Toxic at the Top: Why Self-Reported Workplace Toxicity Is Rising and How Leaders Can Close the Empathy Deficit
Download the article here:
Listen to a review of this article:
Abstract: Businessolver's 2026 State of Workplace Empathy study, fielded with Edelman Data & Intelligence among 300 executives and 1,000 U.S. employees, found that one in three CEOs and 40% of employees now describe their workplace culture as toxic, sharp increases from the prior year, while 61% of CEOs say demonstrating empathy day to day has become difficult. This article treats those vendor findings cautiously, noting that the reported link between toxic cultures and short-term financial growth is associational, and situates them within peer-reviewed research on abusive supervision, destructive leadership, incivility, workplace bullying, psychological contract breach, and psychological safety. Drawing on documented cases from Uber, Wells Fargo, Boeing, Activision Blizzard, Riot Games, Ford, the Confederation of British Industry, and Microsoft, it describes how an empathy deficit between leaders' self-image and employees' experience produces attrition, ill health, silence, and legal and reputational risk. It proposes five evidence-based responses and a long-term agenda centered on board-level culture governance, leadership selection and development, and a humane approach to AI-driven change.
Something unusual happened in the 2026 edition of one of the longest-running workplace surveys in the United States. For eleven years, Businessolver has asked employees and executives how empathetic their organizations are. Executives have usually rated their own cultures more generously than their employees do. This year, the people at the top began describing their own cultures in harsher terms. One in three CEOs surveyed said their workplace culture is toxic, a jump of 25 percentage points from the prior year, while 40% of employees said the same, up 18 points (Businessolver, 2026a). Coverage in October 2026 framed the finding bluntly: a third of chief executives now concede that the cultures they lead are toxic (Fair Play Talks Editorial Team, 2026).
The more uncomfortable part of the report sits underneath the headline. Executives in organizations described as toxic were about twice as likely as those elsewhere to report significant financial growth, 70% compared with 36% (Businessolver, 2026a). Nearly every executive surveyed called themselves an empathetic leader, yet 61% of CEOs said that demonstrating empathy day to day is difficult, a figure that rose 32 points in a year (Businessolver, 2026a, 2026b). Businessolver calls this gap between self-image and lived experience an "empathy paradox." I will call it, less charitably, an empathy deficit: the shortfall between the empathy leaders believe they offer and the empathy employees actually experience when budgets tighten, AI programs roll out, and layoffs loom.
I want to be careful from the start. This is a vendor survey, run by a benefits technology company with a commercial interest in the topic, and its financial-growth finding is a self-reported association rather than evidence that toxicity causes growth. Executives who are pushing hard for short-term results may simply be more willing both to report growth and to admit that the push has costs. Even so, the direction of the data is consistent with a large and much older body of peer-reviewed research. Studies of abusive supervision, destructive leadership, incivility, bullying, and psychological contract breach point the same way: cultures that tolerate disrespect may produce results for a while, but they bleed talent, health, candor, and eventually reputation (Mackey et al., 2017; Nielsen & Einarsen, 2012; Schyns & Schilling, 2013).
This article does three things. It defines what we mean by workplace toxicity and the empathy deficit, and it looks at how prevalent these problems are and why they appear to be rising. It reviews the organizational and individual consequences, drawing on peer-reviewed research and on documented organizational cases from technology, gaming, banking, aerospace, manufacturing, and a national business association. It then sets out five evidence-based responses, followed by a longer-term agenda for building leadership capability that can hold up under pressure. My aim is practical. If one in three CEOs already suspects something is wrong, the question is what they, their boards, and their HR leaders should do next.
The Toxic Culture Landscape
Defining Toxicity, Destructive Leadership, and the Empathy Deficit
"Toxic" is an emotionally loaded word, and survey respondents may mean quite different things by it. Researchers have tried to pin it down. Donald Sull and colleagues, analyzing more than 1.3 million Glassdoor reviews from employees at large U.S. companies, identified what they call the "Toxic Five": cultures that are disrespectful, noninclusive, unethical, cutthroat, and abusive (Sull, Sull, Cipolli, & Brighenti, 2022). Disrespect had the largest negative effect on overall culture ratings of any single topic in their analysis. Their framing is useful because it moves the conversation away from vague unpleasantness toward specific, observable patterns of behavior.
Leadership researchers have their own vocabulary. Abusive supervision refers to subordinates' perceptions that a supervisor engages in sustained hostile verbal and nonverbal behavior, short of physical contact (Tepper, 2000). The emphasis on "sustained" matters: a single bad day does not qualify, while a pattern of belittling, public ridicule, or silent treatment does. Destructive leadership is a broader concept. Einarsen et al. (2007) defined it as systematic and repeated leader behavior that violates the legitimate interests of the organization by undermining its goals, resources, or effectiveness, or by undermining the motivation, wellbeing, or job satisfaction of subordinates. Their model is helpful because it recognizes that a leader can be destructive toward people while appearing loyal to the organization's financial goals, which is precisely the pattern the 2026 Businessolver data seem to capture. Jean Lipman-Blumen's (2005) work on toxic leaders adds a further insight: followers often tolerate or even seek out such leaders because they promise certainty and results in anxious times.
At the milder end of the spectrum sits workplace incivility, which Andersson and Pearson (1999) described as low-intensity behavior that violates norms of mutual respect, where the intent to harm is ambiguous. Rudeness, dismissiveness, interrupting, and ignoring colleagues all count. Their central argument was that incivility can spiral, with each slight inviting a sharper response until behavior escalates into open aggression. Workplace bullying is more severe and more persistent, involving repeated negative acts over time against someone who finds it difficult to defend themselves (Nielsen & Einarsen, 2012).
