This is the accessible text file for GAO report number GAO-03-45 
entitled 'Workplace Safety and Health: OSHA Can Strengthen Enforcement 
through Improved Program Management' which was released on November 22, 
2002.



This text file was formatted by the U.S. General Accounting Office 

(GAO) to be accessible to users with visual impairments, as part of a 

longer term project to improve GAO products’ accessibility. Every 

attempt has been made to maintain the structural and data integrity of 

the original printed product. Accessibility features, such as text 

descriptions of tables, consecutively numbered footnotes placed at the 

end of the file, and the text of agency comment letters, are provided 

but may not exactly duplicate the presentation or format of the printed 

version. The portable document format (PDF) file is an exact electronic 

replica of the printed version. We welcome your feedback. Please E-mail 

your comments regarding the contents or accessibility features of this 

document to Webmaster@gao.gov.



Highlights of GAO-03-45, a report to the Chairman, Subcommittee on 
Workforce 

Protections, Committee on Education and the Workforce, House of 
Representatives



WORKPLACE SAFETY AND HEALTH: OSHA Can Strengthen Enforcement Through 
Improved 

Program Management:



Why GAO Did This Study: 



The United States has made great progress in improving working 
conditions 

since the construction of the Empire State Building. (See graphic.) 
Yet, since 

the early 1990s, over 50,000 workers have died from work-related 
accidents 

and millions experience work-related injuries or illnesses each year.  

The Occupational Safety and Health Administration (OSHA) is the primary 
federal 

agency responsible for protecting workplace safety and health. GAO was 
asked 

to assess how well OSHA was able to target its enforcement resources on 

hazardous worksites, measure its accomplishments, and ensure inspection 
staff 

quality.



What GAO Found:



OSHA has taken important steps toward targeting its enforcement 
resources on 

hazardous worksites, measure its accomplishments, and enhance the 

professionalism of its staff. However, these systems could be 
strengthened 

by better information and mechanisms that would make targeting efforts 
more 

efficient, measurement more precise, and training efforts more 
effective.  



OSHA’s targeting processes have not fully ensured that it identifies 

hazardous worksites for priority inspection because its worksite-
targeting 

programs lack the necessary data to effectively identify high-hazard 
worksites 

or those with hazards under OSHA’s jurisdiction.  Also, OSHA’s 
measurement 

efforts did not accurately demonstrate its impact on workplace safety 
and 

health because, for example, it used national data on injuries and 
illnesses 

to measure its progress in achieving strategic goals even though only 
31 

states are covered by these goals.  Finally, OSHA’s efforts to enhance 
the 

quality of its inspection workforce have the potential to improve 

enforcement, but the anticipated outcomes could be jeopardized by a 
lack 

of necessary mechanisms, such as a training directive that reflects 
current 

plans, or a comprehensive database that tracks training or skills 
obtained 

by inspection staff.  



[See PDF for figure]



[End of figure]



What GAO Recommends:

 

GAO recommends ways that OSHA can get the most out of its targeting 
programs, 

enhance its ability to measure its impact, and help ensure long-term 
success 

of its efforts to enhance inspector quality.



While OSHA expressed concerns about some of the material in the report, 

it generally agreed to act on our recommendations.



www.gao.gov/cgi-bin/getrpt?GAO-03-45.



To view the full report, including the scope and methodology, click on 
the 

link above. For more information, contact Robert E. Robertson at (202) 
512-

7215 or robertsonr@gao.gov.



Report to the Chairman, Subcommittee on Workforce Protections, 

Committee on Education and the Workforce, House of Representatives:



United States General Accounting Office:



GAO:



November 2002:



Workplace Safety and Health: OSHA Can Strengthen Enforcement through 

Improved Program Management:



GAO-03-45:



Contents:



Letter:



Results in Brief:



Background:



OSHA’s Targeting Procedures May Not Effectively Identify Hazardous 

Worksites for Inspection:



OSHA’s Measurement Efforts Have Not Accurately Demonstrated Its Impact:



OSHA’s Efforts to Enhance Inspector Quality Have Potential to Improve 

Enforcement, but Anticipated Outcomes Could Be Jeopardized:



Conclusions:



Recommendations for Executive Action:



Agency Comments and Our Evaluation:



Appendix I: Training Databases Used by OSHA’s Regional and 

Area Offices:



Appendix II: Comments from the Occupational Safety and Health 

Administration:



GAO Comments:



Appendix III: GAO Contacts and Staff Acknowledgments:



GAO Contacts:



Staff Acknowledgments:



Tables:



Table 1: OSHA’s 3-by-5 Strategic Plan Impact Goals (1997 to 2002):



Table 2: OSHA’s Procedures for Identifying Industries, Hazards, or 

Worksites to Target:



Table 3: Training Databases Used By OSHA Regional and Area Offices:



Figure:



Figure 1: Injury and Illness Rates for 3-by-5 Industries, 1993-2000:



Abbreviations:



BLS: Bureau of Labor Statistics:



GPRA: Government Performance and Results Act of 1993:



IDP: individual development plans:



IMIS: Integrated Management Information System:



LEP: Local Emphasis Program:



NEP: National Emphasis Program:



ODI: OSHA Data Initiative:



OSHA: Occupational Safety and Health Administration:



OTI: OSHA Training Institute:



SST: site-specific targeting:



United States General Accounting Office:



Washington, DC 20548:



November 22, 2002:



The Honorable Charlie Norwood

Chairman, Subcommittee on Workforce Protections

Committee on Education and the Workforce

House of Representatives:



Dear Mr. Chairman:



Since the early 1990s, over 50,000 U.S. workers have lost their lives 

in workplace accidents. Although workplace fatality and injury rates 

have improved significantly over the last decade, the Department of 

Labor’s Occupational Safety and Health Administration (OSHA) continues 

to strive to reduce work-related injuries, illnesses, and fatalities 

for the country’s 120 million workers. In carrying out this mission, 

OSHA faces a number of challenges. As with many other organizations, 

OSHA seeks to ensure that it (1) focuses its limited resources on those 

workplaces most likely to cause worker injury, illness, or death; (2) 

shows that its efforts make worksites safer; and (3) has a highly 

trained workforce capable of carrying out its mission effectively. 

OSHA’s efforts to address these challenges are made more difficult by 

the fact that it shares responsibility for workplace safety and health 

with numerous other entities. For example, while OSHA enforces safety 

and health regulations in 31 “federal OSHA states,” it has delegated 

its enforcement authority to protect workplace safety and health to 23 

“state-plan states,” where state agencies have assumed responsibility 

for establishing their own goals and enforcing safety and health 

regulations.[Footnote 1]



As agreed with your office, we assessed the extent to which (1) OSHA’s 

worksite-specific targeting processes ensure that OSHA effectively 

identifies hazardous worksites, (2) the agency’s measurement efforts 

allow it to accurately demonstrate its impact on workplace safety, and 

(3) its human capital efforts to enhance inspector quality are likely 

to improve enforcement. Our work focused on the inspection and 

enforcement activities within the federal OSHA states.



To address the first two objectives, we identified the procedures and 

data OSHA used to develop its targets and measure its results. We then 

obtained and analyzed relevant historical data from (1) the Department 

of Labor’s Bureau of Labor Statistics (BLS) on workplace injuries, 

illnesses, and fatalities;[Footnote 2] (2) OSHA’s integrated management 

information system (IMIS) inspection database, which compiles activity 

information on worksite inspections conducted by OSHA; and (3) the OSHA 

data initiative (ODI) database, through which OSHA targets specific 

worksites for priority inspection. We also reviewed with experts OSHA’s 

procedures and methods to measure its efforts on workplace safety and 

health. For the third objective, we reviewed OSHA’s recent efforts to 

restructure its local offices. We also reviewed its plans for providing 

training to its inspection staff and assessed whether OSHA has in place 

the policy directives and database systems needed to ensure the success 

of these efforts.[Footnote 3] We obtained OSHA personnel’s views on 

these efforts through interviews with officials at 9 of the 10 OSHA 

regional offices and 17 of OSHA’s 85 area offices.[Footnote 4] For all 

objectives, we interviewed OSHA and other Department of Labor 

officials. We conducted our work in accordance with generally accepted 

government auditing standards between December 2001 and October 2002.



Results in Brief:



OSHA’s targeting processes have not fully ensured that it effectively 

identifies hazardous worksites for priority inspection. Specifically, 

OSHA’s efforts to target worksites in the construction industry rely on 

a database that does not permit OSHA to accurately identify smaller 

worksites for inspection. Consequently, OSHA’s area offices tend to 

select larger construction worksites to visit, which experts and 

several OSHA officials believe are generally safer than the smaller 

worksites. OSHA is currently studying ways to improve construction 

targeting. In the meantime, some of OSHA’s local offices are targeting 

small construction sites on their own. Also, OSHA’s efforts to target 

high-hazard worksites across other industries rely on employer-provided 

information that may be unsuitable for accurate targeting. In about 50 

percent of the worksites identified, inspectors were unable to conduct 

inspections or did not find any serious violations. Officials from 

OSHA’s regional and area offices we interviewed expressed concern about 

the ability of this program to efficiently target hazardous worksites.



