Excerpts of Administration’s Proposal to Relax Toxin Rules

by Robert Lewis - July 24, 2008 12:41 pm EDT
Tags: Department of Labor, OSHA, Workplace Safety

Yesterday, we reported about a proposed Department of Labor rule that would make it harder to regulate toxins and chemicals in the workplace. As the Washington Post noted, the proposed rule did not go through the normal process: It was not publicly disclosed and was drafted by a "political deputy in Labor's office of the assistant secretary for policy."

(A Department of Labor spokesman, yesterday, defended the timing of the rule saying "the issue was ripe," and that there would be plenty of time for public comment if and when the administration moves forward with the rule.)

We have been given a copy of the draft rule. The source has asked us not to publish the full document for fear it could identify the person who leaked it. But here are the key excerpts:

Risk assessors must make many assumptions when conducting their assessments. Often assessors will make a more conservative assumption in order to err on the side of safety.

When multiple conservative assumptions are made in the same risk assessment, they compound upon one another to produce an estimate of risk that is considerably inflated over the actual risk…

A common assumption made by both OSHA and MSHA concerns the exposure over a person’s lifetime. The agencies assume that a worker is exposed to a dose for 8 hours a day, 5 days a week, 50 weeks a year for 45 years…This assumption probably overestimates a worker’s exposure. A given worker is likely to move vertically within the organization (being in a new position would presumably change a worker’s exposure) or laterally within the industry, perhaps to a related position that does not include exposure to the hazard over a 45 year period.

Another conservative assumption is found in the adjustment of risk estimates based on uncertainties. There are five sources of uncertainties in any epidemiological or toxicological study: extrapolation from animal studies to humans; variability between animal groups of the same species; extrapolating from acute exposure to chronic exposure; extrapolating from the lowest observed effect level to a no observed effect level; and the uncertainty inherent in any dataset. It is common for risk assessors to adjust their risk estimates to account for uncertainty by dividing metrics by ten.

For example, if a risk assessor estimates that a worker exposed to a substance above a concentration of 50 μg/m³ is at risk, then the assessor might account for uncertainty in the estimation by dividing the concentration by ten, so that the final estimated threshold of risk is at 5 μg/m³. There is no scientific rationale for the practice of dividing by ten; it is merely a conservative adjustment that is made when uncertainty is present and is more the result of historical precedence than scientific reasoning.

Leon Sequeira, the assistant secretary of labor for policy, and David James, the assistant secretary of labor for public affairs, wouldn’t comment on any specifics in the proposed rule.

Need to know more? Get ProPublica headlines delivered by e-mail every day.