Who Lies at Work, and Why

Part 4 of 8 in a series on lies and deception at work.

Most people rarely lie. A few prolific liars tell most of the lies, and targets and hierarchy do the rest. What the research says about who’s lying at work.

On 30 April 2008, a thousand American adults filled in an online survey. It asked them about packaged meals, then cat litter, then lying, then water softeners.

The lying question was simple: how many lies had they told in the past 24 hours? Kim Serota, Tim Levine and Franklin Boster published the answers in 2010. The average was 1.65 lies a day. That fits what most of us would guess, and it’s close to the one or two a day that has been quoted ever since Bella DePaulo’s diary studies in 1996.

The average was also nearly useless. Six in ten people said they hadn’t told a single lie. Half of all the lies came from about 5% of the people surveyed. Just 1% of them accounted for almost a quarter.

Photographers learn not to trust an average. Point a camera’s meter at a snowfield and it will make the snow grey, because it assumes every scene averages out to the middle. Lying works the same way. The average liar doesn’t exist. There’s a large honest majority, and a small group lying often enough to drag the average up for everyone else.

That gives this essay two halves. The first is about the small group: who they are, and what happens to them. The second is about the rest of us, and it’s less comfortable. Most people are honest most of the time. Give them a target they can’t quite hit, or a boss who doesn’t want bad news, and that changes.

The honest majority

The pattern isn’t an American quirk. In spring 2010 the Science Museum in London ran a survey on lying to promote its Who Am I? exhibition, and nearly 3,000 people in the UK answered. Serota and Levine reanalysed the data. They classed 9.7% of respondents as prolific liars.

The gap between the two groups isn’t subtle. Everyday liars reported just over one white lie a day and roughly one big lie a week. Prolific liars reported more than six white lies and nearly three big lies a day. For every big lie from an everyday liar, a prolific liar told about 19.

The same lopsided shape has turned up in South Korea and Japan, and in the Netherlands. The Dutch study surveyed more than 1,000 visitors to a science museum, aged from 6 to 77. It won an Ig Nobel Prize in 2016, more or less for asking a thousand liars how often they lie and then deciding whether to believe them.

It’s a good joke, and a fair question. All of this rests on liars reporting their own lies. When researchers have compared self-reported lying with more objective measures, the two have generally lined up well. Self-reports aren’t perfect, but they’re better than you’d expect.

The 91-day diary study I mentioned in the first essay tracked 632 people and more than 116,000 lies. About three-quarters of the people were consistently low-frequency liars.

I’m not above the small, kind sort of lie myself. I’ve told colleagues I liked their new haircut when I thought it was too short, or just too much of a change. At a company conference, someone asked me for feedback on their presentation. I told them it was good. It was full of holes. I just didn’t want to upset them.

Who the prolific liars are

We know less about them than we’d like. In the UK data, prolific liars tended to be younger and were more often men. They were also slightly more likely to work in senior roles, and the researchers stressed that these differences were modest. Age is the most consistent finding. The Dutch study found lying rises through childhood, peaks in adolescence and falls through adulthood.

Personality says more. In 1996 Deborah Kashy and Bella DePaulo gave 147 students and members of the public a dozen personality measures, then had them keep a diary of every social interaction and every lie for a week. The people who lied most scored higher on manipulativeness and cared more about the impression they made. They were also more sociable. The people who lied least scored higher on a measure called socialisation, and reported better friendships with people of their own sex. Socialisation roughly means how far someone has taken society’s rules on board.

A 2021 study of 305 Japanese university students found that those higher in Machiavellianism and psychopathy, two of the three “Dark Triad” traits, lied more often, apparently because they felt less guilt about it. Machiavellianism roughly means a readiness to manipulate people to get what you want. Narcissism, the third trait, went the other way, and the authors thought that might be specific to Japan.

The most useful idea here is a guess. Serota and his colleagues suggested that people whose lies show probably learn to avoid lying, so the ones who keep going are likely to be those with an unusually honest manner. It’s speculation, not a finding. But if it’s right, the prolific liar at work isn’t the nervous one. It’s the plausible one. That would also explain why the lie-spotting folklore in the first essay fails worst against the people we’d most like to catch.

