The Boeing 737 MAX crisis, which began with two fatal crashes in 2018 and 2019 and whose organizational and legal consequences extended through 2024, offers one of the most thoroughly documented case studies in organizational self-deception available to contemporary observers. It is also one of the most consistently misread. The dominant narrative, that Boeing’s failure was a story of corporate greed overriding safety culture, is not wrong as far as it goes. It does not go far enough. It stops at the most visible and morally legible explanation and does not examine the structural mechanisms that made the outcome possible, predictable, and, in important respects, inevitable given the organizational design that preceded it.

What the subsequent investigations, congressional testimony, Department of Justice findings, and internal document releases revealed was not an organization populated by people who consciously chose to endanger lives for profit. The documentary record does not support that narrative. It reveals something more instructive and more disturbing: an organization that had developed, over years and under competitive and financial pressure, a set of structural features that made it progressively more difficult for accurate technical information to reach the people who needed it, and progressively more costly for the people who possessed that information to deliver it. The catastrophe was not the result of bad people making bad decisions. It was the result of a well-documented organizational process by which good people, operating within a structurally compromised system, made individually defensible decisions that aggregated into institutional failure.

The structural changes that preceded the crashes are by now well documented. The relocation of engineering authority to financial management, which began in earnest following Boeing’s 1997 merger with McDonnell Douglas, progressively reduced the organizational status and decision-making authority of engineers relative to finance and production professionals. The restructuring of the supplier certification process reduced the depth and independence of technical oversight at precisely the points where independent oversight was most needed. The incentive structures governing program management penalized schedule delays and cost overruns more severely than they penalized safety concerns that were difficult to quantify and easy to characterize as manageable. The cultural evolution, from an engineering organization that treated safety concerns as primary information requiring resolution to a production organization that treated them as obstacles to be managed, was documented in the internal communications that became public through congressional investigation.

None of these changes were made by people who intended to build unsafe aircraft. This point is essential and is consistently underweighted in post-mortem analyses that focus on individual culpability. Each structural change, examined individually, was defensible as a rational operational decision in its immediate context. The relocation of authority to financial management reflected legitimate concerns about cost discipline. The restructuring of supplier oversight reflected genuine operational pressures. The incentive structures reflected normal management practice across the aerospace industry. The problem was not any individual decision. The problem was the system that the individual decisions constructed over time, and the progressive degradation of the organization’s capacity to receive, process, and act on accurate technical information that the system produced.

This is the essential feature of organizational self-deception that is most difficult to address in post-mortem analysis and most dangerous to misunderstand in organizational design: it does not require bad actors. It does not require conscious deception. It requires only structures that make it progressively more rational, more comfortable, and more socially rewarding to interpret ambiguous information in the direction of preferred conclusions, and progressively more costly to interpret it in the direction of unwelcome ones. An organization with these structural features will produce self-deceptive outcomes regardless of the quality, integrity, and intentions of the individuals within it. The individuals are not the variable. The structure is.

The MCAS system at the center of the 737 MAX failures was not a secret. Its existence, its function, and the concerns about its behavior under specific failure conditions were known within the organization. What the organizational structure determined was not whether this information existed but whether it moved, whether it traveled from the engineers who understood it to the program managers who needed to act on it, from the program managers to the certification authorities who needed to evaluate it, and from the certification authorities to the pilots and airlines who needed to prepare for it. At each stage of that transmission, the organizational structure that Boeing had built created friction against the movement of unwelcome technical information and momentum in favor of the movement of reassuring assessments. The information did not move as it needed to. People died.

The structural reform that followed has been substantial. Boeing has reorganized its engineering authority structure, rebuilt elements of its safety oversight process, and made significant changes to its program management incentive structures. Congressional oversight has been sustained in a way that is unusual for post-crisis corporate reform. The Federal Aviation Administration has increased the depth and independence of its certification oversight. Whether these reforms are sufficient to prevent a recurrence is a question that cannot be answered in advance. What can be said with confidence is that the reforms that matter are structural rather than cultural. The problem was not that Boeing’s employees lacked commitment to safety as a value. The problem was that the organizational structure Boeing had built made acting on that commitment progressively more difficult and costly over time. Restoring the commitment without redesigning the structure would have produced exactly nothing. The organizations that understand this lesson, not just about Boeing but about their own structural vulnerabilities, are the ones most likely to avoid writing the next case study.

