There Is No Redeeming Value in Harshness
29 Sep 2026 Corporate Culture

Harsh vs Firm Leadership: Why Fear Destroys Corporate Culture

By Andrew Baker · Group CIO, Capitec Bank · 29 September 2026
Sponsored
Is Your WordPress Site at Risk of a Cyber Attack?
2FA · Passkeys · AI threat scan · Brute-force guard · Free
Secure My Site Free →
👁75views

Harshness has no redeeming value because loving firmness achieves the same accountability without the damage. Firmness names the problem, sets the standard and applies consequences. Harshness adds personal attacks that trigger fear, reduce thinking capacity and hide the truth, so it makes fixing the actual problem harder.

Article Summary
  • 1.
    What it is
    This article explains Terry Real's claim that harshness has no redeeming value and that loving firmness does everything harshness does, but better. It tests the claim against research on rudeness and performance.
  • 2.
    Why it matters
    It argues that harshness buys quick compliance but costs trust, information and honesty later, and that those costs are rarely traced back to their cause.
  • 3.
    Key takeaway
    Strip the insults out of a harsh conversation and what remains is a firm one, which shows the insults carried none of the load.
~25 min read
Related Articles
Listen to this article3 plays
CloudScale SEO

Why loving firmness works better than harshness at work, and why harsh organisations slowly go blind

1. A Sentence That Belongs in Every Boardroom

I recently watched a short clip of the couples therapist Terry Real in conversation with Mel Robbins, and one line from it has been rattling around in my head ever since. Real tells Mel that if his clients take only one idea away from their sessions it should be that harshness has no redeeming value whatsoever, and then he compresses the whole argument into nine words: “Harshness does nothing that loving firmness doesn’t do better.” He was talking about marriages, but I have not been able to stop applying it to organisations, because almost every corporate culture I have encountered contains an unspoken theory about harshness, and most of those theories are wrong in exactly the way Real describes.

At first hearing the line sounds almost too absolute to be true. Surely some people need a kick, surely a leader occasionally has to put their foot down, and surely a high performance culture requires a certain hardness. But the word he chose matters: he does not say there is no value in firmness, he says there is no value in harshness, and the distance between those two ideas is enormous. Some of the most expensive failures I have seen in organisations, in technology programmes, in risk functions and in trading businesses, came from leaders treating them as the same thing.

So I went looking for whether the research backs him up in a corporate setting. “No redeeming value” is exactly the kind of absolute that deserves to be tested rather than admired, and behavioural science rarely proves anything absolutely. What it can do is show which direction the evidence keeps pointing, and across laboratory experiments, workplace surveys, hospital wards, technology teams and at least one very public banking scandal, it points in a remarkably consistent direction.

2. Corporate Life Confuses Harshness With Strength

Somewhere along the way corporate life learned that strength has a particular voice. It is loud and unforgiving, it tears apart a presentation in front of the room, it tells people to toughen up, and it treats visible discomfort as evidence that standards are being enforced. Many organisations quietly promote this voice, because it looks decisive in a steering committee and because the leaders who use it often did deliver results at some earlier point in their careers.

Because harshness usually produces an immediate reaction, we assume it works. Someone gets shouted at in a programme review and the next week the plan moves faster, an engineer who is publicly criticised for an outage never raises that particular design concern again, and a team that was humiliated over a missed target hits the next one. We look at the result and conclude that the method worked. Except that sometimes what we changed was not the behaviour but what people were willing to show us. The programme still has the same risks but the plan now hides them, the engineer still has the concern but now keeps it to themselves, and the team still has the same capacity problem but has learned to hit the number by borrowing from next quarter. That is the strange economics of harshness in a company: you buy compliance quickly and pay for it later in lost trust, lost information and lost honesty, and because that cost arrives on a delay and never shows up in a single line item, it almost never gets attributed back to the meeting that caused it.

3. Firmness Is Not Softness

This is where the idea gets misunderstood, particularly by executives who worry that anything short of harshness means lowering the bar, so it is worth being precise. The alternative to harshness is not tolerating poor performance, it is not endless second chances and it is not a culture where nobody is ever told they are wrong. Loving firmness says that this outcome is unacceptable, that it needs to change, that we agreed to a standard, that there will be a consequence and that we are going to fix it together. Harshness adds a second layer on top of all that, one which contributes nothing to the standard being enforced. Sometimes it is explicit, as in how could you be so stupid, what is wrong with this team or you people always do this; just as often it is a contemptuous tone, a public dressing down, shouting, intimidation or pressure out of all proportion to the problem.