I saw a version of this in the criminal courts. Some defendants were repeat offenders with a record of untruthfulness, and in earlier trials judges had allowed juries to hear about their bad character. On the face of it, the vast majority of them were convincing, even when their account went against what the evidence clearly proved. If they’d told the same story in a casual conversation, you’d mostly have had no reason to doubt them. A few did look and sound shifty in the same situation, but they were a minority. Criminal defendants are an extreme group, and I wouldn’t compare them with anyone’s colleagues. But they were the clearest example I’ve seen that lying isn’t spread evenly.

Workplaces have their own version. Many years ago I worked in a call centre with targets for all sorts of things. One of them was time on the phone, which was treated as evidence of productivity. One colleague was always near the top of that chart, yet his sales kept falling short of target. His explanation was bad luck: the area he’d been given couldn’t reliably use the service we were selling. On paper he was a good salesman. He’d passed every sales training course with high scores. Nobody could understand why someone so good, who spent so much time on the phone, sold so little. In the end someone senior listened to some of his calls, which were recorded for compliance. He’d been spending most of that time talking to his girlfriend. He was fired for misusing the company’s phone systems.

Lying a lot tends to catch up with people. The UK prolific liars were almost nine times as likely to say they’d been fired for dishonesty, and four times as likely to say lying had cost them a partner. They’re rarely caught in the moment. They’re caught when their stories pile up and start disagreeing with each other and with the record, which is the subject of the second essay.

People lie for a reason

If most lies come from a few people, the obvious policy is to find them and get rid of them. That’s half right at best.

Tim Levine argues that people don’t lie at random. They lie when the truth gets in the way of what they want. In 2010 he and two colleagues ran three experiments that varied whether the truth served people’s goals. When it did, people were honest. When small changes to the same situations made the truth a problem, around six in ten deceived.

That doesn’t make everyone a liar waiting for an excuse. Most people will pay a lot to stay honest. In 2019 the economists Johannes Abeler, Daniele Nosenzo and Collin Raymond combined 90 studies involving more than 44,000 people in 47 countries. Each used a version of the same test: report the result of a private die roll, or something like it, and get paid more for a better number. Nobody could check any individual. On average people gave up about three-quarters of what they could have made by lying. The authors concluded that the main motives were wanting to be honest and wanting to be seen as honest.

So honesty has a price, and for most people it’s high. It isn’t infinite. Workplaces are very good at raising the cost of telling the truth.

One short of the target

In 2004 Maurice Schweitzer, Lisa Ordóñez and Bambi Douma gave students a word task and let them mark their own work. Some were told to do their best. Others were given a specific target. The students with unmet targets were more likely to overstate their scores. That held whether or not money was attached, and it was strongest among people who had only just missed.

That last detail belongs on every manager’s wall. Being nowhere near a target doesn’t tempt people much. Being one short does.

It’s one lab study with students, so it deserves some caution. The real world has run larger versions.

For years Wells Fargo, one of America’s biggest banks, pushed its staff to sell as many products as possible to each customer. The internal slogan was “eight is great”. In 2016 regulators fined the bank $185 million after finding that staff had opened more than two million bank and credit card accounts that customers may not have authorised. About 5,300 employees had been fired over it. The estimate later rose to 3.5 million accounts. In 2020 the bank agreed to pay $3 billion to settle criminal and civil investigations, and the US Justice Department said the practices ran from 2002 to 2016. The regulator traced the problem to management’s aggressive sales expectations.

Five thousand three hundred is a lot of prolific liars for one bank. The likelier explanation is the one from the word task, scaled up: ordinary people, one short of a target, with their jobs at stake. When that many people tell the same kind of lie, the question to ask is what they were being asked to do.

I’ve watched a smaller version of this in sales meetings. Figures were rounded up as they were reported. Each rounding was small, but they were recorded, and over time they added up to large discrepancies. They were flagged when the reported sales didn’t match the company’s actual income.

I’ve also seen figures change between the team that produced them and the people who presented them, when a grade the organisation wanted was at stake. I’d delivered some of those reports myself, so the auditors were effectively being told I’d said things I hadn’t. It happened on more than one report in the same audit. How it came to light is in the second essay.

Why bad news travels slowly

Targets are one way a workplace raises the price of the truth. Hierarchy is another.

In 2003 Frances Milliken, Elizabeth Morrison and Patricia Hewlin interviewed 40 employees about the things they didn’t tell their bosses. Thirty-four of them said they had felt unable to raise an issue with a supervisor even though they thought it mattered. The most common reason was fear of being seen negatively and damaging relationships. A quarter felt that speaking up wouldn’t change anything.

