A framework for understanding the development of organisational safety culture PDF

Preview

Summary

Safety Science 44 (2006) 551–562 www.elsevier.com/locate/ssci A framework for understanding the development of organisational safety culture a,¤ Dianne Parker , Matthew Lawrie a, Patrick Hudson b a Department of Psychology, University of Manchester, Oxford Road, Manchester M13 9PL, United Kingdom b ...

Description

Safety Science 44 (2006) 551–562 www.elsevier.com/locate/ssci

A framework for understanding the development of organisational safety culture Dianne Parker a

a,¤

, Matthew Lawrie a, Patrick Hudson

b

Department of Psychology, University of Manchester, Oxford Road, Manchester M13 9PL, United Kingdom b Faculty of Social Sciences, Leiden University, 2300 RB Leiden, The Netherlands Received 11 February 2005; received in revised form 28 September 2005; accepted 10 October 2005

Abstract A framework for the development and maturation of organisational safety culture was formulated. The content of the framework was informed by 26 semi-structured interviews with oil and gas company executives, each very experienced in the industry. The form of the framework was based on Westrum’s [Westrum, R., 1996. Human factors experts beginning to focus on organizational factors in safety. ICAO Journal] typology of organisational cultures, which was adapted and extended as proposed by Reason [Reason, J., 1997. Managing the Risks of Organisational Accidents. Ashgate, Aldershot]. The product was a set of short descriptions of each of a number of aspects of organisational safety at each of Wve levels of safety culture advancement. The framework was assessed for face validity. Theoretical implications and possible applications of the framework are discussed.  2005 Elsevier Ltd. All rights reserved. Keywords: Safety culture; Organisations

1. Introduction Organisational culture has been deWned as ‘a complex framework of national, organisational and professional attitudes and values within which groups and individuals function’ (Helmreich and Merritt, 1998). Part of that culture in hazardous industries relates to *

Corresponding author. Tel.: +44 161 275 2570; fax: +44 161 275 2588. E-mail address: [email protected] (D. Parker).

0925-7535/$ - see front matter  2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.ssci.2005.10.004

552

D. Parker et al. / Safety Science 44 (2006) 551–562

safety, which was deWned by Reason (2000) as the “ability of individuals or organisations to deal with risks and hazards so as to avoid damage or losses and yet still achieve their goals”. The beliefs and values that refer speciWcally to health and safety form the subset of organisational culture referred to as safety culture (Clarke, 1999). Important early work on safety culture includes Zohar (1980) study assessing the shared perceptions guiding appropriate and adaptive safety related behaviour based on cues in the environment, and a later study by Cox and Cox (1991) studying the attitudes that employees share in relation to safety. Since then, the concept of safety culture has attracted a great deal of research attention from a range of academic disciplines. Inevitably, the perspectives taken and points of emphasis vary, and there are several ongoing debates in the literature, for example about the distinction between culture and climate (Neal et al., 2000), and about the scope of the concept (Flin et al., 2000). It is not within the scope of this paper to oVer a systematic review of that literature. Several excellent reviews have already been published (Cooper, 2000; Guldenmund, 2000; HSE, 2005). In terms of its application to the issue of accident reduction in high risk industry, the safety culture approach to accident reduction emphasises the role played by social forces within an organisation that act upon its members with respect to safety (Clarke, 1999). It has been suggested that culture reaches equally into all parts of the organisational system and exerts a consistent eVect (for good or ill). For this reason its improvement is more eVective than increased supervision or more rigorous procedures in enhancing safety performance (Reason, 1998). Reason (2000) proposed that an organisation’s safety culture takes on a profound signiWcance at the point where accident rates reach a “plateau”, i.e. where negative outcome data bottom out at some asymptotic value. In order to go beyond this “low but (seemingly) unassailable” plateau and to continue improvement in safety performance, it is necessary to address the hearts and minds of the management and workers (Lee, 1998). This plateau is often reached after requirements for safety “hardware and software” (i.e. barriers and procedures) have been met (Cox and Cox, 1991). In describing how to engineer a safety culture, Reason (1997) identiWed the main elements of such a culture, which was conceptualised as the engine that drives the organisation towards the goal of maximum operational safety. The power of this engine relies heavily on a continuing respect for the potential hazards associated with the organisation’s activities—“not forgetting to be afraid”, termed chronic unease. Reason proposed that an organisation with an eVective safety culture: • has a safety information system that collects, analyses and disseminates information from incidents and near misses, as well as from regular proactive checks on the system; • has a reporting culture where people are prepared to report their errors, mistakes and violations; • has a culture of trust where people are encouraged and even rewarded to provide essential safety-related information, but also in which it is clear where the line between acceptable and unacceptable behaviour is drawn; • is Xexible, in terms of the ability to reconWgure the organisational structure in the face of a dynamic and demanding task environment; • has the willingness and competence to draw the right conclusions from its safety system, and is willing to implement reform when it is required.

