Workplace Safety and Health: OSHA Could Improve Federal Agencies'
Safety Programs with a More Strategic Approach to Its Oversight  
(21-APR-06, GAO-06-379).					 
                                                                 
Federal workers' compensation costs exceeded $1.5 billion in	 
2004, with approximately 148,000 new claims filed that year.	 
Because of concerns for the safety of federal workers, as well as
the costs associated with unsafe workplaces, GAO described the	 
characteristics of federal agencies' safety programs and the	 
implementation challenges they face, and assessed how well the	 
Occupational Safety and Health Administration (OSHA) oversees and
assists federal agencies' efforts to develop and administer their
safety programs.						 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-379 					        
    ACCNO:   A52157						        
  TITLE:     Workplace Safety and Health: OSHA Could Improve Federal  
Agencies' Safety Programs with a More Strategic Approach to Its  
Oversight							 
     DATE:   04/21/2006 
  SUBJECT:   Occupational health and safety programs		 
	     Occupational safety				 
	     Program evaluation 				 
	     Program management 				 
	     Regulatory agencies				 
	     Safety regulation					 
	     Safety standards					 
	     Workers compensation				 
	     Federal employees					 
	     Federal agencies					 
	     Surveys						 
	     Program implementation				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-379

     

     * Results in Brief
     * Background
          * OSHA's Role
               * Enforcement
               * Compliance Assistance
               * Role of the Federal Agencies
          * Six Components of a Sound Safety Program
          * Federal Workplace Trends
          * Injury and Illness Trends in the Federal Government
     * Most Agencies Reported Having Many Safety Program Components
          * Most Agencies Reported Having at Least One Activity for Each
          * Common Implementation Challenges Include Data Management, Ac
               * Data Management
               * Accountability
               * Safety Resources
     * OSHA Provides Inadequate Oversight of Federal Agencies Becau
          * OSHA's Enforcement Strategy Does Not Include a Program for T
          * OSHA Lacks a System for Tracking Disputed Violations
          * OSHA Has Not Conducted Evaluations in More Than 6 Years
          * OSHA Has Not Submitted Timely Annual Reports on Agencies' Sa
          * OSHA Has Used a Variety of Compliance Assistance Programs fo
               * OSHA's Regional Offices Have Struggled to Set Up and Maintai
               * OSHA's Responses to Agency Technical Assistance Requests Are
               * Agencies Have Begun Joining OSHA's VPP
               * OSHA Has Developed Few Strategic Partnerships and Alliances
          * Impact of the SHARE Presidential Initiative Is Unclear
     * Conclusions
     * Recommendations for Executive Action
     * Agency Comments and Our Evaluation
     * GAO Comments
     * GAO Contact
     * Acknowledgments
     * GAO's Mission
     * Obtaining Copies of GAO Reports and Testimony
          * Order by Mail or Phone
     * To Report Fraud, Waste, and Abuse in Federal Programs
     * Congressional Relations
     * Public Affairs

                 United States Government Accountability Office

Report to the Subcommittee on Labor,

GAO

    Health and Human Services, Education, and Related Agencies, Committee on
                          Appropriations, U.S. Senate

April 2006

WORKPLACE SAFETY AND HEALTH

OSHA Could Improve Federal Agencies' Safety Programs with a More Strategic
                           Approach to Its Oversight

GAO-06-379

WORKPLACE SAFETY AND HEALTH

OSHA Could Improve Federal Agencies' Safety Programs with a More Strategic
Approach to Its Oversight

  What GAO Found

Based on a survey of 57 agencies, GAO found that most agencies reported
having at least one activity for each of the six components generally
associated with a sound safety program-(1) management commitment,

(2) employee involvement, (3) education and training, (4) identification
of hazards, (5) correction of hazards, and (6) medical management (which
includes having a return-to-work program for injured employees). However,
agencies faced implementation challenges that cut across the components in
the areas of data management, accountability, and safety resources. The
survey results indicated that many agencies do not have automated systems
for tracking elements of their safety programs, such as training. In
addition, several of the agencies did not demonstrate that their managers
are held accountable for maintaining effective safety programs. Finally,
many agency officials stated that, due to limited resources, they often
must depend on safety officers with limited professional safety
experience.

 Forest Service Smokejumpers Fighting a Blaze at Selway-Bitterroot Wilderness,
                                     Idaho

Source: Buck Nelson. OSHA's oversight of federal agencies' safety programs
is not as effective as it could be because the agency does not use its
enforcement and compliance assistance resources in a strategic manner.
Although inspections are one of OSHA's primary enforcement tools, it does
not conduct many inspections of federal worksites or have a national
strategy for targeting worksites with high injury and illness rates for
inspection. Furthermore, although OSHA is responsible for tracking
violations that agencies dispute and reporting any unresolved disputes to
the President, OSHA does not track these disputed violations or their
resolution. In addition, although OSHA is required to review agencies'
safety programs annually and submit a report on them to the President each
year, as of January 2006, the last report submitted was for fiscal year
2000. Finally, while OSHA has a range of compliance assistance programs
designed to help agencies comply with its regulations and improve safety,
these programs are not being fully utilized.

                 United States Government Accountability Office

Contents

Letter 1
Results in Brief 3
Background 5
Most Agencies Reported Having Many Safety Program Components, but Faced
Common Implementation Challenges 18
OSHA Provides Inadequate Oversight of Federal Agencies Because It Does Not
Use Its Enforcement and Compliance Assistance Resources Strategically 25
Conclusions 36
Recommendations for Executive Action 37
Agency Comments and Our Evaluation 38

Appendix I Scope and Methodology

Appendix II Comments from the Department of Labor

Appendix III Comments from the Department of Homeland Security

Appendix IV Comments from the Department of Interior

Appendix V Data Collection Instrument Sent to the Federal Agencies

Appendix VI Agencies' Responses to the Data Collection Instrument

Appendix VII GAO Contact and Staff Acknowledgments Tables

Table 1: Six Components of a Sound Safety Program and Their

Supporting Activities 12 Table 2: Top Five Types of Injuries Incurred by
Federal Workers,

Fiscal Years 1995 to 2004 16 Table 3: Federal Agencies' Strategic
Partnerships with OSHA 35 Table 4: Agency Responses to the Data Collection
Instrument by

Safety Program Component 59

Figures 
           Figure 1: Number of OSHA Inspections, Fiscal Years 2000 to 2004  7 
           Figure 2: OSHA's 10 Regions                                      8 
               Figure 3: Number of Workers' Compensation Claims of Federal 
                         Workers, Fiscal Years 1995 to 2004                15 
              Figure 4: Percent of Workers' Compensation Costs for Federal 
                     Workers by Age of the Case, Fiscal Years 1995 to 2004 17 
              Figure 5: Workers' Compensation Payments for Federal Workers 
                     by Type of Cost, Fiscal Years 1995 to 2004            18 
            Figure 6: Percent of Inspections of Federal Worksites by Type, 
                              Fiscal Years 2000 to 2004                    26 
              Figure 7: Average Number of Serious Violations at Federal    
                   Worksites by Inspection Type, Fiscal Years 1995 to 2004 27 

Abbreviations

OFAP                Office of Federal Agency Programs                      
OSHA                         Occupational Safety and Health Administration 
OWCP                Office of Workers' Compensation Programs               
OSH Act             Occupational Safety and Health Act                     
VPP                 Voluntary Protection Programs                          
SHARE               Safety Health and Return to Employment                 

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office Washington, DC 20548

April 21, 2006

The Honorable Arlen Specter Chairman The Honorable Tom Harkin Ranking
Minority Member Subcommittee on Labor, Health and Human Services,

Education, and Related Agencies Committee on Appropriations United States
Senate

In the early to mid 1990s, five Yellowstone National Park employees were
killed in on-the-job accidents ranging from a snowmobiling accident to a
drowning. By the time the last death occurred, the Occupational Safety and
Health Administration (OSHA)-the Department of Labor agency responsible
for overseeing federal agencies' safety programs-decided to conduct a
comprehensive inspection and identified over 500 violations at the park.
OSHA also worked with park officials to improve safety. As a result,
Yellowstone transformed its safety program and according to Yellowstone
officials, OSHA has deemed it a success. However, this worksite represents
only one of thousands of locations where federal employees work. During
the past decade, the number of executive branch workers killed or injured
in work-related accidents at federal worksites has fluctuated, although
the number has decreased overall. 1 Over this period, over 800 workers
died from work-related accidents, with 47 deaths occurring in 2004.
Non-fatal work-related injuries declined over the first half of this
period and increased over the subsequent 5-year period, with approximately
148,000 claims for injuries being filed in 2004. While the number of
injuries fluctuated, the costs of the claims for these injuries- adjusting
for inflation-remained constant for most of the decade and rose slightly
in 2004 to $1.5 billion. In view of these injuries and their affiliated
costs, you asked us to answer the following questions: (1) What are the
components of federal agencies' safety programs and what implementation
challenges do they face? (2) How well does OSHA oversee and assist federal
agencies' efforts to develop and administer their safety programs?

1

In this report, we use the term "injuries" to refer to workers' injuries
and illnesses.

To respond to your request, we focused on the executive branch because
OSHA has more oversight responsibilities with respect to these agencies
than it does for judicial and legislative branch agencies. 2 We obtained
information using a data collection instrument from 57 safety managers for
agencies in the eight largest executive branch departments, which
represent about 80 percent of the federal workforce. 3 The instrument
requested information and documentation on six components of sound safety
programs we identified from previous GAO work: (1) management commitment,
(2) employee involvement, (3) education and training, (4) identification
of hazards, (5) following up and correcting hazards, and (6) medical
management. 4 We reviewed the information in the data collection
instrument and documentation that the agencies provided, assessing whether
the documentation supported the agency's responses and identifying the
types of activities the agency conducted for each program component. In
addition, we conducted more detailed follow-up interviews with safety
officials from 12 of the 57 agencies surveyed. We also interviewed
employee representatives from 8 agencies and visited 5 federal worksites,
most of which were identified by OSHA as having strong safety programs. In
addition, we obtained information on workers' compensation claims from the
Department of Labor's Office of Workers' Compensation Programs (OWCP). We
also interviewed all 10 of OSHA's regional administrators and federal
agency program officers and officials

2

The Occupational Safety and Health Act requires all federal agencies,
including those in the executive, judicial, and legislative branches, to
develop and maintain safety programs. However, while Executive Order 12196
and OSHA's regulations provide that OSHA plays a key oversight role with
respect to executive branch agencies, its role is more limited for
judicial and legislative branch agencies. Specifically, the executive
order indicates that OSHA's role with respect to judicial and legislative
branch agencies is to cooperate and consult with them to aid them in
adopting their safety and health programs. Judicial and legislative branch
agencies are not subject to OSHA's regulations unless they have entered
into an agreement to that effect with OSHA. For the purpose of this work
we did not include contract employees on federal worksites. Contract
employees are not covered by federal agency safety programs, but are
instead covered by private-sector procedures under the act.