Two further concepts help explain why these behaviors are so damaging. A psychological contract is an employee's belief about the reciprocal obligations between themselves and their employer, such as fair treatment, development, and security in exchange for effort and loyalty (Robinson & Rousseau, 1994). When employees perceive that the organization has broken its side of the bargain, they experience breach, and the emotional reaction to that breach is called violation. Psychological safety, a term introduced to team research by Edmondson (1999), is a shared belief that a team is safe for interpersonal risk-taking, so people can raise concerns, admit mistakes, or challenge ideas without fear of humiliation or punishment.
Finally, empathy in leadership generally means a leader's ability to understand others' perspectives and feelings and to respond in ways that show that understanding. Muss et al. (2025), in a systematic review of 42 studies, note that the field still lacks a consensus definition, but most work combines a cognitive element (understanding what someone is experiencing) and an affective element (caring about it). The empathy deficit, as I use the term here, is the gap between leaders' stated commitment to empathy and the behaviors employees encounter, particularly when performance pressure rises. The Businessolver data make the gap visible: 98% of executives described themselves as empathetic leaders, rising to 100% among those in self-described toxic cultures, even as 33% of employees reported being intimidated by coworkers (Businessolver, 2026a).
Prevalence, Drivers, and Distribution
How common is toxic leadership? Estimates depend heavily on definitions and methods, but none of them are small. In a representative Norwegian sample, Aasland et al. (2010) found that a large majority of employees reported some exposure to destructive leader behavior over the preceding six months. When they applied stricter thresholds, the share reporting frequent exposure fell to about a third using one classification approach and about three-fifths using another, numbers that have been widely cited in later prevalence work (Lundmark et al., 2021). Porath and Pearson (2013) reported that 98% of workers they had polled over 14 years had experienced uncivil behavior at work, and that by 2011 half said they were treated rudely at least once a week, compared with a quarter in 1998. Robinson and Rousseau (1994), following graduate management alumni over their first two years in a job, found that more than half reported their employer had violated the psychological contract, which led them to describe violation as the norm rather than the exception.
The Businessolver 2026 study adds a recent and specifically executive-level snapshot. Fielded by Edelman Data & Intelligence between March and April 2026, it surveyed 300 C-suite executives (CEOs, CHROs, CFOs, and CTOs) and 1,000 employees across six industries (Businessolver, 2026a, 2026c). Its headline numbers are summarized in Figure 1. The prior-year values in the chart are not published directly; they are derived by subtracting the reported year-over-year increases from the 2026 figures, so they should be read as approximate.
Several features of the data stand out. First, the increase among CEOs was steeper than among employees, which narrows a long-standing perception gap from the top. Second, intimidation is not confined to the front line: 34% of executives overall, and 73% of executives in self-described toxic cultures, reported feeling intimidated by coworkers (Businessolver, 2026a). Employees in toxic cultures were 3.6 times more likely than those in non-toxic organizations to feel intimidated. Third, toxicity clustered with cost-cutting. Executives in toxic cultures were 2.6 times more likely to report layoffs and twice as likely to report benefit cuts (Businessolver, 2026a).
Why might self-reported toxicity be rising now? The report's authors point to three pressures: AI transformation, cost control, and economic uncertainty, which push executives toward efficiency at the expense of employee experience (Businessolver, 2026c). A companion release on AI found that 28% of CEOs said cost savings from headcount reduction were the primary motivation for AI investment, while only 31% of executives named upskilling through personalized learning as a primary goal (Businessolver, 2026b). At the same time, 60% of executives said AI adoption creates a much greater need to demonstrate empathy. Leaders, in other words, know what is required and say it plainly, yet they report finding it harder to deliver.
Broader engagement data point in the same direction. Gallup's State of the Global Workplace: 2026 Report, based on responses from more than 263,000 people in over 160 countries, found that global engagement fell to 20% in 2025, the lowest level since 2020 (Gallup, 2026). Manager engagement dropped further, from 31% in 2022 to 22% in 2025. That matters because managers transmit culture. When managers are depleted, the daily interactions that make up an empathetic or a toxic culture tend to deteriorate. The same report found that 46% of leaders reported experiencing a lot of stress the previous day.
Toxicity is also unevenly distributed. In the Businessolver AI data, employees in toxic cultures were far more likely than employees overall to worry about falling behind on AI (51% vs. 31%) and to say AI had slowed their career progression (54% vs. 35%) (Businessolver, 2026b). Only 33% of employees in toxic cultures said they had received adequate AI training, compared with 87% in empathetic cultures (Figure 2). Women reported lower rates of adequate training than men (45% vs. 56%), and hands-on workers lower rates than computer-based workers (44% vs. 57%). The burden of a toxic transition, in other words, falls hardest on those with the least positional power.
Research on destructive leadership offers a structural explanation for these patterns. Destructive behavior is more likely when leaders face pressure, have wide discretion, and operate in systems that reward outcomes without scrutinizing how they were achieved (Einarsen et al., 2007; Schyns & Schilling, 2013). Sull, Sull, and Zweig (2022) similarly found large differences in culture within the same industries, which suggests that toxicity is shaped by leadership choices rather than being an unavoidable feature of a sector. Figure 3 pulls these strands together as a self-reinforcing cycle: performance pressure leads leaders to deprioritize empathy, which licenses incivility and intimidation, which in turn silences employees and breaches the psychological contract. The resulting costs (attrition, ill health, hidden operational risk, and legal exposure) tend to surface later than the short-term results that pressure produces, which is why the cycle can persist even in organizations that appear to be performing well.
Organizational and Individual Consequences
Organizational Performance Impacts
If toxic cultures are associated with short-term financial growth, why worry? Because the costs tend to arrive later and through channels that quarterly financial reports do not capture well. Businessolver's chief executive made a version of this point himself, arguing that cultural risks show up in people before they show up in profits (Businessolver, 2026a).