Several weaknesses in OSHA’s measurement efforts have affected its 

ability to accurately demonstrate its impact on workplace safety, a 

difficult, but nonetheless important activity. First, OSHA relied on 

national injury and illness statistics compiled by BLS to measure 

progress toward strategic plan goals, despite the fact that OSHA’s 

strategic plan only covers the 31 federal OSHA states. OSHA has not 

made use of BLS data that would allow it to look at injury and illness 

rates for these 31 states combined, although BLS says it could make the 

data available at reasonable cost. Second, the methods OSHA used to 

measure progress toward its strategic goals potentially misstated its 

accomplishments. For example, the way it measured declines in injury 

and illness rates included declines that occurred before its strategic 

plan was implemented. Finally, when assessing its impact, OSHA did not 

fully account for many relevant factors outside its control that may 

have affected changes. For example, in reporting that it exceeded goals 

for reducing fatalities in the construction industry, OSHA did not 

acknowledge that some portion of these fatalities might have been due 

to hazards under the authority of others, such as the Department of 

Transportation. As a result, reductions in such fatalities would have 

no direct relationship to OSHA’s efforts.



OSHA’s efforts to enhance the quality of its inspection workforce can 

possibly improve enforcement, but anticipated outcomes could be 

jeopardized by several factors. OSHA has restructured its local offices 

into multidisciplinary teams (i.e., teams comprised of safety 

inspectors and health inspectors), with team leaders responsible for 

overseeing the work of those teams. While this change has fostered 

greater collaboration between safety and health inspectors, the 

creation of these teams has, in some locations, led to insufficient 

internal controls in the supervisory review process. For example, in 

some cases where team leaders are safety inspectors, they may lack the 

expertise needed to review case files that support health inspectors’ 

citations. This practice could lead to inconsistency in citations or 

citations being made without the proper justification. In addition, 

while OSHA’s plans to upgrade the training provided to its inspection 

force have promise, OSHA faces two primary obstacles to success. First, 

OSHA’s training directive does not reflect the agency’s current 

training plans, so there is no assurance that its current training 

plans will have organizational support over the long term. Second, OSHA 

lacked the necessary data to assess whether training contributes to 

agency performance because it did not use a single, comprehensive 

database to efficiently track the training and skills obtained by its 

inspection force.



We are recommending that OSHA strengthen management of its enforcement 

activities by improving targeting and measurement procedures and by 

helping ensure the long-term success of the agency’s efforts to enhance 

inspector quality. While OSHA expressed concerns about some of the 

material in this report, it generally agreed to act on our 

recommendations.



Background:



The Occupational Safety and Health Act states that it is congressional 

policy to “assure so far as possible every working man and woman in the 

Nation safe and healthful working conditions ..”[Footnote 5] In 

fiscal year 2002, OSHA pursued this mandate with about $443 million and 

2,316 employees, including 1,123 inspectors.[Footnote 6] Despite this 

broad mandate, OSHA does not have complete authority over all worksites 

in this country. For example, through the appropriations process, 

Congress has placed restrictions on OSHA’s enforcement activities 

regarding small farming operations and small employers in low hazard 

industries. Also, pursuant to the act, OSHA has delegated federal 

enforcement responsibility to 23 state and territorial governments that 

carry out their own programs.[Footnote 7] OSHA has provided half the 

funding for these programs to state-plan states, which must have 

program standards that are “at least as effective” as the federal 

program. In other cases, federal agencies other than OSHA have 

jurisdiction over particular workplaces or hazards.[Footnote 8] For 

example, the Department of Labor’s Mine Safety and Health 

Administration is responsible for ensuring the safety of mining 

worksites, while transportation-related hazards are generally within 

the jurisdiction of the Department of Transportation.



OSHA also relies on BLS to provide it with data on injuries, illnesses, 

and fatalities. Since 1992, BLS has surveyed a sample of approximately 

180,000 employers and asked them to report information on the number of 

work-related injuries and illnesses occurring at their worksites. This 

information comes from injury and illness records that private industry 

employers are required to maintain. From this information, BLS 

calculates injury and illness rates. BLS identifies fatalities from a 

census of all 50 states, the District of Columbia, and New York City, 

which report on all work-related fatalities within their jurisdictions. 

BLS requires the reporting entities to corroborate reports on 

fatalities by obtaining multiple sources of information, such as OSHA 

fatality information, death certificates, medical examiners’ reports, 

media reports, and workers’ compensation documents. BLS makes injury, 

illness, and fatality data available at the national as well as at the 

state level.



To ensure that it makes the best use of its resources, OSHA, like other 

federal agencies, established strategic goals that drive agency efforts 

and has begun to measure the attainment of those goals.[Footnote 9] In 

OSHA’s 1997-2002 strategic plan, it identified an overarching goal to 

“[i]mprove workplace safety and health for all workers, as evidenced by 

fewer hazards, reduced exposures, and fewer injuries, illnesses, and 

fatalities.” OSHA characterized this overarching goal as the 

cornerstone of its enforcement program.[Footnote 10] According to OSHA, 

it planned on “focusing . . . [a]gency resources on the most prevalent 

types of workplace injuries and illnesses, the most hazardous 

industries, and the most hazardous workplaces.” OSHA identified 

specific areas that it believed were the most hazardous and by which 

OSHA would measure its progress. These specific areas, as shown in 

table 1, focused on three hazards and five industries or industry 

sectors (which are subsets of larger industries), and are known as 

OSHA’s “3-by-5 goals.” OSHA also identified a goal to reduce fatalities 

in the construction industry by 15 percent and reduce injuries and 

illnesses by 20 percent in 100,000 workplaces where OSHA initiates an 

intervention.



Table 1: OSHA’s 3-by-5 Strategic Plan Impact Goals (1997 to 2002):



Goal: Reduce three of the most prevalent types of workplace injuries 

and causes of illnesses by 15 percent in selected industries and 

occupations.; Targeted areas: Hazards.



Targeted areas: * Amputations.



Targeted areas: * Exposure to lead.



Targeted areas: * Exposure to silica.



Goal: Reduce injuries and illnesses by 15 percent in five industries 

characterized by high-hazard workplaces.; Targeted areas: Industries or 

industry sectors.



Targeted areas: * Construction industry.



Targeted areas: * Food processing.



Targeted areas: * Logging.



Targeted areas: * Nursing homes.



Targeted areas: * Shipyards.



Source: OSHA Strategic Plan for Fiscal Years 1997 to 2002 and 

Department of Labor Annual Report for Fiscal Year 1999.



[End of table]



Because GPRA applies only to federal agencies, the 23 state-plan states 

are not required to adopt these goals. OSHA asked these states to 

establish industry and hazard goals similar to OSHA’s 3-by-5 goals 

relevant to the worksites in their respective states. In some cases, 

the state-plan states selected hazards or industries that mirrored the 

federal ones; in other cases, they did not.



Recognizing that its 3-by-5 goals would probably not include every 

hazard and industry that would prove dangerous to workers over the 5-

year duration of the strategic plan, OSHA established various other 

mechanisms through which it could identify areas that pose hazards to 

workers. Through national emphasis programs (NEP), OSHA headquarters 

has identified industries or hazards deserving priority attention from 

its area offices. In the last several years, OSHA has averaged about 10 

NEPs annually. Some NEPs have reflected areas selected in the strategic 

plan, while others have focused on other areas (e.g., the petrochemical 

industry) not in the strategic plan. While OSHA provides direction to 

its area offices for implementing NEPs, the area offices have 

considerable flexibility in selecting actual worksites. In addition, 

area offices can also use local emphasis programs (LEP) to highlight 

industries or hazards within their jurisdictions that they believe are 

hazardous. (See table 2.):



OSHA also has two national targeting programs that are aimed at 

identifying worksites for priority inspection. First, as shown in table 

2, OSHA has a targeting program that relies on data from the F.W. Dodge 

Report for identifying worksites in the construction industry. Second, 

OSHA has a site-specific targeting program (SST), which identifies a 

list of high-hazard worksites in other industries through its OSHA data 

initiative. OSHA’s targeting efforts must be carried out on a neutral 

and objective basis, in accordance with legal requirements.



Table 2: OSHA’s Procedures for Identifying Industries, Hazards, or 

Worksites to Target:



Targeting effort: NEP; Process followed: Based on national information, 

OSHA headquarters identifies key industries, industry sectors, or 

hazards that regional and area offices may target. (Over the last few 

years, OSHA’s NEPs have covered petrochemical, poultry processing, and 

nursing homes.); Regional or area offices develop a particular strategy 

for targeting workplaces in these areas.



Targeting effort: LEP; Process followed: Based on information for its 

locality, an area or regional office may decide that an industry, 

industry sector, or hazard is particularly dangerous and seek approval 

to identify it as an LEP. (In fiscal year 2001, there were over 100 

LEPs, ranging from residential construction to sawmills.); Area and 

regional offices develop a particular strategy for targeting workplaces 

in these areas.