3

The eight largest departments are the Departments of Defense, Veterans
Affairs, Homeland Security, Treasury, Agriculture, Justice, Interior, and
the Social Security Administration. We excluded the U.S. Postal Service
because, under the Occupational Safety and Health Act, it is considered a
private-sector employer. For this report, "agency" refers to a division
within these federal departments. For example, the agencies we reviewed in
the U.S. Department of the Treasury include the Internal Revenue Service,
the Bureau of Engraving and Printing, and the U.S. Mint.

4

For this report, we use the term "safety program" to mean an agency's
occupational safety and health program.

                                Results in Brief

in 10 area offices including the area director, a compliance safety and
health officer, and, in many cases, a compliance assistance specialist. In
addition, we examined OSHA's inspection data from fiscal year 1995 to 2004
and reviewed, when available, the annual reports that the eight largest
federal departments provided to OSHA from 2000 to 2004. Finally, we
reviewed the reports on federal safety programs that OSHA provided to the
President during this time frame. For a more detailed explanation of our
methodology, see appendix I. We conducted our work between November 2004
and February 2006 in accordance with generally accepted government
auditing standards.

Most agencies reported having at least one activity for each of the six
safety program components. However, we identified common implementation
challenges that cut across the components in the areas of data management,
accountability, and safety resources. Of the 57 agencies surveyed, 54
reported that their safety programs contain at least one activity for each
of the six safety program components. Agencies reported the fewest
activities for the medical management component. For example, 12 agencies
reported they do not have a program to offer injured employees light or
restricted duty to help them return to work more quickly. Moreover, our
analysis of the survey data and interviews with agency officials revealed
a number of challenges agencies face in implementing their safety
programs. The survey results indicated that many agencies do not have
automated systems for tracking elements of their safety programs, such as
training, and agency officials told us that some of their systems are
difficult to use. In addition, many agencies did not demonstrate that
their managers are held accountable for maintaining effective safety
programs. For example, only 16 agencies (28 percent) were able to provide
copies of their performance appraisal review forms citing safety as a
rating element-the remaining agencies provided only their policies or
other written documentation. Agencies we interviewed also reported
difficulties in managing resources for their safety programs because many
do not have a line item for safety in their budgets or face production
goals that compete with safety priorities. In addition, many agencies told
us that, due to limited resources, they often must depend on safety
officers with limited technical or professional safety experience.

OSHA's oversight of federal agencies' safety programs is not as effective
as it could be because the agency does not use its enforcement and
compliance assistance resources in a strategic manner. One of OSHA's
primary enforcement tools is conducting inspections of federal worksites.
However, the agency does not conduct many inspections of federal worksites
or have a national strategy for targeting worksites with high injury and
illness rates for inspection. Instead, OSHA conducts inspections of
federal worksites mainly in response to complaints from employees. In
addition, although OSHA is responsible for tracking violations that
agencies dispute and reporting any unresolved disputes to the President,
it does not track these disputed violations or their resolution, and OSHA
regional officials said they can sometimes remain unresolved for years.
Evaluations, another enforcement tool OSHA has available, entail reviewing
agencies' safety policies and programs and assessing the overall
effectiveness of their safety programs. However, while OSHA is required to
evaluate certain federal agencies annually, it has not done any
evaluations in the last 6 years. Moreover, although OSHA is required to
review federal agencies' safety programs and submit a report on their
programs to the President each year, the last report OSHA submitted was
for fiscal year 2000. Finally, although OSHA has a range of compliance
assistance programs designed to help agencies comply with its regulations
and improve safety, not all of these programs are being fully utilized.
OSHA officials acknowledged these problems with their enforcement and
compliance assistance strategies but noted that they have relatively few
staff dedicated to federal agency oversight.

In order to improve OSHA's oversight of federal agencies' safety programs,
we are recommending that the Secretary of Labor direct the Assistant
Secretary for Occupational Safety and Health to (1) conduct targeted
inspections of federal worksites; (2) track violations disputed by federal
agencies and ensure that unresolved violations are reported to the
President; (3) conduct evaluations of federal agencies as required; and
(4) include in OSHA's annual report to the President an assessment of each
agency's safety program and recommendations for improvements. In
responding to a draft of this report, the Department of Labor generally
agreed with our findings and recommendations. We also received written
comments from the Departments of Homeland Security and Interior. These
agencies generally agreed with our findings and conclusions. Copies of
written comments from these agencies are provided in appendixes II, III,
and IV. In addition, we received technical clarifications from the
Departments of Defense, Justice, and Veterans Affairs, which we
incorporated as appropriate.

  Background

Congress passed the Occupational Safety and Health (OSH) Act in 1970 to
ensure safe and healthful working conditions for working men and women,
including federal employees. 5 While OSHA was created to administer the
OSH Act, the act also gave federal agencies primary responsibility for
providing federal employees with working conditions and workplaces that
are free from safety and health hazards. 6 The act authorizes OSHA to set
mandatory occupational safety and health standards, rules, and regulations
and to enforce their compliance. In turn, each federal agency is required
to establish and maintain a comprehensive and effective occupational
safety and health program that is consistent with OSHA's standards. 7

                                  OSHA's Role

OSHA's Office of Federal Agency Programs within its Directorate of
Enforcement Programs has primary responsibility for overseeing federal
agencies' safety programs. OSHA's regulations and an Executive Order
establish its responsibilities for monitoring federal agencies' programs.
8

OSHA uses two strategies to provide oversight of federal agencies' safety
programs-enforcement and compliance assistance. OSHA's enforcement
strategy includes inspections and evaluations of federal worksites that
help ensure that federal agencies are not violating any OSHA standards and
are complying with the requirements for their safety programs. 9 In

5

29 U.S.C. S:S: 651-678.

6

29 U.S.C. S: 668.

7

Executive Order 12196, issued on February 26, 1980, prescribes executive
branch agencies' and OSHA's responsibilities. 29 C.F.R. Part 1960 contains
OSHA's regulations for federal agency programs.

8

Executive Order 12196 and OSHA's regulations apply to federal executive
departments. Military personnel and uniquely military activities of
executive agencies are not included.

9

OSHA is generally authorized to conduct announced or unannounced
inspections of federal agencies that have not established OSH committees
that conform to the regulatory requirements. OSHA's inspection authority
is somewhat more limited if an agency has established an OSH committee.
Currently, only six departments have such committees: the Central
Intelligence Agency, Securities and Exchange Commission, General Services
Administration, Department of Labor, Tennessee Valley Authority, and U.S.
International Trade Commission. None of the departments we reviewed had
such committees. For those agencies with established OSH committees, OSHA
may only conduct an inspection if (1) half the membership of record of the
federal agency's OSH committee requests an OSHA inspection; or (2) an
employee reports an imminent danger to the agency's OSH committee, but
OSHA determines that neither the committee nor the agency has adequately
responded to the employee's complaint. 29 C.F.R. S:1960.31.

                                  Enforcement

addition, agencies are required to submit annual reports to OSHA on their
safety programs, which OSHA uses to prepare an annual report to the
President on federal agencies' safety programs. OSHA's compliance
assistance strategy consists of a range of programs intended to help
agencies improve their safety programs.

OSHA is authorized to conduct inspections of federal agency worksites but,
as figure 1 illustrates, inspections of federal worksites represent a very
small percentage of OSHA's overall inspections. 10

10

OSHA has enforcement responsibility for all federal worksites in all
states, but has granted authority to about half of the states for their
own enforcement of private-sector and non-federal, public-sector
worksites. At present, 22 states and territories, including Puerto Rico,
have been approved by OSHA to operate their own programs; 4 states and
territories, including the Virgin Islands, are approved for covering
non-federal, public sector employee worksites only. OSHA directly oversees
enforcement for all worksites in the remaining states. The number of
inspections shown includes those conducted by the states and territories.

Figure 1: Number of OSHA Inspections, Fiscal Years 2000 to 2004 Number of
inspections100,000 98,000 96,000 94,000 92,000 90,000

10,000 8,000 6,000 4,000 2,000

0

2000 2001 2002 2003 2004

      Fiscal year

Federal worksites Non-federal worksites Source: GAO analysis of OSHA data.

Between fiscal years 2000 and 2004, less than 1 percent of OSHA's
inspections were of federal worksites in executive branch agencies; the
remaining 99.5 percent were primarily of private-sector worksites. Federal
executive branch workers represented about 1.4 percent of the overall

U.S. workforce between 2002 and 2004.

Inspections are conducted by OSHA's 80 area offices in its 10 regions.
Figure 2 shows the location of OSHA's 10 regions.

Figure 2: OSHA's 10 Regions

AlaskaMaine

Vt.

N.H. Mass.

R.I.

Wash.

Conn.

Mont. N.Dak. Puerto Ore.

Idaho S.Dak.

Wyo.

Minn. N.Y. Rico

Wis. Mich. U.S.

N.J. Virgin Islands

Ill. Ind.

                            Nev. Utah Nebr. IowaOhio

Colo. Calif. Pa.

Kans. Mo.

Del. Md.

W.Va.

Ariz. Va. D.C.

Okla. N.Mex. Ark. Ky.

Guam Hawaii

N.C. Tenn.

Tex. La. S.C.

American SamoaMiss. Ala. Ga.

Fla.

Source: OSHA.

OSHA categorizes inspections as those that are "programmed" and those that
are "unprogrammed." Programmed inspections are those that OSHA plans to
conduct because it has targeted certain worksites for inspection due to
their potential hazards. Unprogrammed inspections are not planned; they
are prompted by actions such as complaints, accidents, and referrals. OSHA
has established a system of inspection priorities that relate to these
categories, with unprogrammed inspections being a higher priority than
programmed inspections. Top priority goes to imminent danger situations in
which death or serious physical harm could occur.

                             Compliance Assistance

The next priority for OSHA inspectors is catastrophes and fatal accidents,
followed by complaints and referrals. Programmed inspections are OSHA's
fourth priority. OSHA's last priority is to perform follow-up inspections,
which are conducted to ensure that hazards identified during previous
inspections have been corrected. From fiscal years 2000 through 2004, only
40 percent of OSHA's inspections of federal worksites were programmed.
During the same period, 54 percent of its inspections of nonfederal
worksites were programmed.

OSHA is required to conduct comprehensive annual evaluations of the larger
or more hazardous federal agencies. 11 Results of these evaluations are
summarized by OSHA in reports that include information from the review of
an agency's safety policies and reports, as well as inspections of the
agency's facilities and interviews with agency personnel.

