The organisational antecedents of individual safety behaviour in the U.K. offshore oil and gas industry
This thesis sets out to identify the organisational antecedents of individual safety behaviour in the workplace, particularly rule violations. The research was set in the UK offshore oil and gas industry as an example of safety management in hazardous work environments. Study la is a review of incident reports collected over one year in a large UK offshore operating company (n=296). The database was examined using Tripod methodology (Groeneweg, 1996) to understand the relative importance of human factors causes within accident causal pathways. This revealed that unsafe acts, in particular procedural violations, were the most frequent immediate cause of accidents. This was followed by Study lb, a review of the company's safety-related internal audit findings and their remedial actions, which again showed the prevalence of human factors issues; whilst the remedial measures were largely addressing non-human factors issues. A problem with historical reports is their reliability; therefore Study 2 took a proactive approach to determine the frequency and type of rule breaking prevalent in the UK oil and gas industry. This was carried out through a questionnaire study on three locations, offshore and onshore (n=279). The instruments were drawn from existing rule breaking scales (HSE, 1995a; Mearns, Flin, Fleming, & Gordon, 1997), but further developed to test Lawton's (1998) violations taxonomy. This revealed a high prevalence of rule breaking, particularly Situational and Routine violations, which provided support for the violations taxonomy. Also, more frequent rule breaking was associated with a higher injury frequency among co-workers and more self-reported injuries. For a better understanding of these results, Study 3 developed and tested a model of the organisational antecedents of individual rule breaking behaviour. A questionnaire study was used, across 12 offshore locations (n=1414) using a combination of established instruments and scales developed specifically for this research. This revealed that the two independent variables, Organisational Features (Work pressure, Job Satisfaction, Safety Discussion and Satisfaction with Safety Management System Activities) and Social Safety Norms (Others Approval of Shortcuts and Relationship with Supervisor), were associated with self-reported rule breaking. However Organisational Features was more strongly associated with individual accident involvement and Social Safety Norms was a stronger predictor of violations. This study also revealed other aspects of rule-breaking behaviour, which has not been studied extensively in previous research. In particular, it highlighted the importance of job satisfaction and perception of one's colleagues attitudes to safety in predicting violations. This study also revealed that the relationship between rule breaking and some of its antecedents (Involvement, Work Pressure and Approval of Shortcuts) are better explained by non-linear relationships. For the Social Safety Norms variable, the direct influences on individual rule breaking were compared for senior management, site management, immediate supervisor and colleagues. This showed that colleagues' approval of shortcut taking was the strongest predictor of individual rule breaking. This was discussed in relation to Social Exchange theory (Blau, 1964) for Organisation Features, and to Social Influence theory (Latane, 1981) for Social Safety Norms. This however does not address the indirect organisational influences on individual behaviour. There is much speculation about the importance of senior managers in safety management (HSE, 1999; Flin et al, 2000) but little previous research specifically on their role. Study 4 was therefore based on appraisals of safety leadership and leadership style by subordinates (n=256) and self-reports (n=59), from the most senior managers in the company. This was done using instruments developed for this study and the Multifactorial Leadership Questionnaire of Transformational, Transactional and Passive Leadership styles (Bass & Avolio, 1995). This revealed that a charismatic leadership style, (Idealised Influence Behaviour) is associated with a stronger perceived commitment to safety; whilst managers with a more passive style (Laissez-Faire and Management by exception passive) were perceived more negatively. For a sub-sample, subordinate perceptions were correlated with safety performance data. This showed that the priority placed on safety was the strongest determinant of organisational safety performance, with those placing a higher priority on safety having a better safety performance. Overall, the evidence presented in this thesis highlights the organisational influences on committing violations and the importance of managing rule breaking to accident prevention. However, to achieve this requires a focus at the workgroup and leadership level, as well as the traditional management system and individual worker-focused approaches.