TY - GEN
T1 - Quantitative Analysis of Occupational Safety and Health Administration Laws and Violation Citation Data for General Industry
AU - Birk, David
AU - Sayama, Hiroki
N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.
PY - 2025
Y1 - 2025
N2 - In 1970, within the United States, an estimated 14,000 workers were killed on the job. This equates to about 38 worker deaths in the United States every day. On December 29, 1970, President Nixon signed the Occupational Safety and Health Act of 1970 (OSH Act) into law, establishing the U.S. Occupational Safety and Health Administration (OSHA). Coupled with the efforts of employers, workers, safety and health professionals, unions and advocates, OSHA and its state partners have dramat-ically improved workplace safety. In 2022, the most recent full-year data available from OSHA, reported 5,486 worker deaths in the United States, a reduction of work-related fatalities by almost 63%. If you consider the worker population has more than doubled since OSHA began, its impact of the reduction in workplace injuries and deaths is clear. However, workplace injuries and deaths are still a major concern. This study explores the relationship between the number of non-compliance (i.e., violation) citations issued by OSHA for Title 29 of the U.S. Code of Federal Regu-lations (CFR) 1910 that apply to General Industry and quantitative measurements of the text contained in that law. The measurements utilized in this study are Shannon entropy, the Hartley function, and the number of characters within each section of the law. Key findings indicate the Shannon entropy and the Hartley function predict the number of citations made by OSHA. Regression results indicate that number of characters is a stronger predictor of the number of citations made by OSHA. This suggests that the length of the regulatory text is more influential in predicting the frequency of citations than the Shannon entropy and Hartley function.
AB - In 1970, within the United States, an estimated 14,000 workers were killed on the job. This equates to about 38 worker deaths in the United States every day. On December 29, 1970, President Nixon signed the Occupational Safety and Health Act of 1970 (OSH Act) into law, establishing the U.S. Occupational Safety and Health Administration (OSHA). Coupled with the efforts of employers, workers, safety and health professionals, unions and advocates, OSHA and its state partners have dramat-ically improved workplace safety. In 2022, the most recent full-year data available from OSHA, reported 5,486 worker deaths in the United States, a reduction of work-related fatalities by almost 63%. If you consider the worker population has more than doubled since OSHA began, its impact of the reduction in workplace injuries and deaths is clear. However, workplace injuries and deaths are still a major concern. This study explores the relationship between the number of non-compliance (i.e., violation) citations issued by OSHA for Title 29 of the U.S. Code of Federal Regu-lations (CFR) 1910 that apply to General Industry and quantitative measurements of the text contained in that law. The measurements utilized in this study are Shannon entropy, the Hartley function, and the number of characters within each section of the law. Key findings indicate the Shannon entropy and the Hartley function predict the number of citations made by OSHA. Regression results indicate that number of characters is a stronger predictor of the number of citations made by OSHA. This suggests that the length of the regulatory text is more influential in predicting the frequency of citations than the Shannon entropy and Hartley function.
KW - Citations
KW - General industry
KW - Hartley function
KW - OSHA 1910
KW - Quantitative measurements
KW - Shannon entropy
KW - Text analysis
KW - Violations
UR - https://www.scopus.com/pages/publications/105019317415
U2 - 10.1007/978-981-96-4880-1_4
DO - 10.1007/978-981-96-4880-1_4
M3 - Conference contribution
SN - 9789819648795
T3 - Lecture Notes in Networks and Systems
SP - 47
EP - 59
BT - Data Science and Security - Proceedings of IDSCS 2024
A2 - Shukla, Samiksha
A2 - Sayama, Hiroki
A2 - Tiwari, Kapil
A2 - Kureethara, Joseph Varghese
PB - Springer Science and Business Media Deutschland GmbH
T2 - International Conference on Data Science, Computation, and Security, IDSCS 2024
Y2 - 7 November 2024 through 9 November 2024
ER -