Boeing and the Architecture of Organizational Self-Deception

When it comes to examining cases of organizational self-deception, the Boeing 737 MAX situation has it all. Two crashes between 2018 and 2019, organizational and legal fall-out extending into 20twenty-four. Never before has so much introspective documentation been made available to outside observers. And rarely has a more complete case of organizational self-deception been so consistently misinterpreted. You’ll find the usual story in popular discussions of the Boeing Crisis: the story of corporate greed overpowering safety culture. It’s true…but only to a point. The story stops short at the most visible, most morally legible explanation for Boeing’s problems. What’s visible tells us little about how Boeing got that way. To see that, we have to look at the structures that shaped its leaders’ decisions.

If you read through the lengthy DOJ report, listen to congressional testimonies, and pore over released internal documents, what you will not find is clear evidence that Boeing was populated by individuals who consciously traded away safety for profit. That narrative is simply not supported by the documentary record. Instead, what you will find is an organization that has, over time, developed incentives, procedures, and policies that have made it harder for accurate technical information to reach decision-makers and made it more costly for individuals who possess that information to deliver it. Bad decisions were made. But they were decisions that individual employees were incentivized to make and could defend as reasonable if challenged by their peers or supervisors.

How did Boeing reach a point where safe flights became riskier than they needed to be? As with most long-term accidents, a series of smaller decisions made sense in their immediate context but designed a system that was bound to produce bad outcomes. Engineering authority was slowly moved under financial authority beginning with Boeing’s merger with McDonnell Douglas in 1997. This change gradually decreased the formal decision-making authority and organizational stature of engineers relative to the roles of finance and production professionals. The supplier certification process was restructured to rely less on independent evaluation, precisely at the points in the production process where it was most needed. Reward structures for program managers penalized delays and budget concerns more than they penalized safety issues that were difficult to quantify. Decisions that deferred safety concerns were culturally safer than ones that raised those concerns. This last point has been well illustrated by the emails and internal discussions released as part of the congressional investigation.

At no point in these changes were decisions made with the goal of making unsafe aircraft. This is critical, and something that is consistently missed by critiques of the organization that focus on individual decisions. Each change in isolation was objectively defensible; they were rational responses to their specific contexts. Shifting decision-making authority to financial professionals was a response to concerns about cost overruns. Restructuring supplier oversight was done for operational reasons. Creating stronger incentives for rapid production was standard across aerospace. Each one, isolated from the larger case, makes sense. The problem is how they fit together. How those decisions designed a system that slowly degraded Boeing’s ability to acquire, process, and act on technical information.

The reason this is the hardest type of organizational self-deception to combat after the fact, and most dangerous to create within your own organization, is that it can happen while everyone is acting with the best intentions. It doesn’t require bad people. It doesn’t even really require deception in the sense of deliberate, malicious intent to mislead. All that is required is a structure that gradually makes it more rational, more comfortable, and socially rewarding to view ambiguous information as confirming your preferred conclusion, and more costly to view it as suggesting something you don’t want to be true. An organization that collects