The Pennsylvania Certified Safety Committee Program: An Evaluation of Participation and Effects on Work Injury Rates
Sep 18, 2008
An Evaluation of Participation and Effects on Work Injury Rates
Published In: American Journal of Industrial Medicine, v. 53, no. 8, Aug. 2010, p. 780-791
Posted on RAND.org on January 01, 2010
BACKGROUND: Since 1994, Pennsylvania, like several other states, has provided a 5% discount on workers' compensation insurance premiums for firms with a certified joint labor management safety committee. This study explored the factors affecting program participation and evaluated the effect of this program on work injuries. METHODS: Using Pennsylvania unemployment insurance data (1996-2006), workers' compensation data (1998-2005), and the safety committee audit data (1999-2007), we conducted propensity score matching and regression analysis on the program's impact on injury rates. RESULTS: Larger firms, firms with higher injury rates, firms in high risk industries, and firms without labor unions were more likely to join the safety committee program and less likely to drop out of the program. The injury rates of participants did not decline more than the rates for non-participants; however, rates at participant firms with good compliance dropped more than the rates at participant firms with poor compliance. CONCLUSIONS: Firm size and prior injury rates are key predictors of program participation. Firms that complied with the requirement to train their safety committee members did experience reductions in injuries, but non-compliance with that and other requirements was so widespread that no overall impact of the program could be detected.
This report is part of the RAND Corporation External publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.
The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.