Targeting effort: Construction Industry; Process followed: OSHA local 

offices ask the University of Tennessee’s Construction Industry Policy 

and Research Center to provide them with monthly lists of randomly 

selected nonresidential construction projects scheduled to start in the 

next 60 days on the basis of particular criteria (usually related to 

project value) developed by the local office.[A]; To develop these 

lists, the center applies a statistical model to data contained in the 

F.W. Dodge Report, which is initially generated by the McGraw-Hill 

Companies. The information provided to local offices includes the 

predicted start and completion dates of the projects. Since 

construction sites are temporary, local offices need this information 

to plan inspections.; Local offices then visit construction sites on 

the list.



Targeting effort: ODI/SST; Process followed: As part of the ODI begun 

in 1998, OSHA identifies various industries that have an injury and 

illness rate above a predetermined level, as well as other industries 

that it believes are hazardous.[B]; OSHA asks Dun and Bradstreet 

Corporation to provide it with specific information (such as employer 

names and addresses) for approximately 140,000 worksites with 40 or 

more employees from these industries.; OSHA selects about 80,000 

worksites that it believes have a strong potential for being hazardous 

based on the relative hazardousness of their industries. OSHA contacts 

each of these worksites and asks them to provide their most recent 

annual injury and illness data.; Of these 80,000 worksites, OSHA then 

selects about 14,000 that it considers to be the most hazardous and 

sends them a letter informing them that they might receive a compliance 

inspection. OSHA provides 3,000 of these worksite names to area offices 

for priority inspection.[C] These worksites constitute the actual SST 

list.



[A] The University of Tennessee’s Construction Industry Policy and 

Research Center also maintains the Dodge Data Lines, which provide OSHA 

with information on residential construction projects. Access to the 

Dodge Data Lines database is included in the price OSHA pays for the 

F.W. Dodge Report data.



[B] In past years, OSHA excluded the construction industry from 

consideration. OSHA now plans to include construction in this process.



[C] OSHA sends this list only to its area offices in federal OSHA 

states. These area offices are expected to visit all SST worksites 

identified; however, any office that does not have the resources to 

inspect all worksites on the original list may use a random process to 

select a subset of these worksites for inspection. OSHA sends 

information on additional worksites in state-plan states to the 

appropriate state agencies. All state-plan states participate in this 

process except Oregon, South Carolina, Washington, and Wyoming.



Source: OSHA.



[End of table]



As another way to help ensure that its resources are well used, OSHA 

has restructured local office operations. Before fiscal year 1997, area 

offices were organized in single-discipline teams that responded to 

worker complaints and conducted planned inspections. Starting in fiscal 

year 1997, OSHA reorganized its local offices into multidisciplinary 

teams comprised of both safety inspectors[Footnote 11] and health 

inspectors. These teams specialized in either responsive activities 

(i.e., responding to worker complaints, accidents, or serious injuries; 

or acting on referrals from other agencies) or planned activities, such 

as conducting planned inspections and providing employers with 

compliance assistance (i.e., various efforts to help employers who 

voluntarily seek to comply with OSHA regulations). To assist these 

teams, OSHA placed a compliance assistance specialist at each area 

office who provides services such as helping employers correct hazards 

identified during inspections. OSHA also instituted new procedures that 

permitted area offices to expedite the process for responding to 

informal worker complaints by allowing inspectors to resolve complaints 

by phone or fax without visiting the worksite.



OSHA has also improved training opportunities for its inspection 

workforce. The agency has expanded course offerings available to 

inspectors at the OSHA Training Institute (OTI) and through satellite 

delivered and web-based training. In addition, OTI is revising its 

curriculum to prepare new inspectors both for their new jobs and for 

professional certification in either safety, health, or as an engineer. 

The agency has also developed a plan to assist experienced inspectors 

to obtain professional certification, should they choose to do so, and 

to retain professional certifications already achieved. The curriculum 

for professional certification will vary considerably depending on the 

type of certification sought by the inspector and the inspector’s 

current experience and level of training. OSHA’s certification 

assistance also includes paying for preparation materials and 

certification examination fees and making time available at work for 

staff to study. Concurrently, OSHA managers are using an individual 

development plan (IDP) process to help inspectors identify training 

needs and select appropriate training opportunities.



OSHA’s Targeting Procedures May Not Effectively Identify Hazardous 

Worksites for Inspection:



The targeting processes that OSHA used have not fully ensured that the 

agency effectively identifies hazardous worksites for priority 

inspection. Specifically, when targeting the construction industry, 

OSHA relied on a database that did not adequately identify the smaller, 

potentially more hazardous worksites. In the meantime, however, OSHA 

area offices have taken actions on their own to target small 

construction sites. Also, the efficiency of OSHA’s efforts to target 

high-hazard worksites across other industries through its SST program 

may be limited by faulty information that caused OSHA to send 

inspectors to worksites that were either not hazardous or that had 

hazards that were outside of OSHA’s control.



Current Construction Industry Targeting Biased toward Larger Worksites:



OSHA’s current construction industry targeting procedure has not 

provided local offices with adequate information on smaller 

construction worksites. OSHA relies on information from the F.W. Dodge 

Report database, provided by the University of Tennessee, to identify 

construction worksites for potential inspection. This database provides 

selected information on each construction site, including the projected 

start and completion dates. However, the start and completion dates, 

which are added to the database at the University of Tennessee, are 

often erroneous for small construction sites.



Since they had more confidence in the information the database provided 

on larger worksites, OSHA’s area offices generally selected larger 

construction worksites to inspect. About half of the area office 

directors we interviewed said they do not request information on 

smaller construction sites through the F.W. Dodge Report process. 

Several local office directors told us that, when relying on the 

database to identify small construction worksites, they would only send 

inspectors to areas where there were multiple worksites in close 

proximity in the hope of finding at least one that would be available 

for inspection. Knowledgeable experts and officials within and outside 

OSHA, including area office officials, saw this as problematic because 

larger construction worksites are generally safer than smaller ones, 

although they acknowledge that conclusive data to demonstrate this are 

unavailable.



OSHA officials acknowledge that the F.W. Dodge process can be improved 

to better identify high-hazard construction worksites and are 

undertaking efforts to identify ways to improve the construction 

targeting process. OSHA has asked the University of Tennessee to study 

all factors, including size, that may lead to injuries and illnesses in 

construction in order to determine the relative level of hazard 

represented by individual construction worksites. As of July 2002, the 

University of Tennessee had yet to initiate work on this study. This 

effort should help OSHA use more sophisticated criteria to select the 

most hazardous construction worksites for priority inspection, but it 

does not aim to address the immediate bias toward targeting larger 

construction worksites. To address current problems, we found that 

several of the local offices were using various methods to supplement 

the F.W. Dodge Report data to better target smaller construction 

worksites. Eight of the 17 area office directors we interviewed stated 

that they relied on more informal criteria and LEP initiatives to 

target smaller construction workplaces. For example, in 

1999-2000, four area offices developed LEPs for residential 

construction worksites because office personnel were seeing increasing 

numbers of fatalities or injuries occurring at these sites. These local 

offices believed their efforts were successful in locating the smaller, 

more hazardous worksites. However, not all area offices had established 

local emphasis programs for smaller construction worksites.



SST Program Does Not Effectively Identify Hazardous Worksites:



The SST program is limited in its ability to effectively identify 

hazardous worksites.[Footnote 12] Our review of OSHA’s own IMIS 

inspection database found that for about half the worksites identified 

through this process, inspectors were unable to do an inspection or, if 

they did, cited no serious violations. While OSHA headquarters 

officials have not analyzed why this occurs, our review of IMIS as well 

as interviews with area office directors indicate that these outcomes 

could result from faulty information that caused OSHA to send 

inspectors to worksites that were either not hazardous or that had 

hazards that were outside of OSHA’s control.



In some cases, OSHA received outdated or incorrect information about 

the establishment itself (i.e., name, location, nature of business, or 

number of employees). As a result, inspectors may have been unable to 

conduct an inspection. In other cases, OSHA received miscalculated 

information about the employer’s injury and illness rate. In these 

situations, inspectors visiting worksites determined from an inspection 

of its records that the actual injury and illness rate was not high 

enough for the employer to qualify for an inspection.[Footnote 13] We 

found, based on inspection data from OSHA’s IMIS database, that 

inspectors performed no inspection or just a records inspection (i.e., 

a review of the employer’s injury and illness records) for about 17 

percent of the worksites identified on the original SST list.



In other cases where the information on the worksite injury or illness 

rates was correct, the data collected may still have been otherwise 

unsuitable for efficiently targeting those high-hazard worksites where 

OSHA can have an effect. In collecting information for this program, 

OSHA asked employers for only 1 year of injury and illness data. Area 

office officials we interviewed said that in some cases, this 1-year 

rate was an outlier that did not reflect general worksite operations. 

Moreover, the data OSHA collected were generally 2 years old before 

inspectors conducted the inspection. As a result, employers might have 

taken actions, such as using OSHA’s consultation program,[Footnote 14] 

to improve working conditions by the time the inspector arrived. Also, 

the injuries and illnesses may have been caused by workplace hazards 

OSHA does not address.[Footnote 15] Again, using IMIS, we found that 

for about 17 percent of worksites on the SST list, inspectors found no 

violations. In another 14 percent, inspectors found no serious 

violations.



