When Leadership Ignores Biology, Systems Break

Why human operating limits should be treated as a core governance constraint

8/23/202616 min read

Author: Trang Phan

Executive summary

Modern organizations have become increasingly sophisticated at managing capital, technology, cybersecurity, supply chains, liquidity, infrastructure, and operational risk, yet they still treat one of the most consequential inputs into enterprise decision-making as effectively unlimited: the human nervous system. Senior leaders are expected to absorb rising information volumes, make high-stakes decisions continuously, manage conflict, travel, respond across time zones, maintain external confidence, remain emotionally composed, and sustain strategic clarity through periods of prolonged uncertainty. The implicit management assumption is that leadership capacity remains stable while demands rise. The evidence does not support that assumption. A meta-analysis covering 61 studies and 71 populations found that restricted sleep significantly impaired overall cognitive performance, including executive function and sustained attention. WHO and ILO estimate that working 55 hours or more per week was associated with approximately 745,000 deaths from stroke and ischemic heart disease in 2016, while roughly 479 million people—9% of the global population—were exposed to those working hours. At the organizational level, Gallup reports that global employee engagement fell to 20% in 2025, while manager engagement declined from 31% in 2022 to 22% in 2025; Gallup estimates that low engagement represented approximately $10 trillion in lost productivity, equivalent to about 9% of global GDP, although that figure is a modeled estimate rather than a direct accounting measure. (World Health Organization) These findings do not prove that fatigued executives directly cause corporate failure. They support a more important and defensible conclusion: human physiological state affects the quality and reliability of cognition, and the enterprise impact of that variability grows as authority becomes more concentrated, decisions become more irreversible, and operating systems move faster. The original leadership thesis correctly identifies this overlooked mechanism: leadership is embodied, stress and fatigue can alter judgment, urgency can change organizational behavior, and the downstream consequences of overloaded leadership can spread through teams and institutions long before they are visible in headline performance. The strategic implication is therefore not a wellness message. It is an operating-model requirement. Organizations should manage leadership capacity, recovery, challenge mechanisms, decision timing, authority concentration, and escalation rights with the same seriousness they apply to any other critical system whose failure can propagate through the enterprise.

1. Leadership begins with biology because strategy is executed through a finite cognitive system

Leadership theory traditionally starts with behavior: decisiveness, vision, communication, judgment, courage, strategic thinking, execution, and influence. That starting point misses the substrate through which every one of those behaviors is produced. A strategic decision is not generated by an abstract management function; it is generated by a human brain operating under a particular combination of sleep, stress, emotional load, attention, metabolic state, time pressure, social pressure, and accumulated fatigue. That distinction matters because the empirical evidence demonstrates that human cognition is state-dependent. The 2017 meta-analysis of sleep restriction found a measurable deterioration in overall cognitive processing, including approximately g=-0.32 for executive function and g=-0.41 for sustained attention, while a more recent meta-analysis of acute sleep restriction found significant increases in attentional lapses and slower sustained-attention reaction times after only one night of restricted sleep. (World Health Organization) The implication is not that every tired leader becomes irrational, nor that an exhausted executive necessarily makes a poor decision. The implication is probabilistic and therefore more relevant to governance: the probability distribution of decision quality changes when the biological state of the decision-maker changes. Consider a board approving a multibillion-dollar acquisition after several weeks of negotiations, travel, investor pressure, legal review, and late-night deliberation. The valuation model may be unchanged, the strategic thesis may be unchanged, and the governance documents may be formally complete, yet the cognitive environment in which ambiguity, downside risk, and contradictory evidence are evaluated may differ materially from the one assumed by the process. Most companies would not run a critical machine outside its operating envelope and then treat changing performance as a motivation problem; they routinely do the equivalent with executives. The central design error is therefore not simply overwork. It is the assumption that the quality of judgment is independent of the biological conditions under which judgment is produced.

