NASA Issues Investigative Report About Starliner Failure

Twenty-three years ago, the Columbia Accident Investigation Board issued its comprehensive report on the 2003 Space Shuttle Accident.  My colleagues Amy Edmondson, Richard Bohmer, and I began an intensive research project about the tragedy, and we published a case study, book chapter, and Harvard Business Review article with our analysis.  We argued that the failed mission demonstrated how and why organizations and their leaders actively downplay ambiguous threats at times.  We examined how the culture and team dynamics made it difficult for technical experts to express dissenting views and share bad news.  

Sadly, it seems that many of these same challenges led to the problems on the 2024 Starliner mission which left astronauts stranded on the space station for approximately nine months.  Yesterday, NASA issued its investigative report on the Starliner failure.   Here is the excerpt from the report that caught my attention: 

A posture of risk acceptance was communicated by CCP (NASA's Commercial Crew Program) and Boeing leadership, creating division within the large working/joint team and eroded trust. During the mission, CCP and Boeing operational leadership consistently conveyed a position of risk acceptance and readiness to undock, which many perceived as premature and dismissive of unresolved technical concerns. This was particularly apparent regarding the Service Module RCS 138thruster anomalies. This posture gave the impression that completing the sortie mission was prioritized over a thorough assessment of crew safety risks.  

One interviewee noted, “People said, ‘Why bother? He’s driving in one direction and that’s what he wants.’”   

Some interviewees also mentioned the shuttle operational background of the SMMT Chair, NOM, and CCP PM, and the possible preconceived notion that accepting risk to return the vehicle and crew was the only real path forward. This mirrors decisions made for the shuttle when no safe haven in LEO or alternative return capability was available. 

This forward leaning approach led to a breakdown in open dialogue. NASA institutional stakeholders, including ISSP, FOD, and Technical Authorities, felt their input was undervalued or ignored, requiring governance intervention to ensure additional data analysis occurred before a final crew return decision. The perception that CCP leadership had formed a position before hearing all viewpoints created organizational silence, resistance to collaboration, and stagnation in decision making. 

Strong personalities within CCP and Boeing were seen as overly optimistic in presenting data, which some interviewees interpreted as lobbying rather than objective analysis. This dynamic discouraged dissenting views and contributed to a growing sense of distrust. As one interviewee described, opposing positions felt like “pushing a rock uphill.” 

The situation improved later in the mission when key personnel changes were made within the Boeing team and there was collective recognition that senior leadership should have played a more active role in facilitating respectful engagement across differing perspectives. These changes allowed for more productive conversations regarding the technical qualification campaign of the hardware and testing at the WSTF. The lack of early intervention to address team dysfunction allowed conflict to overshadow mission objectives and delayed consensus on critical decisions. 

Organizational silence, discouraging of dissenting views, dismissed technical concerns, overly optimistic analysis... the pattern is clear.  Once again, we see ample evidence that leadership did not create a culture in which open and candid dialogue could occur about ambiguous risks.  I'm glad to see a careful after-action review taking place here, with transparency about the organizational problems that have been identified (rather than only focusing on the technical problems).  Having said that, now the challenge is clear: can NASA turn these lessons into action and fundamentally change the way future programs are led?  

Asking Engineers the Right Questions

In our extensive study of the Columbia space shuttle accident, Amy Edmondson, Richard Bohmer, and I examined the culture and leadership at NASA. Specifically, we analyzed the forces that made it difficult for engineers such as Rodney Rocha to speak up regarding their safety concerns, and we looked at how leaders did not ask probing questions to elicit dissenting views. 

When I teach the case to engineers and other technical experts, I often hear them say that management needs to understand how to ask engineers the right questions, and how to interpret their results. They argue that engineers were not going to scream, "There is a safety-of-flight risk!" unless they had conclusive evidence. The engineers' lack of complete certainty might lead to them to give answers that that are interpreted incorrectly by management. Organizational leaders might be looking for a definitive statement expressing alarm and grave concern, but they won't get it if there is scientific uncertainty. 

