Risk and Safety Management in Construction: Examining the Interplay of Organizational Commitment and Project Complexity

Abstract

This study investigated the mechanisms through which safety management (SM), organizational commitment (OC), and risk management (RM) influenced construction performance (CP) in the construction industry, and it also examined the moderating role of project complexity (PC). Grounded in resource-based and contingency theories, a multidimensional model was used to formulate and test hypotheses about the positive direct effects of SM, OC, and RM on CP and also the ways in which PC may exert differential moderating influences. Data from 279 valid questionnaires collected from diverse construction projects in Qingdao, China, were analyzed using partial least squares-structural equation modeling (PLS-SEM). Results confirmed significant positive direct effects: SM (β = 0.360) mitigated accidents and rework, bolstering schedule and cost stability; OC (β = 0.188) enhanced employee engagement and collaborative efficacy; and RM (β = 0.223) proactively addressed uncertainties, curbing delays and overruns. PC negatively moderated the RM-CP relationship (β = −0.092), attenuating RM’s benefits amid heightened uncertainties; positively moderated the OC-CP link (β = 0.119), amplifying OC’s motivational role in complex environments; and exhibited no moderation on SM-CP (β = 0.016), affirming SM’s consistent foundational impact. These findings advanced a theoretical understanding by delineating PC’s nuanced interplays, providing managerial guidance with which firms could institutionalize proactive SM, foster OC via incentives and participation, and implement agile RM strategies in high-complexity contexts to optimize safety, efficiency, and quality synergistically. Limitations could be related to the study’s geographic specificity and any potential biases of self-reporting. Future research should employ cross-cultural, longitudinal designs and incorporate additional mediators such as leadership to enhance model comprehensiveness.