In addition, OSHA is required to submit to the President an annual report
on the status of federal employees' occupational safety and health. 12
OSHA uses reports submitted annually by federal agencies to OSHA on their
safety programs-along with the results of any evaluations it has conducted
of federal agencies' safety programs-to prepare its annual report to the
President. The report should also contain recommendations for improving
agencies' performance.

OSHA's compliance assistance strategy consists of several programs
available to federal agencies, although some programs have only recently
been offered to federal employers. OSHA also provides technical support to
federal agencies, such as conducting studies of accidents and the causes
of injuries and illnesses, and providing training of agencies' safety and
health personnel. Two of OSHA's compliance assistance programs-Field
Federal Safety and Health Councils and Agency Technical Assistance
Requests-specifically target federal agencies, while others are generally
available to both private- and public-sector employers. Compliance
assistance programs for private and public sector employers include the
Voluntary Protection Programs (VPP), alliances, and strategic

11

Executive Order 12196, S:1-401(h).

12

29 C.F.R. S:1960.71(b).

                          Role of the Federal Agencies

partnerships. 13 Approval as a VPP site is OSHA's official recognition of
worksites that have implemented exemplary safety and health programs. 14
The VPP was started in 1982 for private-sector companies but was expanded
to include federal agencies in 1997. The alliance program was started in
2002 and includes organizations that have agreements with OSHA to focus on
training, outreach, and promoting awareness of safety and health issues.
The strategic partnership program was started in 1998 and consists of
agreements between OSHA and employers to address specific safety and
health problems.

OSHA also has been responsible for helping implement Presidential
initiatives, with the most recent initiative being issued in 2004: the
Safety and Health and Return to Employment (SHARE) initiative. This
initiative directs agencies to set and adhere to both safety and workers'
compensation goals. Specifically, the initiative directs federal agencies
to achieve four goals: (1) reduce the overall case rate for these claims,
(2) reduce the lost-time rate-the number of employees who could not return
to work per 100 employees in the workforce, (3) improve the processing
time of workers' compensation claims, and (4) reduce the lost production
day rate-the lost days due to injury or illness per 100 employees. OSHA
works with agencies in addressing the first two goals and helps them
calculate the rates monitored.

OSHA's regulations establish the basic elements of executive agencies'
safety and health programs. According to the regulations, agencies'
programs must include provisions for

     o top management support, participation, and accountability;
     o safety and health policies, procedures, and standards;
     o goals and objectives;
     o worker involvement;

13

OSHA's compliance assistance programs use a mix of different methods
designed to improve worker safety. They target both exemplary worksites
and hazardous ones, and influence employers directly by implementing
safety and health programs and indirectly through collaboration with trade
and professional associations. For more information on these programs, see
GAO, Workplace Safety and Health: OSHA's Voluntary Compliance Strategies
Show Promising Results, but Should Be Fully Evaluated before They Are
Expanded, GAO-04-378 ( Washington, D.C.: Mar. 19, 2004).

14

VPP sets performance-based criteria for a managed safety and health
system, invites sites to apply, and then assesses applicants against these
criteria. OSHA's verification process includes an application review and
rigorous onsite evaluations by a team of safety and health experts.

     o safety and health training of managers and workers;
     o collection of occupational injury and illness data;
     o self-inspection of workplaces and self-evaluation of the programs;
     o abatement of unsafe and unhealthful working conditions; and
     o adequate budgets, staff, and equipment and materials.

In conducting self-inspections, agencies must meet certain requirements.
Inspectors are required to be qualified to recognize and evaluate hazards
and suggest corrections, and they must conduct inspections of every
worksite at least once a year. According to the regulations, agencies
should conduct "sufficient" unannounced inspections and unannounced
follow-up inspections to ensure the identification and correction of
hazardous conditions.

Agencies must also report annually to OSHA on their programs. In November
2004, OSHA issued a final rule amending the injury recordkeeping and
reporting requirements applicable to federal agencies. 15 Prior to this
time, federal agencies were required to collect only injury information
related to workers' compensation claims. OSHA revised the recordkeeping
requirements in order to improve the quality of the federal recordkeeping
system and to increase the utility of the data. 16 Beginning in January
2005, federal agencies were required to record injuries in the same manner
as private-sector employers and to apply new criteria to determine whether
an injury must be recorded. Specifically, a work-related injury must be
reported if, for example, it results in death, 1 or more days away from
work, restricted work, loss of consciousness, or a significant injury or
illness diagnosed by a physician. 17 The regulations do not require that
this data be reported to OSHA. However, the regulations state that agency
heads must submit an annual report to OSHA, containing such information as
OSHA requests. 18 At a minimum, these reports are to describe the agency's
safety program and include, among other things, the agency's required
self-evaluation findings. OSHA uses these reports, along with any
evaluations it has conducted to prepare its annual report to the
President.

15

69 Fed. Reg. 68793, codified at 29 C.F.R. S:S:1960.66-1960.71.

16

69 Fed. Reg. 68796.

17

In addition, unlike prior requirements, agencies must record the annual
average number of employees employed as well as the total hours worked by
all employees. This information is needed to calculate injury and illness
rates.

18

29 C.F.R. S:1960.71(a). The regulations also provide that the Secretary of
Labor must provide the agencies with the guidelines and format for the
reports.

    Six Components of a Sound Safety Program

Safety experts and federal safety agencies agree that, to build an
effective safety program, organizations must take a strategic approach to
managing workplace safety and health. This objective is generally
accomplished by establishing programs built upon a set of commonly
recognized components of sound safety programs, which, together, help an
organization lay out what it is trying to achieve, assess progress, and
ensure that safety policies and procedures are appropriate and effective.
Drawing from our prior work, a review of the literature, and OSHA's
requirements, we identified six components often found in sound safety
programs: (1) management commitment, (2) employee involvement, (3)
education and training, (4) identification of hazards, (5) following up
and correcting hazards, and (6) medical management. 19 Table 1 lists these
components, along with a description of their supporting activities.

     Table 1: Six Components of a Sound Safety Program and Their Supporting
                                   Activities

Componenta               Supporting activities                             
Management commitment o  Establish goals for the program.                  
                         o  Develop activities to communicate the importance  
                            of the safety program to staff, including         
                            management, employees, and contractors.           
                         o  Use management and information systems that allow 
                            for trend analysis, risk analysis,                
                            etc.                                              
                         o  Establish program responsibilities of managers    
                            and employees for safety and health in the        
                            workplace and hold them accountable for carrying  
                            out those responsibilities.                       
Employee involvement  o  Establish procedures for employees to report      
                            job-related fatalities, injuries, illnesses,      
                            incidents, and damage to property or equipment.   
                         o  Establish procedures for employees to report      
                            hazards.                                          
                         o  Provide employee access to the system that        
                            captures information on accidents and hazards.    
                         o  Ensure employee involvement in safety committees  
                            that report on hazards.                           
                         o  Allow employees to provide input on               
                            safety-related training curricula.                
                         o  Ensure employees participate in walkthroughs of   
                            worksites to identify hazardous conditions on a   
                            daily, weekly, or monthly basis.                  
                         o  Allow employee involvement on accident            
                            investigation teams.                              

19See GAO, Architect of the Capitol: Management and Accountability
Framework Needed for Organizational Transformation, GAO-03-231
(Washington, D.C.: Jan. 2003) and Private Sector Ergonomics Programs Yield
Positive Results, GAO/HEHS-97-163 (Washington, D.C.: Aug. 27, 1997).

Page 12 GAO-06-379 Worker Safety

                        Componenta Supporting activities

Education and training

     o Provide general awareness training to all employees so that they can
       recognize hazards and risks; learn procedures for reporting
       job-related fatalities, injuries, illnesses, incidents, and hazards;
       and become familiar with the program (nationallevel training).
     o Provide targeted training to specified groups of employees because of
       the jobs they hold, the hazards they face, or their role in the
       program (agency-level training).
     o Maintain an automated system to track training completed by employees.

Identification of hazards  o  Establish procedures for conducting required
OSHA inspections.

o  Establish procedures for conducting informal walkthroughs of worksites
to identify hazardous conditions.

Following up and correcting hazards  o  Establish procedures for
developing controls for workplace hazards.

     o Establish procedures for following up on inspections to ensure hazards
       are corrected and controls are effective.
     o Maintain an automated system that tracks workplace hazards.

Medical managementb  o  Establish procedures to ensure that an injured or
ill employee is seen within a specified time frame by a medical provider.

     o Maintain an automated system that tracks accident data-including the
       type, nature, and source of injury.
     o Implement a restricted or light duty return-to-work program.
     o Maintain an automated system that tracks the return-to-work status of
       employees.

                            Federal Workplace Trends

Source: GAO and OSHA.

a

Different terminology is often used to describe these components. For
example, "identification of problem jobs" is sometimes referred to as
"hazard identification and assessment" and "analysis and development of
controls for problem jobs" is sometimes referred to as "hazard prevention
and control."

b

Organizations may have a medical management program without necessarily
having a safety and health program.

Over the last 10 years, the federal executive branch workforce has changed
in a number of ways, including its size, demographic characteristics,
experience levels, and types of occupations. During this time, there was a
6 percent decrease in the federal workforce-from 2 million employees in
fiscal year 1995 to 1.9 million in fiscal year 2004. In addition, the
average age of federal workers increased from 44 to 47 years old and the
average length of time in service increased slightly, from 16 to 17 years.
Likewise, the average pay grade level of federal workers increased from
approximately GS-9 to about GS-10. Moreover, the percentage of workers in
professional and administrative positions increased from 85 to 89 percent.

Federal employees encompass a wide range of professions, ranging from
low-risk occupations such as office workers to highly hazardous
occupations such as law enforcement positions. For example, at the U.S.

    Injury and Illness Trends in the Federal Government

Marshals Service, duties of criminal investigators include seizing assets
and apprehending fugitives. In addition, U.S. Forest Service employees are
involved in a variety of potentially hazardous activities such as
developing laboratory products, managing recreational lands, and fighting
wildland fires, while inspectors with the Food Safety Inspection Service
face daily hazards such as exposure to the chemicals used to kill
pathogens in meat. Finally, employees at manufacturing operations such as
at the U.S. Mint and the Bureau of Engraving and Printing use industrial
production equipment such as forklifts and presses.

The impact of demographic changes in the makeup of the federal workforce
on the number of injuries they sustain is unclear. The number of active
workers' compensation claims for work-related injuries declined from
approximately 154,000 claims in fiscal year 1995 to about 137,000 claims
in fiscal year 1999. 20 However, these claims increased from approximately
138,000 claims in fiscal year 2000 to about 148,000 claims in 2004, as
shown in figure 3.

20

For this report, we define "active workers' compensation claims" as claims
made by federal employees who sustained compensable work-related injuries
or illnesses. Workers' compensation benefits provided to covered employees
can include payment for medical treatment, rehabilitation services, death
benefits, and replacement of lost wages.