Generally, officials from OSHA’s regional and area offices we 

interviewed expressed concern about the ability of the SST program to 

reach those worksites with hazards that inspectors can address. Over 

half stated that the program did not identify a sufficient number of 

employers with serious violations to warrant their participation. For 

example, at one local office, we were told that 35 percent of worksites 

on the list were not cited for a violation. They noted that OSHA spends 

significant time and energy to develop the SST list. They also noted 

that significantly fewer resources are spent identifying worksites 

under national or local emphasis programs, yet they appear to be more 

successful in identifying serious violations. Our review of IMIS data 

on the results of LEP inspections found that over 

60 percent of inspections had serious violations.[Footnote 16] Also, 

our review of reports from area offices on the results of their LEP 

efforts identified anecdotal information about the success of LEP 

investigations for reaching the most hazardous worksites.



In contrast to the views expressed by regional and area officials, OSHA 

headquarters officials noted that a 50 percent serious violation rate 

could be acceptable if it meant that employers had actually improved 

working conditions between the time they were notified of a possible 

inspection and the time the inspection actually took place.[Footnote 

17] However, there is insufficient information to determine whether 

this violation rate should be interpreted as a positive sign that 

employers are taking action, or rather an indication that OSHA has not 

reached the most hazardous worksites.



Additionally, there is insufficient information available to know what 

impact the SST program has on reducing injuries and illnesses. First, 

OSHA has little data on injury and illness rates for the period after 

the SST inspections occurred. Having this information could help OSHA 

identify changes that happened after an inspection took place. Such an 

analysis would be imperfect since other intervening factors may have 

influenced injury and illness rates, but the results might still be 

useful. Moreover, OSHA did not establish a comparison group of 

employers whose worksites were equally hazardous, but were not selected 

for inspections. Developing such a comparison group has potential to 

help OSHA address the problem presented by intervening factors. There 

are several possibilities for a comparison group, including employers 

from the original ODI list that were not selected to be on the SST 

list, or similar types of employers located in state-plan states. We 

acknowledge that there are many factors to be considered in developing 

a comparison group. One expert we interviewed suggested that it might 

be difficult to use the ODI database for both targeting and evaluation 

and suggested that OSHA develop a similar database of establishments to 

be used purely for evaluating SST’s effectiveness.



OSHA’s Measurement Efforts Have Not Accurately Demonstrated Its Impact:



Several weaknesses in OSHA’s measurement efforts affected its ability 

to accurately demonstrate its impact on workplace safety. To measure 

progress toward its strategic plan goals, OSHA relied on national 

injury and illness statistics rather than on data specific to those 

states covered by OSHA’s strategic plan. Moreover, the methods OSHA 

used to measure its progress may have misstated its accomplishments. 

Finally, when assessing its impact, OSHA did not account for many 

relevant factors outside its control that may have affected changes in 

the number of work-related injuries, illnesses, and fatalities.



National Data Do Not Provide a Reliable Picture of OSHA’s Impact:



By using national data, OSHA lost the opportunity to understand what is 

happening with regard to injuries and illnesses in the states covered 

by its strategic plan. According to OSHA officials, available data did 

not allow them to measure changes in injury and illness rates for all 

federal OSHA states combined. In about 10 federal OSHA states, the 

amount of data BLS collects about injuries and illnesses is 

insufficient to allow the information to be generalized to the entire 

state. However, BLS uses the information from these 10 states in 

calculating its national estimate of workplace injuries and illnesses. 

OSHA officials told us that the lack of data from these states 

precluded BLS from making injury and illness estimates for all federal 

OSHA states combined. Nonetheless, according to BLS officials, 

available data from these 10 states could be combined with data from 

the other federal OSHA states to provide an overall estimate of injury 

and illness rates for the combined federal OSHA states. They said that 

this could be done at little or no additional cost to OSHA, but it may 

take up to a year to fully generate and test the program needed to 

produce this estimate, although some information could be made 

available sooner.



OSHA’s Methods for Measuring Performance May Have Misstated Its 

Accomplishments:



OSHA used methods to measure its progress in reducing injury and 

illness rates in the industries and hazards highlighted in its 

strategic plan that may have misstated its accomplishments. More 

specifically, to measure its progress in achieving its strategic plan 

goals, OSHA compares the most recent injury and illness data to a 1993 

through 1995 baseline. For example, in its 2001 annual report, OSHA 

compared calendar year 2000 injury and illness data (the latest 

information available from BLS) with the same data for 1993 through 

1995. Based on this comparison, OSHA reported that injury and illness 

rates declined by 26 percent in shipyards; 18 percent in food 

processing; 9 percent in nursing homes; 36 percent in logging; and 23 

percent in the construction industry. Yet, as shown in figure 1, based 

on data reported by BLS, a portion of these declines occurred before 

1997, the first year of the strategic plan’s implementation. While the 

agency may well have contributed to improvements before 1997, those 

downward trends in illness and injury cannot be characterized as an 

indication of the plan’s effectiveness. Further, even using the 1993-95 

point of comparison, two of the five industries highlighted in OSHA’s 

strategic plan did not have changes that were statistically 

significant, according to an OSHA official.[Footnote 18] Comparing 

changes between 1996 (the year before the strategic plan went into 

effect) and 2000, we estimated that three of the five industries 

highlighted in OSHA’s strategic plan did not have changes that were 

statistically significant.[Footnote 19]



Figure 1: Injury and Illness Rates for 3-by-5 Industries, 1993-2000:



[See PDF for image]



[End of figure]



Additionally, OSHA itself has acknowledged that it misstated its 

progress in achieving its goals for two of the three hazards 

highlighted in its strategic plan, those pertaining to reducing 

exposure to silica and lead. Initially, OSHA took reductions in silica 

and lead exposure at worksites it inspected and generalized them to the 

nation as a whole. In its fiscal year 2001 annual report, however, OSHA 

acknowledged that this was inappropriate and that the data and process 

did not satisfactorily measure progress on this goal. OSHA added that 

its methodology did not measure the average exposure severity for lead 

and silica in all workplaces; instead it measured the average exposure 

severity at workplaces that OSHA inspected, which have been specially 

targeted as potentially hazardous.



OSHA Did Not Consider Many Relevant Factors When Assessing Its Impact:



When assessing its impact, OSHA did not consider many of the factors 

outside its control that may have influenced changes in the level or 

type of injuries, illnesses, or fatalities. There is general agreement 

among those we interviewed within and outside of OSHA that other 

factors, such as workers’ compensation programs, have an effect on 

workplace safety and health. However, in presenting its evaluation of 

progress toward strategic goals, OSHA did not account for the potential 

effects of these other influences in its annual reports. Also, in some 

cases, OSHA did not account for hazards causing injuries, illnesses, or 

fatalities not under its full control. For example, while OSHA reported 

that it exceeded its goals for reducing fatalities in the construction 

industry, it did not report that some portion of this reduction might 

have occurred because of declines in transportation-related accidents 

under the authority of the Department of Transportation. For example, 

in 2001, about half of the fatalities in the construction industry that 

resulted from transportation accidents were likely under the authority 

of the Department of Transportation.



OSHA’s Efforts to Enhance Inspector Quality Have Potential to Improve 

Enforcement, but Anticipated Outcomes Could Be Jeopardized:



OSHA’s initial efforts to enhance inspector quality are encouraging, 

but the anticipated outcomes could be jeopardized. First, although 

OSHA’s restructuring efforts, which included the use of 

multidisciplinary inspection teams, have had some positive results, the 

effort may have also led to insufficient internal controls in the 

supervisory review process that could adversely affect the consistency 

of enforcement. Additionally, OSHA’s efforts to increase training 

opportunities for inspection staff hold promise but face two obstacles 

that, if not addressed, may undermine the long-term success of the 

resources invested in training.



Restructuring Is Positive, but May Have Led to Lack of Controls in the 

Supervisory Review Process:



OSHA’s local office restructuring appears to have strengthened 

inspectors’ ability to enforce workplace safety and health standards. 

First, OSHA’s effort to develop multidisciplinary teams has resulted in 

increased opportunities for cross-training among safety and health 

inspectors. Indeed, two of the local office directors we interviewed 

said that inspectors are better able to detect violations, even if the 

violations are outside of their disciplines. Second, having a 

compliance assistance specialist at each area office has provided 

inspectors with a much needed in-house resource for identifying 

techniques to make workplaces safer. Eleven of the 17 area office 

directors we interviewed said the compliance assistance specialist 

position has greatly enhanced inspector quality, helping inspectors 

provide cutting-edge information to employers about how to abate 

identified hazards. Third, OSHA’s new flexible process for responding 

to complaints by phone or fax rather than actual visits has made 

inspectors more efficient and able to focus on priority areas. Some of 

the 17 area office directors we spoke with stated that this more 

flexible complaint process freed up inspectors’ time by as much as 30 

percent or more, allowing them to focus on planned inspections and 

compliance assistance rather than respond to complaints.