2. The business risk appears when biological variability interacts with concentrated authority

Fatigue by itself does not explain system failure. Organizations are designed, in principle, precisely because individual people are fallible. Review processes, boards, independent risk functions, financial controls, legal review, audit, delegated authority, and escalation exist to prevent one person's state from becoming the organization's state. The real risk emerges when human variability is multiplied by concentrated authority and weak correction mechanisms. A fatigued analyst inside a robust review architecture may create a localized error that another person detects. A fatigued chief executive who controls capital allocation, senior appointments, product direction, organizational narrative, and final escalation can create a materially different exposure if challenge mechanisms are weak. Imagine two organizations pursuing the same acquisition under comparable market conditions. In the first, the CEO is powerful but the CFO, board, general counsel, and operating leaders have genuine authority to stop or slow the process; major assumptions are independently challenged and dissent is recorded rather than treated as resistance. In the second, the CEO has become the dominant source of strategy and information, disagreement carries reputational cost, and speed is interpreted as commitment. The biological state of the two CEOs could be identical while enterprise risk differs dramatically because one organization converts human fallibility into bounded variance and the other converts it into systemic exposure. This is why leadership health should not be framed simply as an individual wellbeing matter. Once decision rights become sufficiently concentrated, the operating condition of a key individual becomes analogous to concentration risk elsewhere in the enterprise. Companies routinely diversify a supplier responsible for 60% of a critical input, limit credit exposure to a single counterparty, and build redundancy around critical infrastructure; yet many still permit a single leader or small group to become the indispensable cognitive bottleneck for strategy, culture, capital allocation, crisis response, and executive succession. The organization may appear highly aligned, but the alignment can conceal fragility because too much of the system's adaptive capacity sits inside too few people.

3. Overwork is economically material because it changes both health risk and decision capacity

Long working hours are often culturally interpreted as evidence of commitment, particularly in senior leadership, financial services, professional services, technology, entrepreneurship, and crisis environments. The evidence suggests the variable should be treated more carefully. WHO and ILO estimate that exposure to at least 55 working hours per week accounted for approximately 745,000 deaths from ischemic heart disease and stroke in 2016, including about 398,000 stroke deaths and 347,000 heart-disease deaths; the organizations also estimated 23.26 million disability-adjusted life years attributable to those two conditions and found that the share of the population working at least 55 hours a week increased by roughly 9% between 2000 and 2016. (World Health Organization) Those statistics should not be converted into a simplistic threshold suggesting that 54 hours is safe and 55 hours is dangerous, nor should general workforce epidemiology be mechanically mapped onto CEOs. Their relevance is architectural: workload has measurable biological cost, and organizations that depend on sustained overload are consuming a finite resource rather than creating free productivity. The comparison with infrastructure is instructive. A power system can run a reserve generator continuously during a crisis, but if emergency operation becomes permanent, the organization has stopped using a reserve and started consuming its contingency capacity. Human systems behave similarly. Short periods of extraordinary effort can be rational when the event is bounded and recovery follows. Persistent emergency pacing is different. It converts temporary overload into the operating model. Once that happens, organizations may continue meeting targets by drawing down attention, judgment, health, institutional trust, and recovery capacity that do not appear immediately in accounting statements. The visible output can remain high while the underlying ability to absorb future shocks deteriorates. What looks like exceptional execution may therefore partly be borrowed capacity from the future.