Silence isn’t lying. It runs on the same sum, though. If telling your manager the project is late costs you more than keeping quiet, you keep quiet. When keeping quiet stops working, “it’s slightly behind” is a short step. If every layer of management does a little softening of its own, bad news can reach the top sounding like good news with a caveat.

I’ve watched it happen. Sales reports were softened on their way up. Department heads told directors that everyone in their team was happy and doing well, when morale in those departments was actually quite low.

Why some teams feel safe enough to deliver bad news and others don’t is a question for the sixth essay.

What power does

The view from the top is different. Senior people face fewer checks on what they say. Their reports go to people who depend on them, and the people best placed to spot a problem often work for them.

Power also decides whose story gets believed. I once walked a new senior colleague through a strategy I’d written and was running. Later I was called into the boss’s office and asked why we didn’t have one: my new colleague had said I’d never mentioned it. I sent the boss the strategy, with my name on it, and a colleague who’d been in the room when I explained it confirmed the conversation had happened. The boss was mainly annoyed that it had become one person’s word against another’s. The work was handed over anyway, and my new colleague was later praised for growth I’d built, which had levelled off since the handover.

The record settled who was telling the truth. It didn’t settle who got the credit.

The research on power and honesty is thinner than you’d hope. There’s the UK finding that prolific liars were slightly more likely to hold senior roles. In 2010 Joris Lammers, Diederik Stapel and Adam Galinsky published a set of experiments under the title “Power increases hypocrisy”. People made to feel powerful condemned other people’s cheating more, and cheated more themselves. The cheating was measured by how many lottery tickets people claimed to have won in a dice game. When people felt their power was illegitimate, the pattern reversed and they judged themselves more strictly than others.

That study comes with a warning label. Stapel, one of its authors, was later found to have faked data in dozens of other papers. He was caught by three junior researchers who spotted an entire row of scores appearing identically in two studies. Lammers has said all his work on power was cleared, and that he collected the data himself or with co-authors other than Stapel. It remains a set of small lab experiments, best held loosely.

Which brings me to the awkward part.

The honesty researchers

In 2012 five researchers published a tidy, useful finding. Ask people to sign a form at the top rather than the bottom, and they behave more honestly. One study used real data from a US car insurer. Customers who signed first reported driving more miles. Since more miles means a higher premium, that suggested they were being more honest.

In 2020 a replication attempt failed and the original data was posted. Researchers who asked to stay anonymous spotted problems and took them to the three academics behind the Data Colada blog, Uri Simonsohn, Leif Nelson and Joe Simmons. Half of the starting mileages in the spreadsheet were in the font Calibri and half in Cambria. Each Calibri figure had a close Cambria twin, always within 1,000 miles. Almost none of the mileages were rounded, although people reporting large numbers by hand tend to round them. The bloggers concluded there was very strong evidence the data was fabricated. The paper was retracted in 2021.

The file’s properties named Dan Ariely as its creator. Ariely, the Duke behavioural economist and author of the bestseller Predictably Irrational, has denied wrongdoing and suggested a research assistant or the insurance company may have been responsible. In 2024 he said a Duke investigation had found the data was falsified but no evidence that he knowingly used fake data. Duke declined to comment.

Another of the paper’s authors was Francesca Gino, a Harvard Business School professor and one of the best-known honesty researchers in the world. The Data Colada team alerted Harvard to evidence of fraud in four studies she co-authored, including a separate study in the same 2012 paper. Harvard’s investigation concluded that she had committed research misconduct. The papers were retracted. One of them had argued that dishonesty can lead to greater creativity. In May 2025 Harvard revoked her tenure, something it hadn’t done to any professor since it formalised its tenure rules eight decades earlier.

Gino denies the allegations. She sued Harvard and the three bloggers for $25 million. In 2024 a judge dismissed her claims against the bloggers. Harvard has since sued her, alleging she sent it a falsified dataset to prove she hadn’t committed fraud, which she also denies. Harvard has asked the judge to settle the case without a trial; a trial is scheduled for December 2026.