Page 14 GAO-06-379 Worker Safety

Figure3: Number of Workers' Compensation Claims of Federal Workers, Fiscal
Years 1995 to 2004

      Number of claims

175,000

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fiscal year

Source: GAO analysis of OWCP data.

Although the severity of the injuries changed during this period, the
types of injuries that federal workers incurred remained the same. Despite
the fact that the number of traumatic injury claims decreased
slightly-from 76,633 claims in fiscal year 1995 to 74,322 claims in fiscal
year 2004, as a proportion of total claims, traumatic injury claims
increased slightly over this same period. However, non-traumatic injury
claims decreased by over 30 percent during this period-from 8,508 claims
in fiscal year 1995 to 5,903 claims in 2004. In addition, the top five
types of traumatic injuries incurred by federal workers during this period
ranged from sprains and strains of ligaments, muscles, or tendons to
lacerations. Over this same period, the five most common types of
non-traumatic injuries ranged from hearing loss to back sprain or strain.
See table 2 for a list of the five most common types of traumatic and
non-traumatic injuries federal workers incurred from fiscal year 1995 to
2004.

Table 2: Top Five Types of Injuries Incurred by Federal Workers, Fiscal
Years 1995 to 2004

                   Traumatic injuries Non-traumatic injuries

                   1. Sprain/strain-not back 1. Hearing loss

           2.Traumatic injury-unclassified 2. Carpal tunnel syndrome

1. Back sprain or strain 3. Musculoskeletal condition, other
2. Contusion 4. Conditions of tendons, etc.
3. Laceration 5. Back sprain or strain

Source: GAO analysis of OWCP data.

While the size of the workforce declined, workers' compensation costs for
federal employees remained fairly constant during the most recent 10-year
period, from about $1.54 billion in fiscal year 1995 to about $1.52
billion in 2004. 21 In addition, the compensation per claim filed during
this period increased. For example, while there were about 1,800 fewer new
claims in fiscal year 2004 than there were in fiscal year 1995, the
average compensation per claim increased by 3 percent from fiscal year
1995 to 2004, with the average payment per claim rising from $9,958 in
1995 to $10,242 in 2004. As shown in figure 4, the largest amount of
workers' compensation costs for federal workers paid from fiscal years
1995 to 2004 was for claims that were over 5 years old.

Amounts shown have been adjusted for inflation, with a base year of 2005.

Figure 4: Percent of Workers' Compensation Costs for Federal Workers by
Age of the Case, Fiscal Years 1995 to 2004

        Percentage

100

90

80

70

60

50

40

30

20

10

        0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fiscal year

     64      66      68      68      69      68     67     66     66     65   
     23      21      20      20      20      19     19     21     20     21   
     13      13      12      12      12      13     14     14     14     15   

Over 5 years 1 to 5 years 0 to 1 year Source: GAO analysis of OWCP data.

Finally, the proportion of payments for lost wages, death benefits,
medical costs, and rehabilitation have remained constant, with wage loss
compensation being the largest proportion (approximately 70 percent) of
workers' compensation payments made from fiscal years 1995 to 2004. (See
fig. 5.)

  Most Agencies Reported Having Many Safety Program Components, but Faced Common
  Implementation Challenges

Figure 5: Workers' Compensation Payments for Federal Workers by Type of
Cost, Fiscal Years 1995 to 2004

Dollars1,650,000,000

          1,500,000,000 1,350,000,000 1,200,000,000 1,050,000,000 900,000,000
                  750,000,000 600,000,000 450,000,000 300,000,000 150,000,000

0

    1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fiscal year

Death benefits Medical and rehabilitation benefits Wage loss compensation
Source: GAO analysis of OWCP data.

Information reported by the 57 federal agencies illustrated various ways
in which agencies carry out activities within the six safety program
components-management commitment, employee involvement, training,
identification of hazards, correction of hazards, and medical management.
However, agency officials we surveyed and interviewed reported they face a
number of implementation challenges that cut across the components,
particularly in using automated systems, holding managers accountable for
maintaining an effective safety program, and making the best use of their
limited resources. Officials at these agencies also described measures
they have taken to overcome each of these challenges.

Most Agencies Reported All of the 57 agencies surveyed reported that their
safety programs Having at Least One incorporate activities for the
management commitment component. Activity for Each Safety Activities
supporting management commitment include setting goals for Program
Component the program and communicating from upper management to frontline
staff

Page 18 GAO-06-379 Worker Safety

about the importance of the safety program. Fifty-five of the agencies
surveyed (96 percent) reported that they had established goals for their
safety and health programs, and all 57 agencies reported conducting
activities to communicate the importance of their safety programs to
employees, such as through newsletters and Web sites.

Almost all of the agencies we surveyed reported that they conduct
activities for two other components-employee involvement and training.
Most agencies reported having policies governing employees' participation
in safety committees and reporting injuries and hazards. While 56 (98
percent) of the agencies provided procedures for employees to report
hazards, half of these procedures did not specify the right of employees
to report hazards anonymously, as required by an executive order. 22
Consistent with OSHA regulations, 56 of the 57 agencies reported that they
offer some type of safety training for their employees. 23

While many agencies identified a number of methods for identifying
hazards, fewer had comprehensive procedures for tracking whether hazards
are corrected-two additional components of safety programs. Fifty-five (96
percent) of the agencies reported that they conduct OSHArequired
inspections, which must be performed at least once a year, in order to
identify worksite hazards. However, although an executive order requires
employee representatives to participate in these inspections, seven
agencies (12 percent) reported not having any procedures for informing
employees of their role during safety inspections. 24 Furthermore, while
most agencies reported having some procedures for following up on
inspections and ensuring that hazards are corrected, we found that the
procedures are not always adequate because a third of these agencies did
not specify a reasonable timeframe for correction, as required by OSHA. 25

Agencies reported having the fewest activities for the medical management
component. Eight agencies (14 percent) reported that they

22

Executive Order 12196, S:1-201(h).

23

OSHA regulations require agencies to provide appropriate safety and health
training that must, among other things, inform employees of the agency's
safety program and their rights and responsibilities under the program. 29
C.F.R. S:1960.59.

24

Executive Order 12196, S: 1-201(i).

25

OSHA regulations require agencies to promptly rectify unsafe and
unhealthful conditions, and agencies are required to comply with this
regulation by documenting the seriousness of identified hazards and
providing a reasonable time for correcting them. 29 C.F.R S:1960.26.

    Common Implementation Challenges Include Data Management, Accountability,
    and Safety Resources

Data Management

do not have any procedures designed to ensure that an injured employee is
seen promptly by a physician. In addition, 12 agencies (21 percent)
reported they do not have programs for offering injured employees light or
restricted duty to help them return to work more quickly. Another 11
agencies reported having such programs but did not provide sufficient
documentation of them. For example, two agencies reported having
return-to-work programs, but the documentation they provided showed that
the programs had not yet been implemented. Although federal agencies are
not legally required to include these activities in their safety programs,
the failure to include them may limit the effectiveness of the programs.

We found that agencies face some common challenges in implementing their
safety programs, particularly in using automated systems to manage their
programs, holding managers accountable for workplace safety, and operating
with limited resources.

The use of automated systems presented challenges for many agencies. Some
agencies did not use such systems, while others cited difficulties in
identifying systems that would allow them to collect data relevant to
their safety programs. For example, 23 agencies (40 percent) reported that
they do not have automated systems to collect information on hazards that
have been identified and track whether they have been corrected in a
timely manner. In addition, 14 of the agencies reporting that they have
such a system (41 percent) either indicated that their hazard tracking
systems were not currently operational, or they did not provide sufficient
documentation to support the existence of such systems. Approximately a
quarter of the agencies surveyed reported that they did not have automated
systems for tracking safety training completed by their employees.
Furthermore, 34 agencies (60 percent) reported they did not have an
automated system for tracking the status of employees in light or
restricted duty return-to-work programs, and another 16 agencies did not
provide sufficient documentation of their systems. While federal agencies
are not required to use automated data systems, without such a system,
safety officials would have difficulty tracking broader trends such as
participation rates in light or restricted duty programs and the effect of
program participation on worker's compensation costs.

                                 Accountability

Ten of the 12 agencies we reviewed in more detail reported challenges with
their automated systems, such as ensuring that these systems collected
appropriate data needed to evaluate the effectiveness of their safety
programs. For example, an official from the Tobyhanna Army Depot told us
that their computer technicians were in the process of designing a hazard
tracking program because no agencywide programs were available, and
off-the-shelf programs required too much adaptation to be practical. She
also developed a stand-alone spreadsheet to track all work-related
injuries because the systems available did not capture injuries that were
not recordable on the OSHA log (such as injuries requiring only first aid)
or injuries for which workers' compensation claims are not filed.
Furthermore, a National Park Service official stated that entering safety
meetings and other non-traditional training methods into the agency's
automated system is difficult because the system does not have data fields
for recording these activities. As a result, the agency has difficulty
determining the extent to which employees have been trained on many safety
issues.

Despite these challenges, several agencies told us they have started or
are in the process of implementing automated safety systems that will
allow them to collect and analyze data in order to better manage these
safety programs, including assessing the effectiveness of their programs.
For example, according to a Transportation Security Administration
official, the agency is developing a new injury tracking system that will
link injury and illness data with inspection data, allowing them to
identify trends, such as where injuries commonly occur and demographic
characteristics of injured employees. Similarly, officials with the Bureau
of Engraving and Printing said they are testing a medical management
system that will aggregate data from a number of different sources
including the health unit, safety investigation reports, and their
workers' compensation system. Collecting these data will allow them to
streamline the reporting process and better track injury trends.

Another challenge agencies face is holding managers accountable for
implementing effective safety programs. While 51 agencies (89 percent)
reported having policies that establish responsibility for workers' safety
and health for all employees, 6 reported that they do not have such
policies, despite an OSHA regulation requiring them to establish these
policies for all management officials. 26 Of the 51 agencies reporting
having

26

29 C.F.R. S:1960.11.

such policies, 11 agencies did not provide sufficient documentation of the
policies. For example, one agency provided an Employee Performance Plan,
but there were no performance expectations related to safety anywhere in
the plan. 27 Furthermore, although we asked the agencies to provide copies
of their performance appraisal review forms citing safety as a rating
element, only 16 agencies (28 percent) were able to do so.