4. Burnout at the top becomes enterprise risk when organizations confuse endurance with resilience

The original article correctly argues that burnout in leadership should be understood as more than an individual problem. The deeper management distinction is between endurance and resilience. Endurance describes how long an individual or system can continue operating under stress. Resilience describes whether the system can absorb stress, recover, and continue operating without unacceptable loss of function. A leader working 90-hour weeks for six months may demonstrate extreme endurance; an organization that requires that behavior to function may demonstrate weak resilience. The same principle applies to teams. A company that repeatedly delivers quarterly targets through emergency work, executive intervention, overtime, deferred maintenance, canceled leave, and informal workarounds may appear highly productive while simultaneously revealing that its formal operating model is insufficient for its stated performance level. Gallup's recent data reinforce the management challenge. Global manager engagement fell from 31% in 2022 to 27% in 2024 and 22% in 2025, while non-manager engagement remained comparatively stable at 19% in 2025. (Gallup.com) That does not establish burnout as the sole cause; restructuring, larger spans of control, organizational flattening, economic uncertainty, and technology change are plausible contributors. But it does indicate that the management layer responsible for translating strategy into employee experience and operational execution is itself under increasing strain. Gallup has repeatedly reported that managers account for a substantial share of variance in team engagement; its current global indicator states that managers account for at least 70% of the variance in team engagement. (Gallup.com) In practical terms, degraded manager capacity is therefore not confined to the manager. It can propagate across larger teams through unclear priorities, weaker coaching, inconsistent decisions, poor conflict management, and reduced ability to absorb executive pressure before passing it downward.

5. Leadership pressure propagates through organizations because people adapt to the state of authority

A senior leader rarely needs to order an organization to become anxious, impatient, or silent. Organizations infer priorities from repeated behavior. If the chief executive interrupts frequently, changes direction abruptly, demands immediate responses at all hours, publicly rewards heroic recovery, and treats caution as lack of ambition, the organization adapts. Employees shorten analysis, managers escalate fewer uncertainties, executives pre-emptively align their recommendations, and teams learn that speed carries higher status than careful dissent. This is how individual state becomes institutional behavior. The original article describes unregulated leadership as transmitting stress, normalizing urgency, and distorting incentives. The causal chain should be stated carefully: leader stress does not mechanically create employee stress in every organization, but leadership behavior strongly shapes workload, decision norms, psychological safety, and the perceived cost of disagreement. The organizational example is familiar. A CEO asks why a strategic initiative is six weeks behind schedule. In a healthy system, the program lead can respond that cybersecurity testing identified unresolved issues and that releasing on the original date would create unacceptable exposure. In an unhealthy system, the same leader knows that previous executives who delayed launches were characterized as execution problems; the cybersecurity concern is reframed as manageable, the launch occurs, and the organization subsequently treats the resulting defect as an unforeseeable technical problem. The knowledge existed. The organization failed because its social architecture prevented the knowledge from influencing the decision. This is why leadership state is not simply emotional. It affects the information permeability of the institution.

6. Charisma becomes dangerous only when admiration substitutes for verification

Charismatic leadership can create extraordinary organizational value. It can increase motivation, simplify strategic direction, attract talent, mobilize capital, and coordinate action during periods when ambiguity would otherwise fragment the institution. The risk emerges when charisma changes the evidentiary standard applied to the leader's claims. The original draft argues that charisma can discourage questioning; that proposition is too broad if presented universally, but the mechanism is important. High-status leaders can create authority gradients in which employees unconsciously require more evidence to challenge the leader than the leader requires to make the original assertion. At that point, the organization has changed its epistemic architecture. A forecast presented by an analyst may be interrogated aggressively; the same forecast repeated confidently by a celebrated founder may become strategic truth. This is particularly dangerous in founder-led and highly centralized organizations because charismatic authority can combine with concentrated voting control, cultural identity, and investor narrative. The problem is not personality. It is asymmetric verification. Mature organizations therefore preserve the motivational benefits of charismatic leadership while separating status from evidence. Major assumptions remain challengeable, investment cases remain testable, risk functions retain independence, and material dissent is surfaced before rather than after failure. The institutional rule should be simple: the stronger the authority of the speaker, the more important it becomes that the system does not lower the standard of proof.

