The Quiet Retreat: Why Leaders Are Withdrawing — and What the Science of Agency Erosion Reveals
Psychological withdrawal in leadership is the progressive disengagement of senior executives from the cognitive and emotional demands of their roles, driven not by incompetence but by a sustained erosion of perceived agency. It manifests as reduced initiative, narrowed strategic thinking, and emotional retreat from the responsibilities that define high-stakes leadership — and the neuroscience and behavioural evidence suggest it is accelerating across industries and geographies in ways that organisations are only beginning to understand.
In 2025, chief executive departures across the S&P 500 reached a projected annual rate of thirteen per cent, well above the ten per cent recorded in 2024, while the Challenger tracking firm logged 446 CEO exits in the United States alone — the highest annual total since it began recording such data in 2002. What makes these figures particularly striking is that turnover is no longer concentrated among underperformers. According to The Conference Board, successions at firms in the top three performance quartiles by total shareholder return jumped from seven per cent in 2024 to twelve per cent in 2025. The conventional explanation — that boards are becoming more demanding, that activist campaigns are intensifying (a record 32 CEOs resigned within a year of an activist campaign in 2025), that economic volatility is compressing the grace period afforded to new leaders — captures part of the story. But it does not explain why leaders who are succeeding by every external metric are choosing to leave, nor why those who remain describe an unfamiliar internal experience: a creeping sense that their capacity to shape outcomes has quietly diminished, that the effort required to maintain strategic clarity is no longer matched by the agency they once felt. Merete Wedell-Wedellsborg, adjunct professor of leadership at IMD Business School, characterised this in Harvard Business Review earlier this year as a "pullback driven by lost illusions and a feeling of disempowerment" — a phenomenon she observed not among failing leaders but among high performers who had, over successive quarters of unrelenting demand, started to withdraw from the very decisions their organisations needed them to make.
The behavioural science literature offers a more precise vocabulary for what is happening. Learned helplessness, the concept introduced by Martin Seligman and Steven Maier in 1967 and subsequently refined through decades of experimental and clinical work, describes a state in which repeated exposure to uncontrollable outcomes leads an individual to cease attempting to exert influence, even when circumstances change and control becomes possible. The original research involved animals subjected to inescapable stress, but the translation to human cognition — and specifically to organisational settings — is well established. A body of organisational psychology research demonstrates that when employees perceive their decisions as having no meaningful impact on outcomes, they exhibit reduced initiative, lower creativity, and a measurable decline in problem-solving capacity. What has received less attention, until recently, is the applicability of this framework to the most senior tier of organisational life. The assumption has been that executives, by definition, possess agency — they command resources, set strategy, hire and dismiss. But agency is not merely structural. It is psychological. And when the gap between formal authority and felt influence widens — when a chief executive retains the title and the compensation but experiences a progressive inability to translate intention into outcome — the neurological and behavioural consequences mirror those documented in the helplessness literature with uncomfortable precision.
The neuroscience of chronic stress illuminates the mechanism. Amy Arnsten's laboratory at Yale has demonstrated across multiple studies that the prefrontal cortex, the brain region responsible for executive function — working memory, impulse control, abstract reasoning, and long-range strategic planning — is exquisitely sensitive to sustained stress exposure. Acute stress triggers catecholamine release that, in moderate doses, sharpens attentional focus. But chronic exposure to uncontrollable stress, as Arnsten's group published in Nature Reviews Neuroscience, drives high levels of norepinephrine and dopamine that activate feedforward calcium-cAMP signalling pathways, rapidly weakening synaptic connectivity in the prefrontal cortex. Over weeks and months, this produces measurable dendritic atrophy — a physical shrinking of the neural architecture that supports the highest-order cognitive functions. A 2021 review in Chronic Stress by Woo, Sansing, Arnsten, and Datta confirmed that chronic stress weakens connectivity in the prefrontal cortex through both architectural and molecular changes, with downstream effects on working memory, emotional regulation, and decision-making quality. The practical implication for leadership is direct: the sustained cognitive load borne by senior executives — the relentless context-switching between geopolitical risk, AI transformation, regulatory complexity, talent retention, and stakeholder management — does not merely feel exhausting. It degrades, at the cellular level, the neural substrate required for the strategic foresight and emotional discipline that define effective leadership under pressure. Long-range strategic projection is among the most metabolically expensive functions the prefrontal cortex performs, and it is among the first casualties of resource depletion.
This neurological reality intersects with a more recent development in organisational psychology: the recognition that executive burnout is qualitatively different from burnout at other levels of the organisation. An April 2026 analysis in Harvard Business Review argued that burnout manifests differently across the organisational chart — early-career employees burn out from ambiguity and lack of control, middle managers from responsibility without authority, and executives from value conflicts and moral strain. This last category deserves particular scrutiny. The concept of moral injury, originally developed in military psychology to describe the psychological damage caused by participating in or witnessing acts that violate one's moral code, has been increasingly applied to healthcare leadership and, more recently, to corporate settings. When senior leaders are required to implement strategies they privately believe to be harmful — restructurings that hollow out institutional capability, performance targets that demand behaviours misaligned with stated values, communications that require them to project confidence they do not feel — the resulting internal conflict does not resolve itself through rest or time management. It accumulates. Research published in the Journal of General Internal Medicine found that moral injury in leaders is significantly predicted by perceived lack of management support, poor ethical climate, and reduced empowerment to make role-congruent decisions — and that as moral injury progresses, leaders begin to experience internalised guilt, shame, and perceived powerlessness that contributes to withdrawal from decision-making responsibilities.