Agency officials and employee representatives at 7 of the 12 agencies
selected for follow-up interviews cited further difficulties in
maintaining accountability throughout all levels of their organizations.
For example, a Veterans Health Administration employee representative
reported that, while there is a high level of commitment to safety at the
headquarters level, the message is diluted as it reaches lower levels of
the agency. In another example, the Defense Commissary Agency implemented
a program that requires regional safety managers to evaluate stores'
safety programs. Agency officials stated that regional officials are
expected to follow up to ensure that stores make timely corrections, but
are not required to document when hazards are corrected. As a result, the
agency has little assurance that the safety of store employees is
adequately protected. In addition, according to an employee representative
from the Commissary, it is not always clear who is responsible for
ensuring that hazards are corrected.

Several agencies reported that they had developed ways to help ensure that
employees and managers are held accountable for agency safety programs.
For example, in order to address accountability issues within the agency,
the Veterans Health Administration initiated a program that ties agency
safety goals to performance ratings. Moreover, instead of simply including
safety as a general element of performance review, the agency selects two
to three specific safety program goals that change every few years
according to agency needs. Past goals have included submitting workers'
compensation claims on time and reducing the occurrence of needle stick
injuries. According to agency officials, bonuses for executive staff
members are provided based on their progress in meeting these goals. When
significant improvement has been made in

We have reported that aligning individual performance expectations with
organizational goals, such as workers' safety and health, can help hold
individuals accountable for contributing to organizational results. GAO,
Results-Oriented Cultures: Creating a Clear Linkage between Individual
Performance and Organizational Success, GAO-03-488 (Washington, D.C.: Mar.
14, 2003).

Page 22 GAO-06-379 Worker Safety

Safety Resources

these areas, safety officials set new goals-enabling continuous
improvement.

Both agency and OSHA officials cited challenges in funding their safety
programs, although OSHA regulations require agencies to provide adequate
resources to implement and maintain these programs. 28 One agency official
we interviewed reported difficulty identifying funding for the agency's
safety program because safety funding is not specifically designated as a
line item in its budget. This lack of information on available resources
makes it particularly difficult to plan for long-term safety issues, such
as developing and providing training. For example, officials with the
National Park Service, which employs a large cadre of seasonal workers,
reported that the lack of itemized safety funds within its budget makes it
hard to develop their training plans. OSHA officials cited the federal
budget process itself as problematic because it requires federal agencies
to budget months in advance for safety-related equipment purchases or
other safety devices, long before they may have identified the need for
this equipment. This was corroborated by a U.S. Mint official who reported
that it is difficult to correct hazards that require a lot of capital
investment and planning.

In addition, a potential consequence of operating with limited resources
is the use of collateral duty safety officers-employees whose primary
responsibilities do not involve safety. Nearly all of the agency officials
we interviewed reported relying on these positions, which are typically
filled by employees who volunteer or are assigned by the agency. While
some agency officials reported their collateral duty officers were
appropriately trained, as required by OSHA, others reported that these
officers have limited knowledge or experience in safety. 29 Some agency
officials we interviewed said that this lack of experience, as well as the
limited amount of time collateral duty officers are allotted for safety
duties, has made it difficult for these officers to learn all of the
safety program requirements. For example, according to a National Park
Service official, collateral duty officers at this agency typically spend
about 10 percent of their time on their safety responsibilities, and this
may inhibit their ability to respond effectively when safety concerns
arise. Moreover, one Forest Service official told us that collateral duty
officers questioned the feasibility of

28

29 C.F.R. S:1960.7.

29

29 C.F.R. S:1960.58.

building safety programs with collateral duty officers, and was concerned
that the safety duties might detract from their primary job
responsibilities.

Finally, 11 of the 12 agencies selected for follow-up interviews reported
that competing priorities make it difficult to manage their safety
programs. For example, a Food Safety Inspection Service official noted
that completing safety forms and fulfilling data requests can be a burden
to the agency's overall mission of meat and poultry inspections.
Similarly, an employee representative at the Forest Service told us that,
because safety achievements are not typically recognized or rewarded-even
though such recognition is encouraged by OSHA regulations-supervisors
focus on meeting production targets rather than working safely. 30

Agencies identified a number of techniques for addressing the difficulties
associated with managing resources. For example, an official with a
National Park Service regional office said that they host monthly
conference calls with the collateral duty safety officers at several
national parks, which gives these individuals a chance to ask technical
questions of the regional safety officer and share effective practices
among the parks. These monthly calls also enable their collateral duty
officers, who have limited backgrounds in safety, to gain knowledge and
experience over time. Other agencies maximized their resources by
collaborating with each other. For example, one official with the Forest
Service said that they have an informal partnership with the Bureau of
Land Management that allows them to pool their resources by pursuing joint
activities and sharing offices and staff. One activity involved jointly
developing and teaching an accident investigation course and an
off-highway vehicle course.

29 C.F.R. S:1960.11.

  OSHA Provides Inadequate Oversight of Federal Agencies Because It Does Not Use
  Its Enforcement and Compliance Assistance Resources Strategically

OSHA's oversight of federal agencies' safety programs is not as effective
as it could be because it does not use its enforcement and compliance
assistance resources in a strategic manner. First, OSHA does not routinely
conduct inspections that target federal worksites with high injury and
illness rates. In addition, OSHA lacks procedures for tracking and
resolving violations disputed by federal agencies. Third, OSHA has not
conducted required evaluations of the larger or more hazardous agencies in
the last 6 years. Fourth, OSHA has not submitted its own annual reports to
the President in a timely manner, and they have not included an assessment
of each agency's safety program, as required. Finally, while OSHA has a
range of promising programs for assisting agencies in complying with its
regulations and improving worker safety, not all of these programs are
being fully utilized.

    OSHA's Enforcement Strategy Does Not Include a Program for Targeting
    Worksites with High Injury Rates for Inspection

Unlike its enforcement strategy for private-sector employers, OSHA's
oversight of federal worksites does not include a national program that
targets federal worksites with high injury and illness rates for
inspection. According to its internal guidance, OSHA is supposed to
develop a list that targets federal worksites for inspection. However,
OSHA's Office of Federal Agency Programs has not developed such a list in
over 5 years. In the past, OSHA used workers' compensation claims data
collected by OWCP to identify federal worksites with high numbers of
injuries and illnesses. Because of limitations in the data, however, it
was difficult to identify where each injury occurred and, therefore, use
this information to target federal worksites for inspection. 31 OSHA
officials at the national office reported that they are working to start a
new targeting effort but are still facing the same difficulties in using
workers' compensation data to select federal worksites for inspection.

31

According to OSHA officials, the database does not include a code for
worksite location, and the zip code recorded often reflects where the
claim was processed instead of where the injury occurred. In addition,
OSHA officials said that they had to rely on the numbers of injuries
rather than injury rates because federal agencies were unable to provide a
list of the number of employees who worked at each worksite-information
needed to calculate these rates.

Page 25 GAO-06-379 Worker Safety

As shown in figure 6, OSHA primarily conducts inspections of federal
worksites as a result of complaints. 32

Figure 6: Percent of Inspections of Federal Worksites by Type, Fiscal
Years 2000 to 2004

Area Offices without Targeted ProgramsArea Offices with Targeted
Programs(Total number of inspections =1,658) (Total number of inspections
= 826)

Source: GAO analysis of OSHA data.

OSHA's inspection data of federal worksites show that complaint
inspections generally result in few violations compared to targeted
inspections, which generally identify a greater number of serious
violations (see fig. 7). 33 For example, over the last 10 years,
unprogrammed inspections, which are generally initiated by complaints,
uncovered an average of one serious violation per inspection, in contrast
to an average of four serious violations for programmed (targeted)
inspections. The

32

OSHA regulations require that federal agencies have provisions for
responding to employee reports of unsafe or unhealthy working conditions.
However, employees may also report hazardous conditions directly to OSHA.
If OSHA receives a complaint from an employee who works for a federal
agency that lacks an established OSH committee, it may initiate an
inspection or other appropriate action. 29 C.F.R. S:1960.28(e).

33

According to OSHA's internal guidance, a serious violation is defined as
one in which there is substantial probability that death or serious
physical harm could result, and the employer knew or should have known of
the hazard.

small average number of violations for unprogrammed inspections is driven
by the fact that over half of these inspections result in no violations
being identified.

Figure 7: Average Number of Serious Violations at Federal Worksites by
Inspection Type, Fiscal Years 1995 to 2004

      Number of serious violations

9

8

7

6

5

4

3

2

1

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

      Fiscal year

Unprogrammed inspections

Programmed inspectionsSource: GAO analysis of OSHA data.

The new recordkeeping rule, which was implemented in January 2005,
requires federal agencies to begin collecting the same injury and illness
data as private-sector employers and could help OSHA develop its targeting
program, according to OSHA officials. Since the new rule requires federal
worksites to keep logs that include information that can be used to
calculate injury and illness rates, OSHA officials said these data would
be more useful in creating an effective targeting program than the
workers' compensation data. While the new rule does not require federal
agencies to report injury and illness data to OSHA, OSHA officials said
they could target federal worksites for inspection in the same way they
target private-sector employers in industries with high injury and illness
rates for inspection. For its targeting program of private worksites, OSHA
surveys a sample of worksites in industries with the highest injury and
illness rates. The survey form requires employers to report on (1) the
average number of employees who worked for them during the previous
calendar year, (2) the total hours the employees worked during the
previous year, and (3) summary injury and illness data from their OSHA
logs. OSHA then uses this information to compute the worksites' injury and
illness rates and sends those with relatively high rates a letter
informing them that they may be inspected. Finally OSHA develops a list of
worksites with high injury and illness rates to be targeted for
inspection. As an alternative to conducting a survey of federal worksites,
OSHA has the option of requiring federal agencies to report this
information in their annual reports to OSHA.

One of OSHA's regional offices-which includes four area offices-and an
area office in another region developed their own targeted programs of
federal agency worksites using the workers' compensation data. While
officials reported using the data has been difficult, they said that these
efforts have resulted in improved safety at federal worksites. In
addition, they reported that the agencies that were inspected have become
more aware of OSHA's role and, in turn, have sought OSHA's assistance in
improving their safety programs. Furthermore, agency officials whose
worksites have been selected for inspection have focused more attention on
safety and shared information, resulting in further improvements. For
example, at one worksite in Montana, Forest Service officials reported
that, after colleagues in Idaho told them OSHA had targeted federal
worksites in the state for inspection, they were reviewing their safety
programs and OSHA's requirements in preparation for possible OSHA visits.

Officials with OSHA's national office said that they have encouraged
regions to develop their own programs targeting federal agencies for
inspection, but we identified some challenges that need to be addressed
before more regions can successfully develop these programs. For example,
one regional OSHA official reported requesting workers' compensation data
from the national office to start a targeting program, but was told the
national OSHA office did not have enough time to provide the data
requested. In addition, regional and area office OSHA officials said that
the ability to develop and maintain targeting programs depends on the
resources available. Besides the time and effort required to identify
worksites, they said the availability of inspectors is also a factor.
According to OSHA's policies, OSHA inspectors' top priority is responding
to imminent danger situations, followed by accidents, followed by
responding to complaints; conducting targeted inspections is a lower
priority.