The first deals with the problem, while the second is aimed at the person, and humiliating or frightening the person rarely improves anyone’s ability to solve the problem. If you strip the humiliation and contempt out of a harsh conversation, what remains is almost always a firm one, which tells you they were not carrying any of the load. That is the core of Real’s claim, and in an organisation it is testable: if harshness adds something that firmness does not, we should be able to see it in performance, in retention, in error rates or in delivery. What the data mostly shows is that the extra layer subtracts.

4. Culture Is an Information System

As a technologist, I find the most useful way to think about corporate culture is as an information system: it determines what signals travel, how fast they travel, how much they are distorted along the way and whether they ever reach the people who can act on them. The sociologist Ron Westrum made essentially this argument decades ago when he classified organisations by how they handle information. In what he called pathological cultures, messengers are shot and failure leads to scapegoating; in generative cultures, messengers are trained and failure leads to inquiry.

Seen that way, the argument of this whole post fits into a single chain. Harshness creates fear, fear changes what people are willing to say and show, that distorts the information flowing up the organisation, and leaders then make worse decisions on worse information while believing they are in control. The evidence in the sections that follow is essentially evidence for each link in that chain.

That typology has since been tested at scale in technology organisations. The DORA research programme, which grew into Google’s State of DevOps work, found that a high trust, generative culture that emphasises information flow is predictive of both software delivery performance and wider organisational performance, and its 2019 analysis found the same for a climate of psychological safety. This is survey research with self selecting respondents, so it shows prediction and association rather than a controlled causal effect, but it is notable that one of the most rigorous long running studies of engineering performance keeps finding that culture sits underneath the technical practices rather than beside them.

Engineers already understand this intuitively. We spend fortunes on telemetry, logging and alerting precisely because unobserved failure is the most expensive kind of failure, and a harsh leader is effectively switching off the monitoring and then congratulating themselves on how quiet the alerts have become. The practice of blameless postmortems exists for exactly this reason: it is completely firm about the failure, the timeline, the root cause and the remediation, while deliberately removing the incentive to hide any of it. That is loving firmness expressed as an operating process, and I would argue it is one of the most important cultural inventions of the last twenty years, and one that most non technology functions have still not adopted.

5. Harshness Is a Tax on Thinking

After somebody makes a serious mistake there is useful information that needs to travel between two people: what happened, why it happened, what they believed at the time, what they missed, what should change and how we prevent it happening again. Harshness introduces noise into that signal. The other person stops thinking about the mistake and starts thinking about you, about how angry you are, whether their job is at risk, how to defend themselves, what to say to end the conversation and who else might share the blame. Their brain has moved from learning to survival, and while you may feel that you are holding somebody accountable, you have made the actual process of accountability harder, because accountability requires the truth and fear is remarkably good at editing the truth.

This is not just intuition. Christine Porath and Amir Erez ran three experiments, published in the Academy of Management Journal in 2007, in which participants experienced rudeness from an authority figure, rudeness from a third party, or were simply asked to imagine a rude incident, and then completed anagram puzzles and a brainstorming task generating uses for a brick. In all three setups rudeness reduced performance on both the routine task and the creative one and made people less willing to help someone afterwards, and the authors found that disruption to cognitive processing fully mediated the effect on performance. A follow up paper in 2009 found that merely witnessing rudeness directed at somebody else reduced observers’ performance on the same kinds of task. These were laboratory studies, largely with students, measuring short term task performance, so I would not stretch them further than they go; but they describe a mechanism that every leader should recognise, which is that contempt occupies the mental bandwidth people need in order to fix the problem, and it taxes everyone else in the room as well.

The workplace survey data points the same way. In a poll of 800 managers and employees across 17 industries, reported by Porath and Christine Pearson in Harvard Business Review in 2013, 48% of those who had been on the receiving end of incivility said they intentionally decreased their work effort, 38% intentionally decreased the quality of their work, 80% lost work time worrying about the incident, 78% said their commitment to the organisation declined, 12% said they left their job because of it, and a quarter admitted taking their frustration out on customers. These are self reported figures, so they measure what people say they did rather than what was independently observed, but it is worth sitting with the fact that people were willing to admit deliberately doing worse work. Following that research, Cisco reportedly ran its own internal analysis and estimated that incivility was costing it around $12 million a year. Harshness is not free; it is simply billed to a cost centre nobody is watching.

