed the units. Where the most errors were reported, nurses said mistakes were normal to write down and that they were never afraid to tell the nurse manager. Where the fewest were reported, nurses described an unforgiving place where “heads will roll”. The better teams weren’t making more mistakes. They were admitting more.

Edmondson went on to call the difference psychological safety: a shared belief in a team that it’s safe to take interpersonal risks, like admitting a mistake or disagreeing with the boss. She set the idea out in 1999, and it has since become one of the most studied ideas in management.

For this series, her first finding matters most. An error report measures the errors. It also measures how safe it is to report them. A department whose figures look spotless might be the best in the building. It might just as easily be the most frightened. From the top, the two look identical.

I’ve worked in both kinds of place. In one job, my manager’s view was that as long as you were acting in good faith, a mistake wasn’t a problem. You could own up and learn from it.

In another, the opposite was true. I forgot to run a backup. Nobody knew until someone accidentally deleted some files from the server and we couldn’t recover the latest versions. The newest backup we had was a week old. We restored it, and a week’s work was lost. It was treated as something to punish, not something to learn from. I was given a stern talking-to.

A telling-off deals with one mistake. What it teaches everyone watching is to keep the next one quiet.

A caution about a number you’ll find in almost every talk on this subject. The “iceberg of ignorance” says frontline staff know about 100% of an organisation’s problems, supervisors 74%, middle managers 9% and senior leaders 4%. It’s usually credited to a consultant called Sidney Yoshida and dated to 1989, but the study behind it is hard to find, and the people quoting it rarely cite one. The idea may well be right. The exact figures are a line to take.

The awareness asymmetry paradox

Everything so far explains how an organisation can end up telling a lie. None of it explains the strangest thing about these lies, which is how visible they are.

I call it the awareness asymmetry paradox. The people at the top believe the lie is contained. The people underneath can see straight through it. The ones who know most about the lie know least about whether it’s working.

The people underneath usually see it for plain reasons. They do the work, so they’re the first to notice when results don’t match what’s being reported. They hear the evasive answer at the team meeting and see the policy that changes without a reason. And they talk to each other. As the fifth essay showed, the grapevine is usually accurate about working life, and it pools what each person has noticed into a picture nobody could put together alone.

There’s a neat reversal in the research on individual liars. In 1998 Thomas Gilovich, Kenneth Savitsky and Victoria Medvec had people tell lies in front of an audience, then guess how many of the audience had seen through them. The liars consistently overestimated it. The researchers called this the illusion of transparency: we feel our inner state is leaking out when mostly it isn’t. As the first essay explained, they had less to worry about than they thought.

Organisations make the opposite mistake. The individual liar assumes everyone can see. The organisation assumes no one can. The individual feels the lie from the inside, as a racing pulse and a hot face. An institutional lie is spread so thinly across so many people that nobody feels it. And the one signal that might warn the top, people below saying so, is exactly what a climate of silence switches off. So silence gets mistaken for belief.

This is where the Emperor’s New Clothes needs updating. In Hans Christian Andersen’s story, everyone pretends because each person fears they’re the only one who can’t see the cloth. At work it’s the other way round. On the floor, people know that everyone knows. They’ve talked about it at the coffee machine. It’s the people at the top who are on their own, reading a quiet room as a convinced one.

I saw a version of this in the criminal courts, which I described in the second essay. Defence witnesses stuck to a story after CCTV had already disproved it, because they’d waited outside the courtroom and hadn’t seen the footage. A leadership team defending an institutional lie is often in the same position, defending a story it doesn’t know has already collapsed. The difference is that the witnesses couldn’t have known. Management usually could, if anyone had felt safe enough to tell them.

Sometimes the view from the ground is written down and then removed on its way up. After the Hillsborough disaster in 1989, in which 97 Liverpool supporters died, senior South Yorkshire Police officers vetted statements written by junior officers before they went to the official investigations. In 2012 the Hillsborough Independent Panel found that 164 statements had been significantly altered, 116 of them to remove or change comments critical of the police. One constable had written that there didn’t appear to be any leadership. That sentence was deleted.

The insult

Of all the damage an institutional lie does, I think the worst is what it tells staff about how they’re seen.

A lie told to people who can see through it is more than a lie. It’s a statement that the teller thinks they won’t notice. Staff can put up with a lot: bad news, cuts, mistakes, even an employer that falls short of its own values. What’s harder to forgive is being treated as if they can’t count.

In 2017 Bishop James Jones wrote up the experiences of the Hillsborough families. He gave his report a title that has stuck: The Patronising Disposition of Unaccountable Power. The families had been told a false story about how their relatives died, one that blamed the fans. Then they spent decades being managed and talked down to by institutions that assumed they would never find out. Patronising is the right word for this. It assumes the other person can’t see.

In the audit I described in the second essay, figures were changed between the team that produced them and the auditors. The people who changed them were editing data that came from a team who knew exactly what it said, and who had the records to prove it.

It felt as though the IT team were seen as a room full of geeks who dealt with error messages. It bothered all of us. They’d assumed we wouldn’t notice, which meant they didn’t really understand what we did.

The Post Office

Most of what this essay describes turns up in one case.

From 1999 the Post Office rolled out Horizon, an accounting system built by ICL, a company owned by Fujitsu, across its branches. Sub-postmasters began reporting shortfalls they couldn’t explain. Many were told they were the only one having problems, a lie I looked at in the fifth essay. The Post Office’s position was that Horizon was robust and its figures could be relied on. Between 1999 and 2015 around a thousand people were wrongly prosecuted and convicted.

In July 2025 the first volume of Sir Wyn Williams’s final report found that a number of senior people at the Post Office knew the original system could produce errors. Yet for its whole working life, the Post Office “maintained the fiction that its data was always accurate”.