    OSHA Lacks a System for Tracking Disputed Violations

OSHA's procedures for tracking violations disputed by federal agencies
differ from those for the private sector. Whereas private-sector employers
can dispute OSHA violations cited during inspections by requesting that
the violations be reviewed by an independent administrative law judge,
federal agencies must seek resolution with OSHA officials. In these
situations, federal agencies may first request an informal conference with
OSHA area office officials to discuss the violation in question. If the
dispute is not resolved, it is referred to the relevant OSHA region for
review and, if necessary, to OSHA's national office.

While OSHA's internal instructions require that area office and regional
officials be consulted in decisions made by national office officials and
an Executive Order requires OSHA to submit unresolved violation disputes
to the President, neither of these things appears to be occurring. 34
Although national office officials reported that there have not been any
unresolved disputed violations, and they have not had to report any
unresolved violations to the President in over 3 years, area office and
regional staff told us some unresolved disputed violations from federal
agencies have lingered for years. For example, a regional OSHA official
reported that, in another region, a federal agency was cited for violating
a safety standard that did not apply to that particular agency. 35 The
agency challenged the violation, and the dispute reached the national
office, where no decision was made-leaving the violation unresolved for 7
years. Another OSHA official reported a case in which the Bureau of
Prisons refused to have guards wear special gloves as required while
conducting cell searches because the guards thought the gloves would not
provide them with enough sensitivity to feel for objects hidden by
prisoners. According to this official, it was important for the guards to
wear gloves during these searches because of the danger of receiving
needle sticks or cuts from sharp objects. The case reached OSHA's national
office, but it chose not to act on the case-leaving the guards at risk and
the violation unresolved.

OSHA could not provide us with a list of all violations disputed by
federal agencies or the status of their resolution because it does not
have a system for tracking these disputed violations. OSHA officials at
the national office indicated that part of the reason the agency has not
developed such a

34

Executive Order 12196, S: 1-401(k).

35

According to the OSHA official, the violation involved a military maritime
facility that was held to a safety standard for general industry; there is
no such standard for the maritime industry.

Page 29 GAO-06-379 Worker Safety

    OSHA Has Not Conducted Evaluations in More Than 6 Years

system is because few federal agencies dispute violations. In addition,
according to these officials, disputed violations are resolved in a timely
manner. These officials reported that they seek to review cases in a
similar manner to the manner in which administrative law judges review
privatesector employers' cases and have considered using either a
permanent or ad hoc panel to ensure consistency in their review of
violations disputed by federal agencies. However, without a system for
tracking violations disputed by federal agencies, OSHA cannot ensure that
all disputes have been resolved or that they are resolved in a consistent
manner.

Although OSHA is required to conduct annual evaluations of the larger or
more hazardous federal agencies, and less frequent evaluations for smaller
and less hazardous federal agencies, it has not conducted any evaluations
since 1999. OSHA officials reported that because evaluations are so
resource intensive, they did not have enough staff to support doing them.
Evaluations are another element of OSHA's enforcement strategy and include
both a national-level review of an agency's safety program and
site-specific assessments. In the past, OSHA's national office identified
federal worksites for evaluations and the area offices inspected them.
OSHA's policies require agencies to correct any violations identified
during inspections conducted as part of its evaluations. In addition,
OSHA's internal guidance encourages its officials to coordinate
evaluations with targeted inspections in order to use its resources more
efficiently.

The last evaluation that OSHA conducted, at the Veterans Health
Administration, resulted in a report that agency officials said they still
use to improve their safety program. While some OSHA officials told us
that evaluations are resource intensive and ineffective because agencies
have not always corrected the problems identified, other OSHA and agency
officials said OSHA's evaluation of the Veterans Health Administration
helped bring management and union officials together for discussions
during the evaluation process. According to these officials, this improved
relationship continued after the evaluation was completed.

    OSHA Has Not Submitted Timely Annual Reports on Agencies' Safety Program
    as Required

As of February 2006, OSHA had not submitted its annual report to the
President that summarized and assessed the status of federal agencies'
safety programs since 2000 nor provided recommendations of ways for
federal agencies to improve their safety programs, as required. 36 OSHA is
working to reduce the backlog for these reports, according to the
officials we interviewed. In addition, OSHA officials told us that they
could not assess the effectiveness of these programs or make
recommendations because they do not collect original data on agencies'
safety programs but, instead, rely on the reports agencies provide to them
on an annual basis. According to these officials, they cannot assess or
evaluate agencies' programs without collecting independent information on
their programs. However, we believe that OSHA could use the information
provided by the agencies in their annual reports to assess agencies'
safety programs, including whether they are meeting OSHA's requirements.
For example, OSHA could use the agencies' reports to determine what types
of safety and health training they are providing to their managers and
workers, the number and types of self-inspections they are conducting of
their workplaces, and the measures used to correct unsafe and unhealthful
working conditions identified during these inspections. In addition, OSHA
could use these reports to make recommendations for improvement.

OSHA requires agencies to summarize their injury and illness rates and
provide information on new initiatives they have started and their
accomplishments in their annual reports. However, OSHA officials told us
that they do not systematically review these reports over time to ensure
that agencies are making progress. Our analysis of the agencies' reports
for fiscal years 2000 through 2004 showed that agencies generally
described the accomplishments of their safety programs but sometimes
repeated their safety goals across years. For example, one agency reported
in 2 consecutive years that it had "launched a new e-training program"
that included safety modules. In addition, agencies generally did not
provide any follow up information on their prior years' goals or
challenges. For example, one agency reported having a goal to develop a
database for tracking injury and illness trends but made no mention of the
system in the following year's report. One OSHA regional official
suggested that the national office could use regional staff more
effectively by requiring each region to review selected federal agencies'
annual reports. In this way, regional staff could become more familiar
with specific agencies'

36

29 U.S.C. S: 668(b).

    OSHA Has Used a Variety of Compliance Assistance Programs for Federal
    Agencies, but None Is Widespread

OSHA's Regional Offices Have Struggled to Set Up and Maintain Field
Federal Safety and Health Councils

programs, which would allow them to more readily identify discrepancies
and deficiencies in their annual reports.

Federal agencies can receive compliance assistance from OSHA through
programs developed especially for federal agencies as well as programs
initially developed for private-sector employers. The two compliance
assistance programs developed specifically for federal worksites-Field
Federal Safety and Health Councils and Agency Technical Assistance
Requests-have generally been helpful, according to OSHA officials, but
they are not consistently available to all federal agencies. Some of the
programs that OSHA initially developed for private-sector employers and
later expanded to federal agencies-the VPP, strategic partnerships, and
alliances-have not all been widely used by federal agencies. As of January
2006, only 14 federal worksites had joined the VPP and OSHA had
established few strategic partnerships and alliances with federal
agencies. However, although only a limited number of federal worksites
have used these programs, OSHA officials told us many of these efforts
have been successful and they are encouraging more agencies to
participate.

Regions have anywhere from 2 to 13 active Field Federal Safety and Health
Councils, depending on the effort regional OSHA officials have made to
develop and maintain them. These councils, established by OSHA to
facilitate the exchange of ideas and information about occupational safety
throughout the federal government, consist of management and employee
representatives from local federal agencies. 37 OSHA officials reported
that the councils are intended to provide a networking and training forum
for safety officials from different agencies in a given area, but all
agreed that maintaining the councils has been a struggle.

Both OSHA and agency officials cited challenges in maintaining the
councils. Some OSHA officials reported that federal agencies do not always
give their representatives time to attend the meetings. Other OSHA
officials raised concerns that federal agencies have failed to properly
train their collateral duty safety officials, which has inhibited their
contributions to the councils. In addition, some officials reported that
distance makes it difficult for council members to attend meetings. One

37

OSHA regulations, which established Field Federal Safety and Health
Councils to fulfill one of the requirements in Executive Order 12196,
state that OSHA will charter these councils and maintain a liaison with
agency heads to facilitate participation in the councils. 29 C.F.R.
S:S:1960.84-1960.90.

Page 32 GAO-06-379 Worker Safety

OSHA's Responses to Agency Technical Assistance Requests Are Sometimes
Delayed

Agencies Have Begun Joining OSHA's VPP

OSHA area director used the state's library videoconferencing system to
bring together council members from different areas and suggested that
OSHA consider similar methods to encourage collaboration. On the other
hand, a couple of agency safety managers and OSHA officials told us the
councils are not necessarily an effective tool for agencies because the
safety concerns are so different among the agencies. For example, a
Department of Veterans Affairs' safety manager might be focused on
preventing needle sticks and identifying violent patients, while National
Park Service safety staff might be concerned about snake bites and heat
exhaustion.

The councils also have limited financial resources. Funding is provided
solely by OSHA's regional offices and is not a line item in their budgets.
While regions attempt to provide training to the councils, any budget
constraint can quickly eliminate their ability to do so. Until last year,
OSHA's national office sponsored an annual conference and the regions
provided the travel funds for the council presidents to attend the
conference. However, the conference was canceled in fiscal year 2005,
partly because the national office did not have the funds to set up the
meeting and partly because the regions reported not having the travel
funds required.

OSHA officials said they sometimes are reluctant to respond to Agency
Technical Assistance Requests, which can delay this assistance, because
they consume their limited enforcement resources. An agency can request
OSHA to provide advice on hazard abatement, training, or program
assistance. OSHA cannot cite agencies for violations during this process
but, in making the request, agencies understand they are expected to
correct any violations OSHA observes. While these requests for technical
assistance are considered part of OSHA's compliance assistance strategy,
rather than enforcement, OSHA area offices and regions must use their
enforcement budgets and staff to conduct them. Because these offices have
limited enforcement resources, a regional OSHA official told us that,
although OSHA responds to all of these requests, this assistance may be
delayed.

As of January 2006, there were 14 federal worksites among the more than
900 private-sector worksites in OSHA's VPP, which promotes effective
worksite safety and health. In general, OSHA and agency officials told us
the program is beneficial for federal agencies and they expect more
worksites to join. An agency official also said that having one federal
worksite join often is an impetus for others to consider applying to join
the program. For example, since the U.S. Mint in Philadelphia became a VPP

OSHA Has Developed Few Strategic Partnerships and Alliances with Federal
Agencies

site in 2005, other agencies within the Department of Treasury have
considered joining. In addition, some OSHA field staff reported that they
are in the process of assisting agencies with their VPP applications.
While a few agency officials told us that the VPP was not feasible for
agencies because of the resources required, many told us they had
worksites seeking to join the program.

