Measuring safety performance of construction employees using data envelopment analysis: A case in Australia

Ruixue Zhu, Xiancun Hu, Amy Wei, Wei Yang, Fanrong Ji

Research output: Contribution to journalArticlepeer-review

Abstract

INTRODUCTION: Safety is one of the critical and persistent challenges in the construction industry. Measuring safety performance could allow decision-makers to check safety production processes and enhance the health and safety environment.

METHOD: This study developed a total-factor framework based on the global Data Envelopment Analysis (DEA) method to measure safety performance. The performance trend and influencing factors of pure technical efficiency and scale efficiency were separately investigated. The safety performance of construction employees in Australia was measured as a case study.

RESULTS: The results demonstrate that the safety performance in the Australian construction industry has been evidently enhanced, which is mainly promoted by the progress of pure technical elements. The scale factors did not play a positive and important role in driving the performance.

CONCLUSIONS: The increasing regional differences in safety performance could be reduced by learning the practices from the benchmark construction industry, such as a young worker program, small business safety program, workplace mental health program, and advanced construction technologies.

PRACTICAL APPLICATIONS: The method can be utilized to measure safety performance and investigate the pathways to enhance performance without influencing production inputs and outcomes. The model and experiences of how to promote safety performance on the policymakers and employers were recognized.

Original languageEnglish
Pages (from-to)293-302
Number of pages10
JournalJournal of Safety Research
Volume88
DOIs
Publication statusPublished - Feb 2024

Fingerprint

Dive into the research topics of 'Measuring safety performance of construction employees using data envelopment analysis: A case in Australia'. Together they form a unique fingerprint.

Cite this