Important: Please read the introduction and disclaimer below.

2002 > Zip file (4 meg) containing 4 Excel files (9.7 meg)

2001 > Zip file (5 meg) containing 5 Excel files (12 meg)

2000 > Zip file (4.6 meg) containing 4 Excel files (11.5 meg)

1999 > Zip file (4.3 meg) containing 4 Excel files (10.8 meg)

1998 > Zip file (4.3 meg) containing 4 Excel files (11 meg)

1997 > Zip file (4.5 meg) containing 4 Excel files (11.3 meg)

1996 > Zip file (4.7 meg) containing 5 Excel files (12.1 meg)

1996-2002 > Zip file (31.5 meg) containing all 30 Excel files (79 meg) [hosted by the Internet Archive]

Database > Zip file (22 meg) containing the original MDB/Access file from OSHA (57 meg) [hosted by the Internet Archive]

Earlier on The Memory Hole: Companies With Highest Levels Of Employee Injury and Illness

Associated Press: Court Rules OSHA Must Reveal Injury Rates [04 Aug 2004]

WORKSITES BEHAVING BADLY:
INJURY/ILLNESS RATES FOR TENS OF THOUSANDS OF COMPANIES

by M.R.

================

The Department of Labor's Occupational Safety and Health Administration (OSHA) has for several years tracked lost work day injury and illness (LWDII) rates at about 80,000 worksites in selected high-hazard industries across the country. OSHA kept secret its LWDII database, claiming that the data are confidential. Reporter David Barstow of the New York Times, seeking the names and rates of the most injury- and illness-prone worksites, requested the data under the Freedom of Information Act in October 2002. His request was rejected, and when his appeal was also turned down, he sued for the release of the data in mid-2003.

As a preemptive move, in February 2004 OSHA released a list of the workplaces with the highest rates, but the list didn't contain the rates themselves. That list can be found on The Memory Hole here.

In August 2004, a federal judge ruled that OSHA had to release the names and rates. Eventually the agency did so. (Some background on the case is found in court papers here.)

The complete data - for the years 1996 through 2002 - were obtained under a Freedom of Information Act request and are presented at the links above.

================

OSHA collects the data by requiring surveyed worksites to complete and return a form, or enter information on OSHA's website, including (1) the total number of hours worked by all employees and (2) data from illness and injury records.

OSHA calculates the injury/illness rate by using the equation:

LWDII = (N/EH) x 200,000

Where N is the total number of incidents of lost workday injuries and EH is the total number of hours worked by all employees at the location. The ratio is multiplied by 200,000 to convert it to a rate per hundred full-time employees, assuming that each full-time employee works 2,000 hours per year.

For several years, OSHA contacted companies with LWDII values exceeding 8.0 and asked them to take steps to reduce their injury and illness rate, noting that this value is "considerably greater than the national average." OSHA has stated elsewhere that the national average is 2.8.

While the Times reporter was successful, and published articles using the data, the newspaper decided not to disseminate the raw OSHA data itself. OSHA released the data to a requester in response to an independent FOIA request. Because the courts had ruled against OSHA and ordered release of the data, and because the Times had already obtained the data, OSHA is obliged to release it to anyone else who asks for it.

More recently, OSHA has shifted its focus to worksites in the industries with the very worst overall injury/illness records and worksites with LWDII rates above 14.

This database includes the LWDII rates for all worksites, including both the sites with lower injury/illness rates, as well as the 13,000 higher rate sites.

Nursing homes are quite prominent in the listings.

================

The original data file supplied by OSHA in MDB format proved too large to be downloaded easily. Tech guru Brett Milner converted the database into Excel files for each year. Thus, each year is available as a zipped file above.

Also above, you'll find a large zip file containing all 30 Excel files and another zip with the Microsoft Database file (MDB/Access) that OSHA originally sent. Both of these are graciously hosted by the Internet Archive.

DISCLAIMER

The following disclaimer from OSHA's Directorate of Evaluation and Analysis accompanied this database:

=================

Because of the nature and sensitivity of this data, I ask that you forward the following explanatory notes to anyone that will evaluate and use this information:

1. Scope of the data: For each data collection cycle, OSHA only collects data from a small portion of all private sector establishments in the United States (80,000 out of 7 million total establishments). Therefore, these data are not representative of all businesses and general conclusions pertaining to all US business should not be drawn.

2. Data quality: While OSHA takes multiple steps to ensure the data collected is accurate, problems and errors invariably exist for a small percentage of establishments. OSHA does not believe the data for the establishments with the highest rates on these files are accurate in absolute terms. It would be a mistake to say these establishments are the "most dangerous" or "worst" establishments in the Nation.

Calendar year 2002 data for Minnesota, Michigan and North Carolina are not included in this file. These three states have informed the Department that disclosure of the rates of establishments surveyed in their states for 2002 would have adverse effects on state enforcement proceedings because of the way in which their state-specific enforcement
targeting lists are prepared and administered. Accordingly, establishment rates for these states are exempt from disclosure in accordance with Exemption 7(A) of the Freedom of Information Act and are not included in the data being provided to you. Establishment rates for 2002 for all other states that participate in the ODI have informed the Department that disclosure of the data for establishments in their states for 2002 would not have a similar adverse effect on enforcement proceedings and are, therefore, being provided to you.

All calendar year 2003 data is withheld from this file. The 2003 injury and illness rates are to be the basis for OSHA's upcoming Site-Specific Targeting program (SST05). Disclosure of the 2003 establishment specific rates would have an adverse effect on the agency's enforcement proceedings. Specifically, release of the establishment rates would be a partial release of the SST05 enforcement targeting list. Therefore, I have determined that the 2003 establishment rates are exempt from disclosure in accordance with Exemption 7(A) of the Freedom of Information Act.

front page | index + search
about | contact | blog | donate