D. Parker et al. / Safety Science 44 (2006) 551–562

553

From this brief description of some key studies and concepts, it is clear that a ‘desirable’ safety culture does not just emerge, fully-formed, within an organisation. Given the size and complexity of many modern organisations, within one speciWc organisation it is likely that there are areas in which the safety culture is less well developed than in others. Moreover a ‘desirable’ safety culture would take time to establish, and development may proceed more quickly in some areas than in others. This being the case it can be argued that general measures of safety culture yielding one overall index or score, provide at best a crude indication of this complex phenomenon. A more useful way of conceptualising the safety culture must acknowledge that it is likely to vary within a single organisation. A recent empirical demonstration of this came in Zohar’s (2000) study of 53 work groups in a single manufacturing company, demonstrating both within-group homogeneity and between-group variation in safety-related perceptions within the organisation. These groups, operating at the level of supervisory units, diVered in terms of their perceptions of safety over a range of facets, or aspects, of the organisation. Moreover, safety-related perceptions were predictive of behaviour, in terms of accidents at work requiring Wrst aid or more over a Wve-month period. At the same time, research has indicated that safety culture is best considered as a multidimensional concept. For example, studies of safety culture invariably stress the importance of safety-related attitudes and actions among management (e.g. Clarke, 1998). Zohar (1980) and Cox and Cox (1991) both discuss the importance of commitment to safety by top management. It is crucial that the workforce sees that managers have the attitudes and enact the behaviours that support safety. Perceptions of senior managers’ attitudes and behaviours in relation to safety will form the basis for the safety behaviour of workers, and hence the safety performance of the organisation (Clarke, 1999). Negative perceptions of management commitment to safety can erode employees’ safe behaviours (Clarke, 1998). While positive safety attitudes at the senior management level are essential for positive safety culture (GriYths, 1985), it cannot be assumed that such attitudes will cascade through the organisation. There is a need to regulate how these attitudes are transmitted to subordinates to ensure that managerial commitment to safety is accurately perceived (Clarke, 1999). The importance of communication, including the reporting of hazards, incidents and accidents, in the development of safety culture is well established. In his classic study, Zohar (1980) found that frequent and open contact between managers and workers was related to good safety performance. More recently, Clarke (1999) has advised that the role of supervisors in staV-management communications needs more careful attention, as supervisors are often the intermediaries between management and workforce. Clarke suggested that negative stereotyping of senior management attitudes and actions may be reXective not only of a lack of direct contact, but also of a more fundamental lack of conWdence in management commitment to safety. At the heart of a safety culture is the way in which organisational intelligence and collective imagination regarding safety issues are deployed (Pidgeon and O’Leary, 1994), i.e. the safety information system of an organisation, which is an essential aspect of an informed culture (Reason, 1997). Some studies have taken the use of, and conWdence in, an incident reporting system as an indicator of employees’ perceptions of management commitment to safety: as a dimension of safety culture (Zohar, 1980; Mearns et al., 2001a); as a direct inXuence on employees’ perceptions in relation to safety (Rundmo, 1992), and as an inXuence on safety behaviours (Clarke, 1998). The willing participation of the workforce in a safety