6. Fear Edits the Truth

Some of the best evidence that fear distorts what an organisation can see comes from a study that was designed to prove something else entirely. In the early 1990s Amy Edmondson, then a doctoral student at Harvard, joined a larger research programme tracking medication errors on hospital units, where a team of nurses reviewed patient charts over a six month period. Her hypothesis was simple: units with better teamwork should make fewer errors. When the data came back, the better teams, as measured by her team survey, had higher recorded error rates rather than lower ones.

Her explanation, which her later work supported, was that the better teams were probably not making more mistakes; they were more willing and able to talk about them. One of the ways the units differed was whether people were punished for errors, and where they were, mistakes were far more likely to go quietly unrecorded. She went on to formalise the idea as psychological safety, and in her 1999 paper studying 51 work teams in a manufacturing company she found that psychological safety predicted learning behaviour, which in turn predicted team performance. Google’s Project Aristotle, which examined 180 of its own engineering and sales teams, later reported psychological safety as the strongest of the five factors it associated with team effectiveness.

I want to be careful with the hospital finding, because it is often retold as “better teams had better patient outcomes”, and that is not what the original study measured. The narrower and more important claim is about observability: a punitive climate did not necessarily reduce errors, it reduced the reporting of errors. For anyone running a regulated business, that distinction should be alarming, because it means that the harshest parts of your organisation may well be the ones producing the cleanest looking risk reports.

7. Harsh Leaders Create Excellent Actors

That observability problem scales directly into corporate culture. A leader who responds harshly to bad news eventually stops receiving bad news, not because the bad news disappears but because people learn. They learn which numbers should be softened before a meeting, which risks are better discussed privately, which issues should be parked as “amber, recovering” for another month, and the difference between an executive who wants disagreement and one who merely says they want disagreement. Eventually the organisation becomes a theatre production in which everyone understands their part: the executive committee asks whether everything is under control, everyone nods, the dashboard is green and the programme is on track, while three floors down somebody is wondering whether they should mention that the database appears to be on fire.

The research literature on this is now very large. A 2025 meta analysis by Liu and Ling in the Baltic Journal of Management pooled 354 articles covering more than 169,000 employees and found that authoritarian leadership and abusive supervision suppress employee voice, working through reduced psychological safety and a reduced belief that speaking up will achieve anything. A study by Christian Kiewitz and colleagues, published in the Journal of Applied Psychology in 2016 and following employees over several waves, found that fear helped explain why abused employees fell defensively silent, and described a reinforcing loop in which abusive supervision increased fear, fear produced defensive silence, and that silence was in turn followed by more abusive supervision. Most of this work relies on employee surveys, so it describes associations and perceived behaviour rather than controlled experiments, but the direction is consistent across a very large body of evidence, and the loop it describes will be familiar to anyone who has watched a leader grow angrier precisely because nobody tells them anything.

The irony is wonderful. The harsh leader believes fear gives them greater control, when fear quietly switches off the very telemetry that control depends on.

8. When Pressure Becomes Policy

The clearest corporate illustration I know comes from my own industry. In 2016 Wells Fargo was found to have opened large numbers of accounts that customers had not authorised, and in April 2017 its independent directors published the findings of their investigation. Their central conclusion was that the root cause was a “distortion of the Community Bank’s sales culture and performance management system”, which, combined with aggressive sales management, pressured employees to sell unneeded products and in some cases to open unauthorised accounts. The report described sales goals that leadership in many instances knew were unattainable, a performance system that exerted significant and in some cases extreme pressure on employees, and staff who feared being penalised for missing their numbers.

I want to be careful not to overclaim here, because the Wells Fargo failure had many causes, including incentive design, a decentralised structure and control functions that the report found were too deferential to the business. It is not a study of rudeness, and the report’s limited review did not identify a pattern of retaliation against employees who complained. But it is a very expensive demonstration of the pattern Real describes: when the organisation’s response to falling short is pressure and fear rather than firm and honest inquiry, people do not stop falling short, they find ways to make the numbers look right, and the leaders at the top end up managing a set of figures that no longer describe reality.