Some OSHA officials reported that federal agencies face unique challenges
in joining the VPP. For example, in order to participate, agencies must
have an injury and illness rate below the average within their given
industry. However, some agencies do not fit within a particular industry
code or definition. This was the case for Yellowstone National Park when
the worksite first applied to join the VPP. The park was required to
classify itself in an industry category that included amusement parks and
miniature golf courses, worksites with much lower injury and illness rates
than the park. The industry codes were recently changed and now include a
code for national parks, but Yellowstone is still challenged because its
injury and illness rates are higher than those of other parks such as
national monuments with many fewer hazards and injuries.

OSHA has developed relatively few strategic partnerships and alliances
with federal agencies, although OSHA officials said those that have been
formed have generally been beneficial to the agencies in improving their
safety programs. Strategic partnerships are agreements that employers make
with OSHA to address specific safety and health problems, while alliances
are agreements organizations make with OSHA to focus on training,
outreach, and promoting awareness of safety and health issues. OSHA has
created a limited number of strategic partnerships with federal agencies
at the national and regional level. At the national level, OSHA has one
partnership-an agreement with the Army created in October 2004 aimed at
increasing awareness of safety, reducing ergonomic injuries, and sharing
best practices. At the regional level, OSHA has 7 current and 10 completed
partnerships with federal agencies. (See table 3.)

Table 3: Federal Agencies' Strategic Partnerships with OSHA
            Current Federal agency         Completed Federal agency           
National       1 Army                           0 
Region                                            
1              0                                0 
2              1 National Park Service          1 National Park Service    
                    San Juan                         Fire Island National     
                                                     Seashore                 
3              0                                0 
4              1 National Park Service          0 
                    Mammoth Cave                     
5              0                                2 National Park            
                                                     Service-Isle Royale &    
                                                     Sleeping Bear Dunes      
6              1 Federal Bureau of              2 National Park Service    
                    Prisons, Three Rivers            Padre Island             
                                                     Federal Interagency      
                                                     Training Council         
7              0                                0 
8              3 National Park                  1 Veterans Health          
                    Service-Grand Teton,             Administration Cheyenne  
           Yellowstone, & Glacier                    
9     0                                         2 National Park            
                                                     Service-Yosemite &       
                                                     Golden Gate              
10             1 Bureau of Land                 2 Defense Commissary       
                    Management-Fremont               Agency, Ft. Lewis        
                    National Forest                  Forest Service           
Subtotal       7                               10 
regional                                          
partnerships                                      
Total       8                                  10 

Source: OSHA.

In general, OSHA officials said that these partnerships have helped
agencies reduce their injury and illness rates by helping them to develop
stronger safety programs. However, in two instances OSHA terminated its
strategic partnerships with federal agencies prior to their completion,
either because the agency could not agree on the terms of the partnership
or because the agency lacked the commitment to make the changes needed to
improve their safety programs.

Federal agencies have joined two national alliances and formed a total of
10 regional or local alliances. While most of the alliances have focused
on general safety issues, more recently Region 10 signed an alliance with
the Fort Lewis Army Garrison that focuses on improving the training and
communication for emergency response efforts. According to one OSHA
official, this alliance has leveraged both agencies' resources well. OSHA
has gained training from Fort Lewis on emergency response techniques, and
Fort Lewis has utilized OSHA's expertise in properly fitting staff members
for personal protective equipment to be worn during an emergency response.

OSHA assists federal agencies with SHARE, the Presidential initiative
begun in 2004 and intended to encourage federal agencies to improve their
safety programs and reduce federal workers' compensation costs, but the
impact of the initiative on agencies' safety programs is not clear.
Specifically, OSHA officials reported coordinating with OWCP to provide
training to the agencies about SHARE, but they had different views on the
effectiveness of the SHARE initiative. According to some OSHA officials,
the initiative has encouraged agencies' national offices to pay more
attention to safety issues than they otherwise would have. Other officials
said that they thought SHARE was a paper exercise rather than a tool for
agencies to improve their safety programs, or that this type of program
might encourage underreporting of injuries. OSHA's national office uses
workers' compensation data to calculate agencies' injury and illness rates
to determine whether they have met their SHARE goals related to workers'
safety, but it has not conducted any agency reviews to determine whether
underreporting has increased, according to OSHA officials.

OSHA officials at the national office said that they would like to use the
SHARE data to develop a list of agencies to target for inspection. By
focusing on agencies that are not meeting their SHARE goals, these
officials said they thought they could assist agencies in reducing their
injury and illness rates. OSHA officials said the agency will continue to
use workers' compensation data to calculate agencies' injury and illness
rates through 2006, but would consider using injury and illness data
collected under the new recordkeeping requirements after this time. Using
this new information would allow OSHA to identify trends for each federal
agency worksite and set more specific goals for improving agencies' safety
programs.

    Impact of the SHARE Presidential Initiative Is Unclear

                                  Conclusions

OSHA faces a number of challenges in monitoring federal agencies' safety
programs and, over time, has adapted its methods to try to make the most
of its resources. However, OSHA's oversight could be further strengthened
if it took a more strategic approach. Because targeted inspections
generally uncover more workplace hazards than its other inspections, by
not targeting its inspection efforts to the most hazardous federal
worksites, OSHA is not using its limited enforcement staff and resources
in the best way possible. Now that federal agencies are collecting injury
data that would make targeting more feasible, OSHA is missing a critical
opportunity to identify and correct hazards. OSHA could require, as part
of the federal agencies' annual reports, that each agency submit certain
portions or summaries of the data that agencies are required to collect
under the new recordkeeping requirements. This information could be used
to target federal worksites for inspection in the same way it targets
private-sector employers in industries with high injury and illness rates
for inspection. Alternatively, as OSHA does with private employers, OSHA
could develop its targeting program using the newly-required data that
federal agencies are collecting by surveying selected agencies and
worksites.

In addition, OSHA is not tracking violations disputed by federal agencies
or how they are resolved. As a result, hazardous worksite conditions may
remain uncorrected for years and OSHA may be limiting its ability to
address challenges agencies are facing in complying with OSHA's standards
and to provide additional assistance to the agencies.

While inspections are specific to individual federal agency's worksites,
evaluations allow OSHA to make thorough, agencywide assessments of their
safety programs. These evaluations require a lot of time and staff, but,
in the past, OSHA has been able to maximize its resources by strategically
combining evaluations of entire agencies with inspections of federal
worksites. By not conducting evaluations of the larger or more hazardous
federal agencies, OSHA is missing a critical opportunity to provide
agencies valuable feedback and assistance to agencies for improving their
safety programs in a more systematic way.

OSHA could also more effectively assess federal agencies' safety programs
if it ensured that the agencies complied with the requirements for filing
annual reports and used the reports, as well as OSHA's evaluations and
inspection data, to assess their safety programs and develop
recommendations for improvement. Because OSHA does not provide an
assessment of agencies' safety programs in its annual report to the
President or recommendations for improvement as required, its ability to
ensure the effectiveness of these programs is limited.

The Secretary of Labor should direct OSHA to

  Recommendations for

Executive Action  o  develop a targeted inspection program for federal
worksites based on the new worker injury and illness data federal agencies
are required to collect by requiring that relevant portions or summaries
of that data be

Page 37 GAO-06-379 Worker Safety

                       Agency Comments and Our Evaluation

included in agencies' annual reports to OSHA or by obtaining the data from
agencies or worksites through periodic, selected surveys;

     o track violations disputed by federal agencies to their resolution and
       ensure that unresolved disputes are reported to the President;
     o conduct evaluations of the largest and most hazardous federal agencies
       as required; and
     o use evaluations, inspection data, and annual reports submitted by
       federal agencies to assess the effectiveness of their safety programs,
       and include, in OSHA's annual report to the President, an assessment
       of each agency's worker safety program and recommendations for
       improvement.

We provided a draft of this report to the Secretaries of the Departments
of Labor, Agriculture, Defense, Homeland Security, Interior, Justice,
Treasury, and Veterans Affairs and the Commissioner of the Social Security
Administration. Officials from Agriculture, Treasury, and the Social
Security Administration informed us that their agencies did not have any
comments on our draft report. We received written comments from the
Departments of Labor, Homeland Security, and Interior. These comments are
reproduced in appendixes II, III, and IV. The Departments of Defense,
Justice, and Veterans Affairs provided technical clarifications, which we
incorporated as appropriate.

Labor generally agreed with all of our recommendations. In responding to
our first recommendation, OSHA explained that, for the immediate future,
it would use OWCP data to identify federal worksites for inspection. It
did not support the use of the annual reports to collect data on injury
and illness recorded by the agencies to use in targeting federal worksites
for inspection, but thought the use of surveys to collect these data was
noteworthy.

In regard to our second recommendation, OSHA reported that it will create
a database to track the status of OSHA citations disputed by federal
agencies. In responding to our final two recommendations, OSHA reported
that it would begin evaluations and a more rigorous review of agencies'
annual reports once staffing had increased.

The Departments of Homeland Security and Interior noted that Labor could
provide more assistance to agencies in addressing the challenges we
identified. While we believe agencies should seek assistance from OSHA on
ways to overcome these challenges, we also believe that these challenges
will require agencies to work internally to build support for worker
safety programs.

In addition, the Department of Homeland Security suggested that our
recommendations to Labor to increase OSHA's enforcement activities may not
appreciably lower the incidence of injuries and illnesses and may indeed
reduce agencies' requests for OSHA's assistance. We continue to believe
that increased enforcement activities would provide OSHA with a balanced
strategy for ensuring workplace safety. In addition, inspections will
allow OSHA to review federal agencies' injury and illness logs to ensure
that underreporting is not occurring-another concern that Homeland
Security raised in its comments. Finally, Homeland Security suggested that
OSHA should take the lead on developing a governmentwide safety
information system. We agree that it is important to have a governmentwide
safety information system and note that Labor has made some effort in that
direction.

We will make copies of this report available upon request. In addition,
the report is available at no charge on GAO's Web site at
http://www.gao.gov.

If you have any questions about this report, please contact me at (202)
512 9889 or at [email protected]. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. Key contributors to this report are listed in appendix
VII.

Robert E. Robertson

Director, Education, Workforce, and Income Security Issues

                       Appendix I: Scope and Methodology

We sent a data collection instrument to 57 agencies within the 8 largest
departments. The instrument requested information and documentation on six
components of sound safety programs we identified from previous GAO
reports: (1) management commitment, (2) employee involvement, (3)
education and training, (4) identification of hazards, (5) following up
and correcting hazards, and (6) medical management. We chose the eight
departments because they represented 80 percent of the federal executive
branch workforce-excluding the U.S. Postal Service, which under the OSH
Act is considered a private sector employer. We contacted officials with
each of the 8 departments to obtain the names of their operational
agencies and they provided us with the names of 57 agencies. We reviewed
the documentation supplied by the agencies to support their answers to
selected questions on the data collection instrument. In reviewing the
documentation, we made two assessments: (1) whether the documentation
supported the agency's responses and (2) what types of activities the
agency conducted for each program component. We examined each document
provided by the agencies in support of their responses and assessed each
as either "supporting" or "not supporting" the agencies' responses. Each
document was reviewed by two people to ensure that our assessment of the
sufficiency of the documents provided by the agencies was consistent. Of
the 57 agencies that completed the data collection instrument, two did not
provide any supporting documentation.