However, the move to multidisciplinary teams may have undermined the 

internal control process for supervisory review of inspectors’ case 

files. Some area office directors we interviewed said that some team 

leaders (who generally have backgrounds in safety) do not have the 

expertise needed to review the health inspectors’ case files that 

accompany and support proposed violations. About half of the 26 

regional administrators and area directors we interviewed expressed 

concern about this issue. These officials explained that when team 

leaders are unable to review case files, they generally ask someone 

else in the local office with the appropriate expertise to review them, 

thus resulting in additional review time and a greater potential for 

mistakes because of the additional individuals involved in the process.



Area office officials we interviewed have attempted to address this 

problem with varying degrees of success. Some have sought to resolve 

this issue working within the multidisciplinary team structure. For 

example, those offices that were large enough to establish multiple 

teams for carrying out planned activities selected team leaders from 

both the safety and health disciplines. They can support each other and 

provide technical guidance to inspectors regarding both safety and 

health issues. Other area offices did not follow headquarters 

directives and chose not to restructure into multidisciplinary teams. 

Instead, they maintained separate teams of safety inspectors and health 

inspectors. These area office directors told us that they essentially 

attain the goals of a multidisciplinary team approach by creating 

temporary multidisciplinary teams whenever the need arises. They stated 

that, in this way, their offices obtain the benefits of 

multidisciplinary teams while minimizing the problems other offices 

have had with team leaders lacking necessary expertise. Two area office 

directors we spoke with expressed a desire to go back to separate teams 

of safety inspectors and health inspectors but believed they needed 

permission from OSHA headquarters to do so. OSHA has yet to address 

these concerns at the national level.



Training Plans Hold Promise, but Lack of an Appropriate Training 

Directive and Data System Pose Obstacles:



OSHA’s plans to upgrade inspector training have the potential to 

improve the professionalism and capabilities of the inspection staff. 

OSHA officials we interviewed stated that they are upgrading the 

curriculum of the OSHA Training Institute (OTI), which will provide 

inspectors with training opportunities that give them a broader 

understanding of the issues surrounding worker safety and health and 

will improve their skills to conduct inspections. Furthermore, the 

officials added that OSHA’s individual development plan process should 

help ensure that all inspectors and their managers identify the kind of 

training they need to maintain skills and expand expertise. A majority 

of area office officials we interviewed were encouraged by OSHA’s plan 

to offer inspectors opportunities to become professionally certified. 

Eight of the 17 area directors we interviewed stated that inspectors 

with professional certifications would appear more credible to 

employers and be better able to assist these employers to correct 

hazards. Over three-quarters of area office directors we spoke with 

said they saw value in promoting professional certifications among the 

inspector staff. Professional association officials we spoke with 

supported continuing education and/or certification training for OSHA 

inspectors, which they believed would enhance inspector quality.



However, OSHA’s training plans face obstacles that may jeopardize 

long-term success. First, OSHA’s official training directive fails to 

reflect OSHA’s new commitment to training. It states, “at a minimum, 

each [OSHA inspector] is required to attend a safety and health related 

course once every 3 years.” This directive is inconsistent with current 

training practices and the planned training for professional 

certification. Two of the area office directors we spoke with said that 

they could not rely on the directive as leverage to encourage those 

inspectors that may be unwilling to take needed training to improve 

their skills. Moreover, without having a formal directive that reflects 

what OSHA is trying to accomplish, area office directors expressed 

concern that current management focus and initiatives with regard to 

training could be lost in the event of a budget cut. This is especially 

important given that OSHA does not plan on finalizing its assessment of 

the level and type of resources that will be necessary to carry out 

this training until February 2003.



The second obstacle that may affect OSHA’s long-term success is the 

lack of a comprehensive database that tracks training or skills 

obtained by inspectors. With regard to training, OSHA’s local offices 

currently access or maintain 18 incompatible national, regional, and 

local (including formal and informal) databases for inspector training 

information. For example, 7 of the 9 regional offices we contacted each

maintain their own training database and 2 regional offices use a human 

capital database operated by the Department of Labor.[Footnote 20] Of 
the 

17 area offices we interviewed, 8 of them use the relevant regional 

database, while 4 area offices developed their own database system and 

5 used paper files. (See table 3 in app. I.) In addition, the OTI 

operates a separate database that tracks only the training that it 

provides. None of these databases track OSHA inspectors’ work-related 

skills. Area office directors we interviewed stated that OSHA often 

hires inspectors due to expertise in a particular area (e.g. crane 
safety) 

that they obtained prior to joining OSHA. This expertise is not 
reflected 

in any of these databases. As a result, OSHA managers would be unable 
to 

easily identify an inspector hired from the construction industry who 

has experience with crane safety developed from another job.



Because OSHA managers do not have reliable information on training and 

skills, they cannot readily identify inspectors with expertise in key 

hazards. In addition, OSHA managers seeking to determine whether the 

agency can meet certain future contingencies have no formal mechanism 

for identifying the skills of inspectors but must rely instead on 

personal knowledge or informal paper records held by individuals within 

the organization. Furthermore, we have reported that, without reliable 

data to assess the level and type of training and skills available, an 

agency cannot assess the extent to which training contributed to agency 

performance.[Footnote 21]



Conclusions:



OSHA has taken important steps toward targeting its enforcement 

resources on high-hazard worksites, measuring its impact, and enhancing 

the professionalism and quality of its personnel. However, OSHA’s 

enforcement efforts could be strengthened by better information and 

procedures that would make targeting efforts more efficient, 

measurement more precise, and training efforts more effective.



OSHA may not be getting the most out of its targeting programs because 

of data problems that limit OSHA’s ability to inspect high hazard 

worksites. For example, the current process for targeting construction 

worksites may not allow OSHA to systematically identify potentially 

more hazardous construction worksites. OSHA’s long-term research may 

ultimately address this problem. However, in the meantime, there is 

little assurance that area offices are inspecting smaller worksites. 

Data problems also limit the effectiveness of OSHA’s site-specific 

targeting program. Without improving the suitability of these data, 

OSHA will continue to expend significant resources on this program with 

little certainty that it is identifying the most hazardous worksites 

and making the best use of its inspection resources. Moreover, OSHA did 

not set up the site-specific targeting database in a manner that would 

allow it to evaluate the program’s impact on reducing injury and 

illness rates. As a result, it lacks sufficient information to 

determine whether program outcomes justify the resources expended.



The data and methodologies OSHA has used to measure its progress toward 

achieving strategic goals do not offer sufficient assurance that its 

efforts to measure its accomplishments produce a true picture of the 

agency’s impact on workplace safety and health or that they offer an 

appropriate assessment of agency progress in meeting its own goals. 

OSHA, as the federal agency responsible for overall workplace safety 

and health, understandably has an interest in tracking national trends 

in workplace injuries, illnesses, and fatalities. However, these data 

and methodologies may not reflect what is happening in those states or 

for those hazards that OSHA is responsible for under its strategic 

plan. As a result, OSHA lacks valuable management oversight information 

concerning the impact of those inspection activities for which it is 

most directly accountable.



Finally, OSHA has taken significant actions and plans to improve the 

quality of its inspection staff. However, unless area offices can share 

best practices regarding supervisory review, OSHA may not be able to 

ensure that the move to multidisciplinary teams does not adversely 

affect internal controls. Moreover, unless its training directive is 

updated to reflect OSHA’s current training strategy, the agency cannot 

ensure that its current strategy becomes institutionalized and 

implemented. And, OSHA currently lacks reliable information on the 

training and skills of its inspection workforce. This information is 

fundamental to improving the quality of OSHA’s workforce.



Recommendations for Executive Action:



To better ensure that OSHA gets the greatest benefit out of its 

targeting programs, we recommend that the Secretary of Labor direct 

OSHA to:



* encourage area offices to supplement inspections of large 

construction worksites with locally planned efforts to inspect smaller 

worksites,



* strengthen the validity of the data used to identify worksites in the 

site-specific targeting program by addressing the data weaknesses 

identified in this report, and:



* assess the site-specific targeting program’s impact on workplace 

injuries and illnesses in light of the resources expended.



To enhance OSHA’s ability to more precisely measure its impact from the 

strategic planning process, we recommend that the Secretary of Labor 

encourage OSHA and BLS to work together to obtain the necessary data to 

understand those injuries, illnesses, and fatalities occurring in areas 

covered by the strategic plan or under OSHA’s authority. This could 

include exploring additional ways of analyzing existing BLS data or 

exploring the costs of collecting additional information that would 

allow state-level estimates.



To help ensure that OSHA’s efforts to improve inspector quality achieve 

their potential outcomes, we recommend that the Secretary of Labor 

direct OSHA to:



* review area office efforts to develop alternative supervisory review 

procedures in order to identify promising practices and disseminate 

results to other offices,



* update OSHA’s training directive to reflect its current training 

strategy, and:



* work with Labor’s Office of the Assistant Secretary for 

Administration and Management to develop an information system to track 

and assess training and skills obtained by the inspection staff. This 

could include developing a new system or adapting existing systems.



Agency Comments and Our Evaluation:



We provided a draft of this report to Labor for comment. Overall, OSHA 

said that our report provides useful recommendations to consider as it 

moves forward in its efforts to improve the working conditions 

throughout the nation. The agency also pointed out a 30-year decline in 

occupational-related fatalities that took place despite huge increases 

in the U.S. workforce. Further, OSHA highlighted its belief that its 

enforcement system has achieved striking results, noting recent 

declines in injury and illness rates.



