Developing Leading Safety Indicators for the New Zealand Construction Industry: A Systems Thinking Approach

Abstract

Construction remains one of the top contributors for workplace injuries and fatalities in many countries. Due to the inability of lagging indicators (e.g., accident rates) to provide early warnings of accidents, the development of leading safety indicators has been a topic of increasing concern for both academics and practitioners in recent years. A review of existing leading indicators in the construction industry reveals that they have the following limitations: (1) ambiguous definitions, (2) problematic simplification process, and (3) a lack of development method. Therefore, the overarching goal of the research is to develop a set of descriptive safety leading indicators that can be used to (1) simplify complex safety phenomena, (2) measure safety performance, and (3) predict the trend in safety. To achieve the goal and address the limitations of existing construction leading indicators, this research first develops a pragmatic method for developing leading indicators, which consists of four steps: conceptualization, operationalization, indicators generation, and validation and revision. The development method provides a systematic process for developing leading indicators. In order to offer systemic insights into simplification process of complex safety realities, this research then explores the dynamics and complexity of construction safety management at the industry and project level. Eight construction safety archetypes are identified, which capture the common behaviour patterns of construction safety management. In addition, a system dynamics model is developed and simulated to monitor the dynamics of safety level at the project level. Furthermore, this research develops and tests an integrative model of safety behaviour using structural equation modelling (SEM). The results improve an understanding of safety behaviours shaping mechanisms and thus help to determine leading indicators with predictive validity. The validity of the integrative model is tested across small and large construction companies. Results suggest that the relationships among safety climate factors and safety behavior were equivalent across the two groups. Finally, a pressure-state-practice (PSP) model is develop to provide an overall framework for developing leading indicators. The safety level of a construction project is conceptualized as a high-level abstract construct that can be assessed by state indicators, pressure indicators, and practice indicators. Criterion validity (i.e., concurrent validity and predictive validity), practicability and cost-effectiveness of the leading indicators were qualitatively tested and supported by the empirical evidence collected from three construction projects. Overall, the research adds to the body of scientific knowledge of leading safety indicators. It improves the understanding of complexity and dynamics of safety management in the construction industry. In addition, the safety leading indicators developed in this paper provide the construction industry with a promising tool to measure safety performance proactively and facilitate safety assessment.