STATEMENT OF JOSEPH A. DEAR
ASSISTANT SECRETARY OF LABOR
FOR OCCUPATIONAL SAFETY AND HEALTH
FOR THE HEARING OF THE HOUSE SCIENCE COMMITTEE
ON RISK ASSESSMENT AND TITLE III OF H.R. 9

FEBRUARY 3, 1995

Thank you for this opportunity to make a statement on H.R. 9, the Job Creation and Wage Enhancement Act, and particularly on Title III, which deals with risk assessment and risk characterization as these pertain to regulatory programs. The Occupational Safety and Health Administration was established in 1970 to assure safe and healthful working conditions for working men and women. The Agency's mission is to save lives, prevent workplace injury and illness, and protect the health of America's workers. This mission continues to be compelling today because, despite considerable progress, there are more than 6,000 fatal workplace injuries, tens of thousands of fatal work-related illnesses, and millions of non-fatal illnesses and injuries every year. The major tools OSHA uses to address this needless waste of human resources include the development and enforcement of regulations and the provision of consultation and technical assistance to employers and employees.

OSHA believes that risk analysis is a necessary and appropriate tool for linking sound policy decisions with sound science. The Agency has nearly 15 years of experience in conducting risk analyses to support occupational safety and health regulations. The result of OSHA's approach to risk assessment has been the development of a body of regulations that address and reduce significant risks and are both technologically feasible and economically justifiable. OSHA regulations, such as those for asbestos, vinyl chloride, cotton dust, cadmium, bloodborne pathogens, confined space entry, lockout/tagout, and fall protection, have saved thousands of lives and even more job-related injuries and illnesses.

OSHA's principal concern with Title III of H.R. 9 is that it would be harmful to America's working men and women because it would create a procedural obstacle course for OSHA's risk assessment process and would prevent the more flexible case-by-case approach now in use. Far example, H.R. 9 describes a heavily layered approach to risk assessment that outlines the various levels of peer review and economic analysis that must be conducted. However, it is OSHA's experience that in many cases such complex procedures are not required and would, in fact, be costly and counter-productive. For example, OSHA often conducts simplified rulemakings to streamline or update standards that were issued more than twenty years ago. OSHA performed a qualitative analysis of the comparative risks of non-asbestiform actinolite, tremolite, and anthophyllite and concluded that a less stringent permissible exposure limit was appropriate for these substances, thus lightening the regulatory burden on the industry which produced and used them. If OSHA had been required to conduct formalized peer review of the risk assessments involving these substances it would have been much more time-consuming and difficult to lift the burden on industry.

Much of the risk assessment debate focuses on risks at the margins (i.e., on reducing risks with regulations that impose major regulatory burdens to achieve minimal reductions in risk). However, OSHA routinely regulates risks that pose clear and evident dangers to American workers. A complicated risk assessment is not needed to determine that hundreds of U.S. workers lose their lives every year as a result of falls. On the health side, OSHA regulations often address risks that threaten the lives of as many as 1 in 100 exposed workers. For example, OSHA estimates the risk associated with a working lifetime exposure to ethylene oxide at the limit existing prior to OSHA regulation ranged from 63.4 to 109.3 excess cancer deaths per 1000 exposed workers. These risks would not be considered marginal by any responsible policy maker.

OSHA supports most of the goals of H.R. 9. We believe, as do the authors of the bill, that regulations should be based on a realistic consideration of risk, and that regulatory priorities must be based on scientifically sound and unbiased risk assessments. We share your goals of reducing paperwork burdens on employers. However, after careful analysis of the bill and the manner in which various sections of the bill will interact with one another, we have concluded that H.R. 9 will undermine the bill's stated intent. The bill will add multiple bureaucratic layers, increase costs, decrease flexibility, inhibit consensus building, and create delays. These unintended consequences will also impose a human cost in terms of increased numbers of preventable occupational fatalities, injuries, and illnesses. For example, OSHA estimates that, if the Agency's recent Process Safety Management standard had been delayed by even one year, an estimated 132 fatalities and 767 injuries would have occurred.

OSHA's risk analyses routinely involve many of the elements that would be required by H.R. 9, such as evaluations of some comparative and some substitution risks, presentations of alternative analytical models and assumptions, and descriptions of the range of uncertainty bounding its best estimates of risk. Currently, OSHA can apply these techniques to the specific circumstances of the particular risk at issue. H.R. 9 would force the Agency to use a "one-size-fits-all" approach to risk assessment in every case. It would prohibit OSHA from tailoring its analysis to fit the particular analytic, scientific, and policy challenges posed by job-related problems that vary from chemical carcinogens, to fire and explosion hazards, infectious diseases to workplace violence, and reproductive toxins. H.R. 9 will layer analysis upon analysis, force or encourage endless loops of review and revision, and effectively prevent interested parties from bringing issues to closure. It would make the most complex, lengthy and costly model of risk assessment the required option in every situation, even in those circumstances where a simple and inexpensive analysis would suffice. In addition, a "one-size-fits-all" approach may have undesirable effects by discounting the importance of risks that are high but for which there is relatively more scientific uncertainty than for other, lower risks. Using such a rigid approach, risk assessment may lead to the unwise concentration of resources in a few areas.

