Rand Corp. Study on OSHA Inspections, Injury Rates
Key Finding: Inspections conducted from 1999 through 2005 lowered
injury rates among manufacturers 19 to 24 percent, a far greater improvement
than was found in a study of inspections performed from 1979 through 1998.
What's Next: Researchers plan to examine why the size of employers
makes a difference in how much injury rates decline.
Injury rates at manufacturers declined an average of 19 to 24 percent
annually in the two years after receiving visits by federal safety inspectors,
according to a new study of the impact of Occupational Safety and Health
Administration enforcement actions from 1998 through 2005.
After two years, injury rates tended to rebound to what they had been before
inspectors arrived, the study found.
The results contrast with a report by some of the same researchers issued
seven years ago that found inspections had much less impact on injury rates
during the 1980s and 1990s. In 1998, the average annual injury rate decreased by
5 percent following an inspection.
John Mendeloff, co-author of both reports and director of Rand Corp.'s Center
for Health and Safety in the Workplace, told BNA May 24 the results do not
provide clear answers to why the more recent inspections prompted larger
“The article considers whether site specific targeting (SST) may have been a
factor, but we can't tell because the sample of SST inspections in our sample is
too small,” Mendeloff said.
The study, “A New
Estimate of the Impact of OSHA Inspections on Manufacturing Injury Rates
1998-2005,” was published online May 7 by the American Journal of
Industrial Medicine. It is a follow-up study to a 2005 report, “The
Declining Effects of OSHA Inspections in Manufacturing, 1979-2005,” that first
appeared in the July 2005 edition of Industrial & Labor Relations
While the two studies used different sources for data, Mendeloff said, that
does not account for the improvement.
For the new study, the researchers used data for Pennsylvania manufacturers.
To determine injury rates, the team checked numbers from Pennsylvania's workers'
compensation program and state employment figures. Inspection information came
The researchers used Pennsylvania data because they already had access to the
information from other studies they were conducting, and three of four authors
were based in Pittsburgh.
The prior study focused on the 29 states where federal OSHA inspected private
employers; injury rates were obtained from the Bureau of Labor Statistics and
inspection results from OSHA.
Using workers' compensation data instead of BLS numbers meant more small
manufacturers--those with fewer than 250 workers--were included in the new
study, making the sample more representative, Mendeloff said.
The final sample group had 8,645 Pennsylvania manufacturers, all of whom have
only a single site and 11 or more employees.
While the sample groups are different, Mendeloff believes the results can be
compared because in all cases federal OSHA conducted the inspections.
Several different inspection scenarios were examined, including:
impact of inspections that produced penalties,
happened when there were no sanctions,
effectiveness of programmed inspections compared with inspections prompted by
difference employers' size made.
Overall, the research found that during the two years after an inspection,
injury rates declined about 19 to 24 percent annually. The visits with the
greatest impact were programmed inspections that produced penalties, Mendeloff
After two years, an inspection's impact faded, and the rate returned to about
what it had been before OSHA arrived.
While the study did not look at why injury rates rebounded, Mendeloff said
there could be several reasons. Workplaces may change procedures or buy new
equipment, but those improvements could be countered by old practices that
reemerge or new manufacturing processes that create unfamiliar hazards.
There were some scenarios in which inspections did not lead to a significant
difference in injury rates, Mendeloff said.
Inspections did not change injury rates for establishments with more than 250
workers or 11 to 20 workers. The rates also did not change if no penalty was
Mendeloff said the fact that a large employer did not show an impact could be
because the inspection did not cover the full range of activities at the
business, leaving some functions untouched by inspections. Also, large employers
are more likely to have safety programs in place and be inspected more often
than smaller businesses, factors that could reduce the impact of an
The reason for the lack of impact on small employers is not clear. “That's
one thing we hope to do more research on,” Mendeloff said.
By Bruce Rolfsen
The new article is available for a fee at http://onlinelibrary.wiley.com/doi/10.1002/ajim.22062/abstract.
The 2005 report is available at http://op.bna.com/env.nsf/r?Open=jstn-8urra3.