Randomized government safety inspections reduce worker injuries with no detectable job loss

Science

Volumn 336, Issue 6083, 2012, Pages 907-911

  Levine, David I ^a   Toffel, Michael W ^b   Johnson, Matthew S ^c  

^a UNIVERSITY OF CALIFORNIA   (United States)

^b HARVARD BUSINESS SCHOOL   (United States)

^c USA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EMPLOYMENT; HEALTH RISK; INJURY; OCCUPATIONAL EXPOSURE; STATE ROLE; SURVIVAL; WORKPLACE;

ARTICLE; GOVERNMENT REGULATION; HEALTH CARE COST; HUMAN; OCCUPATIONAL ACCIDENT; OCCUPATIONAL SAFETY; PRIORITY JOURNAL; RISK ASSESSMENT; RISK REDUCTION; UNEMPLOYMENT; UNITED STATES;

CALIFORNIA; UNITED STATES;

EID: 84861219936     PISSN: 00368075     EISSN: 10959203     Source Type: Journal    
DOI: 10.1126/science.1215191     Document Type: Article

References (25)

1. A. Haviland, R. Burns, W. Gray, T. Ruder, J. Mendeloff, J. Safety Res. 41, 339 (2010).
2. J. Feldman, OSHA Inaction: Onerous Requirements Imposed on OSHA Prevent the Agency from Issuing Lifesaving Rules (Public Citizen's Congress Watch, Washington, DC, 2011): www.citizen.org/documents/osha-inaction.pdf.
3. S. Pelley, Is enough done to stop explosive dust? 60 Minutes (CBSnews.com), June 8, 2008, www.cbsnews.com/stories/2008/06/05/60minutes/ main4157170.shtml.
4. MSDSonline, New safety poll: Do OSHA regulations kill jobs or stop jobs from killing people? October 13, 2011, http://blog.msdsonline.com/2011/10/new- safety-poll-do-osha-regulations-kill-jobs-or-stop-jobs-from-killing-people/.
5. J. Sherk, Opportunity, parity, choice: A labor agenda for the 112th Congress (Heritage Foundation, Washington DC, July 14, 2011), www.heritage.org/research/reports/2011/07/opportunity-parity-choice-a-labor- agenda-for-the-112th-congress.
6. Public Citizen's Congress Watch, Sen. Coburn is dead wrong on worker safety [press release] (1 August 2011), www.citizen.org/pressroom/ pressroomredirect.cfm?ID = 3394.
7. R. S. Smith, Ind. Labor Relat. Rev. 32, 339 (1979).
8. The supplementary materials provide a formal model of these results.
9. R. S. Smith, J. Hum. Resour. 14, 145 (1979).
10. W. K. Viscusi, Bell J. Econ. 10, 117 (1979).
11. J. W. Ruser, R. S. Smith, J. Hum. Resour. 26, 212 (1991).
12. W. B. Gray, J. T. Scholz, Law Soc. Rev. 27, 177 (1993).
13. J. Mendeloff, W. Gray, Law Policy 27, 219 (2005).
14. A. Haviland, R. M. Burns, W. B. Gray, T. Ruder, J. Mendeloff, Am. J. Ind. Med. 10.1002/ajim.22062 (2012).
15. W. B. Gray, Am. Econ. Rev. 77, 998 (1987).
16. C. Dufour, P. Lanoie, M. Patry, J. Prod. Anal. 9, 233 (1998).
17. J. W. Ruser, J. Risk Insur. 62, 67 (1995).
18. J. P. Leigh, J. P. Marcin, T. R. Miller, J. Occup. Environ. Med. 46, 10 (2004).
19. California Department of Industrial Relations, 2005 Report on the High Hazard Enforcement Program and High Hazard Consultation Program (Division of Occupational Safety and Health, Sacramento, CA, 2007), www.dir.ca.gov/dosh/ enforcementpage.htm, accessed September 2011.
20. To assess the impact of our handling of outliers, we reestimated our models on variables that were neither top-coded to correct for large outliers nor corrected to account for very small outliers (i.e., we added only 1 before taking the log of Injury count, Injury cost, Sales, Employment, and Payroll). The results, presented in tables S9 and S10 in the supplementary materials, continue to indicate that inspections lead to statistically significant reductions in Injury count and Injury cost. The results also continue to yield no evidence that inspections affected Employment, Payroll, or Sales. The magnitude of these estimated effects on injury rates and injury costs exceeded those yielded by our primary model results. This confirms the conservative nature of our primary estimates and suggests the importance of mitigating the influence of outliers.
21. Cal/OSHA would have had some data on injury rates for workplaces they had recently inspected. Because their procedures were to avoid randomized inspections for workplaces with any inspection in the previous 2 years, we dropped potential treatments and controls that had been inspected within 2 years before the match year. Cal/OSHA only had inspected 7% of treatments in the 4 years before the random inspection year; results were unchanged when we dropped treatments or controls with inspections in the prior 4 years.
22. In the supplementary materials, we show that the cost of reported injuries in medical care and lost wages, not counting pain and suffering, is very roughly $8400 per employee in high-hazard industries in California. If an inspection reduces all costs by the same 26% that we estimated for workers' compensation costs (Table 1, column 3) and if there is an average of 33 employees per employer in our sample, then a Cal/OSHA inspection leads to roughly $71,000 in lower medical costs and lost wages per year. If the effect lasts from the inspection year through the next 4 years (as in Table 1, column 4), the total value to society of an inspection is very approximately on the order of $355,000. This estimate is very rough and ignores the underreporting of injuries (23, 24), safety benefits lasting more than 4 years, the reduction in pain and suffering, and (working in the opposite direction) the discounting of future benefits.
23. K. D. Rosenman et al., J. Occup. Environ. Med. 42, 25 (2000).
24. J. Biddle, K. Roberts, K. D. Rosenman, E. M. Welch, J. Occup. Environ. Med. 40, 325 (1998).
25. The figure of $221,000 is the lower bound of the 95% CI of our estimate on payroll (-0.021) times mean payroll ($2,101,500) times 5 years.