One unintended consequence of the requirements for risk assessment in Title III would be demands for greater amounts of data from the private sector. OSHA would have to obtain more data from employers, trade associations, the insurance industry and other sources in order to meet the more stringent risk assessment requirements of the bill.

Every regulatory action taken by OSHA to address significant occupational safety and health risks and their prioritization would be subject to the requirements of H.R. 9. OSHA notes that Title III is captioned, "Risk Assessment and Cost/Benefit Analysis for New Regulations" and therefore does not appear to apply to non-regulatory agency activity such as enforcement or consultation. If Title III were applicable to enforcement or consultation activity, its requirements might apply to hundreds of additional documents such as inspection reports, abatement agreements, consultation reports, and compliance directives that may be interpreted to be risk assessments or risk characterizations under H.R. 9.

OSHA's risk assessments vary in complexity depending upon the nature of the risk to be addressed. The risk assessment for a safety hazard, such as fatal falls from roofs, differs substantially from those for toxins, such as cadmium, which can cause many forms of disease. In developing risk assessments for toxic substances, hazardous physical agents, and safety hazards, OSHA follows established scientific principles and nationally recognized guidelines, such as those of the National Academy of Sciences. For health standards, OSHA also carefully explains and justifies its choice of risk assessment models and discusses the weight of the evidence in a comprehensive manner. For safety standards, the Agency describes all relevant injury and fatality data and any other information relevant to the assessment of risk. OSHA takes these steps to ensure that its risk assessments and risk characterizations are as clear and understandable as possible. OSHA invites comment on all aspects of its risk assessments at the proposal, public hearing; and final rule stages of standards development. By prescribing what assessments must be done and dictating how they must be conducted, OSHA'S ability to perform simple risk assessment will be replaced by the requirement to conduct complex and costly analyses in every case.

The proposed bill is also filled with vague language that will invite litigation and lead to subsequent gridlock. This trend will be accelerated by the many apparent inconsistencies in the bill (e.g., Title III's requirement that benefits, justify costs for major rules and Title VII's requirement that benefits outweigh costs for rules affecting more than 100 people), which will force litigants to go to court to obtain interpretations and resolve a host of technical issues.

The numerous procedural steps and analytical factors that H.R. 9 would add to OSHA rulemaking would greatly increase the scope of judicial review. Extending judicial review in these ways is likely to lead to countless lawsuits and will make each standard and its supporting analysis vulnerable to invalidation on grounds that are not necessarily reflective of the standard's essential efficacy or feasibility. In addition, Title VIII creates a new private cause of action against agencies and agency employees based on claims that the agency or employee took action against a private person because the private person disclosed certain information.

The exposure scenario requirement in Title III appears to require an analysis of the risk posed to every separately identifiable population of workers. If this is the case, OSHA would be required to conduct extensive employer surveys to obtain industry-specific employment turnover data. This requirement alone would add an average of $1 million (the approximate cost of a 5,000 establishment multi-sector survey) to the cost of each rulemaking. It would also mean a minimum of one year delay while OSHA analyzed and tabulated survey results. For the rulemaking on cadmium this means that 17 workers exposed to the substance would have lost their lives and another 78 would have developed progressive kidney disease.

The adverse impact of the bill on worker safety and health would be compounded at many steps in the risk analysis process. Title III would require OSHA to "include a statement of any significant substitution risks" for every chemical that could potentially be substituted for the chemical being regulated. If this provision is interpreted to require full blown risk assessment for these substitutes, OSHA's burden would be enormous; this could substantially delay important protection for workers if the provision is interpreted to require risk assessments for all substitute chemicals. OSHA can and does conduct risk assessment of substitutes when appropriate. However, as we have interpreted the substitution requirements of Title III, these requirements alone would have delayed six proposed rules published by OSHA in the last few years by an average of four years for each regulation. This would have led to the deaths of 377 workers whose lives could have been saved by the regulations. In addition, in order to issue the standard for glycol ethers proposed by OSHA in 1994, the agency would have been required to conduct 21 additional risk assessments, which would have led to a delay of 7 years. During the time that OSHA would have been gridlocked in an endless cycle of risk assessments, between 1,800-7,800 male workers would have experienced adverse reproductive effects and between 14-87 working women would have given birth to children with developmental abnormalities.

