The commercial airline era started after the First World War when the pilots of fighter aircraft used their ability to fly aircraft for business purposes (Hanlon, 1999; Wells, 1997). The growth of the aviation industry itself was launched after the Second World War and supported by the enactment of the United States Airline Deregulation Act of 1978. The value of the air transport market is very high, in fact, some occupations and trades would simply not be possible if air transport were not available, development of air transport was soon followed by other countries (Gillen & Morrison, 2005; Wells, 1997).

At the beginning of the twenty first century, the growth of the world aviation business started to collapse, triggered by the 9/11 tragedy in 2001. However, over-capacity in the market had been signalled in 2000 because the launch of Low Cost Carriers (LCC) in the early 2000s (Doganis, 2006). This condition was also supported by increased operational costs that resulted in stagnancy in the air transport industry (Doganis, 2006). In contrast with the aviation business, the developments of improvements in aviation safety in aircraft design and airline operation were not affected. The awareness of aviation safety is also supported by the uniqueness of this industry. Different from other types of industry, aviation is categorised as a seeking-reliability organization or a high hazard environment organization because if a crisis occurs, it will not only affect the aircraft, the victims or the particular airline, but also the entire aviation organization. Therefore, safety is essential in aviation. Studies related to aviation safety are needed because previous studies proved that developments in design and system safety made significant contributions to reducing the aviation accident rate (Liou, Tzeng, & Chang, 2007).


Aviation Safety

Safety is a crucial element in any complex socio-technological organization. Safety means dealing with risks, which is not easy but possible to achieve. It is highly desirable to improve safety to anticipate unpredictable events in the future. In commercial aviation, safety is one of the priorities. However, ever since “the Glider King” Otto Lilienthal invented the glider in 1891 and was killed in one of his experiments, world air transport has faced the same problem: accidents.

The first commercial aviation accident occurred on 10 January 1954. The aircraft, a de Havilland DH-106 Comet, operated by BOAC, crashed off the coast of Italy, killing 39 passengers and crew. This fatality warned the aviation community of the need to improve safety. Early research in aviation safety focused on mechanical or technical failures and the new inventions in aircraft technology succeeded in reducing the number of accidents caused by technical matters. Lately, human error has been seen as a new perspective in aircraft accidents.

Human error issues first emerged at the Flight Safety Foundation 28th Annual Seminar in 1974 and, in the following year, the IATA 20th Technical Conference focused on the theme of human error and human factors in aviation safety. At this point, human error was defined as a deviation from expected human performance (Allnutt, 1976; Chapanis, 1972; Goldberg, 1984; Peters, 1966; Senders & Moray, 1991) which results in accidents. Later, Rasmussen & Pedersen (1984), Perrow (1984) and Reason (1990, 1997) argued that although human error is a cause of accidents, it is also introduced as a consequence and product of a system in an accident. In addition, the contribution of humans to aviation accidents does not act as a single failure but as a final act of failure in a complex system. Research has shown that human contributions to aircraft accidents is approximately 70% (Boeing Commercial Airplanes, 2007; McFadden, 1996; McFadden & Hosmane, 2001; McFadden & Towell, 1999) and sometimes up to 80% (Gaur, 2005; D. A. Wiegmann, Neil, & Paul, 2001; Douglas A. Wiegmann & Schappell, 2003).


Aviation Accident Theory

In contrast with safety, accident is a word that is used to indicate the failure in an operation. In aviation, accident is associated with an occurrence in aircraft operation and the cause of fatality or serious injury, the aircraft sustaining damage or structural failure, the aircraft is missing or is completely inaccessible (ICAO, 2001b). Thus, an aviation accident is always related to victims, damage or failures and much work is needed to try to minimize this by understanding the origin of the accident through investigation.

Studies of the origin of aviation accident had determined the Primary Causal Factors as: human errors, mechanical failures, hazardous weather, sabotage and military operations (Janic, 2000; Wells, 1997; Wells & Rondigues, 2003), as the factors most closely related to aircraft accidents and incidents (Wells, 1997). On the other hand, NTSB classifies commercial aviation accident causations into three levels: Causal factors (weather, traffic environment, pilot capabilities, aircraft capabilities and unpredictable acts), operational effects (federal aviation, commercial aviation, and manufacturers), and broad effects (economics, values and policies). These factors mostly act together because even though it is possible, it is rare for an accident to be caused by a single factor. The interactions of causal factors were called unsafe acts (Heinrich, Petersen, & Roos, 1980), or unsafe condition (Bird & Loftus, 1976) or active failures by Reason (1990, 1997).

Unsafe acts or active failures are defined as related to latent conditions by Reason (1990, 1997) who introduced a "Swiss Cheese Model” to explain both terms that work together to generate an accident. Reason explained that latent conditions (e.g. decisions of top management), like a pathogen in a human body, will adapt to the organization's environment and weaken the organization's immunity into a state of vulnerability. Thus, the latent condition will hide for a period of time while it deteriorates an organization so an organizational accident does not occur suddenly but slowly over time. Hence, Reason clearly stated there is an organizational influence in an accident.