The clearest organizational cost is attrition. Sull, Sull, and Zweig (2022) analyzed 34 million online employee profiles to estimate attrition at more than 500 large U.S. companies during the early months of the so-called Great Resignation in 2021. Toxic corporate culture was by far the strongest predictor of industry-adjusted attrition, roughly ten times more predictive than compensation, which ranked sixteenth among the topics they examined. In a follow-up analysis, the same research group estimated that turnover attributable to toxic culture cost U.S. employers nearly $50 billion a year even before the pandemic-era surge in quits (Sull, Sull, Cipolli, & Brighenti, 2022). The Businessolver survey points in a similar direction from a different angle: 66% of employees said they would leave for a more empathetic workplace even with a pay cut of up to 20%, rising to 73% among those in toxic cultures (Businessolver, 2026a).
Peer-reviewed meta-analyses reinforce this. Schyns and Schilling (2013), synthesizing 57 studies, found that destructive leadership was positively related to turnover intentions, resistance toward the leader, and counterproductive work behavior, and negatively related to individual performance. Its strongest association was with negative attitudes toward the leader, followed by counterproductive behavior. Tepper's (2000) foundational study found that subordinates of more abusive supervisors were more likely to quit, and that those who stayed showed lower commitment and higher distress, effects largely explained by perceptions of injustice. Mackey et al. (2017) consolidated two decades of abusive supervision research into a meta-analysis that positioned justice perceptions as a central mechanism linking abusive behavior to poor outcomes.
Psychological contract breach has similar consequences. Zhao et al.'s (2007) meta-analysis found that breach is strongly related to mistrust of management and to lower job satisfaction, moderately related to lower commitment and higher intention to leave, and modestly related to lower in-role performance and citizenship behavior. When an organization talks about empathy and then cuts benefits or announces AI-driven layoffs without explanation, it is creating exactly the kind of breach this research describes.
A subtler organizational cost is the loss of candor. Edmondson's (1999) study of 51 manufacturing teams showed that psychological safety predicted learning behavior, such as seeking feedback and discussing errors, which in turn predicted team performance. When employees fear their leaders or colleagues, they stop raising problems. In most industries that means lost improvement ideas. In safety-critical industries, it can mean catastrophe.
Boeing illustrates this connection between culture and operational risk. In February 2024, the Federal Aviation Administration released the report of an expert panel convened under Section 103 of the Aircraft Certification, Safety, and Accountability Act, which had reviewed Boeing's safety management system and safety culture after the two fatal 737 MAX crashes (Forecast International, 2024). Drawing on roughly 4,000 pages of documents, seven surveys, and more than 250 interviews, the panel issued 27 findings and 53 recommendations. It reported a disconnect between senior management and other employees on safety culture, found that employees hesitated to report concerns for fear of retaliation, and noted that the structure for employees with delegated FAA authority still allowed opportunities for retaliation (CNN, 2024). Boeing responded that it had taken steps to strengthen its safety culture but that there was more work to do. Boeing's case does not show that every toxic culture produces safety failures. It shows how quickly suppressed voice can become an enterprise-level risk.
Legal and regulatory costs are a further channel. Wells Fargo agreed in February 2020 to pay $3 billion to resolve criminal and civil investigations by the U.S. Department of Justice and the Securities and Exchange Commission into sales practices in its Community Bank between 2002 and 2016, conduct the company tied to product-based sales goals (Wells Fargo & Company, 2020). Four years earlier, the Consumer Financial Protection Bureau had fined the bank $100 million after finding that employees had opened roughly 1.5 million unauthorized deposit accounts and around half a million unauthorized credit card accounts to hit sales targets and earn bonuses; about 5,300 employees had been fired over such behavior (Doerer, 2016). The pressure that drove this conduct came from the top of the system, while the firings fell mainly on front-line staff.
In professional and membership organizations, the cost can be existential. The Confederation of British Industry (CBI), the UK's most prominent business lobby, dismissed its director general in April 2023 after an independent investigation by the law firm Fox Williams into workplace misconduct complaints, and acknowledged "serious failings" in how it had acted as an organization (Confederation of British Industry, 2023). Within weeks, major members including Aviva, John Lewis, and EY had cut ties, others paused membership, and government ministers stopped engaging with the organization (City A.M., 2023). The CBI's president conceded that the organization had been complacent and that senior leaders had tried to resolve harassment cases internally rather than remove offenders.
Individual Wellbeing and Stakeholder Impacts
For individuals, the evidence on toxic work environments is consistent and sobering. Nielsen and Einarsen's (2012) meta-analysis of workplace bullying, covering 66 independent cross-sectional samples with more than 77,000 participants, found that exposure to bullying was associated with mental and physical health problems, symptoms of post-traumatic stress, burnout, higher intentions to leave, and lower job satisfaction and commitment. Their longitudinal analysis, based on 13 samples, found that bullying predicted later mental health problems. It also found that baseline mental health problems predicted later reports of bullying, a reminder that vulnerable employees can become targets and that the relationship runs in both directions.
Tepper's (2000) work showed that abusive supervision spills beyond work, being linked with lower life satisfaction and greater work-family conflict. Schyns and Schilling (2013) likewise found that destructive leadership was negatively associated with follower wellbeing. These are not marginal effects that disappear once someone leaves the office. They follow people home.
The Businessolver AI data suggest a newer source of strain. Among employees who were concerned about AI's effect on their future, 64% reported a mental health issue in the past year, compared with 50% of those who felt optimistic (Businessolver, 2026b). Because employees in toxic cultures were also much more likely to worry about their future with the organization because of AI (51% vs. 39% overall), the AI transition and the culture problem appear to compound each other.
Executives are not immune. The finding that 73% of executives in toxic cultures reported feeling intimidated by coworkers suggests that toxic cultures harm the people nominally running them (Businessolver, 2026a). Gallup (2026) found that 31% of leaders reported loneliness on the previous day. A culture of intimidation at the top is unlikely to produce leaders who have the emotional bandwidth to model empathy for others.