554

D. Parker et al. / Safety Science 44 (2006) 551–562

information system was one of Reason (1997) critical subcomponents of safety culture, as applied to limiting organisational accidents. The organisational safety literature supports the inclusion of several aspects of an organisation to describe the components of safety culture (Zohar, 1980; Reason, 1997; Hale, 2000; Mearns et al., 2001a). In addition to these studies, Cox and Cox (1996) stated that few studies have been designed to capture multiple components (i.e. aspects) of safety culture as it has been deWned in the literature; they typically look at a few in detail and acknowledge the need to research the remainder (Sorensen, 2002). In a similar vein, Mearns et al. (2001b) proposed that an organisation’s eVorts to develop a safety culture should involve identifying areas of strength and weakness. A third issue of relevance to research on safety culture is the extent to which it is amenable to change. While some aspects of culture may be relatively static, there are nevertheless some levers for change and development. For example, a change in the top leadership, or the introduction of a revised safety management system might be expected to have an eVect on the culture of an organisation. Thus it can be argued that a useful framework ought to include developmental aspects of safety culture. The thinking of sociologist Ron Westrum (1993, 1996, 2004) is pertinent here. He suggested that one way to distinguish between organisational cultures was according to the sophistication of the way that safety-related information was handled in the organisation. He developed a typology of cultures, each reXecting a characteristic way of handling information Xow, and representing increasing levels of advancement. The three types he identiWed were labelled pathological, bureaucratic and generative (see Fig. 1). It is noticeable that many of the characteristics of the generative organisation suggested by Westrum echo the characteristics of the high reliability organisation (HRO) as described by Karl Weick (1987). HROs are those few organisations (e.g. air traYc control, aircraft carriers) with high risk technologies which nevertheless cope well with the associated hazards and have good safety performance records. Weick proposed that the key element of HROs was simultaneous centralisation and decentralisation. Compliance is ensured without surveillance, because members of the organisation are intrinsically motivated for safe working behaviour. This can only be achieved through a highly developed, mature safety culture. Hale (2000) stated his belief that the essence of safety culture lies in issues dealt with by Westrum (1993), Reason (1997) and Hudson (1999), and that an

Pathological

Bureaucratic

Generative

Information is hidden

Information may be

Information is actively

ignored

sought

Messengers are “shot”

Messengers are tolerated

Messengers are trained

Responsibilities are

Responsibility is

Responsibilities are shared

shirked

compartmentalised

Bridging is discouraged

Bridging is allowed but

Bridging is rewarded

neglected Failure is covered up

Organisation is just and

Failure causes inquiry

merciful New ideas are actively

New ideas create problems

New ideas are welcomed

crushed

Fig. 1. How diVerent organisations respond to information concerning safety (from Westrum, 1996).

D. Parker et al. / Safety Science 44 (2006) 551–562

555

organisation has only achieved a “true” safety culture when it has attained the generative level of culture put forward in the Westrum typology. Westrum’s thinking about levels of safety culture advancement has not, to our knowledge, previously been directly used in an organisational setting. His idea of diVerent ‘levels’ of safety culture provides the basis for developing a framework of safety culture that takes into consideration both a range of areas in which safety culture might manifest itself, and a range of levels of sophistication. In the study reported here a framework of safety culture was developed that was adapted from Westrum’s three levels and covered several of the aspects of safety culture shown to be key in previous studies. The key objective of the study was to develop a theory-based tool that could be used by organisations in the oil industry to self-assess their current level of safety culture advancement. Building on the three levels originally suggested by Westrum (1993), Wve levels of safety culture were included and characterised in the following phrases: • • • • •

Pathological; Who cares about safety as long as we are not caught? Reactive; Safety is important: we do a lot every time we have an accident. Calculative; We have systems in place to manage all hazards. Proactive; We try to anticipate safety problems before they arise. Generative; HSE is how we do business round here.

