Download as pdf
Download as pdf
You are on page 1of 24
In: Crisis Management ISBN: 978-1-60876-570-6 Editors: P. Alvintzi et al. pp. 187-210 © 2010 Nova Science Publishers, Ine. Chapter 6 THE LINATE AIR DISASTER: A MULTILEVEL MODEL OF ACCIDENT ANALYSIS Maurizio Catino” Associate professor in Sociology of organization ~ University of Milan, Bicoeca - Taly. ABSTRACT “This chapter uses a multilevel organizational model to analyse the Linate air disaster (8 October 2001 Milan, Italy), in which a SAS MD-87 and « Cessna collided with the Joss of 118 fives, The cause of the accident was a mistake by the Cessna pilots, who took a taxiway different from the one authorized by the ground controller and moved their plane onto the runway from which the MD-87 was about to tke off But the disaster was Mot caused by human error alone. Analysis of the disaster reveals multiple failures in Communication, coordination, and the technologies used. This chapter develops @ felational and multilevel model for the analysis and prevention of accidents in complex ‘Srganizational systems. This multilevel analysis model highlights the different levels of failure that provoked the disaster: individual (pilots and sir-traffic controllers) organizational (Linate Airport), and inter-organizational (the various organizations involved in the air transport system). The analysis centres on the problem of coordination. 1. INTRODUCTION On Monday October Sth 2001, at 8.10 am., two aigplanes collided at Linate Airport (titan) causing the deaths of 118 people. After the Tenerife disaster in 1977, in which 583 people were killed, the Linate disaster is the most serious ground accident in the history of sir traffic. An SAS aircraft was taking off when it collided with a Cessna, a small-sized private airplane with four people on board. The Cessna took the wrong taxiway, entered the main runway and was destroyed by the impact with the SAS airplane, which was taking off. * Conesponsting author E-mail: maurziocatinogiunimiit 188 Maurizio Catino ‘Analyses of major organizational accidents have revealed the irsporlatlt of organizational factors in the etiology of such events (Perrow, 1984; Weick, 1990; Turner and Pidgeon, 1997; Vaughan, 1996; Reason, 1990; 1997; Snook, 2000) Several approaches and see pave been developed in order to explain accidents: the Latent Factors, Theory (Reason, 1990; 1997); the Normal Accidents Theory (Perrow, 1984; Sagan 1993); the High Reliabtio, Theory (Roberts and Rousseau, 1989; Roberts, 1990; 1993; Weick, 1990, ba Porte seaee alah, 1994, Weick et al. 1999; Roe and Schulman, 2008); the Socio-Technical Theory andlor Resilience Engineering (Rasmussen, 1997; Hollnagel, 2004; Holinagel al. 2006), Various theories and case studies of accidents (see Weick, 1990: Vauehan, 1996; 1909, Snook, 2000) underine the role played by different levels in the production of the accident, huraas, organizational, enviroamental and organizational network factors inleraet aeevontribute to the genesis of accidents, Disasters can only be understood from a relational perspective because they are fundamentally relational phenomena (Gabrosky and Roberts, 1990), This chapter aims to develop a relational and organizarional multilevel model for the analysis of accidents in complex organizational systems, A multileve} and relational perspective is needed 10 investigate on one organizational phenomenon shat weakens an Gruanization’s ability to avert disasters: coordination neglect (Heath and Staudenmaes, 2000, Sook, 2000; Roberts, Madsen and Desai, 2005). A basic assumption ofthis chapter is that re cannot fully understand the Linate disaster (nor others) by treating it as an isolated event By providing a detailed description of the disaster, I shall focus on how the accident took place, and how it as affected by thee inter-related levels of falures: A) the individual tevel ceva errors commitied by the people directly in charge of the event, ie, the air-ttafie vesntrotler and the Cessna pilots B) the organizational level: the weaknesses of the defense System, erroneous managerial decisions and/ar ambiguous conditions that produced the sr C) the fnter-organizational level: the differentiation, coordination and integration of sMrrcent actors with different roles, all involved in the correct functioning of the air waffic system, The Linate accident was caused by human error; but human errors in organizations ae sovially organized and systematically produced (Vaughan, 1996). The alm of this chapter je to demonstrate the multilevel nature of accidents and the problem of coordination failure in complex high-risk systems. Analysis of the Linate disaster shows that risk and salety a properties which depend on the interaction and coordination among different levels