Today, I ran across a good quote (in Fast Company) from Google's long-time Chairman and CEO, Eric Schmidt, regarding this issue. Scmidt explains how you have to ask multiple questions, in different ways, to make sure you are getting the whole picture when working with technical experts: 

They are taught to think logically. If you ask engineers a precise question, they will give you a precisely truthful answer. That also tends to mean that they’ll only answer the question that you asked them. If you don’t ask them exactly the right question, sometimes they’ll evade you — not because they’re lying but because they’re being so scrupulously truthful.”

Lessons from the Columbia Accident

10 years ago tomorrow (February 1st, 2003), the Columbia shuttle accident occurred.  After the tragedy occurred, my colleagues and I embarked upon a lengthy research project culminating in the creation of an award-winning multi-media case study, a book chapter, a Harvard Business Review article, and other materials.  What lessons did we derive from this tragedy?

1.   NASA had a culture in which many people did not feel safe speaking up.  As a result, people downplayed their concerns about the foam strike, rather than sharing them openly.   Leaders did not do enough to cultivate dialogue and dissent.  They tended to be very passive, assuming that bad news would surface on its own.  It rarely does.  Leaders need to draw out the bad news, as it does not usually rise to the top.  

2.  Framing the shuttle program, from the very beginning in the 1970s, as a routine, operational endeavor turned out to have long-lasting detrimental effects.   The shuttle program should have been framed as an experimental initiative.  Promising dozens of flights per year, naming the vehicle a "shuttle" and describing space travel as "routinized" pushed the organization away from its roots during Mercury and Apollo as a laboratory of exploration.   The adoption of the production/routine mindset contributed to an obsession with schedules and deadlines, the creation of rigid organizational processes and protocols, and an insufficient emphasis on the imperfect state of knowledge in the organization. In short, it meant that the organization did not have a sufficient learning orientation. 

3.  Managers at NASA exhibited what Rebecca Wohlstetter has described as a "stubborn attachment to existing beliefs."  They had become convinced that foam strikes could not damage the shuttle.   The confirmation bias exacerbated the problem.   They tended to look for data that confirmed this pre-existing belief, and they discounted signs that contradicted the conventional wisdom about foam. 

Lessons from the Challenger Accident

Yesterday marked the 27th anniversary of the Challenger space shuttle accident.  Still today many leadership programs and business schools study this awful tragedy.   We analyze it because it offers profound lessons about leadership, risk management, and decision-making.  

Let me offer three points that we should take away from this tragedy as we reflect on it today:

1.  Normalization of deviance.    Diane Vaughan wrote the seminal book on the Challenger accident in 1996.  In that book, she explained how organizations gradually come to accept more and more risk.   She described this normalization process whereby the unexpected gradually becomes the expected and then becomes the accepted.   All organizations should be wary of moving down that slippery slope.

2.   Advocacy vs. Inquiry.    Many people initially described the eve-of-the-launch teleconference as an example of groupthink.   I don't think that's correct.   Engineers from Morton Thiokol clearly were arguing against the launch.  The meeting did not lack dissent and disagreement.   The real problem was that both sides (for and against launch) were operating purely in advocacy mode.  Each side was pushing its position.  Neither side did much to truly learn from the other.   The entire group did not move into collective inquiry mode, seeking to understand the problem of O-ring erosion more deeply or the reasons for each side believing what it did.   Instead, they argued their positions with regard to the next morning's launch and became more and more entrenched and polarized.   Effective teams balance advocacy with inquiry, so as to have a more constructive dialogue and debate.

3.  Learning from our failures.   During the Columbia shuttle accident investigation, astronaut Sally Ride said, ""I think I'm hearing an echo here."  She meant that NASA had repeated many of the failures of the past.  While they had fixed the technical errors after Challenger (no more o-ring erosion), NASA executives had not fixed the leadership and cultural problems that contributed to the tragedy.  Those problems continued and contributed to the Columbia disaster many years later.  Successful organizations take a hard look at their failures, and they examine the broad range of causes for those failures...not just the technical issues.