In September 2026 the journal Psychological Science retracted Ariely’s much-cited 2002 paper on procrastination and deadlines, written with Klaus Wertenbroch, after a failed replication and a Data Colada analysis. The bloggers found duplicated observations and values that should have been rounded but weren’t. Wertenbroch says he never had access to the data, and asked for the retraction. Ariely said the records he still had were insufficient to answer the questions raised.

This isn’t here to gloat. It’s here because it fits almost everything above.

It was a few people, not a whole field. Behavioural science is full of careful researchers, and some of them spent years unpicking this. The incentives were enormous: striking findings about honesty led to bestselling books, TED talks and consultancy. Gino was at one point the fifth highest-paid Harvard employee. A dull result leads nowhere. Status worked as a shield, too. The insurance data wasn’t made public until eight years after the paper appeared. And none of it was caught by anyone’s face or manner. It was caught by fonts, rounding and file properties, by people patient enough to open the data. For their trouble, they were named in a $25 million lawsuit.

So who lies at work?

Not most of your colleagues. Most people tell few lies, and will give up real money to avoid telling more. Believing them by default, as I argued in the first essay, is usually the right call.

A few people lie a great deal. They tend to be more manipulative and more concerned with how they come across. If Serota’s guess is right, they’re also better than average at seeming trustworthy. You won’t spot them by watching them. You find them the way lies usually come out: by noticing that their stories and the record keep disagreeing. Over time the record tends to win.

Even in the lopsided American figures, though, half the lies came from everyone else. A good share of the lying at work doesn’t come from liars at all. It comes from honest people in positions where the truth costs more than they can afford: one short of a target, or carrying bad news to someone who doesn’t want it. Hunting for liars is the smaller job. The bigger one is lowering the price of the truth. What it costs an organisation when it doesn’t is the subject of the seventh essay.

Personally, I think a workplace where people are afraid to tell the truth does more damage than the few prolific liars. A company has to be able to accept hard truths: the ones it doesn’t want to hear, but desperately needs to.

Sources and further reading

Abeler, J., Nosenzo, D. & Raymond, C. (2019). Preferences for truth-telling. Econometrica, 87, 1115–1153.

Ariely, D. & Wertenbroch, K. (2002). Procrastination, deadlines, and performance. Psychological Science, 13, 219–224. Retracted 2026.

Daiku, Y., Serota, K. B. & Levine, T. R. (2021). A few prolific liars in Japan. PLOS ONE, 16, e0249815.

Debey, E. et al. (2015). From junior to senior Pinocchio. Acta Psychologica, 160, 58–68.

DePaulo, B. M. et al. (1996). Lying in everyday life. Journal of Personality and Social Psychology, 70, 979–995.

Halevy, R., Shalvi, S. & Verschuere, B. (2014). Being honest about dishonesty. Human Communication Research, 40, 54–72.

Kashy, D. A. & DePaulo, B. M. (1996). Who lies? Journal of Personality and Social Psychology, 70, 1037–1051.

Lammers, J., Stapel, D. A. & Galinsky, A. D. (2010). Power increases hypocrisy. Psychological Science, 21, 737–744.

Levine, T. R. (2020). Duped: Truth-Default Theory and the Social Science of Lying and Deception. University of Alabama Press.

Levine, T. R., Kim, R. K. & Hamel, L. M. (2010). People lie for a reason. Communication Research Reports, 27, 271–285.

Milliken, F. J., Morrison, E. W. & Hewlin, P. F. (2003). An exploratory study of employee silence. Journal of Management Studies, 40, 1453–1476.

Schweitzer, M. E., Ordóñez, L. & Douma, B. (2004). Goal setting as a motivator of unethical behavior. Academy of Management Journal, 47, 422–432.

Serota, K. B., Levine, T. R. & Boster, F. J. (2010). The prevalence of lying in America. Human Communication Research, 36, 2–25.

Serota, K. B. & Levine, T. R. (2015). A few prolific liars. Journal of Language and Social Psychology, 34, 138–157.

Serota, K. B., Levine, T. R. & Docan-Morgan, T. (2022). Unpacking variation in lie prevalence. Communication Monographs, 89, 307–331.

Shu, L. L., Mazar, N., Gino, F., Ariely, D. & Bazerman, M. H. (2012). Signing at the beginning makes ethics salient. PNAS. Retracted 2021.

Simonsohn, U., Simmons, J. & Nelson, L. (2021). Evidence of fraud in an influential field experiment about dishonesty. Data Colada [98].