The effects extend to stakeholders beyond employees. Customers bore the direct costs of Wells Fargo's unauthorized accounts. Airline passengers and regulators have a stake in Boeing's safety culture. In the video game industry, toxic cultures have become public controversies affecting players, partners, and investors. Activision Blizzard was sued in 2021 by the California Civil Rights Department, which alleged that the company paid women less than men for substantially similar work and denied them promotion opportunities; the company agreed in December 2023 to a settlement of about $54.9 million (California Civil Rights Department, 2023). Separately, a federal court approved an $18 million consent decree in March 2022 resolving U.S. Equal Employment Opportunity Commission (EEOC) allegations of severe or pervasive sexual harassment, pregnancy discrimination, and retaliation (Golden, 2021). Riot Games, maker of League of Legends, agreed in December 2021 to pay more than $100 million to resolve state allegations of systemic sex discrimination and harassment affecting roughly 1,065 women employees and 1,300 women contractors (California Civil Rights Department, 2021).
Ford Motor Company provides an example from manufacturing. In 2017 the EEOC found reasonable cause to believe that female and African American employees at two Chicago plants had been subjected to sexual and racial harassment and that the company had retaliated against those who complained. Ford agreed to pay up to $10.125 million and to five years of training, monitoring, and reporting obligations, without admitting liability (U.S. Equal Employment Opportunity Commission, 2017).
The pattern across these cases is that the people with the least power absorb the most harm, the organization absorbs legal and reputational costs later, and leadership often learns the scale of the problem last. That is the empathy deficit in its most concrete form.
Evidence-Based Organizational Responses
There is no single intervention that turns a toxic culture into a healthy one. The research and the cases suggest instead a set of mutually reinforcing responses that address measurement, incentives, voice, leadership capability, and independent oversight. Table 1 summarizes them; the subsections that follow discuss each in turn.
Table 1: Evidence-Based Responses to Toxic Culture and the Empathy Deficit
Response | Evidence Base | Key Practices | Illustrative Examples |
Diagnose culture at the level where harm occurs | Sull, Sull, Cipolli, & Brighenti (2022); Sull, Sull, & Zweig (2022); Robinson & Rousseau (1994) | Granular, segmented culture measurement; listening exercises; text analysis of employee feedback | Uber (Holder review); CBI (listening exercise) |
Reward how results are achieved, not only what | Schyns & Schilling (2013); Einarsen et al. (2007); Mackey et al. (2017) | Remove harmful targets; behavioral criteria in reviews and pay; leader accountability | Wells Fargo (ended product sales goals); Activision Blizzard (EEO criteria in manager reviews) |
Protect voice and make speaking up safe | Edmondson (1999); Andersson & Pearson (1999); Porath & Pearson (2013) | Independent reporting channels; anti-retaliation safeguards; early response to incivility | Boeing (FAA panel findings); CBI (permanent confidential channel) |
Develop empathy as a leadership skill | Kock et al. (2019); Muss et al. (2025); Van Bommel (2021) | Perspective-taking practice; empathetic communication of hard news; role modeling | Microsoft (Nadella-era culture shift) |
Install independent oversight and structural repair | Zhao et al. (2007); Tepper (2000); Nielsen & Einarsen (2012) | External monitors; pay-equity audits; time-bound remediation commitments | Riot Games (monitor and equity expert); Ford (EEOC monitoring) |
Diagnose Culture at the Level Where Harm Occurs
Most organizations measure culture through annual engagement surveys and report the results as company-wide averages. The research suggests this is exactly where toxicity hides. Sull, Sull, Cipolli, and Brighenti (2022) found, for example, that only around 0.8% of employees described their manager as abusive in Glassdoor reviews, but those who did rated their overall culture markedly lower. Averages wash out small, intense pockets of harm. Their recommendation is to measure culture at the level of the individual and the team, to look for the specific elements that make a culture toxic, and to focus remediation on the issues that cause the most pain rather than trying to fix every irritant at once.
The Businessolver data illustrate another measurement problem: the perception gap between executives and employees. When 98% of executives describe themselves as empathetic while a third of employees report intimidation, self-assessment by leaders is clearly not a reliable instrument (Businessolver, 2026a). Robinson and Rousseau's (1994) work on psychological contracts reinforces the point that what matters is the employee's perception of obligations and breaches, not management's account of its intentions.
Effective approaches:
Segmented measurement: Report culture and engagement results by team, manager, level, gender, race, and job type so that pockets of harm are visible rather than averaged away.
Behavior-specific items: Ask about concrete behaviors, such as being belittled, excluded, or pressured to cut ethical corners, rather than relying only on global satisfaction scores.
Unstructured feedback analysis: Use text analysis of open comments, exit interviews, and grievance data to identify recurring themes, as Sull and colleagues did with public reviews.
Executive-employee gap tracking: Compare leaders' self-ratings of empathy with employees' ratings of those same leaders and treat large gaps as a risk indicator.
Listening sessions with protection: Run facilitated, confidential listening exercises, especially after a crisis, with a commitment to report back what was heard.
Uber offers an early and instructive example. In 2017, amid allegations of sexual harassment and broader cultural problems, the company retained former U.S. Attorney General Eric Holder and his colleagues at Covington & Burling to investigate its workplace culture. Uber's board unanimously adopted all of the resulting recommendations in June 2017 (Associated Press, 2017). The recommendations included clearer anti-harassment policies, better HR record-keeping and tracking of complaints, using performance reviews to hold leaders accountable, increasing workforce diversity, and revising cultural values that had been used to justify poor behavior. Founder Travis Kalanick announced an indefinite leave of absence the same week, acknowledging that he needed to become a better leader. The Holder review mattered because it treated culture as something to be investigated with evidence, much like a financial irregularity, rather than as a matter of opinion.
