The Toxic-Employee Statistic Nobody Can Actually Source | The Bad Apple Effect

Nobody could point to the moment it started. One member of a team, nothing dramatic, no shouting, no complaint. Just someone who did less than their share, who was always first to say an idea wouldn't work. Within months, the best people on that team were quietly heading for the door. For most of the 20th century, organisational psychology assumed a good team could absorb one difficult member without much cost.

In 2006, a study put that assumption to the test. What's happened to that finding since is what this video is actually about. Thank Here's an assumption worth calling out, because understanding the main point of this video depends on you knowing about it. We assume groups are resilient, that the combined weight of everyone else at the table will absorb, dilute or quietly correct one difficult member.

That assumption has real pedigree. Decades of research on group dynamics show individuals adjusting toward the norms of the group they're in, not the other way around. If that's the whole story, one difficult colleague should be a minor irritant, smoothed over by the much larger pull of everybody else's good nature.

The bad apple research asks a sharper question. What if, under the right conditions, the arrow points the other way? What if one person's persistent negativity predicts what happens to the group better than everyone else's disposition combined? The paper usually credited with putting this on a scientific footing is Phelps, Mitchell and Byington's 2006 review, How, When and Why Bad Apple Spoiled the Barrel, in the peer-reviewed annual volume of Research in Organisational Behaviour.

It's worth being precise about what kind of paper this actually is, because accuracy matters for what can honestly be claimed from it. It's a review and a theoretical model. Phelps and his colleagues synthesized roughly two dozen existing studies into a single framework. It wasn't a brand new experiment with its own reported result.

Their definition is the paper's real contribution. A negative group member on this model persistently does one or more of three things. They withhold effort, so the workload shifts onto everyone else. They express negative effect, chronic pessimism, complaint, hostility. Or they violate the group's basic norms of respect.

Those three behaviours generate specific reactions in teammates. A sense of unfairness, falling trust, an unconscious drift toward matching the negative member's own mood, just to avoid feeling exploited by comparison. From the outside, the result looks like a team that simply isn't very good. The actual cause is the sustained negative effect of just one person.

Thank you. Thank Now, here's the part almost every article on this topic gets slightly wrong, and I want to slow down on it because getting it right is why this channel exists. The number you've probably heard is that a single bad apple drops a team's performance by 30-40%. It comes from a follow-up experiment Phelps ran.

A trained actor, Nick, secretly planted in 44-person teams, playing the jerk, the slacker or the depressive pessimist. The qualitative story checks out completely. Negative behaviour spread. Teams argued more and shared less, and one team with an unusually skilled member neutralised Nick entirely. What doesn't check out is the evidence for this number.

A search was conducted for a peer-reviewed publication supporting the 30-40% figure. It isn't there. It lives in a 2008 public radio interview, and in every blog and business book that's copied it since, never in a citable dataset with a stated method. That's not a reason to dismiss the finding. It's a researcher's own consistent, detailed account of the study he designed and ran, and the mechanism behind it is independently proven, which is where we're going next.

But if you've quoted a 30-40% number as an established statistic, you are quoting an interview, not a paper, so attribute it to Phelps' own reporting, not to a study. With If the number's shaky, why does the underlying idea hold up? Because a separate line of research proves the actual mechanism, twice over, both in the lab and in the field.

Seagal-Barsad's 2008 study Embedded a Trained Confederate coached to display one of four moods into small groups completing a realistic management exercise. Because the mood was assigned experimentally, the result, participants' moods measurably shifting to match it, confirmed by outside coders who didn't know the condition, could only be contagion, not some third factor explaining why difficult people or difficult groups turn up together.

Robinson and O'Leary Kelly's 1998 field study found the same pattern outside the lab, across 187 employees in 20 real organisations. People whose co-workers behaved antisocially became significantly more likely to behave the same way themselves, independent of their own baseline disposition. The most striking part of Phelps' own Confederate study points at the same mechanism.

Teammates didn't just tolerate the negative behaviour, they began, without noticing, adopting it, more critical when paired with the jerk, effort quietly dropping alongside the slacker. Negative mood also has a head start. It's stickier, more attention-grabbing and physiologically longer-lasting than positive mood.

One reliably negative person and eight positive ones will not settle near mostly positive. The negative signal outweighs its numerical share of the room. Most of this behaviour, in most people, is circumstantial. A bad quarter, burnout or a genuinely dysfunctional environment that provokes a reasonable person into unreasonable conduct.

But a smaller, more consequential group displays this pattern chronically, largely independent of circumstance. The relevant construct here is the dark triad. Machiavellianism, narcissism and subclinical psychopathy, with everyday sadism now generally added as a fourth trait in the dark triad. The best population estimate for subclinical psychopathy from a representative household survey of Great Britain puts it at 0.6% of the general public.

That rate doesn't hold in the corporate world. Babiak, Neumann and Hare's landmark study of 203 corporate professionals, people their own companies had selected for management development, found 3.9% scoring above the clinical threshold and a further 4.4% close behind it. Roughly six to seven times the general population rate, concentrated specifically among people already judged worth investing in.

Separate survey work finds that somewhere between 1 in 20 and 1 in 8 employees report having worked directly with someone they perceived as psychopathic, a wide range and one worth treating with caution rather than repeating as a precise figure. That spread is itself informative. It reflects genuine disagreement in the field about where to draw the line between difficult and psychopathic using lay judgment.

This is exactly why a single decimal point version of this figure deserves more scepticism than the confidence it's usually repeated with. It refers Here's the finding that should change how we think about this entirely, and it comes from the largest study ever done on the subject. In 2025, Lenky Roth and Uta Christina Kleha published a meta-analysis in the Journal of Applied Psychology, pooling 166 independent samples and over 49,000 employees.

