Safety climate factors and its relationship with accidents and personal attributes in the chemical industry M.N. Vinodkumar a, * , M. Bhasi b a School of Engineering, Cochin University of Science and Technology, Kochi, Kerala 682 022, India b School of Management Studies, Cochin University of Science and Technology, Kochi, Kerala 682 022, India article info Article history: Received 15 October 2004 Received in revised form 17 September 2008 Accepted 18 September 2008 Keywords: Safety climate Safety management Safety perceptions Safety performance Safety in chemical industry abstract Safety in the chemical industry is a major issue in a thickly populated country like India. The study was carried out to determine the safety climate factors in the chemical industry in Kerala, India. A survey using a questionnaire was conducted among 2536 employees in eight major accident hazard chemical industrial units in Kerala. The study population included workers and first line supervisors at the lowest end of the management. 75% of the data collected was subjected to principal component factor analysis with varimax rotation using SPSS program. This revealed 8 factors which together explained 52.15% of the total variance. Internal consistency (Cronbach Alpha) across items in each of the 8 factors and that of the total scale were found acceptable. The model was tested with the remaining data by running con- firmatory factor analysis using the AMOS 4.0 structural equation modeling program and was found to produce a good fit. The safety climate scores calculated were found to have significant negative correla- tion with self-reported accident rates revealing good predictive validity. One way ANOVA results show that companies’ mean safety climate scores differ significantly from each other indicating that organiza- tions have different safety climate levels. Tests were also conducted to find out the effects of qualification, age, job category and experience of respondents on their perceptions and attitudes about safety. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction The Bhopal gas tragedy that occurred in India in 1984 was an eye-opener for the chemical process industry not only in India but also throughout the world. Investigations and studies of this incident revealed the negligence of human life by company man- agement, ignorance and complacency of workers and a totally irre- sponsible attitude from regulatory agencies (Gupta, 2002). He also opines that such an accident could have happened in advanced countries too in the presence of any of the above conditions. Learn- ing lessons from this incident and realizing the large compensatory expenses, production losses and loss of morale of workers, com- pany managements in India have shown interest to invest for safety infrastructure in all types of industries, especially, major accident hazard industries (ILO, 1996). Even after two decades since the Bhopal disaster, the fatal acci- dent rate in the Indian manufacturing sector stay close to 20 fatal- ities per hundred thousand employees against 1–6 in advanced countries (DGFASLI, 2006). In India, the management of occupa- tional safety and health of workers at organizational level is gov- erned by The Factories Act, 1948 (Central Act 63 of 1948) and Rules formed there under by the state governments. To measure the state of safety in an organization, safety climate studies are a much better option as it overcomes many of the limitations of tra- ditional safety measures such as reporting biases and after the fact measurement (Kennedy and Kirwan, 1998). 1.1. Safety climate Safety climate can be defined as employees’ shared perceptions of safety policies, procedures, practices, as well as the overall importance and the true priority of safety at work (Griffin and Neal, 2000; Zohar, 1980). It is a multi dimensional factor and is regarded as an important antecedent of safety in the workplace. Measuring safety climate can be compared to taking the ‘‘safety temperature” of an organization (Budworth,1997), which provides a snapshot of that organization’s ‘‘state of safety” at a discrete point in time (Cheyne et al., 1998; Huang et al., 2007). Zohar (1980) developed the first measure of safety climate, based on an Israeli sample in 1980 using a 40 items questionnaire covering metal fabrication, chemical, textile and food processing industries. After factor analysis, his final model included eight dimensions with workers’ perceptions of the importance of safety training, management attitude towards safety, effects of safe con- duct on promotion, level of risk at workplace, effects of work pace on safety, status of safety officer, effects of safe conduct on social 0925-7535/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ssci.2008.09.004 * Corresponding author. Tel.: +91 484 2331698; fax: +91 9446606906. E-mail addresses: mnvinodkumar@cusat.ac.in (M.N. Vinodkumar), mbhasi@ cusat.ac.in (M. Bhasi). Safety Science 47 (2009) 659–667 Contents lists available at ScienceDirect Safety Science journal homepage: www.elsevier.com/locate/ssci