The CBI, in its April 2023 response to the Fox Williams investigation, committed to a review of culture, governance, and processes that would begin with a listening exercise among staff, overseen by board members (Confederation of British Industry, 2023). Whether that listening exercise ultimately rebuilt trust is a separate question, but the sequencing was right: listen first, then redesign.
A common pitfall: Treating a culture survey as a referendum on leadership's popularity rather than as a diagnostic tool. When results are bad, leaders sometimes question the methodology, delay publication, or announce generic programs. Employees read that response as further evidence that speaking up is pointless, which makes the next survey less honest.
Reward How Results Are Achieved, Not Only What
The most robust finding in the destructive leadership literature is that bad leadership behavior is rarely an individual quirk. It is sustained by systems that reward outcomes and ignore methods. Einarsen et al. (2007) explicitly allow for leaders who are "pro-organizational" in their pursuit of financial goals while being anti-subordinate in their treatment of people. Schyns and Schilling (2013) called for more research on what triggers destructive leadership, and their review points toward situational pressures and organizational tolerance as recurring factors. The 2026 Businessolver finding that toxic cultures cluster with reported financial growth (Businessolver, 2026a) is a warning that incentive systems may be rewarding the very behaviors that will erode performance later.
Effective approaches:
Audit targets for harm: Review sales, productivity, and cost targets for whether they can realistically be met without misconduct, coercion, or burnout.
Behavioral performance criteria: Include how a leader achieves results, such as team retention, upward feedback, and complaint patterns, in performance reviews and incentive pay.
Clawbacks and consequences: Ensure that leaders who deliver numbers through abusive or unethical means face real consequences, including the loss of variable pay.
Promote for character: Make respectful, inclusive leadership an explicit requirement for promotion to people-management roles.
Close the "brilliant jerk" exemption: Apply standards consistently to high performers, since exceptions teach everyone else what the organization actually values.
Wells Fargo is the clearest example of what happens when targets overwhelm conduct, and of the structural changes that followed. In announcing its 2020 settlement with the Department of Justice and the SEC, the company listed the changes it had made since 2016: a new CEO and largely new operating committee, eight new independent directors, the elimination of all product-based sales goals for retail bankers, a new incentive structure tied to customer outcomes and risk accountability, enhanced customer consent processes, and more than 800,000 hours of learning and development for retail employees (Wells Fargo & Company, 2020). The scale of those changes reflects how deeply the incentive system had shaped behavior.
Activision Blizzard's 2022 EEOC consent decree required the company to add an equal employment opportunity component to performance reviews for managers, supervisors, and HR staff, alongside an internal EEO coordinator, a third-party consultant, and unannounced harassment audits (Golden, 2021). Building conduct into the performance system turned an aspiration into an evaluated expectation.
A common pitfall: Firing front-line employees for misconduct while leaving the targets and incentives that drove it untouched. Wells Fargo's dismissal of thousands of employees before structural reform shows the limit of that approach: it treats symptoms while the system continues to generate the same behavior.
Protect Voice and Make Speaking Up Safe
Toxic cultures are self-concealing. People who are belittled, harassed, or pressured into misconduct often stay silent because they fear retaliation or doubt anything will change. Edmondson's (1999) research established psychological safety as a precondition for learning: teams that feel safe raise errors and concerns, and those that do not, bury them. Andersson and Pearson (1999) showed how incivility, left unchecked, can spiral into more serious aggression, which argues for responding to low-level disrespect early rather than waiting for formal complaints. Porath and Pearson's (2013) reporting on the rising frequency of rudeness suggests that many organizations are failing to intervene at that early stage.
Effective approaches:
Multiple, independent channels: Offer reporting routes outside the line management chain, including confidential and anonymous options managed by an independent party.
Anti-retaliation safeguards: Monitor the careers, pay, and assignments of people who raise concerns and investigate any adverse changes.
Close the loop: Tell reporters what happened as a result of their concern, within the limits of confidentiality, so speaking up feels worthwhile.
Respond to small signals: Train managers to address incivility, exclusion, and disrespect when it first appears, before it becomes a formal case.
Separate investigation from line authority: Ensure that the people investigating complaints are not managed or evaluated by those being investigated.
Boeing shows what happens when these safeguards are weak. The FAA expert panel found that employees hesitated to report safety concerns for fear of retaliation, and that employees holding delegated FAA authority still had their salaries and furlough rankings controlled by management, which the panel said left opportunities for retaliation (CNN, 2024; Forecast International, 2024). Among other things, the panel's 53 recommendations aimed to make roles in safety reporting clearer and to strengthen protections for those exercising independent judgment. For any organization, the lesson is that a reporting system is only as good as the protections around the people who use it.
The CBI's response illustrates the repair side. Its board announced in April 2023 that the independent, confidential channel set up during the crisis would become permanent, and it created a new chief people officer role reporting directly to the board on workplace conduct and culture (Confederation of British Industry, 2023). Its president also acknowledged that senior leaders had previously tried to handle harassment complaints internally, which allowed offenders to feel they would not be caught (City A.M., 2023). The structural change, a reporting line that bypasses the executive team, addressed that specific failure.
A common pitfall: Launching a hotline and then measuring success by a falling number of reports. In a culture that is improving, reporting often rises at first because people begin to trust the system. Falling reports in a culture with known problems may mean that people have given up.
Develop Empathy as a Leadership Skill
Empathy is sometimes treated as a fixed personality trait that leaders either have or lack. The evidence suggests it is better understood as a set of behaviors that can be practiced, especially in how leaders communicate. Kock et al. (2019) examined empathetic leadership in a study of leaders' language and found preliminary causal support for a model in which leaders' empathetic communication improved follower performance through job satisfaction and innovation. Muss et al. (2025), reviewing 42 studies, concluded that empathy in leadership is predominantly associated with positive outcomes, including follower wellbeing, lower burnout, more favorable attitudes, perceptions of fairness when bad news is delivered, and better performance. They also noted mixed findings, such as the risk that excessive empathy is perceived as weak in crises.