The two additional levels, reactive and proactive, were initially proposed by Reason (1997) as extensions of Westrum’s original typology. They were thought to add depth to the framework, allowing for more subtle classiWcation and increasing the accessibility of the framework to industry employees by including terms they would be familiar with, and so clarifying the idea that organisations would be expected to progress through the levels of safety culture with increasing maturity. Organisational safety research has not generally focused on an integrative framework that includes both formal safety management systems, and the safety-related behaviour of individuals, including managers and frontline workers. For example, the eVects of the content of safety management systems on safety performance have largely been ignored (Cooper, 2000). Mearns et al. (2000) stated that auditing was a key requirement in any eVective safety management system, and Isla Díaz and Díaz Cabrera (1997) found company policies towards safety to be critical, e.g. compliance with safety standards, feedback on performance, assignations of funds and resources to safety areas, importance of safety training. Other inXuences on safety culture that have been suggested are the status within the organisation of safety oYcers (Zohar, 1980), the extent and quality of safety training and the demonstrable impact of safe behaviour on promotion (Cooper and Phillips, 1994). The aim of this exploratory study was to generate a theory-based framework that could be used by organisations to understand their own safety culture. The framework was designed to reXect the multidimensional, dynamic nature of safety culture by providing descriptions of an organisation with respect to a range of key aspects of safety culture, at each of Wve levels of safety culture advancement developed from Westrum’s initial ideas (1993). The aspects of safety culture included were chosen to reXect the key features of safety identiWed in the organisational safety culture literature. It was anticipated that the resulting framework would be meaningful and useful to those working in the petrochemical industry who are concerned to improve the safety culture of their organisation.

556

D. Parker et al. / Safety Science 44 (2006) 551–562

2. Method In depth interviews were carried out with a purposive sample of 26 senior oil and gas company executives, working at a range of multinational oil companies and contracting companies each of which had oYces in Houston, Texas. The aim was to interview very senior employees who had extensive Wrst hand experience of how HSE systems work in the oil industry at all hierarchical levels of an organisation. Although all the interviewees were senior executives at the time of interview, they had all worked their way up through the industry from much lower levels through their careers. They therefore had all had the opportunity to see HSE from a number of diVerent perspectives. The interviews lasted between 60 and 90 min, and were tape recorded with the permission of the interviewee. A second researcher simultaneously made a written summary of the interviewee’s key responses. This summary was read back to them to check that the scribe had captured the essence of what the interviewee wanted to convey. Interviewees were asked to consider in turn several aspects of a large, worldleading organisation’s HSE safety culture. These aspects were grouped under headings, which were the elements of the organisation’s documented HSE management system (HSE-MS). This strategy was adopted because the elements of an HSE-MS are familiar and meaningful to the interviewees, ensuring their engagement with the interview process. Although one speciWc organisation’s HSE-MS was used to frame the interviews, the elements it contained were familiar to all those interviewed, and were present in the management systems of all the companies involved. Following a distinction pointed out by Zohar, both concrete and abstract aspects of safety culture were investigated. Tangible, or concrete, aspects included the system for benchmarking and auditing safety performance, and the way in which work is formally planned. These were the elements of the HSE management system with which the interviewees were all familiar. Less tangible, or abstract, aspects involving the perceptions of the workforce, thought to be important by Zohar (2000) were also considered, as they are likely to be crucial indicators of organisational safety culture. The interviews covered 11 tangible and seven less tangible aspects of safety culture, as shown in Figs. 2 and 3. Interviewees were asked to describe how an oil company would function in terms of each of the aspects covered at each of the Wve levels of safety culture, from pathological through to generative. Once it had been explained to them, Westrum’s framework for levels of safety culture was quickly understood by the interviewees, and they had little diYculty in relating an organisation’s aspects to safety culture at all levels. 3. Results Over the course of the interviews, a consensus quickly emerged in terms of the descriptions oVered by the interviewees, and saturation was achieved in that new and diVe...