The behavioural expression of this withdrawal is subtle but consequential. A leader in the early stages of agency erosion does not typically announce disengagement. Instead, the pattern follows a recognisable trajectory that behavioural psychologists would identify as consistent with the cognitive reappraisal failures documented in the helplessness literature. Decision-making becomes more reactive and less proactive. Strategic discussions narrow from possibility-scanning to risk-avoidance. Delegation increases not as a function of trust but as a mechanism of retreat. The leader's communication style shifts toward what organisational psychologist Robert Kegan might describe as a socialized mind operating within a self-authoring role — deferring to consensus not because consensus is wise but because the internal resources required to hold and defend an independent position have been depleted. CNBC reported in May 2026 that 71 per cent of leaders in the DDI Global Leadership Forecast reported increased stress levels, up from 63 per cent in 2022. But the more telling finding may be qualitative: a growing number of executives describe not simply feeling overwhelmed but feeling that the relationship between their effort and its outcomes has fundamentally changed. This is the phenomenology of eroded agency — not exhaustion per se, but the collapse of the perceived causal link between action and consequence that sustains motivation and strategic engagement.
What the Evidence Demands of Organisations and Leaders
The implications extend well beyond individual resilience. If the phenomenon described here were confined to a handful of temperamentally vulnerable individuals, it could be addressed through executive coaching or sabbatical programmes. But the data suggest something structural. Russell Reynolds Associates' Global CEO Turnover Index for 2025 reported that the average tenure of departing CEOs had fallen to 6.3 years, below the 7.1-year global average, while external hires in the S&P 500 nearly doubled from 18 per cent to 33 per cent — the highest level in eight years. Organisations are not merely replacing leaders; they are cycling through them at an accelerating rate, importing outsiders on the assumption that fresh perspective will compensate for the conditions that depleted their predecessors. The behavioural science evidence suggests this is a category error. If the operating environment systematically erodes agency — through chronic cognitive overload, compressed decision timelines, value-incongruent performance demands, and the progressive attenuation of the link between executive action and organisational outcome — then replacing the individual without addressing the system will produce the same result, repeatedly. A 2026 study published in the British Journal of Social Psychology argued for a fundamental reconceptualisation of power, noting that traditional definitions of power as asymmetric resource control fail to capture its real-world complexity — power operates through multiple intersecting mechanisms, and its meaning varies dramatically across contexts. For leadership, this means that the felt experience of agency cannot be inferred from the formal authority of a title.
What, then, might a more serious organisational response look like? The evidence points in several directions simultaneously. First, the neuroscience of chronic stress suggests that cognitive load management at the senior executive level should be treated with the same rigour as financial risk management — not as a wellness initiative but as a governance priority. The prefrontal cortex cannot sustain indefinite demand without degradation, and organisations that treat executive cognitive capacity as an inexhaustible resource are making an empirically false assumption with strategic consequences. Second, the moral injury literature implies that organisations must attend not only to what they ask leaders to do but to the alignment between those demands and the leader's internalised values. Value-incongruent demands do not merely cause discomfort; they produce measurable psychological damage that impairs the very judgment the organisation depends upon. Third, the learned helplessness research suggests that restoring agency requires more than motivational exhortation. It requires the re-establishment of credible causal links between executive decisions and their outcomes — which may mean simplifying reporting structures, reducing the number of concurrent strategic priorities, and creating decision environments in which senior leaders can observe the consequences of their choices within timeframes that sustain the perception of influence. Leaders operating in environments where decisions carry significant consequence may wish to treat the preservation of their cognitive and psychological agency not as a personal indulgence but as a professional obligation — one that the organisations around them have a corresponding responsibility to support.
Frequently Asked Questions
What is psychological withdrawal in leadership?
Psychological withdrawal in leadership is the progressive disengagement of senior executives from the cognitive and emotional demands of their roles. It manifests as reduced initiative, narrowed strategic thinking, and retreat from decision-making — driven not by incompetence but by sustained erosion of perceived agency and the collapse of the felt link between effort and outcome.
Why are so many CEOs leaving their roles in 2025 and 2026?
CEO departures hit record levels in 2025, with 446 exits in the US alone. Contributing factors include activist campaigns, compressed grace periods, economic volatility, and executive burnout from chronic cognitive overload. Notably, turnover is rising even among high-performing firms, suggesting systemic rather than individual causes.
How does chronic stress affect a leader's brain?
Chronic stress triggers sustained catecholamine release that weakens synaptic connectivity in the prefrontal cortex — the brain region responsible for working memory, strategic planning, and emotional regulation. Over weeks and months, this produces measurable dendritic atrophy, physically degrading the neural architecture required for high-quality executive decision-making.
What is moral injury in executive leadership?
Moral injury in executive leadership occurs when senior leaders are repeatedly required to implement decisions that conflict with their internalised values — such as restructurings, misleading communications, or ethically ambiguous performance targets. The resulting psychological damage includes guilt, shame, and withdrawal from decision-making responsibilities.
How can organisations prevent leadership agency erosion?
Organisations can address agency erosion by treating executive cognitive load as a governance priority, ensuring alignment between strategic demands and leaders' values, simplifying reporting structures, reducing concurrent priorities, and creating decision environments where leaders can observe meaningful consequences of their choices within reasonable timeframes.
