Estimating the number of fatal accidents and investigating the determinants of safety performance in indian construction

Construction workers and employees are extremely vulnerable to injuries and fatalities due to the hazardous conditions of their work, unsafe acts, and deficient safety management. These lead to poor safety performance. The study was conceptualized and implemented in four phases to meet the following objectives: (1) the estimation of the number of fatal accidents in the Indian construction sector; (2) the development of a model to predict the safe work behavior of employees and co-employees working in a construction project; (3) the formulation of a model to predict the safety climate of a construction project; (4) the creation of a framework to evaluate the project hazard level; and (5) the assessment of some hypotheses on determinants of safety performance and their interrelationships. The research methodology which was used included the artificial neural network (ANN), fuzzy set theory, fuzzy measure, fuzzy integral, multi attribute decision making (MADM) methods, and the structural equation modelling (SEM). An important finding of the study is that on an average, 38 fatal accidents occur daily in the Indian construction industry. Safety climate constructs including the supervisory environment, work pressure, employees’ involvement, personal appreciation of risk, supportive environment, commitment, appraisals of physical work environment, work hazard, and communication are sensitive to safe work behavior while management commitment, supervisory environment, appraisal of physical work environment and work hazard are sensitive to safety climate of the projects. The study contributes to theory and practice in the form of models and methodologies to: (1) predict the safe work behavior of employees and safety climate in construction projects; (2) to discover the outliers in projects based on the safety climate and the safe work behavior of employees working in them; and (3) to evaluate a project’s hazard level. The study lays emphasis on the systematic and functional accident data recording system at the state and national level. It also advocates the use of sensitive constructs for speedy improvement in safe work behavior and safety climate of projects. The study offers valuable resources for the improvement of safety performance.