Although OSHA generally agreed to take action on the report’s 

recommendations, it expressed a number of concerns about material 

presented in the report. More specifically, OSHA raised questions about 

certain aspects of our analysis of the construction worksite targeting 

program. For example, it noted that we had recommended the use of local 

emphasis programs to help target small construction sites while 

acknowledging an absence of definitive data showing that smaller 

worksites are more hazardous than larger sites. We made this 

recommendation because (1) knowledgeable experts--including some at 

OSHA--told us that smaller sites were more hazardous and (2) the 

current construction targeting system does not adequately encompass 

these sites. We continue to believe that supplementing inspections of 

larger worksites with inspections of smaller ones is a prudent approach 

to take until OSHA completes its study of factors to help identify 

better ways to identify the most hazardous construction worksites.



OSHA also expressed several concerns about our analysis of the SST 

program. For example, it pointed out that, in apparent contrast to our 

findings, the agency’s quality control reviews indicated a high level 

of accuracy regarding the employer-submitted injury and illness data 

the program uses for targeting inspections. In our view, there is not 

necessarily an inconsistency between the quality review findings and 

ours. For example, the data may have been accurate at the time an 

employer submitted it to OSHA but out of date by the time it was used 

for targeting purposes. Additionally, OSHA noted that interventions, 

other than from the SST program (e.g., consultation visits), could have 

caused an employer to correct unsafe conditions and help explain the 

lack of citations during the SST visits. We agree. However, the fact 

remains that there are insufficient data to determine the validity of 

this explanation versus other possible explanations. In the meantime, 

the SST program--a targeting program intended to identify high-hazard 

worksites--continues to direct inspection resources to large numbers of 

sites that have no serious violations.



Finally, OSHA made several comments on our examination of the agency’s 

ability to demonstrate impact on workplace safety and health. For 

example, it reaffirmed its use of 1993 through 1995 data as a 

reasonable baseline from which to measure its strategic plan’s 

accomplishments, noting, among other things, that a baseline is by 

definition arbitrary. We continue to believe that, by selecting the 

baseline it did, OSHA took credit for declines that occurred before 

1997, the year when the strategic plan was implemented.



Labor’s comments in their entirety as well as our responses to their 

comments are shown in appendix II. Additionally, both OSHA and BLS 

offered technical comments, which we incorporated throughout the 

report.



We are sending copies of this report to the Secretary of Labor and the 

Assistant Secretary for Occupational Safety and Health. We will also 

make copies available to others upon request. In addition, the report 

will be available at no charge on the GAO Web site at http://

www.gao.gov. If you have any questions about this report, please call 

me on (202) 512-7215. Major contributors are listed in appendix III.



Sincerely yours,



Robert E. Robertson

Director, Education, Workforce and Income Security



Signed by Robert E. Robertson



[End of section]



Appendix I: Training Databases Used by OSHA’s Regional and Area 
Offices:



Table 3: Training Databases Used by OSHA Regional and Area Offices:



[See PDF for image]



Source: OSHA.



[End of table]



[End of section]



Appendix II: Comments from the Occupational Safety and Health 

Administration:



Note: GAO comments supplementing those in the report text appear at the 

end of this appendix.



This text file was formatted by the U.S. General Accounting Office 

(GAO) to be accessible to users with visual impairments, as part of a 

longer term project to improve GAO products’ accessibility. Every 

attempt has been made to maintain the structural and data integrity of 

the original printed product. Accessibility features, such as text 

descriptions of tables, consecutively numbered footnotes placed at the 

end of the file, and the text of agency comment letters, are provided 

but may not exactly duplicate the presentation or format of the printed 

version. The portable document format (PDF) file is an exact electronic 

replica of the printed version. We welcome your feedback. Please E-mail 

your comments regarding the contents or accessibility features of this 

document to Webmaster@gao.gov.



U.S. Department of Labor:



Assistant Secretary for Occupational Safety and Health, Washington, 
D.C. 

20210:



Mr. Robert E. Robertson:



Director, Education, Workforce, and Income Security Issues:



United States General Accounting Office, 441 G Street NW, Room 5930, 

Washington, DC 20548:



Dear Mr. Robertson:



Thank you for the opportunity to respond to your report entitled 

“Workplace Safety and Health: OSHA Can Strengthen Enforcement Through 

Improved Program Management.”:



The report provides useful recommendations for the Agency to consider 

as OSHA moves forward in our efforts to improve the working conditions 

throughout the Nation. We wish to emphasize, however, that OSHA’s 

primary goal is to reduce workplace fatalities, injuries, and illnesses 

- not merely to inspect worksites. The report should also acknowledge 

the demographics of the American workplace and how OSHA has chosen to 

target its enforcement actions, as well as the results those targeting 

efforts have produced. OSHA’s responsibilities extend to over 6 million 

establishments, including 700,000 construction worksites, employing 

more than 120 million people. To cover this workforce, federal OSHA has 

fewer than 1,200 inspectors, supplemented by inspectors from the 23 

states that operate their own OSHA programs.



Despite the huge increases in the U.S. workforce over the past 30 

years, the number of occupational-related fatalities continues to 

decline. Last year (excluding those fatalities attributed to the events 

of September 11, 2001), fewer than 6,000 workers were killed on the job 

and over 40% of those were related to traffic accidents, events over 

which OSHA has limited responsibility. OSHA’s enforcement system-which 

relies on a combination of site-specific targeting of establishments, 

National and Local Emphasis Programs, and local area office 

construction prioritization-has achieved striking results. 

Specifically, from 1992 to 2000, the nations total injury/ illness rate 

has decreased by 31.5%, from 8.9 to 6.1 cases per 100 workers. However, 

in the cover letter to Congress, GAO aggregates all the fatalities over 

the last 21 years to state that there were over 100,000 worker deaths 

since 1980, overlooking the dramatic decreases that have occurred in 

both fatality and injury/illness rates.



Even with the dramatic reduction, we acknowledge that the targeting 

system can be improved. Therefore, in Fiscal Year 2003, the Agency 

plans to evaluate the Site Specific Targeting Program and its Special 

Emphasis Programs.



However, there are some statements in GAO’s report that OSHA believes 

need to be corrected. For example, GAO claims the Agency does not 

effectively identify worksites for inspections, asserting OSHA’s 

databases do not permit accurate identification of smaller worksites 

for inspection. GAO’s recommendations for improving the national 

Construction Targeting System are based on the presumption that smaller 

construction sites are more hazardous than larger sites. Yet, GAO 

acknowledges that there is no conclusive data to demonstrate this and 

supports the concept of the Construction Project Survey that will allow 

OSHA to rate hazards by project size. GAO further states that prior to 

obtaining the results of the survey, OSHA should rely on Local Emphasis 

Programs (LEPs) and “informal criteria” used by area offices to 

identify and target smaller construction projects.



Bureau of Labor Statistics (BLS) data pertaining to construction 

employer size show that medium-size employers experience the highest 

rates of injuries and illnesses, not the smallest employers (a lost 

workday case incidence rate of 4.8 for 50-249 employees versus 2.8 for 

1-10 employees). The BLS quartile data also show that more than 75% of 

the smallest construction employers experience no injuries or illness 

during a given year.



Also, in its analysis of the site-specific targeting (SST) program, GAO 

analyzed IMIS data and equated the number of serious violations cited 

to the relative hazardousness of the establishment. GAO’s analysis did 

not control for OSHA interventions other than the SST inspection that 

may have contributed to the removal of hazards. Worksites with high 

injury/ illness rates may experience multiple interventions from OSHA 

(receipt of high injury/ illness letters, consultation visits, other 

inspections, including complaint or fatality). We believe these other 

interventions would cause an employer to correct unsafe conditions and 

help explain the lack of OSHA citations during SST visits.



GAO also claims OSHA acts on faulty and inaccurate employer injury and 

illness information. However, OSHA performs ongoing evaluations of the 

quality of employer injury and illness records in its Audit and 

Verification Program of Occupational Injury and Illness Records. This 

program has been in place for five years and will continue. The audit 

program is designed to estimate the percent of establishments that 

maintain accurate records (at-or-above the 95 percent confidence 

threshold). In all five years of the program, the audit findings have 

shown that approximately 90 percent of employers accurately record 

injuries and illnesses for both total recordable and lost workday 

injury and illness cases. Based on the audit findings, OSHA has 

concluded and reported to OMB that the accuracy of employer injury and 

illness recordkeeping represents reasonable quality for OSHA’s 

targeting and performance measurement purposes. In fact, GAO, in an 

earlier report entitled, “Program Evaluation: Studies Helped Agencies

Measure or Explain Program Performance,” cited the recordkeeping audit 

program as an example of an excellent quality control program.