Higher psychopathic traits reliably predicted lower task performance, less cooperative behaviour and substantially more counterproductive conduct. But the practical important finding is this. Personality researchers distinguished primary psychopathy, cold, calculating low anxiety, dispositionally stable, from secondary psychopathy, impulsive, poorly regulated, reactive to stress.

The popular image of the corporate psychopath is entirely the primary type. Roth and Claire's modelling found that primary psychopathy's direct links to workplace harm were mostly weak or negligible. The damage, to performance, to cooperation, to conduct, was overwhelmingly driven by secondary psychopathy instead.

Think about that for a second, because it inverts the story almost everyone tells. The person doing the most measurable damage to your team is less often the cold, controlling operator from the films, and more often the poorly regulated, stress-reactive colleague whose difficulty manifests as a persistent inability to manage their own frustration.

And the same analysis found something else worth knowing. This pattern predicted counterproductive behaviour more strongly among supervisors than among individual contributors. Authority doesn't restrain it. If anything, it licences it. What does this actually cost an organisation? Michael Hussman and Dylan Miners' Harvard Business School analysis of more than 50,000 employees defined a toxic worker conservatively as someone eventually terminated for serious misconduct.

On that definition, avoiding a single toxic hire saved an organisation around $12,500 in avoided turnover among colleagues who would otherwise have quit. In contrast, hiring a top 1% superstar only added about $5,300 in value. Their analysis showed that avoiding one bad apple was worth roughly 2.5 times as much as landing one exceptional performer.

And the traits that predicted who'd eventually be fired for misconduct also cut against the obvious assumption. On average, they were more productive than their peers beforehand, more confident, more self-centred and more likely in pre-hire assessment to insist that rules should always be followed without exception.

That last one is the counterintuitive part. The rigid rules-first respondent isn't who managers would flag as a future risk. Clive Boddy's 2017 case study follows what happens when nobody does. A UK charity, one senior manager, two consecutive chief executives, one independently assessed as scoring at the very top of a validated psychopathy measure, one showing essentially none of the same traits.

Under the new CEO, staff sickness absence rose by the equivalent of 8,000%. 86% of the workforce left within two years. Within three years, staff turnover exceeded 100%. Some of the people hired to replace the first wave of levers had themselves already gone. None of this happens without the quiet cooperation of people who are not themselves difficult.

Tembran Sola-Messick's concept of ethical fading explains why. Unethical behaviour rarely involves a conscious trade-off. People engage in self-deception that removes the ethical dimension from view entirely, reframing a bullying pattern as having high standards, or an aggressive manager as being results-driven.

A related concept, Albert Bandura's moral disengagement, explains what happens next, after someone has already witnessed it. It's the specific recognisable language colleagues use to explain why they didn't intervene. It wasn't really my place. They've had a hard time. I try to give them the benefit of the doubt.

Each is a textbook mechanism operating quietly, without the speaker believing they're doing anything other than being reasonable. A complementary framework, Hogan and Kaiser's Toxic Triangle, explains why destructive behaviour takes root in the first place. It requires three things converging at once.

A leader with the traits, followers who comply or actively collude, and an environment with weak checks and balances. Remove one destructive individual without addressing the followers and the environment, and a genuinely conducive environment will, in time, tend to produce or attract another one. This is the single most correctable of the three.

Teton Burnett's trait activation theory explains precisely why a weak environment, ambiguous metrics, minimal oversight, and no real consequence for norm violation, activates exactly the traits this video describes. A strong environment constrains them, regardless of individual's underlying dispositions. following An honest caveat before moving on to what actually helps.

This kind of finding is easy to misuse, turning one person's character into the whole story when the environment that let it happen matters every bit as much, maybe more. A 2005 paper claimed the precise ratio of positive to negative effect predicting whether teams flourish was cited everywhere, even in a US Army resilience programme, until a 2013 critique showed the maths was invalid, and the authors withdrew it.

Precision and popularity are not evidence of validity. And even at the higher corporate estimates, most difficult colleagues meet no clinical thresholds. This is a tool for recognising a real pattern, not a licence to diagnose someone from a distance. So, what actually helps? Certainly not a personality label.

Structure. Document behaviour, not character. What was said or done, when and who saw it. Address the environment alongside the individual. Has dissent stopped reaching decision makers? Has one account gone unchallenged for too long? Weight oversight more heavily as seniority rises, not less, because the evidence says authority doesn't restrain these patterns of behaviour, it licences it.

And where the pattern looks reactive and stress-linked rather than calculating, ask whether workload, stress and support are the most useful lever. That's a reasoned inference from the pattern, not a guaranteed fix. Escalate on a documented, corroborated pattern, not a single bad week or an instinct.

So, the team from the start of this video. Now he can probably describe what actually happened more precisely than someone changed the mood. One person, persistently doing one of three things. A mechanism, contagion, that's been demonstrated in a lab and in real organisations. A minority who does this chronically.

And within that minority, a specific reactive profile doing most of the measurable damage, not the calculating one you'd expect. The bad apple effect is real and it's large. What deserves more scepticism is the specific number most people have heard attached to it, not the underlying phenomenon, which the evidence supports more than the folklore usually gives it credit for.

If you want more information on this topic, you can read the source article and find the references on my website. The web address is on the screen now. Thank you for watching. If you found this video interesting, please like and subscribe to my channel. It helps the algorithm find the channel and ensures you find about the next episode.

Episode 2 was The Dark Empath. If you haven't seen it, it's the other half of this story. The link is on the end screen. All