H.R. 9's requirements for formal peer review panels and detailed peer review procedures would mean considerable delay and added expense to complete a regulation. If applied to the six regulations recently proposed by OSHA, the peer review provisions alone would have caused a delay of approximately 15 months in issuing each one of these rules. During that time approximately 539 workers whose lives could have been saved by these rules would die and an even greater number would become injured or sick from on-the-job hazards.

In many cases, H.R. 9's requirements for formal peer review would be unnecessary and wasteful because the procedures for estimating risk are uncomplicated. Requiring formalized peer review of the risk assessments for these kinds of situations would cause extensive debate and polarization around minor issues. OSHA would be hindered from responding to changing workplace conditions and, in some cases, reducing the regulatory burden on industry. For example, industry and many members of Congress have urged OSHA to issue a regulation on training for those who operate powered industrial trucks. A Resolution signed by 272 members of the House of Representatives last year urged that this rule be issued. The provisions for peer review and other risk assessment requirements in Title III would delay this rule by 15 months at a projected cost of 26 workers' lives and 9,582 lost workday injuries.

Subtitle C of Title III is written in mandatory terms. If OSHA issued a rule without complying completely with the requirements for peer review, the rule could be overturned by a reviewing court. The peer review panels would probably be considered advisory committees under section 3 of the Federal Advisory Committee Act and would have to be chartered. The Administration has sought to reduce the number of Federal advisory committees as part of its reinvention effort to streamline government. Congress has supported the reinvention effort. H.R. 9 would contravene the reinvention goals by creating the need for more advisory committees.

Before issuing any safety or health regulation, OSHA must demonstrate that the regulation addresses a significant risk, that the regulation will substantially reduce that risk, and that it will do so in a feasible and cost-effective manner. The analyses OSHA conducts to demonstrate that the regulation meets these tests include detailed examinations of regulatory costs, economic impacts, and benefits.

OSHA routinely evaluates the costs and benefits of its regulations before taking final action. However, if H.R. 9's requirements for strict cost/benefit analysis were applied to OSHA, they would adversely affect OSHA's ability to issue regulations protecting workers. H.R. 9 would require a mathematical balancing of the costs and benefits of each regulation. OSHA opposes such an approach because it will result in regulations that are not fully protective of workers. The use of "average" estimates of risk required by H.R. 9 would compound the problem. Reliance on an "average" estimate of risk and a strict cost/benefit equation would have meant that OSHA could not have reduced the risk of sterility and impotence in working men exposed to the herbicide dibromochloropropane in a timely fashion. No definitive risk assessment methodology exists to estimate reproductive risks and no definitive method is currently available to quantify and monetize the benefits of these frightening effects. The need to regulate on the basis of a certified cost/benefit analysis could also prevent OSHA from addressing emerging but difficult-to-quantify risks such as violence in the workplace and infectious tuberculosis.

H.R. 9's strict requirement for cost/benefit analysis, unlike the corresponding requirements contained in E.O. 12866, would preclude OSHA from giving adequate consideration to factors such as equity (should workers have to accept greater risks than the general public because they need a paycheck?) in developing regulations. Issues such as the need to develop human resources to enhance the nation's future competitiveness, avoidance of premature mortality, and the reductions in quality of life from diseases such as cancer would all be ignored. If H.R. 9's cost/benefit requirements become law, OSHA could be forced to overlook particularly vulnerable subpopulations of workers, such as women of childbearing age, or to value younger workers more highly than older ones.

Other provisions of H.R. 9 interact with the provisions on risk assessment and cost/benefit analysis to make it more difficult for OSHA to protect the lives of working men and women. For instance, while compliance with Title III would demand increased use of surveys and other data-gathering techniques, Title V would reduce the information collection efforts of the Federal government. It would even require agencies who want to gather information by survey to receive public comment on their request. Agencies would be placed in a "Catch 22" posture of needing more information to justify their risk estimates while being forced to overcome formidable procedural hurdles to gather such information.

Title VII would further complicate rulemaking efforts by requiring OMB to certify that all rules and impact analyses have met certain standards of clarity, grammar, organization and syntax. OSHA supports clarity in regulations and is committed to writing standards that are easy to understand and comply with. However, under H.R. 9, OMB would, in effect, become the chief grammarian for the Federal Government but would have no clear standards upon which to base its critique of the agencies' work. Regulations designed to save human lives could be delayed for months while Federal employees debate fine points of grammar and syntax.

The most important criterion which OSHA used in evaluating H.R. 9 is this: Will the bill help or hinder the agency in protecting the working men and women of this Nation in a feasible and cost effective manner. For the reasons that I have described in this statement we believe that the legislation would not be helpful to OSHA in accomplishing the mission established by the Congress more than twenty-four years ago.