The return-to-office mandates that proliferated through 2022 and into 2023 were almost universally justified on the grounds of productivity, collaboration, and culture. Almost universally, the evidence cited in support of these justifications was weak, selectively presented, or directly contradicted by the organizations’ own performance data from the preceding two years.

This is not an argument that remote work is superior to office work. It is an observation that the stated rationale for RTO mandates was, in most cases, not the actual rationale. Organizations that had demonstrably maintained or improved productivity during remote operations nonetheless mandated returns, dismissing their own evidence in favor of managerial intuition and, in some documented cases, explicit pressure from real estate commitments and middle management anxiety about relevance.

The behavioral mechanism at work is one that organizational research has documented extensively: motivated reasoning in the service of self-concept maintenance. Senior leaders who had built their careers in office environments, whose social identities were bound up with the culture of physical presence, and whose intuitions about productivity were formed in a world where presence and output were structurally conflated, could not process evidence that challenged those intuitions as genuine evidence. They processed it as noise to be explained away.

This dynamic was compounded by the structural position of middle management. Remote work had exposed, with uncomfortable clarity, the degree to which middle management value had been built on proximity functions: information relay, activity monitoring, coordination of physically co-located teams, and the social authority that comes from visible presence in organizational space. When those functions became either unnecessary or performable by technology, the relevance of the middle management layer became a legitimate question. RTO mandates restored the conditions under which that relevance was unquestioned.

The research on what actually drives organizational performance in knowledge work environments does not support the proposition that physical co-location is a significant independent variable. What the research does support is that certain specific activities, creative problem-solving that benefits from spontaneous interaction, onboarding of new employees, and the repair of damaged working relationships, are better performed in person. These activities represent a fraction of most knowledge workers’ time. Building an organization-wide mandate around them, at substantial cost to employee satisfaction and retention, reflects a choice to optimize for managerial comfort rather than organizational performance.

The employees who resigned in response to RTO mandates, in numbers sufficient to register in workforce data, were not primarily resigning because they preferred working from home. Many were indifferent to location as such. They were resigning because they had watched their organizations construct elaborate justifications for decisions that were plainly made on grounds other than the ones stated. The RTO debate, for many workers, was not about location. It was a legibility test. Organizations that stated one rationale while acting on another demonstrated, with clarity, the degree to which their stated values and their operational values were disconnected. Many employees drew the rational conclusion that an organization that could not be honest about why it wanted them in the office could not be trusted on questions of greater consequence.

The organizations that navigated this period most successfully were those that were honest about their actual reasons for wanting employees present, made specific rather than universal demands, and structured those demands around demonstrable organizational needs rather than managerial preference. Honesty about motivation, even when the motivation is not entirely flattering, produces better organizational outcomes than sophisticated justifications for decisions made on undisclosed grounds. This is not a moral observation. It is an empirical one.

When organizations face existential pressure, they do not become different. They become more fully what they already were. The COVID-19 pandemic, arriving in March 2020 with the force of a structural stress test no business school had designed, did not create organizational dysfunction. It illuminated it.

The organizations that failed workers in the early weeks of the pandemic, that withheld safety information, that misrepresented supply chain stability to shareholders, that told employees operations were safe when leadership privately knew otherwise, were not suddenly corrupted by crisis. They were organizations that had already developed robust internal capacities for self-deception. The pandemic simply removed the ambient noise that had allowed that self-deception to pass unnoticed.

What made the early pandemic period so instructive was the speed at which the gap between organizational communication and organizational reality became visible. Companies that had spent years constructing narratives of employee-centered culture were revealed, within weeks, to have no structural mechanisms for actually centering employees when doing so carried a cost. The narrative was real. The structure was not.

This is the distinction that most post-mortems miss. The question is never whether organizational leaders intended to deceive. In most cases, they did not. The question is whether the organization was structurally capable of receiving, processing, and acting on accurate information when that information was unwelcome. Most were not. The pandemic did not make them dishonest. It made their existing dishonesty consequential in ways that could no longer be absorbed.

The behavioral economics literature has long established that motivated reasoning operates below the threshold of conscious intent. Decision-makers do not experience themselves as filtering information. They experience themselves as exercising judgment. The pandemic compressed the feedback loop between motivated reasoning and organizational consequence to a degree that made this process visible in real time, across thousands of organizations simultaneously.

What should organizations take from this? Not that their leaders are dishonest people, but that their structures reliably produce dishonest outcomes regardless of the character of the people within them. The structural question is the only question that matters. Everything else is noise.