Lately, Reason’s “Swiss Cheese Model” was developed further by Wiegmann and Shappel (2003) to include organizational influence in an aviation accident. Their model is named Human Factors Analysis and Classification System (HFACS) and is able to measure organizational influence in an accident. HFACS also proved Reason’s theory of organizational influence in creating an accident by applying this model in FAA and NASA research (Netjasov & Janic, 2008). Recently, several studies applying the HFACS model to investigate aviation accidents in different countries showed significant contributions of organizations to aviation accidents.

Gaur (2005) examined Indian aircraft accidents by analysing 48 accident cases and stated there were 52.1% organizational influences in Indian aviation accidents while Li, Harris and Yu (2008), who examined 41 commercial aviation accidents in the Republic of China, concluded a significant contributory relationship of operational errors within higher levels of organizations. In Germany, Dambier and Hilkenbein (2006) stated that organizational factors in aviation accidents are much lower in India, at only 15%. However, they assumed it might be caused by differences in recording data between the two countries.


Aviation Accident Investigation

Investigation in aviation is essential as the rapid growth of air transport has often been followed by a series of accidents (Janic, 2000). Owen (1998) stated there were difficulties about aircraft accident investigations because, generally, they arise from a complex system of mutually dependent and without improvement in aviation safety, the occurrences will be high and would likely to continued to occur. Thus, investigations of accident causation in fatal accidents, assessment of their risk and setting-up a risk standard consistent with society’s preference function, are used to improve aviation safety (Sage & White, 1980). In addition, investigation of an accident also provides recommendations to prevent the same event happening again (Wells, 1997). For those reasons, the aviation accident investigation shall be used only for preventing the occurrence in the future without any intentions to apportion blame or liability (ICAO, 2001b) because the investigation should not only focus on immediate or direct causes but also consider the historical background and unfavourable organizational contexts (Dien, Llory, & Montmayeul, 2004).


Organizational Learning

Learning after an accident has been studied from various perspectives and has even been combined with different disciplines, for example, socio-engineering, psycho-engineering and socio-psychology. Study after accidents is used to consider the level of analysis, before a suitable approach can be developed. Sagan (1994) mentioned that investigation should not only focus on organizational systems after accidents have occurred but also investigate the organizations’ safety systems. Therefore, Le Coze (2008) proposed an approach to accident investigation based on different levels of analysis, whether the focus will be at the micro level (human error), meso level (collective and organizational level), or macro level (inter-organizational and social level). These levels of analysis will affect the type of data that will be collected, whether it will be from a normative perspective (data that should fit the model) or use a descriptive strategy (model that should fit the data).


Barriers in Organizational Learning

Perrow (1984) stated the influence of power in high-risk systems in which he implied that power tends to beneficial in only for some people. The influence of power in organizational learning is also stated by Sagan (1993) in addressing four barriers that affect an organization’s failure to learn. He stated there is a high ambiguity of perceptions as the result of differences in interpretations according to the information available. Another barrier is the politic of blame because this is the easiest way to clean up the mess by asking someone to pay the bills. The other two barriers are faulty reporting as the result of someone’s interests to prevent the truth and the last one is secrecy of internal organization due to restrictions on information flow.

The restricted information or information difficulties in organizational learning was also implicitly implied by Vaughan (1996). She found this barrier when she examined the Challenger Accident. She implied that an organization tends to cover the failure within the organization for several reasons, such as protecting the reputation of the organization and preventing more losses.

Lately, Pidgeon and O’Leary (2000) also found two barriers to organizational learning from a disaster. The first barrier is information difficulties which in highly uncertain and ill-structures undermine attempts to learn because misunderstandings resulting from wrong assumptions, inherent difficulties of handling information, ambiguity of regulations and denying danger will delay preventive action. The second barrier is blame, organizational politics and cover-up which organizational politics tend to deny, cover up or even blame others to protect organisational secrecy.

A recent study from Ballesteros (2007) addressed four barriers to organizational learning. His conclusions are based on his experiences and research in aviation organizations. The barriers in organizational learning are regulatory hypertrophy, emphasis on the determination of responsibility, excessive automation, learning difficulties for the human operator and the dysfunctional role of the regulatory bodies.

Moreover, the contribution of power in an accident also stated plays significant roles in organizational learning. It has been stated by Perrow (1984), Sagan (1993, 1994), Pidgeon and O’Leary (2000). They believe that power, whether inside or outside organization, influences accidents before and after they occur. Before the accident occurs is related to operational processes and decision making while after the accident is related to failure in learning from the occurrence. Thus, the influence of power in an accident should be considered since it makes a significant contribution to the accident.

Hence, power inside or outside an organization possibly leads to hiding, ignoring or covering-up the existence of accidents by interested individuals or organizations. It seems that the powerful actors will play a significant role in controlling the interpretation of accidents and lessons learnt from them. Although it is hard, but by having willingness to improve safety, an accident can be prevented.

No body wants an accident to happen. Toward investigation and organizational learning, we can prevent it but also able to create recurrence. We have a choice, to prevent it or to make it happen again!

Penulis adalah PhD candidate di Lincoln University, New Zealand