findividval, organizational, interorganizational) of the complex organizational system ‘The analysis also draws some coaclusions conceming the relationship between individual and ‘organizational responsibility in the ease of such disasters in the risk society 2. THEORETICAL DOMAIN 2.1. Organizational Accidents: Theories Compared Accidents in complex organizations cannot be atributed to any single cause. Analyses of major organizational accidents have revealed the relevance of organizational factors in He ttiology of such events (Perrow, 1984; Weick, 1990; Vaughan, 1996; Tumer and Pidgee, 1997. Reason, 1990; 1997; Snook, 2000), The importance of organizational factors in the penesis of aocidents was frst asserted by Barry Tumer (1976; Turner and Pidgeon, 1997) “The Linate Air Disaster: A Multilevel Model of Accident Analysis 189 ‘with the disaster incubation model. According to Tumer, accidents are events characterized by a period of incubation during which a series of signals are emitted. f these signals are recognised in advance, they can prevent disasters or reduce their damage, On the basis ofa large number of empirical cases, Tuer analyses accidents as produced by organizational tovon (organization-made disasters) and its bounded rationality (Simon, 1947; 1955; 1956) ‘The disaster incubation model identifies six stages of disaster: 1) starting point, 2) incubation period, 3) precipitaing event, 4) onset, 5) reseue and salvage, 6) fll cultural readjustment Since Turer's theories gained currency (belatedly and to an insufficient extent), several approaches and theories have been developed to explain accidents, a brief outline of which is provided below ‘According 10 the Latent Factors Theory (Reason, 1990; 1997), human errors activate accidents, but these errors are embedded in latent conditions that make the disaster possible scatent conditions - such as poot desiga, gaps in supervision, undetected manufacturing “Jefects or maintenance failures, unworkable procedures, shortfalls in training, .. may be present for many years before they combine with Tocal circumstances and active failure (6 Ponetrate the system's many layers of defences... Active failures are committed by those a the human-system interface (the front-line or sharp-end personnel). Latent conditions, on the ‘ther hand, are spawned in the upper echelons of the organization and within relaed manufacturing, contracting, regulatory and governmental agencies» (Reason, 1997: 10-11). Onpanizational accidents are caused by a rare conjunction of a sat of breaches in successive defences, These windows of opportunity are rare because of the multiplicity of defences and te mobility of the holes. "The Normal Accidents Theory (Perrow, 1984; Sagan, 1993) emphasizes the inevitability of accidents in systems characterized by high complexity and tight coupling. In these ‘organizations, accidents are normal, not in the sense of being frequent or expected, but rather dn the sense thatthe occasional experience of such an interaction is an inherent property of the ‘pater (Perrow, 1984). According to NAT. accidents in high-risk technologies contexts are bound to happen because of their complex and interactive systems, Complex interactions are those of unfumiliar sequences, or unplanned and unexpected sequences, and either not visible ‘or aot immediately comprehensible (Perrow, 1984), Tight coupling, on the other hand, is in place when each part ofthe system is ihtly connected to the others, so that a change at one point produces an immediate change in other parts of the system, with rapid and uncontrolled propagation of is effecis, I'there is interactive complexity and tight coupfing, the system will Inevitably produce an accident, «The odd term normal aceident is meant to signal tht, given the system characteristics, multiple and unexpected interactions of failures are inevitable» (Perrow, 1984: 5). From this it follows that neither better organization nor technological innovation are able (o reduce the likelihond of systemic aceidents in the majority of systems. Two further sets of conditions add to the previous elements (degree of interaction and type of connection) in characterizing high-tisk systems (Sagan, 1993): #) the notion of bounded rationality (Simon, 1947; 1955; 1956) and the garbage can model (Cohen, March, and Olsen +1072: March, 1988): b) the power theory, the notion of organizations as tools to he used in the interests of their masters. “High Reliability Theory (Roberts, 1990; 1993; Schulman, 1993; LaPorte and Consolini, 1994; Weick, 1987; 1990; Weick Sutcliffe and Obstfeld, 1999; Roe and Schulman, 2008) studies unusually successful complex organizations in search of clues for enhanced management of accidents. Researchers seek 10 detect cognitive and management processes

You might also like