9. The Voice Between Your Ears Becomes the Voice of the Organisation

This was the part of Real’s comments that interested me most, and I think it is the part that matters most for executives. He describes carrying a harsh voice inside his own head for decades, which he traces back to a violent father; the father no longer needed to be physically present because the criticism had been internalised. Many leaders run a less extreme version of the same program. They make a bad call and the voice calls them an idiot, they miss a deadline and it tells them they have no discipline, and when something fails it concludes that they are simply not good enough. The problem is that this voice rarely stays inside. The way a leader speaks to themselves after a failure tends to become the way they speak to their team after a failure, and the way a leadership team speaks to itself tends to become the culture everyone else lives inside.

We justify this internal brutality with a seductive argument, which is that if we stop being hard on ourselves we will become complacent, and many high achievers genuinely believe their inner critic is the source of their success. The research is not kind to that belief. Juliana Breines and Serena Chen ran four experiments, published in 2012, comparing people prompted to treat a failure or weakness with self compassion against people given a self esteem boost, a pleasant distraction or no intervention at all. In one of them, participants who had struggled with a difficult test and then received a self compassionate framing spent more time studying for a retest than those in the comparison conditions, and across the other experiments the self compassion groups reported more motivation to make amends for a moral lapse and to change a personal weakness. Note what was measured: effort and motivation to improve, not eventual scores. The honest conclusion is that self compassion did not make people complacent, which is precisely what the harsh inner voice insists would happen.

The longer running evidence points the same way. Theodore Powers, Richard Koestner and David Zuroff, in prospective studies in the United States and Canada, found that self criticism was negatively associated with progress on people’s own goals, and a later set of five studies from the same research group found that negative association held consistently, while self oriented perfectionism, essentially holding high personal standards, was positively associated with goal progress once self criticism was controlled for. That distinction is the whole corporate argument in miniature: the high standard was helping, and the attack bolted onto it was dragging it down. These studies are correlational, so they cannot prove that self criticism causes worse progress, but they make it very difficult to argue that it helps. In my experience harshness also has a remarkable ability to reproduce itself down a management hierarchy, because people tend to manage the way they were managed.

10. Kindness Without Standards Is Not Leadership

There is an equally dangerous extreme on the other side, and it would be dishonest to write this without acknowledging it. Many organisations that have heard the message about psychological safety have overcorrected into a culture where nothing difficult is ever said, where underperformance is tolerated for years and where every review is a gentle exercise in mutual reassurance. That is not safety, it is avoidance, and it is just as corrosive. Leadership without standards eventually becomes neglect wearing a friendly face, and teams led that way tend to lose their best people first, because strong performers notice very quickly when nobody is holding the line. Edmondson herself has been consistent that psychological safety is not about lowering standards, and that the highest performing environments combine high safety with high accountability.

You can enforce every standard without humiliation. You can tell someone that this cannot happen again, that they committed to something and did not deliver, that if it continues there will be consequences, and even that you do not think the role is working for them. None of those statements requires cruelty; the standard does the work, and you do not need to add venom. The same is true of the “brilliant jerk” that so many organisations tolerate because of their individual output. Firmness means dealing with that person’s behaviour as directly as you would deal with anyone else’s missed target, because the evidence above suggests that the cost of their harshness is being paid, quietly, by everyone around them.

11. Surprise Is the Cruellest Form of Harshness

One form of harshness rarely gets called harshness at all, because it arrives politely and on schedule: the surprise. A leader who says little all year, lets frustration accumulate privately and then delivers a poor rating at the year end review has been far harsher than one who restated the standard every week, even if the review itself is delivered in a calm and measured voice. The person on the receiving end has been denied the one thing that would have let them respond, which is time, and they will reasonably conclude that the organisation’s real standards are hidden and that its feedback only arrives once it is too late to act on.

Loving firmness is therefore repetitive, almost boringly so. It states the goals, the outcomes that matter and the cost of failing them, and then states them again, long before anything has gone wrong. I recently sent my own leadership team a short message of exactly this kind about the few remaining priorities that will define our year. It was not gentle. It reminded them that we are a performance based organisation, that these priorities had been flagged repeatedly, that the outcome would shape everyone’s assessment including my own, and that consequences would follow naturally if things were allowed to drift. What it was designed to prevent was surprise; nobody reading it could claim at year end that they had not known what mattered or what failure would cost.