Catalyst's survey of nearly 900 U.S. employees found that empathy from leaders was an important driver of innovation, engagement, and inclusion, particularly during crises (Van Bommel, 2021). In a summary of the research, 76% of employees whose boss showed empathy reported being engaged at work, compared with 32% of those whose boss did not (Broom, 2021). This is still survey evidence, but it aligns with the more rigorous studies.
The Businessolver finding that 61% of CEOs now find empathy difficult to demonstrate day to day deserves a sympathetic reading (Businessolver, 2026b). Leaders are being asked to make harder trade-offs, often about headcount and AI, while also being expected to show care. Empathy development, in that context, is less about feeling more and more about learning to deliver difficult decisions in ways that preserve dignity and trust.
Effective approaches:
Perspective-taking practice: Build structured opportunities for leaders to hear directly from employees at different levels, including through shadowing and reverse mentoring.
Empathetic delivery of hard news: Train leaders to explain the reasons for difficult decisions, acknowledge their impact, and offer concrete support, since perceived fairness in delivery shapes reactions.
Language and listening skills: Coach leaders on emotionally attuned communication, which Kock et al. (2019) identified as the mechanism for performance effects.
Role modeling from the top: Senior executives' behavior in meetings, emails, and crises sets the norm more powerfully than any training program.
Support for leaders themselves: Address leader stress and isolation, since depleted leaders have less capacity for empathy.
Microsoft is the example most often cited for an empathy-centered culture change, and it is worth examining carefully. When Satya Nadella became CEO in 2014, Microsoft was widely seen as losing relevance and hampered by internal conflict. Drawing on Carol Dweck's research on mindset, Nadella encouraged a move from a "know-it-all" to a "learn-it-all" culture (Ibarra, 2016). According to an account of the transformation published in Fortune, he asked top executives to read Marshall Rosenberg's Nonviolent Communication early in his tenure to encourage compassion rather than internal competition, and he refused to tolerate yelling in executive meetings (Wadhwa et al., 2021). The same account describes him treating empathy as central to revamping the company's culture.
Microsoft's story should not be romanticized. A culture shift does not remove the need for difficult decisions about strategy, structure, and jobs. What the Microsoft case shows is that a CEO can redefine what counts as good leadership, model it consistently, and connect it to business strategy so that empathy is seen as a source of performance rather than a distraction from it.
A common pitfall: Running one-off empathy workshops while the incentive system, the layoff process, and the behavior of senior leaders all signal something different. Employees quickly recognize when empathy language is decoration. Indeed, the gap between leaders' self-perception and employees' experience in the Businessolver data suggests that many organizations already have empathy rhetoric without empathy practice.
Install Independent Oversight and Structural Repair
When a culture has caused serious harm, internal good intentions are often not enough. Employees who have experienced breach of the psychological contract are, according to meta-analytic evidence, strongly inclined to mistrust management (Zhao et al., 2007). Promises from the same leadership that presided over the harm may carry little weight. Independent oversight, external monitoring, pay-equity audits, and time-bound remediation commitments provide a way to rebuild credibility because they do not depend on employees trusting management's word. The individual health consequences of bullying and abuse documented by Nielsen and Einarsen (2012) and Tepper (2000) also argue for remediation that addresses harm done, not only future prevention.
Effective approaches:
External monitors: Appoint an independent monitor with access to complaint data and authority to report to the board or a regulator.
Pay and promotion equity audits: Commission independent analyses of pay, assignment, and promotion patterns, and commit in advance to remedy unexplained gaps.
Remediation for those harmed: Provide compensation, restored opportunities, or career support to people affected by past conduct.
Time-bound public commitments: Set specific, measurable targets with deadlines and report progress publicly.
Board-level ownership: Assign a board committee explicit responsibility for culture and conduct, with regular reporting.
Riot Games agreed, as part of its December 2021 settlement with California, to an independent expert who would conduct annual gender-equity analyses of pay, job assignments, and promotions for three years, with an obligation to remedy unjustified disparities; an independent monitor to audit workplace protections and complaint investigations annually; a $6 million annual reserve for pay adjustments and diversity programs; and offers of 40 full-time positions to qualified former contract workers (California Civil Rights Department, 2021). Those terms go beyond financial compensation to change the underlying systems that produced inequity.
Ford's 2017 conciliation agreement with the EEOC similarly required five years of regular training, continued policy distribution, workforce monitoring for harassment, and reporting to the EEOC on complaints at its Chicago plants (U.S. Equal Employment Opportunity Commission, 2017). Activision Blizzard's December 2023 settlement with California required an independent consultant to evaluate and recommend changes to compensation and promotion policies and training (California Civil Rights Department, 2023).
A common pitfall: Treating oversight as a legal burden to be minimized rather than as a credibility asset. Organizations that negotiate monitors down to the narrowest possible scope, or that comply with the letter but not the spirit of remediation, often find that the cultural problems resurface once the oversight period ends.
Building Long-Term Empathic Leadership Capability
The responses above address immediate risks. Sustaining a healthy culture requires durable capability: governance that treats culture as a strategic risk, leadership pipelines that select and develop for empathy, and a deliberate approach to the AI transition, which is now one of the biggest tests of whether empathy survives under pressure.
Treating Culture as a Board-Level Risk
Boards routinely oversee financial, cyber, and regulatory risk. Culture deserves the same attention, because its failures can be just as costly and are usually visible earlier to those who know where to look. The Businessolver data suggest that self-reported financial growth can coexist with rising intimidation and declining psychological safety (Businessolver, 2026a). A board that looks only at financial dashboards would see success; a board that also looks at segmented culture data, attrition by manager, and the pattern of complaints would see risk accumulating.
