The toxic workplace crisis
Three-quarters of American workers have experienced a toxic workplace. What's going on?
Eighty per cent say their current one is toxic, up from sixty-seven per cent the year before.
Those numbers come from two large 2025 surveys. iHire interviewed 1,781 employees and 504 employers across fifty-seven industries. Monster surveyed 1,100 American workers about their mental health at work. The two studies used different methodologies, asked slightly different questions, and arrived at the same place.
What I want to do in this piece is move past the headline figures and into the more interesting question underneath: What is actually happening inside these organisations? What is the mechanism?
Much of the public conversation about toxic workplaces stops at the level of bad bosses and poor culture, as though these were the explanation rather than the symptoms. They aren’t. The real story lies one layer deeper, in the territory of how power is held and what happens to those who hold it.
The good news is that this territory has been studied carefully for decades by researchers who don’t usually appear in shadow work conversations. Their findings are striking, well-evidenced, and almost completely aligned with what we see when we look at organisational life through a depth psychology lens.
Let me bring three of them in.
The power paradox
Dacher Keltner is a professor of psychology at the University of California, Berkeley, where he runs the Berkeley Social Interaction Laboratory. He has spent more than twenty-five years studying what happens to human beings when they acquire power. His findings are summarised in his 2016 book The Power Paradox, and the central observation is unsettling.
People rise to positions of power, Keltner argues, because of qualities we would all recognise as “good”. They are perceptive, generous, attentive to others, and willing to share. They listen well. They make those around them feel seen. These are the qualities that win you the trust of a team, the respect of peers, and the recognition of senior leaders. They are the qualities that get people promoted.
And then, once people have power, those same qualities begin to disappear.
In Keltner’s experiments, simply placing people in a position of authority changes their behaviour in measurable ways. They become more impulsive. They interrupt more. They take more risks. They are more likely to behave inappropriately, to disregard rules, to feel entitled to things they would not have felt entitled to before. In one famous study, his team brought groups of three strangers into a lab, randomly assigned one of them to be the “manager,” and then, twenty minutes later, brought in a plate of cookies. The managers were significantly more likely to take an extra cookie, to eat with their mouths open, to leave crumbs on the table. The smallest experience of power, in a stranger, was enough to alter behaviour.
In another study, Keltner’s team observed drivers at pedestrian crossings in California, where pedestrians have legal right of way. They coded the cars by status, from old beaters to luxury vehicles. The drivers of expensive cars were significantly more likely to break the law and refuse to yield. Power, even in the form of a nicer car, eroded the basic civility of stopping for someone trying to cross the street.
The most important finding from this body of work is what Keltner calls the power paradox itself. The very practices that allow us to gain influence, namely empathy, generosity, and attention to others, vanish in the experience of having it. Power makes people less able to do the things that earned them the power in the first place. This is a near-universal effect that appears in almost every population studied.
A leader who began their career by listening carefully and making people feel seen will often, after a few years in a senior role, become someone who interrupts, dismisses, and assumes their own perceptions are correct. They will not notice this happening. Their team will, of course, notice it immediately, but they will often not be able to say so.
When you hear the iHire figure that 78.7% of workers cite poor leadership as the top reason their workplace is toxic, this is a large part of what they are pointing at. It’s not that their leaders were always bad. Their leaders changed, in ways the leaders themselves cannot see.
Abusive supervision and what it does
The second research thread I want to bring in is more clinical and more specific. Bennett Tepper, a professor of management at Georgia State University, coined the term “abusive supervision” in a foundational paper in the Academy of Management Journal in the year 2000. He defined it as the sustained display of hostile verbal and non-verbal behaviour from supervisors, excluding physical contact.
Note the precision of that definition. Abusive supervision is a sustained pattern of behaviour over time that people on the receiving end experience as hostile. This looks like belittling, public criticism, withholding of information, ignoring, taking credit for others’ work, blaming and ridicule.
In the twenty-five years since Tepper’s original paper, abusive supervision has become a heavily researched area in organisational psychology. There are now multiple meta-analyses, hundreds of empirical studies, and a 2017 Annual Review synthesis by Tepper and his colleagues that summarises what the field has learned. The findings are clear and depressing. Abusive supervision is reliably associated with lower job satisfaction, lower commitment to the organisation, higher turnover, work-family conflict, psychological distress, depression, anxiety, and physical illness in the people who experience it. It also damages team performance, reduces innovation, and increases counterproductive work behaviour. The cost to organisations runs into the billions.
But the most interesting question Tepper’s research has tried to answer is not what abusive supervision does. It is where abusive supervision comes from. Why do supervisors behave this way?
The answer the literature has converged on is not what most of us would expect. Abusive supervisors are not, in the main, sadistic personalities. They are not bullies who enjoy hurting people. The dominant predictors of abusive supervision in the empirical research are things like the supervisor’s own perceived stress, their own experience of being mistreated by their boss, their own sense of being threatened by the people they manage, and their own inability to regulate the anxiety that comes with the role.
Abusive supervision, in other words, is largely the behaviour of frightened people in positions of power. The fear comes first. The behaviour comes after. And the people on the receiving end carry the cost of the fear that the supervisor cannot acknowledge in themselves.
This matches Keltner’s findings. The same person who was generous and attentive on the way up becomes impulsive, defensive, and self-protective once the role is theirs. And when their own anxiety is high enough, that self-protection takes the form of behaviour that meets Tepper’s definition of abusive supervision.
Dying for a paycheck
The third research thread I want to bring in is the most sobering. Jeffrey Pfeffer is the Thomas D. Dee II Professor of Organisational Behaviour at the Stanford Graduate School of Business. He has been studying organisations for more than four decades. In 2018, he published a book titled Dying for a Paycheck, which examines how modern management practices affect the bodies and minds of workers.