The awareness asymmetry in this story is almost total. The sub-postmasters knew their own tills. Fujitsu’s support staff knew they could get into branch accounts remotely, and one of them, Richard Roll, said so on BBC Panorama in 2015. Second Sight, the forensic accountants brought in to investigate, told MPs that Fujitsu had told them in 2013 that remote access was possible and was happening. Computer Weekly had been reporting the sub-postmasters’ stories since 2009.

In 2015, the year of the Panorama programme, the Post Office told MPs in written evidence that neither a branch, the Post Office nor Fujitsu could edit or remove transaction data once it had been recorded. It gave ground only when it had to. By 2019, when more than 500 sub-postmasters were suing it in the High Court, it had accepted that Fujitsu could change branch data remotely. In December 2019 Mr Justice Fraser described the Post Office’s defence of Horizon as “the 21st century equivalent of maintaining that the earth is flat”.

Look at who knew. The people at the counter. The engineers in the support centre. The investigators. The trade press. The people still insisting the data was always accurate were the ones furthest from it.

I’ve worked in IT, and it wouldn’t have surprised me to learn that a support team could get into branch systems remotely. I’d have been shocked if they couldn’t.

The cost is still being counted. The Inquiry found a real possibility that 13 people took their own lives because of what they went through. The remaining volumes of the report, on how it happened and who knew what, are still to be published.

Exit, voice and loyalty

People who can see through the story have choices, and an economist named them more than 50 years ago.

In 1970 Albert Hirschman published Exit, Voice, and Loyalty, a short book about how people respond when a firm, organisation or state starts to decline. They can leave, which he called exit. They can complain and push for change, which he called voice. Loyalty is what keeps people in long enough to try voice instead of simply leaving.

Voice is expensive. The fourth essay looked at why individuals keep quiet, and at the Post Office the people who said the system was wrong were treated as the problem. Exit is quiet. To the organisation it looks like ordinary staff turnover. Nobody’s exit interview says they left because they could see the story wasn’t true and management couldn’t.

Hirschman also pointed out that the people most sensitive to decline can be the first to leave, taking with them the complaints the organisation most needed to hear. Carry that forward, and an organisation that keeps telling a story its staff can see through will slowly lose the people who can’t live with it and keep the ones who can. Over time, the asymmetry gets worse.

With the audit, I used both: voice first, then exit. The IT team reported what we’d found to the IT director. Then I left.

Making the lie expensive

The law has started to treat institutional honesty as a duty rather than a virtue. Since November 2014, following the inquiry into Mid Staffordshire hospital, NHS bodies in England have been under a statutory duty of candour. When something goes wrong and a patient is harmed, they have to tell the patient or their family and apologise.

The Hillsborough families have campaigned for years for the same principle across public life. The Public Office (Accountability) Bill, known as the Hillsborough Law, would put public authorities and officials under a legal duty of candour and create an offence of misleading the public. It passed the Commons in July 2026 and is now in the House of Lords.

Neither reaches most private employers. Most workplaces rely on something harder to legislate for: whether the people at the top actually want to know.

Edmondson’s research suggests what that looks like. Leaders who talk about their own mistakes and ask for bad news get more of it. A rising count of reported problems can be a sign of a team getting healthier, and should be treated that way. And a figure is only as good as someone’s ability to check it against the work it describes. That’s why the record mattered so much in the second essay.

Feynman ended his report by saying that for a successful technology, reality has to come before public relations, because “nature cannot be fooled”.

If I could say one thing to a senior team who think their staff don’t know, it would be that they do. With that audit, I wasn’t the only one who could tell something was wrong. The results didn’t match what people saw every day, so rumours went round the company, and staff would casually ask the IT team what was going on. We didn’t say anything. They already knew enough to ask.

The shuttle flew 135 missions. Two were lost: Challenger in 1986 and Columbia in 2003. That’s about one in 68. The engineers had been far closer to the truth than their managers. The people doing the work usually are.

Sources and further reading

Andersen, H. C. (1837). The Emperor’s New Clothes.

Bates v Post Office Ltd (No 6: Horizon Issues) [2019] EWHC 3408 (QB).

Edmondson, A. C. (1996). Learning from mistakes is easier said than done. Journal of Applied Behavioral Science, 32, 5–28.

Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44, 350–383.

Edmondson, A. C. (2018). The Fearless Organization. Wiley.

Fast, N. J., Burris, E. R. & Bartel, C. A. (2014). Managing to stay in the dark. Academy of Management Journal, 57, 1013–1034.

Feynman, R. P. (1986). Personal observations on the reliability of the Shuttle. Appendix F, Report of the Presidential Commission on the Space Shuttle Challenger Accident.

Gilovich, T., Savitsky, K. & Medvec, V. H. (1998). The illusion of transparency. Journal of Personality and Social Psychology, 75, 332–346.

Hillsborough Independent Panel (2012). Hillsborough: The Report of the Hillsborough Independent Panel.

Hirschman, A. O. (1970). Exit, Voice, and Loyalty. Harvard University Press.

Infected Blood Inquiry (2024). The Report. Chair: Sir Brian Langstaff.

Jones, J. (2017). ‘The Patronising Disposition of Unaccountable Power’. Home Office.

Morrison, E. W. & Milliken, F. J. (2000). Organizational silence. Academy of Management Review, 25, 706–725.

Post Office Horizon IT Inquiry (2025). Final Report, Volume 1. Chair: Sir Wyn Williams.

Public Office (Accountability) Bill, UK Parliament.

Tucker, A. L. & Edmondson, A. C. (2003). Why hospitals don’t learn from failures. California Management Review, 45, 55–72.

Vaughan, D. (1996). The Challenger Launch Decision. University of Chicago Press.