7. Urgency is useful in crises but destructive when it becomes the permanent operating regime

Urgency can be rational. Cyberattacks, liquidity shocks, industrial accidents, geopolitical disruptions, product recalls, and immediate safety issues all require compressed decision cycles. The problem identified in the original article is not emergency response itself but the institutional normalization of urgency. Once everything becomes urgent, the organization loses the ability to discriminate between events requiring immediate response and work that would benefit from deliberation. This degrades both execution and attention. Teams receive too many simultaneous priorities, managers interrupt planned work continuously, employees reduce investment in long-horizon improvement because today's escalation always dominates tomorrow's prevention, and strategic thinking becomes episodic because operating pressure consumes available cognitive capacity. A crisis system can tolerate this temporarily because its objective is survival. A permanent crisis system eventually produces its own crises because prevention, maintenance, reflection, talent development, and redesign are repeatedly postponed. The managerial analogy is to emergency medicine: triage works because not every patient receives the highest emergency category. If every case were classified as critical, the classification would become operationally meaningless. Leadership systems need the same differentiation. Certain decisions should move in minutes, others in days, others over months. An organization that equates leadership quality with universal speed is not necessarily becoming more decisive; it may be destroying its ability to allocate attention according to consequence.

8. Silence is one of the most dangerous outputs a high-pressure leadership system can produce because silence looks like alignment

Organizations are especially vulnerable to information failures that resemble success. Visible conflict is uncomfortable but detectable. Silence is easier to misread because it resembles agreement. The original article's distinction between compliance and alignment is therefore strategically important. Consider a 12-person executive committee evaluating a major restructuring. Eight executives believe the proposed timetable will damage customer service, but only one raises the concern because the CEO has already described the change as essential and urgent. Minutes record broad consensus; the board observes an aligned management team; implementation begins rapidly. The organization has not actually achieved consensus. It has deleted seven independent warning signals from the decision process. This reframing matters because dissent is often discussed as a cultural or interpersonal topic when it is better understood as a sensing capability. Engineers see technical risk, finance sees liquidity exposure, legal teams see regulatory issues, salespeople see changes in customers, and frontline employees see operational failures long before executive dashboards capture them. If those signals do not travel upward, leadership can become increasingly confident while the institution becomes progressively less informed. This is why psychological safety, protected escalation, independent risk functions, direct board access for control roles, and pre-mortems should be understood as information infrastructure rather than employee perks. Their purpose is to preserve the organization's ability to perceive conditions that leadership would prefer not to be true.

9. Responsibility is prospective; accountability alone arrives too late

The original article distinguishes responsibility from accountability by describing accountability as retrospective and responsibility as prospective. This distinction should sit at the center of the governance model. Accountability asks who should answer after an outcome occurs. Responsibility asks what structures should exist before the outcome to prevent predictable harm. Many organizations are stronger at the first than the second. They have investigations, performance reviews, disciplinary systems, audit processes, and board committees capable of identifying who failed. Yet they may have weak mechanisms for testing whether the person assigned responsibility possessed sufficient authority, information, capacity, and time to discharge it. A regional leader may be held accountable for margin while headquarters controls pricing. A plant manager may be accountable for safety while production incentives penalize stoppages. A product leader may be accountable for quality while executive deadlines eliminate testing time. In each case, formal accountability exists but operational responsibility has been fragmented. The organization can therefore appear well governed because every outcome has an owner while still creating conditions in which no one has sufficient control to prevent failure. Mature leadership governance should test four questions before assigning responsibility: Does the person control the relevant decision? Do they possess the information required? Do they have sufficient capacity to act? Can they stop or escalate when conditions become unsafe? When the answer is no, accountability risks becoming a mechanism for transferring system design failure onto individuals.

