This brook, downstream from an abandoned paint factory's dumpsite, contains high levels of lead and arsenic.

The recommendation to change the arsenic standards came from a subcommittee called the National Research Council (NRC), an NAS affiliate. Congress established the NAS in 1863 to provide science-related recommendations to the federal government. It progressed into a membership organization for the best U.S. scientists, whose members elect new members based on recognition of distinguished achievement in original research.

The NRC provides these advisory duties on a for-hire basis, since the NAS does not receive federal appropriations directly in the budget process. The U.S. Environmental Protection Agency (EPA) contracted with the NRC to review and report on the science relevant to arsenic in drinking water. The current arsenic standard, which was set in the 1940s, is going on 60 years old.

Short Cut Science?

Instead of issuing reports describing the differing scientific views of subcommittee members, it seems that the NRC's reports may be presenting compromise consensus on the arsenic issues. No one disputes the fact that studies of U.S. populations with high levels of arsenic in drinking water have higher rates of cancer. However, most areas of the U.S. have low levels of arsenic in their drinking supplies.

Some scientists advocate the use of mathematical models to predict cancer risk from lower levels of arsenic exposure. Assumptions used in these models are critical to conclusions resulting from these models. The EPA uses a linear model that assumes any exposure to arsenic will increase cancer risk - the more one is exposed to arsenic, the more the risk increases. There are other models, called sub-linear models, which assume there is a "safe" level of exposure, concluding that increases in cancer risk are negligible at low doses. Most toxicology processes follow a sub-linear model.

Based on biochemical, toxicological and human study findings, the subcommittee stated in its report to EPA that only the sub-linear models were plausible. But because they could not agree on which sub-linear model was correct, the consensus was to opt for the EPA's linear model.

Resolved Quickly?

Whitman said she would delay the decision on a final standard until February 2002 but will require compliance by 2006. Many democrats are contending that the delay is illegal, citing previous legislation that states, "The (EPA) Administrator shall promulgate a national drinking water regulation for arsenic no later than June 22, 2001." The NRC announced that it intends to file a lawsuit on June 22 against the EPA if Whitman does not start the arsenic rule by then.

We will have to wait to see whether the new standard for arsenic is reduced to within 3 to 20 ppb. All scientists agree that arsenic in drinking water can cause some form of cancer - the debate is in the numbers and how those numbers are calculated to protect Americans. If costs of removing arsenic from drinking water are added to the equation, we may never get a consensus on what the new standard should be.

Sidebar: Relevant Arsenic Facts

EPA Contamination Levels
The Safe Drinking Water Act usually allows contamination only at levels that would cause one additional case of cancer per 10,000 people exposed. However, the National Academy of Sciences (NAS) found that at 50 ppb, there was a chance of one cancer case per 100 people exposed. The assumption is that if one person in 100 is at risk at 50 ppb then there would be one case per 500 people exposed at 10 ppb. This level is still 20 times higher than the normally acceptable level. To achieve the normal standard of one case per 10,000, levels of arsenic would need to be reduced to 0.5 ppb.

Where Is Arsenic?
Arsenic is both a naturally occurring substance and an industrial byproduct, entering the water supply from natural deposits and pollution. It is found at high concentrations in states with mining, electric-power generation, and other areas heavily involved with coal-burning and copper smelting. Arsenic, which was once used widely in pesticides, is among the most prevalent pollutants at toxic waste sites across the United States.

Approximately 15 percent of New Hampshire's drinking water exceeds 10 ppb of arsenic, as do 3,300 water systems in Michigan and 25 percent of New Mexico's water districts. All totaled, approximately 15 millions Americans drink arsenic-tainted water.

Cost Estimates
The American Water Works Association estimates the national cost of meeting a 10 ppb arsenic standard at $4.5 billion for special treatment facilities, plus approximately $500 million per year in increased extraction costs. That burden would fall mainly on ratepayers in small communities in the Southwest and parts of the upper Midwest and New England, where arsenic leaches naturally from rocks into water supplies.