Furthermore, GAO misunderstood the constraints placed on OSHA’s 

targeted inspection system by law. According to the Supreme Court 

decision, Marshall v. Barlow, Inc., 436 US 307 (1978), OSHA must have 

either a reasonable belief that cause for inspection exists (e.g., a 

fatality, a complaint, a referral) or an unbiased administrative plan 

for selecting construction work sites for inspection. The key element 

of an acceptable plan under law is a neutral and objective selection 

system, not a capricious and arbitrary procedure. In the late 1980’s, 

the Agency, after considerable research and study, decided to implement 

a system of random, neutral, computergenerated selections from among 

the universe of active construction sites contained in the F.W. Dodge 

Reports (the most complete listing of active construction work sites 

known to the Agency) to be performed by an objective third party.



The draft report also inadequately distinguishes general industry from 

construction, obscuring differences in how work is done in these two 

major industry subdivisions and detracting from the report’s usefulness 

as a guide for future efforts to improve inspection targeting. 

Methodology and data sources OSHA uses for targeting inspections of 

fixed workplaces, where working conditions change relatively little 

from year to year, are far less useful in construction.



The majority of construction projects are completed in less than a 

year. And, also, there are no requirements for a project log. Moreover, 

the employment relationship is more fluid in construction, where 

workforce turnover, the roster of active subcontractors on a particular 

project, and changes in the corporate identity of employers make it 

much more difficult to identify unsafe construction establishments in 

advance.



We recognize, as does GAO, that the Dodge System may not be the most 

effective way to identify dangerous worksites. The Agency has 

investigated alternatives to the Dodge Report. We are considering 

possible alternative schemes for construction that seek to identify 

problem worksites based on characteristics such as the size and cost of 

particular projects, the construction methods and materials used, the 

locality of the project, and other such variables. OSHA’s Construction 

Project Survey is intended as a first step in developing such an 

alternative.



GAO also reports that OSHA misstates its accomplishments by using the 

wrong baseline for measurements. Yet, in this case, it is impossible 

for OSHA to avoid setting an arbitrary baseline. Given that the 

baseline is by definition arbitrary, and keeping in mind that 1995 

injury/illness data are not available until December 1996 (i.e., Fiscal 

Year 1997), it was reasonable to choose a time period immediately 

before implementation of the strategic plan. Moreover, the three-year-

average baseline was chosen for the valid statistical purpose of 
avoiding 

year-to-year fluctuations.



Injury/ illness rates did not decline consistently in the chosen 

industries between 19931995 and actually increased or stayed the same 

in 1997 compared to 1996, for all industries except shipyards. Finally, 

GAO seems to conclude that the decline between 1993-1995 would have 

continued even without OSHA’s strategic plan efforts. This assumes that 

the future will be like the past, which is a common assumption, but one 

which is often incorrect. It is impossible to say what would have 

happened without OSHA’s strategic plan efforts.



GAO also states in its report that OSHA takes credit for reductions in 

injury and illness rates without acknowledging that its efforts were 

just one of many factors that affect workplace safety. This statement 

is inaccurate. OSHA’s annual reports have never claimed that OSHA’s 

efforts alone caused the reductions. Instead, OSHA has also given 

credit to employers and workers for the decline.



RECOMMENDATIONS:



Response to Recommendations:



1.Encourage all area offices to supplement inspections of large 

construction with locally planned efforts to inspect smaller worksites.



OSHA agrees with this recommendation. The mechanism and procedures 

already exist and occur for area offices to supplement the programmed 

inspection targeting lists provided by the National Office with 

worksites identified through an approved local emphasis program (LEPs). 

LEPs must be implemented in a manner consistent with the neutral 

administrative selection of worksites required by the Supreme Court’s 

Barlow decision. A new Construction Targeting Task force, initiated 

last spring, will give guidance to area offices on how to better use 

the Dodge Report and to ensure that LEPs are consistent with the Barlow 

decision. In the new strategic plan (2003-2008), OSHA plans to include 

approaches that emphasize local efforts to inspect smaller worksites. 

The plan is expected to be published in March 2003.



2.Strengthen the validity of the data used to identify worksites in the 

site-specific targeting program by addressing the data weaknesses 

identified in this report.



OSHA systematically addresses the validity of the data collected for 

targeting purposes. The collection process has quality controls. Audits 

are conducted to ascertain the accuracy of employer recordkeeping. 

However, OSHA plans to evaluate the effectiveness of the SST in FY 

2003. As part of that evaluation, OSHA will examine data validity 

issues raised in the GAO report.



3.Assess the site-specific targeting program’s impact on workplace 

injuries and illnesses in light of the resources expended.



OSHA recognizes the importance of evaluating the effectiveness of its 

enforcement programs and plans to evaluate and consider changes to the 

SST program in Fiscal Year 2003.



4.To enhance OSHA’s ability to more precisely measure its impact from 

the strategic planning process, we recommend that the Secretary of 

Labor encourage OSHA and BLS to work together to obtain the necessary 

data to better understand those injuries, illnesses, and fatalities 

occurring in areas covered by the strategic plan or under OSHA’s 

authority. This could include exploring additional ways of analyzing 

existing BLS data to derive estimates for federal OSHA states, or 

exploring the costs for collecting additional information that would 

allow state-level estimates.



OSHA has discussed these ideas with BLS. Improving the data would 

require BLS to redesign the Annual Survey sample at considerable cost. 

In addition, lead-time would be required to implement changes. BLS 

officials have stated that the current BLS industry estimation software 

for the injury and illness survey was not designed to produce estimates 

for State aggregates below the National level. A major software 

development effort would be required to modify the industry estimation 

system to support industry estimates for State aggregates. OSHA will, 

however, work with BLS to identify precise costs for producing state 

estimates in order to determine the costeffectiveness of such action.



5.Update OSHA’s training directive to reflect its current training 

strategy.



OSHA plans to update its training directive to reflect its current 

training strategy after the details of its new training strategy are 

developed, staffed and approved. The Agency is currently analyzing its 

safety specialist, health specialist and safety engineer positions to 

identify specific job competencies. These will be merged with 

competencies required for professional certification and clustered into 

courses for new inspectors. Additionally, these new courses will be 

delivered by traditional classroom training, satellite delivered 

training, and by web-based training technologies. Guidance will be 

established for specific areas such as workplace violence, ergonomics, 

silica, lead, construction and other strategic plan and high hazard 

areas that will allow inspectors to share inspection practices and 

experiences across the Agency. OSHA anticipates this analysis and 

revision will be completed in FY 2003, at which time the training 

directive will be updated.



6.Develop an information system to track and assess training and skills 

obtained by the inspection staff. This could include developing a new 

system or adapting existing systems.



OSHA recommends this be written to have the Secretary of Labor direct 

the Office of the Assistant Secretary for Administration and Management 

(OASAM) at the Department of Labor to develop an information system to 

assess training and skills obtained by Departmental personnel. Other 

organizations such as MSHA and OFCCP also have inspectors and would 

benefit from such an information system.



Thank you for the opportunity to comment on your draft report. If you 

have questions, please contact Frank Frodyma, Acting Director for the 

Directorate of Evaluation and Analysis, on (202) 693-2400.



Sincerely,



John L. Henshaw



[Signed by John L. Henshaw]



The following are GAO’s comments on the OSHA letter dated November 14, 

2002.



GAO Comments:



1. We clarified the language in the report to better highlight recent 

improvements in work-related fatality and injury rates.



2. Our draft report acknowledged there are no definitive data showing 

that smaller worksites are more hazardous than larger worksites. 

However, knowledgeable experts--including some at OSHA--believe that 

the smaller sites are indeed more dangerous than the larger ones. 

Furthermore, OSHA’s current construction targeting efforts do not 

include these smaller sites. Accordingly, we continue to believe that 

supplementing inspections of larger worksites with inspections of 

smaller ones is a prudent approach to take until OSHA completes its 

study of factors that will help it better identify the most hazardous 

construction worksites.



3. While OSHA provided data to show that medium-sized employers may be 

more hazardous than smaller ones, employer size and construction 

worksite size are two different measures. As a result, these statistics 

may not reflect the level of hazards at small construction sites.



4. We believe that using the rate of serious violations cited is a 

valid measure of a worksite’s hazardousness, although we acknowledge it 

is not the only one. This approach has been used by researchers and 

OSHA itself to identify whether OSHA is focusing its inspection 

resources in the right places. Additionally, we agree with OSHA that 

interventions, other than from the site-specific targeting (SST) 

program (e.g., consultation visits), could have caused an employer to 

correct unsafe conditions and help explain the lack of citations during 

the SST visits. However, the fact remains that there are insufficient 

data to determine the validity of this explanation versus other 

possible explanations. In the meantime, the SST program--a targeting 

program intended to identify high-hazard workplaces--continues to 

direct inspection resources to large numbers of sites that have no 

serious violations.



5. We did not assess the results of OSHA’s efforts to verify the 

accuracy of employer-submitted injury and illness data. However, even 

if the data are as accurate as OSHA suggests, our report points out 

other data limitations that hinder its usefulness in targeting 

inspections. For example, the data collected only reflects what 

happened during a single year at a particular employer and that data 

may have been an outlier that did not reflect general worksite 

operations. Additionally, the data may not be current--it may be 2 

years old before an SST inspection is conducted.



6. Our earlier report, U.S. General Accounting Office, Program 

Evaluation: Studies Helped Agencies Measure or Explain Program 

Performance, GAO/GGD-00-204 (Washington, D.C.: Sept. 29, 2000) did not 

independently assess the quality of OSHA’s recordkeeping audit program. 