10. The strongest leaders are not those who ignore limits but those who allocate capacity according to consequence

The heroic leadership model equates strength with endurance: work longer, decide faster, absorb more pressure, suppress doubt, remain available continuously, and never visibly lose momentum. This model can produce extraordinary short-term performance. It can also create fragile organizations because the system becomes dependent on behaviors that cannot scale indefinitely. The alternative is not passive or cautious leadership. It is capacity-aware leadership. High-capacity leaders understand that not every decision deserves equal cognitive attention, that some decisions should be delegated, that others require protected reflection, that important assumptions should be challenged by people who were not involved in generating them, and that recovery after sustained pressure is part of performance rather than time stolen from it. The comparison with other capital-intensive systems is again useful. Airlines do not schedule aircraft based only on maximum possible flight hours; maintenance constraints are incorporated into fleet economics. Banks do not maximize asset deployment without capital buffers because the buffer is part of the institution's ability to survive uncertainty. A leadership system that schedules every executive hour, fills every cognitive margin, and interprets unused capacity as inefficiency is making the same error as an electrical grid operating permanently at 100% utilization. The apparent efficiency comes from eliminating the margin required when the environment stops behaving as forecast.

11. Self-regulation matters because leaders shape the organization's interpretation of uncertainty

A leader's emotional state matters not because executives are expected to become emotionless but because authority amplifies signaling. A visibly anxious CEO can cause employees to infer that conditions are worse than formal communications indicate. A leader reacting aggressively to bad news can teach the organization that negative information is dangerous to report. A leader changing direction impulsively can cause teams to defer long-term investment because priorities are expected to change again. Conversely, a leader who can acknowledge uncertainty without panic, distinguish urgency from importance, and remain receptive to contradiction can preserve organizational information quality during stress. The original article frames this as leadership self-regulation. A more precise management interpretation is that senior leaders need the capacity to prevent their transient internal state from becoming an uncontrolled organizational input. This does not mean hiding emotion. It means understanding the multiplier created by authority. A frustrated comment from a junior employee may affect a small group; the same comment from a chief executive can redirect organizational behavior because thousands of employees infer strategic significance from it. Leadership therefore requires not merely control over decisions but awareness of the second-order effects created by how decisions and uncertainty are communicated.

12. AI will increase the importance of biologically informed leadership because machines remove human friction faster than institutions redesign governance

Artificial intelligence is often positioned as a solution to managerial overload because it can summarize information, generate analysis, automate administration, retrieve knowledge, monitor workflows, and increasingly execute tasks. Those capabilities can reduce cognitive burden, but they can also create a new problem: AI increases organizational velocity while human governance remains biologically constrained. A senior leader who previously reviewed ten strategic alternatives may soon receive fifty machine-generated alternatives. A risk team that monitored hundreds of transactions may oversee millions of automated decisions. An executive who once waited days for analysis may receive recommendations continuously. The bottleneck therefore moves upward from information production toward judgment, prioritization, verification, and authority. If organizations respond by demanding that leaders simply process more machine-generated intelligence faster, AI can intensify rather than solve the human-capacity problem. The correct architecture is selective. Machines should absorb search, synthesis, monitoring, routine modeling, and well-bounded execution; humans should concentrate limited cognitive capacity on unresolved trade-offs, novel conditions, values, high-consequence exceptions, irreversible decisions, and governance. The leadership advantage of AI is therefore not that executives can work at machine speed. It is that organizations can deliberately reserve human cognition for the decisions where human judgment remains economically most valuable.

13. Boards should treat leadership capacity as part of operational resilience rather than private executive behavior

Boards traditionally assess leadership through succession, performance, compensation, conduct, strategic delivery, and key-person risk. The biological dimension belongs within this existing governance mandate rather than in a separate wellness agenda. A board does not need access to private medical information to ask whether decision architecture is overconcentrated, whether executive schedules permit adequate preparation for consequential decisions, whether crisis pacing has become permanent, whether succession depth is sufficient, whether senior teams can challenge the CEO, whether the organization depends excessively on after-hours intervention, or whether major decisions routinely occur under conditions of compressed review. This is similar to assessing any other critical operational dependency. The objective is not to monitor whether an executive sleeps eight hours. It is to determine whether the system has become dangerously dependent on one person's sustained biological overperformance. Key indicators might include repeated emergency decision cycles, executive turnover, span-of-control expansion, excessive escalation to the CEO, chronic reprioritization, declining management engagement, reduced succession depth, rising decision reversal, and recurring execution problems requiring senior intervention. None of these proves biological overload independently. Together, they can signal that the leadership operating model is consuming more capacity than it regenerates.