The organizational cases bear this out. Uber's board adopted independent directors and an independent chair as part of its cultural reset (Associated Press, 2017). Wells Fargo added eight new independent directors (Wells Fargo & Company, 2020). The CBI created a chief people officer reporting directly to the board (Confederation of British Industry, 2023). In each case, governance changed after a crisis. The opportunity for other organizations is to make these changes before one.
Practically, this means boards should receive regular culture reports that include leading indicators: upward feedback scores for senior leaders, regretted attrition by team, speak-up and retaliation data, and the gap between executive self-assessment and employee experience. Sull, Sull, and Zweig's (2022) finding that toxic culture predicts attrition far more strongly than pay suggests that culture metrics belong alongside compensation benchmarks in board discussions of talent risk.
Selecting and Developing Leaders for Empathy
Leadership pipelines tend to reward technical excellence, decisiveness, and the ability to deliver results. Those qualities matter, but they are not sufficient. The destructive leadership literature shows that leaders who deliver results through harmful behavior impose costs on followers' wellbeing and on the organization's long-term health (Schyns & Schilling, 2013). Gallup's (2026) finding that manager engagement fell to 22% suggests that many managers are themselves struggling, which makes it harder for them to support others.
Long-term capability requires changes at every stage of the pipeline. Selection processes should assess how candidates treat people, using structured references, upward and peer feedback, and behavioral interviews focused on how they have handled conflict and bad news. Development programs should treat empathy as a practiced skill, building on evidence that leaders' empathetic communication affects follower outcomes (Kock et al., 2019; Muss et al., 2025). Promotion decisions should include evidence about team climate, not only about results.
Organizations should also invest in managers' own wellbeing and support. Leaders who are stressed, lonely, or intimidated, as many in the Businessolver and Gallup data report, are poorly placed to model empathy. Peer support, coaching, realistic workloads, and spans of control that allow time for people management all contribute to leaders' capacity to lead well.
Making the AI Transition Humane
The AI transition is where the empathy deficit is now most acute. The Businessolver data show a sharp divide: employees in empathetic cultures were far more likely to report adequate AI training, while employees in toxic cultures were more likely to report anxiety, disrupted careers, and extra clean-up work created by AI (Businessolver, 2026b). Employees with adequate training reported markedly better outcomes, including greater confidence in keeping up with AI trends (80% vs. 57%) and more optimism about their future at the organization (73% vs. 49%).
These findings suggest that how organizations introduce AI is itself a culture decision. Framing AI primarily as a headcount-reduction tool, as 28% of CEOs said was their primary motivation, risks creating exactly the kind of psychological contract breach that research links to mistrust and withdrawal (Businessolver, 2026b; Zhao et al., 2007). Framing it as a capability investment, with fair access to training and honest communication about how roles will change, offers a more sustainable path.
Practical steps include ensuring equitable access to AI training across gender, job type, and level; being transparent about which roles are likely to change and how affected employees will be supported; involving employees in redesigning workflows rather than simply imposing tools; and monitoring whether AI is adding hidden work, such as checking and correcting outputs, that increases strain. Leaders who handle the AI transition with empathy may find that it strengthens trust rather than eroding it.
Conclusion
The 2026 Businessolver survey captured a striking moment: a third of CEOs describing their own cultures as toxic, two-thirds of employees saying they would trade pay for empathy, and a clear majority of chief executives admitting that showing empathy has become harder (Businessolver, 2026a, 2026b). It is a vendor survey with obvious limitations, and its association between toxicity and short-term financial growth should not be read as causal. What makes it important is that its direction matches decades of peer-reviewed research on abusive supervision, destructive leadership, incivility, bullying, and psychological contract breach, all of which link harmful leadership to attrition, ill health, silence, and organizational risk (Mackey et al., 2017; Nielsen & Einarsen, 2012; Schyns & Schilling, 2013; Zhao et al., 2007).
The organizational cases show how these costs eventually arrive. Wells Fargo paid billions after sales pressure produced widespread misconduct. Boeing's expert panel found employees hesitant to report safety concerns. The CBI lost members and government access within weeks. Activision Blizzard, Riot Games, and Ford faced settlements and years of oversight. Uber needed an external investigation to begin resetting its culture. In each case, the warning signs were visible to employees long before they were acknowledged by leadership.
The empathy deficit is not mainly a shortage of good intentions. Nearly every executive in the Businessolver sample considered themselves empathetic. The deficit is a gap between belief and practice, widened by performance pressure, cost-cutting, and the AI transition, and sustained by systems that reward results without asking how they were achieved. Closing it requires measuring culture where harm occurs, aligning incentives with conduct, protecting voice, building empathy as a skill, and accepting independent oversight where trust has been damaged.
Measure where harm happens: Replace company-wide averages with segmented culture data, and treat gaps between leaders' self-ratings and employees' experience as a risk indicator.
Pay for how, not only what: Remove targets that cannot be met without misconduct and build conduct, retention, and upward feedback into leaders' performance reviews and incentives.
Make speaking up safe and worthwhile: Provide independent channels, track the careers of those who report, and act early on incivility before it escalates.
Treat empathy as a practiced skill: Train leaders to deliver hard decisions with explanation, acknowledgment, and support, and model this behavior from the top.
Invite independent oversight when trust is broken: Use monitors, equity audits, and public commitments to rebuild credibility that internal promises cannot.
Handle AI with care: Provide equitable training, communicate honestly about role changes, and avoid framing AI only as a way to cut headcount.
Put culture on the board agenda: Give boards regular leading indicators of culture health alongside financial metrics, before a crisis forces the issue.
Research Infographic
References
Aasland, M. S., Skogstad, A., Notelaers, G., Nielsen, M. B., & Einarsen, S. (2010). The prevalence of destructive leadership behaviour. British Journal of Management, 21(2), 438–452.