Instead, we discussed the role of the program in helping OSHA evaluate 

its performance.



7. We anticipate that, when implementing our recommendation to 

encourage local offices to supplement inspections of larger 

construction worksites with inspections of smaller construction 

worksites, OSHA will ensure that these inspections are conducted in 

accordance with all legal constraints.



8. We believe that the report adequately distinguished general industry 

from construction targeting. We reported that the SST program to date 

has focused on general industry worksites while the Dodge system 

focuses on construction worksites. Nonetheless, we do believe they both 

have data limitations that affect their ability to effectively identify 

hazardous worksites.



9. We understand, based on OSHA’s comments, that more current data 

would have been available at the time when OSHA began to evaluate 

progress toward the plan’s strategic goals. We continue to believe that 

OSHA’s selection of a 1993-95 baseline allowed it to take credit for 

declines that occurred prior to the implementation of the strategic 

plan. As we note in the report, while OSHA may have had some effect on 

these changes, the changes cannot be seen as an indication of the 

plan’s success.



10. Our report did not conclude that declines between 1993 and 1995 

would have continued without OSHA’s efforts. We agree that it would be 

difficult to determine what would have happened in the absence of 

OSHA’s strategic plan efforts.



11. We changed the language in our report to clarify that OSHA, in 

presenting its evaluation of its progress toward strategic goals, did 

not account for the influence of other factors that affect workplace 

safety and health.



12. We altered the language of the recommendation to include Labor’s 

Office of the Assistant Secretary for Administration and Management. We 

believe that both OSHA and the Assistant Secretary’s office have a role 

in developing the kind of data system necessary to accurately track and 

assess inspectors’ training and skills.



[End of section]



Appendix III: GAO Contacts and Staff Acknowledgments:



GAO Contacts:



Lori Rectanus (202) 512-9847

Joseph Natalicchio (202) 512-5897:



Staff Acknowledgments:



Dennis Gehley, Kris Trueblood, H. Brandon Haller, and Catherine Hurley 

made significant contributions to this report. In addition, Richard 

Burkard and Julian Klazkin provided legal support, while Patrick 

DiBattista, Barbara W. Alsip, and Susan Bernstein provided writing 

assistance.



FOOTNOTES



[1] Both “federal OSHA states” and “state plan states” include 

jurisdictions, such as the District of Columbia and Puerto Rico. For 

this review, we included Connecticut, New Jersey, and New York as 

federal OSHA states. Although these states have been delegated 

authority for public sector workers, OSHA maintains responsibility for 

private sector workers in these states.



[2] Throughout this report, we use nonfatal injury and illness “rates,” 

which represent the number of work-related nonfatal injuries and 

illnesses per 100 workers. 



[3] OSHA has not yet developed an overall human capital strategic plan 

that these efforts would fit into. OSHA’s overall human capital 

planning efforts were outside the scope of our review.



[4] We did not contact OSHA’s regional office in San Francisco because 

it has only state-plan states within its jurisdiction. State-plan 

states are not governed by OSHA’s human capital policies and 

procedures. We selected the 17 area offices to reflect a mix of 

characteristics, including size of office, the level of employment 

within the geographic area covered by the office, the average length of 

service time for the staff in each office, and whether the office 

obtained a best practices designation from OSHA. 



[5] 29 U.S.C. 651(b).



[6] These include safety inspectors, who seek to prevent workplace-

related accidents resulting from unsafe machinery or procedures; and 

health inspectors, who seek to prevent illnesses resulting from toxic 

exposures. 



[7] 29 U.S.C. 667. Delegations of authority occur only after a state or 

territory submits a plan that is determined to be consistent with the 

requirements of this section. 



[8] See our January 31, 2000, report, U.S. General Accounting Office, 

Occupational Safety and Health: Federal Agencies Identified as 

Promoting Workplace Safety and Health, GAO/HEHS-00-45R (Washington, 

D.C.: Jan. 31, 2000), which states that OSHA works with 14 other 

federal agencies that have limited jurisdiction over workplace safety 

and health.



[9] These efforts were primarily driven by the enactment of the 

Government Performance and Results Act of 1993 (GPRA). OSHA reported 

its goals for 1997-2002 separately but reported on their attainment as 

part of the overall efforts of the Department of Labor.



[10] OSHA’s strategic plan includes a variety of goals, many of which 

specify the activities that the agency will pursue.



[11] For purposes of this report, the term “safety inspector” refers to 

both OSHA safety inspectors and safety engineers. Safety engineers have 

more education than safety inspectors. 



[12] As shown previously in table 2, through OSHA’s ODI process, OSHA 

obtains information on approximately 140,000 worksites in relatively 

hazardous industries, obtains injury and illness data from about 80,000 

of them, and narrows down that list to the 3,000 it believes are the 

most hazardous.



[13] OSHA generally does not contact employers to give them advance 

notice of the specific date when an inspection will occur. Because 

inspectors do not contact employers, inspectors may not have the 

opportunity to know in advance that the visit may not be worthwhile. 

Also, according to OSHA management officials, in order to ensure that 

every worksite identified through this program is inspected, OSHA 

headquarters sends the names of all identified worksites to area 

offices even if the injury and illness rate appears to be incorrect. 

OSHA withholds information on such worksites only if the rate is out of 

meaningful range. 



[14] Through this program, OSHA provides assistance to small employers 

in hazardous industries who voluntarily request a consultation. See our 

October 12, 2001, report, U.S. General Accounting Office, Workplace 

Safety and Health: OSHA Should Strengthen the Management of Its 

Consultation Program, GAO-02-60 (Washington, D.C., Oct. 12, 2001). OSHA 

procedures do not permit state agencies operating consultation programs 

to share information on participating employers with OSHA inspectors. 



[15] This situation was most prevalent in the nursing home industry 

because of OSHA’s difficulties in promulgating an ergonomics standard 

to address a series of musculoskeletal hazards. OSHA targeted nursing 

homes, whose workers frequently experience ergonomic-related injuries, 

with the expectation that it would have a standard for use by 

inspectors in citing these hazards. Before the inspections could be 

conducted, however, OSHA’s ergonomics standard was legislatively 

overturned. Inspectors visiting nursing homes had no standard for 

citing ergonomics hazards and were discouraged by OSHA from using 

OSHA’s general duty clause to cite violations. Under the Occupational 

Safety and Health Act’s general duty clause, OSHA has cited employers 

for recognized hazards for which the agency has no specific standard. 

See 29 U.S.C. 645. According to OSHA officials, the agency plans to 

remove nursing homes from future targeting efforts.



[16] OSHA’s IMIS system permits individual inspections to be coded as 

representing more than one OSHA initiative (i.e., SST, construction, 

LEP, and NEP). In deriving this statistic, we identified all 

inspections coded as LEP, even if they were also coded for other 

initiatives. 



[17] In comments on this report, OSHA noted that there are numerous 

interventions that may contribute to the removal of hazards, such as 

inspections in response to a worker complaint.



[18] Injury and illness data are collected through a survey. As a 

result, there are sampling errors associated with estimates of injury 

and illness rates. In some cases, differences in injury and illness 

rates are not large enough to determine whether or not changes in 

injury and illness rates were real or due to sampling errors. In such 

cases, changes are not statistically significant. OSHA performed 

statistical significance evaluations of its trend data in response to a 

draft of this report.



[19] Both GAO and OSHA analyses calculated statistical significance at 

a 95 percent confidence level.



[20] The Department of Labor’s “People Power” database, which is 

maintained by the Office of the Assistant Secretary for Administration 

and Management, contains a broad range of information, including 

training, on its employees and serves as a system of record for all 

personnel actions. Of the 26 regional and area offices we interviewed, 

only 2 regional offices used the People Power system. Regional and area 

office staff generally do not have the special training needed to fully 

use this system. 



[21] See our March 15, 2002, exposure draft, U.S. General Accounting 

Office, A Model of Strategic Human Capital Management, GAO-02-373SP 

(Washington, D.C.: Mar. 15, 2002).



GAO’s Mission:



The General Accounting Office, the 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.



Obtaining Copies of GAO Reports and Testimony:



The fastest and easiest way to obtain copies of GAO documents at no 

cost is through the Internet. GAO’s Web site ( www.gao.gov ) contains 

abstracts and full-text files of current reports and testimony and an 

expanding archive of older products. The Web site features a search 

engine to help you locate documents using key words and phrases. You 

can print these documents in their entirety, including charts and other 

graphics.



Each day, GAO issues a list of newly released reports, testimony, and 

correspondence. GAO posts this list, known as “Today’s Reports,” on its 

Web site daily. The list contains links to the full-text document 

files. To have GAO e-mail this list to you every afternoon, go to 

www.gao.gov and select “Subscribe to daily E-mail alert for newly 

released products” under the GAO Reports heading.



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. General Accounting 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, Waste, and Abuse in Federal Programs:



Contact:



Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov



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



Public Affairs:



Jeff Nelligan, managing director, NelliganJ@gao.gov (202) 512-4800 U.S.



General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C.



20548: