After declining for decades, the U.S. airline fatal accident rate began to “plateau” in the early 1990s. Because airline flying volume was projected to double in 15-20 years, the industry was concerned that the number of accidents from doubling the volume without reducing the rate could generate undue public concern. 

Commercial aviation is a complex system that consists of several subsystems – airlines, manufacturers, airports, pilots, flight attendants, air traffic controllers, maintenance personnel, and regulators – that must all work together in order for the entire system to function. Because the subsystems are coupled, changes to one subsystem can have effects on the other subsystems.  One of the major challenges of improving the safety of a complex system is the possibility of unintended consequences. 

‘System Think’

One way to minimize unintended consequences is referred to in this article as “System Think” – understanding how a change in one subsystem may affect other subsystems. The airline industry accomplished System Think by using a voluntary system-wide collaborative process known as the Commercial Aviation Safety Team (CAST) that included all of the industry participants, including the regulator.  

CAST exceeded expectations in several ways – first, by improving the stubbornly flat fatality rate, which many safety experts thought was unlikely to improve, by more than 80 percent in less than ten years. Second, CAST improved not only safety but also productivity, which flew in the face of conventional wisdom. This is crucial because safety improvements that hurt the bottom line are not generally sustainable. Third, CAST initiatives have resulted in very few unintended consequences.

Workplace safety improvement efforts may involve collaboration within a facility or within a company, but they do not generally employ the breadth of collaboration that CAST is using. For example, workplace safety improvement efforts do not generally include either the manufactures of the tools and equipment or the regulator, the Occupational Safety and Health Administration (OSHA).

Broader collaboration

CAST collaboration, on the other hand, includes, among others, the aircraft manufacturers and the regulator. Inclusion of aircraft manufacturers gives them a better understanding of how effective, efficient, and safe their aircraft are in actual operations, and that knowledge can help them improve the design. Inclusion of the regulator is crucial because actions or inactions by the regulator may provide “links in the chain” to an accident, so participation by the regulator in the collaborative improvement process can help the regulator identify and address those links in the chain.1 Moreover, the collaborative process often generates very sensitive information, especially if it involves employee mistakes, so the regulator must clarify that the information will not be used for enforcement unless it reveals criminal or intentional wrongdoing.

The theory of broader collaboration is very simple – everyone who is involved in a problem should be involved in the solution. CAST has demonstrated how powerful system-wide collaboration can be to improve safety.

Focusing on safety rather than regulatory compliance

Although the airline industry is highly regulated, David Hinson, who was the Federal Aviation Administration (FAA) Administrator when CAST was created, realized that the way to reduce the fatality rate was not more regulations or more enforcement, but a focus on improving the safety of a complex system. Hence, CAST focused on improving safety rather than upon regulatory compliance, and it generated a level of safety that is far above the “regulatory floor of compliance.”

When the focus is on compliance, compliance is the best that is achieved. When the focus is on improving safety, the resulting level of safety is far higher. Drawing from the CAST experience, workplace safety efforts could be far more effective if the focus were on improving safety rather than compliance. 

Prompt implementation

Collaboration represents a major paradigm shift from the normal regulatory model, in which industry participants often do not agree with the regulator’s proposed changes, so they oppose them vigorously, which can take years. In the collaborative process, on the other hand, all of the industry participants are involved in identifying problems and developing solutions. Thus, the final solutions reflect and respond to all participants’ concerns more effectively, and they are implemented willingly and promptly.2

Getting started:  A beta test

Instead of pursuing an industry-wide approach, as aviation did, a suggested first step toward trying a collaborative process would be to conduct a beta test on a problem that has existed for a long time. If a problem has existed for a long time, despite several attempted remedies and the normal employee turnover, the problem probably results from the procedures and equipment that are being used rather than from the involved employees. Thus, the beta test would involve the creation of a collaborative process improvement team that includes everyone who is involved in the problem – management, labor, the manufacturers of equipment and tools, and the regulator – to study the problem and develop remedies.3

If the beta test can develop remedies for the most challenging problems, then it will be apparent that (a) collaboration can help improve safety and productivity in your workplace, not just in aviation; and (b) if collaboration can address the most challenging problems, it can address all problems.


The U.S. airline industry has demonstrated the power of collaboration to improve process safety in a very complex system. In theory, this collaborative process can also help prevent slips, trips, and falls. CAST has shown that the path to improving safety well beyond regulatory compliance is not more regulations and more enforcement, but using collaboration that involves all of the participants to determine how to operate a complex system more safely, effectively, and efficiently.


  1. In general, more NTSB safety improvement recommendations go to regulators than to any other single participant in the transportation process.
  2. Moreover, if the solution is not quite right, all participants willingly engage in tweaking it. In the normal regulatory process, tweaking often requires another rulemaking.
  3. As the name “process improvement team” suggests, the objective is to improve the process rather than to punish. It is critical that this objective be made clear to the employees in order for them to be willing to participate. 
  4. Member and Vice Chairman of the NTSB from 2009-2014, Acting Chairman from 2014-2015, Chairman from 2015-2017, and Member from 2017 until leaving the NTSB in January 2018.