14. The economic case is not softer leadership; it is better allocation of scarce decision capacity

The strongest objection to biologically informed leadership is that business is competitive and senior roles are inherently demanding. That objection is correct but does not invalidate the argument. The relevant choice is not between demanding leadership and comfortable leadership. It is between using scarce cognitive capacity intelligently and wasting it indiscriminately. If a chief executive spends four hours resolving a routine internal escalation that should have been handled two layers lower, the cost is not merely four executive hours; it is the opportunity cost of the strategic work displaced and the signal that the organization has failed to distribute authority appropriately. If a board schedules six major capital-allocation decisions into one extended session because calendars are difficult, administrative efficiency may be purchased at the cost of decision quality. If executives receive hundreds of pages of material immediately before meetings, the organization has technically supplied information while practically reducing the probability that the information can be processed deeply. Biologically informed management therefore resembles capital allocation: scarce high-quality decision capacity should be directed toward the decisions with the highest consequence and weakest substitutability. Routine cognition should be delegated, automated, standardized, or removed. High-consequence cognition should be protected.

15. The leadership standard should move from heroic endurance to system-safe performance

A modern leadership system should therefore evaluate more than whether an individual is decisive, inspiring, intelligent, ambitious, and capable of sustained output. Those characteristics remain relevant, but they are incomplete. The stronger questions are whether the leader preserves information quality under pressure, distinguishes genuine emergencies from manufactured urgency, maintains challenge mechanisms when the stakes rise, delegates without abandoning accountability, slows irreversible decisions when evidence is weak, prevents personal stress from automatically becoming organizational stress, and builds a system that can continue functioning without constant heroic intervention. The original article describes these capabilities as maintaining coherence, preserving dissent, slowing decisions when required, and protecting people and institutions from avoidable damage. These should not be characterized as “soft skills.” They are closer to operational control variables because they influence the organization's capacity to perceive risk, challenge assumptions, and recover from error.

Conclusion

Leadership is often presented as the ability to overcome constraints. The more mature view is that leadership is the ability to understand which constraints can be changed and which must be governed. Human biology belongs to the second category. Cognitive capacity fluctuates. Attention is finite. prolonged overload has measurable health consequences. Sleep restriction can impair executive function and sustained attention. Long working hours are associated with substantial cardiovascular disease burden. Manager engagement has deteriorated materially in recent years, with Gallup reporting a decline from 31% in 2022 to 22% in 2025, while global engagement fell to 20%. (World Health Organization) None of those facts means that organizations should become slower, less ambitious, or less demanding. They mean that organizations should stop building management systems on the assumption that human decision capacity is infinitely expandable.

The original thesis by Trang Phan is strongest when understood not as a wellness argument but as a systems-governance argument: when leadership systems ignore biological limits, they increase the probability that fatigue, stress, authority, urgency, and organizational silence will interact in ways that reduce error detection precisely when consequential decisions require it most.

The strategic answer is therefore structural. Distribute authority before overload occurs. Preserve independent challenge before urgency rises. Protect recovery before exhaustion becomes visible. Separate routine decisions from irreversible ones. Use AI to remove low-value cognitive load rather than increase the rate at which leaders must absorb information. Design boards and executive teams so that no individual's transient biological state can become an uncontrolled enterprise state. Treat dissent as sensing infrastructure, not disloyalty. Treat recovery as maintenance, not weakness. And judge leadership strength not by how long an individual can override limits, but by whether the organization can continue to make high-quality decisions without requiring anyone to behave as though those limits do not exist.

That is the deeper law:

Human limits are not an obstacle to organizational performance. They are one of the design conditions that determine whether performance can remain reliable under pressure.