That is the distinction I care about. Constantly restating goals and consequences is not nagging, it is building a platform for dialogue, because it gives people repeated opportunities to say that something is off track, that a dependency is blocked or that a target is unrealistic while there is still time to act. That last point matters, and Wells Fargo is the warning: repetition only works if the goals are achievable and people are genuinely allowed to say when they are not. The alternative is frustration that builds silently and then spills over into a shock performance review, which is harsh in exactly the sense Real means, because it inflicts the pain without having delivered the information when it could still have helped. The discipline for the leader, and I include myself in this, is to keep those repeated messages about the outcome and the standard, and to resist letting accumulated frustration leak into the wording, because the moment a message starts characterising people rather than stating the goal, it stops inviting dialogue and starts inviting silence.

There is another uncomfortable possibility, which is that harshness in the workplace often is not really about improving the other person at all but about discharging the leader’s own frustration. Something went wrong, the executive feels disappointed, frightened or embarrassed, perhaps because they now have to explain it to their own board, and their nervous system wants somewhere to put that emotion. Conveniently, there is a person on the call who owns the failing workstream. So the emotion gets unloaded onto them, the leader feels better for a moment, and everyone calls it accountability.

Accountability should be judged by what happens next rather than by how satisfying the meeting felt. The useful questions are whether the person understood, whether their behaviour changed, whether the system improved, whether trust survived and whether anybody learned anything. If the answer to most of those is no, then perhaps all the leader accomplished was moving their discomfort into somebody else, and given what Porath and Erez found about bystanders, into everyone else who happened to be in the room or on the call as well.

12. Retiring the Survival Mechanism

Toward the end of the clip Real says something else that stayed with me. At this stage of his life he has essentially made himself a deal that if something is not kind he is not interested in it, and that applies not only to how other people speak to him but to how he speaks to himself. It does not mean the tough part of him has disappeared. He describes a younger, fighting part of himself that helped him survive difficult circumstances, and rather than trying to destroy it he recognises what it did for him; it simply no longer gets to run the place.

Organisations have exactly the same pattern. Many of the behaviours a company eventually needs to change once protected it: the aggressive founder energy that got it through its first near death experience, the command and control style that stabilised it during a crisis, and the relentless pressure that drove extraordinary growth in its early years. Those behaviours worked, right up until the organisation became too large and too complex for any single leader to see everything, at which point its survival started to depend on information flowing freely rather than on fear driving effort. Growing up, for a leader and for a company, sometimes means thanking an old survival mechanism for getting you this far, and then quietly taking away its admin privileges.

13. What Loving Firmness Looks Like in Practice

If this is right, then culture change is less about slogans and more about a small number of repeated behaviours at the moments that matter most. The first is how a leader reacts to the first piece of bad news they receive in a new role, because everyone is watching and that single reaction will set the price of candour for years. The second is how incidents and failures are reviewed, where the test is whether the process is fiercely rigorous about causes and remediation while being explicitly neutral about blame. The third is whether the organisation actually rewards the messenger, by visibly thanking the person who surfaced a risk early rather than quietly marking them as negative. The fourth is whether performance conversations separate the standard from the person, stating plainly what was not good enough and what must change, without the commentary on character. And the fifth is whether leaders are prepared to measure culture as seriously as they measure delivery, by asking teams directly whether mistakes are held against them and whether difficult issues can be raised, and then treating a poor answer as an operational risk rather than an HR statistic.

None of these practices is soft. Every one of them demands more discipline from a leader than simply losing their temper, which is why they are rarer than they should be.

So the next time you feel the urge to be harsh with somebody at work, there is a wonderfully simple question to ask, which is what outcome you are actually trying to produce. If somebody needs to understand a boundary, make the boundary clear; if they need a consequence, apply the consequence; if they need feedback, give the feedback; and if they need to improve, help them see exactly what needs to improve. When the person who failed is you, do exactly the same thing: look at the failure, own it, understand it and change something, but do not confuse adding cruelty with adding effectiveness.

There is nothing inherently weak about kindness and nothing inherently strong about harshness, and the evidence, imperfect as behavioural evidence always is, keeps pointing in the same direction as Real’s nine words. Harsh organisations do not have fewer problems; they simply have fewer problems that anyone is prepared to tell them about. Sometimes the strongest sentence in the room is simply this: “This is not good enough. I know you can do better. Let’s work out what changes now.” It is clear, firm and human, and no scar is required.

References

You Might Also Like