His estimate, drawn from a wide range of public health and organisational research, is that workplace stress is responsible for approximately 120,000 excess deaths per year in the United States alone. Job stress costs American employers more than $300 billion a year in healthcare expenses, absenteeism, and lost productivity. In one survey he cites, 61% of employees said workplace stress had made them physically sick, and 7% said they had been hospitalised because of it. In China, where the culture of working from 9 am to 9 pm, six days a week, is standard in many industries, an estimated one million people a year are dying from overwork.
Pfeffer’s central argument is that this is the result of specific management choices that organisations have made and continue to make. These choices result in long hours, layoffs, job insecurity, health insurance tied to employment, the removal of worker autonomy, and the destruction of work-life boundaries. None of these things is inevitable. They are the consequences of how power is held within organisations, and most produce no measurable performance improvement. Pfeffer’s evidence suggests that the toxic management practices that are killing people are also bad for the bottom line. The lose-lose is the common case, not the exception.
The cost of these practices shows up in the physical, and not just the psychological. It shows up in cardiovascular disease, diabetes, addiction, suicide, and early death.
The 80% of workers who reach for the word “toxic” are not exaggerating. They are describing an environment that is literally taking years off their lives.
The shadow underneath
So we have three converging lines of research from three of the most respected organisational researchers in the world. Keltner shows that power changes the people who hold it, eroding the very qualities that earned them the role. Tepper shows that supervisors under stress become hostile to their teams in patterns that produce measurable harm. Pfeffer shows that the cumulative effect of these dynamics is killing people.
What none of these researchers quite say, but all of them are circling, is what shadow work names directly:
Power, when held by a wounded part of a person, becomes harmful. The wound is what does the harm. And the wound is invisible to the person carrying it.
Every leader has parts. The seven-year-old who learned that being right was the only way to get love. The teenager who decided that letting anyone see weakness was unsafe. The young professional who built their identity around being the most capable person in the room. The middle manager who learned, painfully, that the way to survive in this organisation was to align upwards and offload pressure downwards. None of these parts is bad. All of them are the result of someone trying to survive a system that demanded a certain kind of self.
The problem is that when these parts are not integrated, they do not stay quiet simply because the person has been promoted. They take the wheel. The seven-year-old who needed to be right is now managing performance reviews. The teenager who could not show weakness is now refusing to take responsibility for the failed product launch. The capable young professional is now micromanaging the team because letting go feels like falling. The middle manager is now passing impossible deadlines down the chain because pushing back feels like a death sentence.
The leader does not usually know that any of this is happening. From inside the wounded part, usually taking an inflated form, the behaviour feels like leadership. It feels like decisiveness, high standards, and protecting the company. The leader cannot see that the team is being damaged because the part doing the damage cannot see itself. This is what shadow means. The parts of us we cannot see.
Keltner’s research tells us this happens to most people who acquire power. Tepper’s research tells us what it produces for the people on the receiving end. Pfeffer’s research tells us the cumulative cost, both in monetary terms and in years of life. And shadow work tells us why none of it can be addressed by the usual interventions, because those interventions are aimed at the behaviour, not at the part that is producing it.
What this means
If you take all of this seriously, it changes how you think of a toxic workplace.
A toxic workplace is not a place run by bad people. It is a place where the people in positions of power have not done the inner work required to hold that power well.
It is a place where unintegrated parts of leaders are being given the authority to shape the lives of the people who work for them. It is a place where the leader's wound becomes the team's daily experience.
This is not a counsel of despair. Quite the opposite. The inner work that would change things is real, available, and well understood. Leaders can do it. Organisations can support it. The shadow work that would help a leader hold power without inflation is not exotic or mysterious. It is the practice of getting to know the internal parts that are running the show without conscious permission, and learning to lead them rather than be led by them.
What that requires is something most modern organisations have actively discouraged in their leaders: the willingness to acknowledge fragility, to admit not knowing, to feel the fear that comes with the role rather than acting it out on the team. The willingness to let the adult self take the microphone away from the seven-year-old.
Only when organisations and leaders wake up to what’s really going on under the surface will there be lasting change in toxic workplaces. But this is hard for them to see because they were promoted, in part, for being good at hiding the parts that need to be integrated. The system that produced them rewarded the hiding.
This is why the numbers keep climbing. Eighty per cent and rising is a sign that the structural mechanism producing the harm has not changed. We keep promoting people for their public face, putting them in positions of power, and then watching the power activate the parts of them they have spent their lives keeping hidden. It is always the team that feels it.
The team eventually reaches for the only phrase that captures what they are feeling:
This place is toxic.
The themes explored here are central to my forthcoming book, The Shadow at Work. Please subscribe to support this project.
Sources: iHire, 2025 Toxic Workplace Trends Report (1,781 workers, 504 employers, December 2024). Monster, 2025 Mental Health in the Workplace poll (1,100+ U.S. workers, April 2025). Keltner, D., The Power Paradox: How We Gain and Lose Influence (Penguin, 2016), and 25+ years of research from the Berkeley Social Interaction Laboratory. Tepper, B.J., “Consequences of Abusive Supervision,” Academy of Management Journal 43:178-190 (2000); Tepper, B.J., Simon, L. and Park, H.M., “Abusive Supervision,” Annual Review of Organizational Psychology and Organizational Behavior 4:123-152 (2017). Pfeffer, J., Dying for a Paycheck: How Modern Management Harms Employee Health and Company Performance (HarperBusiness, 2018).