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? The spiraling effect of incivility in the workplace. Academy of Management Review, 24(3), 452–471.
Associated Press. (2017, June 13). Uber recommendations show company was "out of control." KATV.
Broom, D. (2021, October 11). This is the most important skill for a leader to have right now. World Economic Forum.
Businessolver. (2026a, June 23). New Businessolver research reveals alarming trend: Toxic workplaces are winning—for now [Press release].
Businessolver. (2026b, August 11). 3 in 10 CEOs say cutting headcount is primary driver of AI investment even as most say AI demands more empathy [Press release].
Businessolver. (2026c). 2026 State of Workplace Empathy report [Report overview].
California Civil Rights Department. (2021, December 27). Riot Games, Inc. agrees to $100 million settlement and systemic reforms to resolve allegations of workplace sex discrimination and harassment [Press release].
California Civil Rights Department. (2023, December 15). Civil Rights Department announces settlement agreement to resolve employment discrimination and equal pay lawsuit against Activision Blizzard [Press release].
City A.M. (2023, April 24). CBI admits group failed to root out "toxic" workers amid sexual misconduct crisis. City A.M.
CNN. (2024, February 26). Aviation safety panel finds Boeing culture included safety gaps, fear of retaliation. CNN Business.
Confederation of British Industry. (2023, April 11). Announcement from the CBI Board [Press release].
Doerer, K. (2016, September 8). Wells Fargo fined for creating fake accounts, other illegal practices. PBS NewsHour.
Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383.
Einarsen, S., Aasland, M. S., & Skogstad, A. (2007). Destructive leadership behaviour: A definition and conceptual model. The Leadership Quarterly, 18(3), 207–216.
Fair Play Talks Editorial Team. (2026, October 9). One in three CEOs now describe their own workplace culture as toxic. Fair Play Talks.
Forecast International. (2024, February 27). ODA review panel issues report on Boeing's safety culture. Flight Plan.
Gallup. (2026). State of the global workplace: 2026 report. Gallup, Inc.
Golden, R. (2021, October 4). Judge approves Activision Blizzard's $18M EEOC settlement over harassment, discrimination. HR Dive.
Ibarra, H. (2016). Satya Nadella at Microsoft: Instilling a growth mindset (Case No. CS-18-008). London Business School.
Kock, N., Mayfield, M., Mayfield, J., Sexton, S., & De La Garza, L. M. (2019). Empathetic leadership: How leader emotional support and understanding influences follower performance. Journal of Leadership & Organizational Studies, 26(2), 217–236.
Lipman-Blumen, J. (2005). The allure of toxic leaders: Why we follow destructive bosses and corrupt politicians—and how we can survive them. Oxford University Press.
Lundmark, R., Stenling, A., von Thiele Schwarz, U., & Tafvelin, S. (2021). Appetite for destruction: A psychometric examination and prevalence estimation of destructive leadership in Sweden. Frontiers in Psychology, 12, Article 668838.
Mackey, J. D., Frieder, R. E., Brees, J. R., & Martinko, M. J. (2017). Abusive supervision: A meta-analysis and empirical review. Journal of Management, 43(6), 1940–1965.
Muss, C., Tüxen, D., & Fürstenau, B. (2025). Empathy in leadership: A systematic literature review on the effects of empathetic leaders in organizations. Management Review Quarterly. Advance online publication.
Nielsen, M. B., & Einarsen, S. (2012). Outcomes of exposure to workplace bullying: A meta-analytic review. Work & Stress, 26(4), 309–332.
Porath, C., & Pearson, C. (2013). The price of incivility. Harvard Business Review, 91(1–2), 114–121.
Robinson, S. L., & Rousseau, D. M. (1994). Violating the psychological contract: Not the exception but the norm. Journal of Organizational Behavior, 15(3), 245–259.
Schyns, B., & Schilling, J. (2013). How bad are the effects of bad leaders? A meta-analysis of destructive leadership and its outcomes. The Leadership Quarterly, 24(1), 138–158.
Sull, D., Sull, C., Cipolli, W., & Brighenti, C. (2022). Why every leader needs to worry about toxic culture. MIT Sloan Management Review.
Sull, D., Sull, C., & Zweig, B. (2022). Toxic culture is driving the Great Resignation. MIT Sloan Management Review.
Tepper, B. J. (2000). Consequences of abusive supervision. Academy of Management Journal, 43(2), 178–190.
U.S. Equal Employment Opportunity Commission. (2017, August 15). Ford Motor Company to pay $10.125 million to settle EEOC harassment investigation [Press release].
Van Bommel, T. (2021). The power of empathy in times of crisis and beyond. Catalyst.
Wadhwa, V., Amla, I., & Salkever, A. (2021, December 21). How Microsoft made the stunning transformation from Evil Empire to Cool Kid. Fortune.
Wells Fargo & Company. (2020, February 21). Wells Fargo reaches settlements to resolve outstanding DOJ and SEC investigations related to historical Community Bank sales practices [Press release, Exhibit 99.1 to Form 8-K].
Zhao, H., Wayne, S. J., Glibkowski, B. C., & Bravo, J. (2007). The impact of psychological contract breach on work-related outcomes: A meta-analysis. Personnel Psychology, 60(3), 647–680.
Jonathan H. Westover, PhD, Chief Research Officer (Nexus Institute for Work and AI); Co-Founder & Chief Workforce and Learning Officer (Future State University); Founder & CEO (Human Capital Innovations); Professor of Organizational Leadership & Change (UVU). Read Jonathan Westover's executive profile here.
Suggested Citation: Westover, J. H. (2026). Toxic at the Top: Why Self-Reported Workplace Toxicity Is Rising and How Leaders Can Close the Empathy Deficit. Human Capital Leadership Review, 40(1). doi.org/10.70175/hclreview.2020.40.1.4