The practical difficulties of conducting any survey may introduce errors,
commonly referred to as nonsampling errors. For example, differences in
how a particular question is interpreted, the sources of information
available to respondents or in how the data are entered into a database or
were analyzed can introduce unwanted variability into the survey results.
We took steps in the development of the survey instrument, the data
collection, and the data analysis stages for the purpose of minimizing
such nonsampling errors. For example, a survey specialist designed the
survey instrument in collaboration with GAO staff with subject matter
expertise. We pre-tested this survey at two agencies and, based on the
results and comments received during pre-testing, made appropriate
revisions. We also independently verified the entry of all survey
responses entered into an analysis database as well as data analyses
procedures.

We conducted follow-up interviews with safety managers and when possible,
employee representatives from the largest agency within each department,
and with the agencies with the highest lost-time or injury and illness
rates. In some cases, the largest agency had both the highest losttime and
injury and illness rates. In total, we conducted follow-up interviews with
safety managers in 12 agencies, as well as employee

Appendix I: Scope and Methodology

representatives in 8 agencies. The interview questions were based on how
each agency originally responded to the data collection instrument and
their supporting documentation.

We also visited five federal agencies' worksites: the Tobyhanna Army Depot
in Tobyhanna, Pennsylvania; Yellowstone National Park in Wyoming; the U.S.
Mint in Philadelphia, Pennsylvania; the U.S. Forest Service's Gardiner
District Office (Gallatin National Forest) in Gardiner, Montana; and the
Veterans Health Administration's Rocky Mountain Network Office (a Veteran
Integrated Service Network site) in Glendale, Colorado, and Eastern
Colorado Health Care Center in Denver, Colorado. The first three worksites
are OSHA recognized VPP sites. The Forest Service site bordered
Yellowstone National Park, and the Veterans Health Administration site was
recognized by OSHA as having a good safety program. At each of these
locations, we interviewed safety officials and discussed the challenges
and solutions they faced in developing their safety programs.

We obtained information on claims filed by federal workers for injuries
they incurred from fiscal years 1995 through 2004 from OWCP. We used data
from two of OWCP's data systems to tabulate basic descriptive statistics
provided in this report. One system provides injury and case status
information on all individuals who have filed claims with OWCP while the
other is used to bill agencies for the actual amount of workers'
compensation payments made on the agencies' behalf. These systems were
used to develop our tables including the number of new cases filed; the
types of injuries incurred; and the actual amounts paid, by the age of the
case and types of payment. To assess the reliability of these data, we
interviewed OWCP and OSHA officials, reviewed published reports based on
these data (including reports from Labor's Office of the Inspector
General), and performed our own tests for consistency and completeness. We
found that certain data elements had high levels of missing information
and thus could not be used in this report. For the elements we used,
although small data discrepancies were found, we determined that the data
were sufficiently reliable for providing the basic descriptive statistics
reported.

In reviewing OSHA's role, we analyzed inspection data of federal agencies
for fiscal years 1995 through 2004 from OSHA's Integrated Management
Information System. We interviewed OSHA officials at the national office
and all of its 10 regional administrators and federal agency program
officers. For each region, we interviewed the director of the area office
that had the largest number of inspections of federal worksites in the
last 5

Appendix I: Scope and Methodology

years. We also interviewed a compliance safety and health official in each
of these offices identified by the area director and, where possible, the
compliance assistance specialist, although not every area office had a
compliance assistance specialist. In addition, we interviewed two OSHA
officials about the local emphasis program for federal worksites that one
region had implemented.

Finally, when available, we examined agencies' annual reports to OSHA from
2000 to 2004 and asked to review OSHA's annual reports to the President
for the same time period. However, as noted in the report, OSHA had not
completed its annual reports to the President for fiscal years 2001
through 2004 as required. We reviewed the report to the President that
OSHA had completed for fiscal year 2000.

Appendix II: Comments from the Department of Labor

Note: GAO comments supplementing those in the report text appear at the
end of the appendix.

See comment 1.

Appendix II: Comments from the Department of Labor

                                 See comment 2.

Appendix II: Comments from the Department of Labor

Appendix II: Comments from the Department of Labor

Appendix II: Comments from the Department of Labor

The following are GAO comments on Labor's letter dated March 17, 2006.

  GAO Comments

1. We reordered the SHARE goals as OSHA requested and identified those
       goals for which OSHA is responsible.
2. OSHA suggests that our finding-that violations disputed by federal
       agencies were not being tracked-was confusing because the agency has
       an inspection database that it uses to track the status of all
       violations. However, as noted in its comments, when OSHA generated a
       report to identify unresolved violations at federal agencies, staff
       could not determine the status of 11 violations. In addition, OSHA
       acknowledged that the Office of Federal Agency Programs (OFAP) does
       not have a formal tracking system for cases it receives for
       resolution. We reviewed the report language and believe that it
       accurately explains the process in place.

        Appendix III: Comments from the Department of Homeland Security

Appendix III: Comments from the Department of Homeland Security

Appendix III: Comments from the Department of Homeland Security

Appendix IV: Comments from the Department of Interior

Appendix IV: Comments from the Department of Interior

Appendix V: Data Collection Instrument Sent to the Federal Agencies

                        Page 54 GAO-06-379 Worker Safety

                        Page 55 GAO-06-379 Worker Safety

                        Page 56 GAO-06-379 Worker Safety

                        Page 57 GAO-06-379 Worker Safety

                        Page 58 GAO-06-379 Worker Safety

Appendix VI: Agencies' Responses to the Data Collection Instrument

The following table summarizes agencies' responses to the data collection
instrument. In addition, the last column summarizes whether the
documentation agencies provided to support their responses for selected
questions was sufficient.

 Table 4: Agency Responses to the Data Collection Instrument by Safety Program
                                   Component

                        Page 59 GAO-06-379 Worker Safety

                                                                   Percent of 
                                                            selected agencies 
Survey question MANAGEMENT                                responding "Yes" 
COMMITMENT Q1 Department           Number of  Number of      that provided 
goals established for the           agencies   agencies         sufficient 
occupational safety and      responding "No" responding         supporting 
health program.                            2   "Yes" 55  documentationa 91 
Q2 Activities designed to                                                  
communicate the importance                              
of the OSH program to staff.               0         57                 93
Q3 Management information                                                  
system that allows for trend                            
analysis, risk analysis,                                
etc.                                       8         49                 84
Q4 Policies that establish                                                 
responsibility for workplace                            
safety and health for all                               
staff through performance                               
reviews.                                   6         51                 78
EMPLOYEE INVOLVEMENT                                    
Q5 Procedures for employees                0         57                 89 
to report accidents.                                    
Q6 Procedures for employees                1         56                 91 
to report hazards.                                      
Q7 Employee access to the                                                  
system capturing information                            
on accidents and/or hazards.              19         38                N/A
Q8 Employee involvement in                                                 
safety committees/teams.                   3         54                 83
Q9 Employee                                                                
input/involvement in the                                
safety- related training                                
curriculum.                                8         49                N/A
Q10 Employee participation                                                 
in walkthroughs of worksites                            
to identify hazardous                                   
conditions.                                7         50                N/A
Q11 Employee involvement on                                                
accident investigation                                  
teams.                                    14         43                N/A
IDENTIFICATION OF HAZARDS                               
Q12 Procedures for                         2         55                 93 
conducting required OSHA                                
inspections of worksites by                             
safety personnel.                                       
Q13 Procedures for                         7         50                 80 
conducting informal                                     
walkthroughs of worksites to                            
identify hazards.                                       
CORRECTION OF HAZARDS                                   
Q14 Procedures for                         4         53                N/A 
developing controls for                                 
workplace hazards.                                      

Appendix VI: Agencies' Responses to the Data Collection Instrument

                                                          Percent of selected
                                                          agencies responding
                                                          "Yes" that provided
                        Number of agencies Number of               sufficient 
                                           agencies                supporting 
Survey question         responding "No"  responding "Yes"  documentation a 
Q15 Procedures for                                                         
following up on                       4                53               87
inspections to                                            
ensure hazards are                                        
corrected.                                                
Q16 Automated system to              23                34               59 
track workplace                                           
hazards.                                                  
TRAINING                                                  
Q17 Departmental OSH                  1                56              N/A 
training program.                                         
Q18 National training                                     
initiatives targeted by                                   
headquarters staff.                  12                45              N/A 
Q19 Automated system to              14                43               63 
track employee                                            
training.                                                 
MEDICAL MANAGEMENT                                        
Q20 Procedures to                                         
ensure that an injured                                    
or ill                                                    
employee is seen within                                   
a specified time frame                                    
by a                                                      
medical provider.                     8                49              N/A 
Q21 Automated system                  9                48              N/A 
tracking accident data.                                   
Q22 Department                                            
restricted or light                                       
duty return-to-                                           
work program.                        12                45              76% 
Q23 Automated system                                      
tracking return-to-work                                   
status of employees.                 34                23              30% 

Source: GAO analysis.

a

Fourteen questions were selected for in-depth documentation review.

Appendix VII: GAO Contact and Staff Acknowledgments

Robert A. Robertson (202) 512-9889

  GAO Contact

Revae E. Moran (Assistant Director) and Margaret A. Holmes (Analyst in
Charge) managed all aspects of the assignment. Jessica A. Lemke, Sheila

R. McCoy, and Kris Trueblood made significant contributions to this
report. Other key contributors to this report included Kyle Browning,
Richard Burkard, Nina E. Horowitz, Tovah Rom, Beverly Ross, Jeremy D.
Sebest, John G. Smale Jr., Rachael C. Valliere, and Eric A. Wenner.

(130388)

  GAO's Mission

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( www.gao.gov) . Each weekday, GAO posts GAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To

have GAO e-mail you a list of newly posted products every afternoon, go to
www.gao.gov and select "Subscribe to Updates."

                             Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

  To Report Fraud, Contact:

Web site: www.gao.gov/fraudnet/fraudnet.htm

  E-mail: [email protected]

Federal Programs Automated answering system: (800) 424-5454 or (202)
512-7470

Gloria Jarmon, Managing Director, [email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Relations
Washington, D.C. 20548

Paul Anderson, Managing Director, [email protected] (202) 512-4800

  Public Affairs

U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

    PRINTED ON

RECYCLED PAPER
*** End of document. ***