Scorecard Hits Home: Web Site Confirms Internet's Reach

Chemical Week, Vol. 160, No. 21

June 3, 1998

By Andrea Foster, Peter Fairley, and Rick Mullin

Whether the forum was CMA's board meeting in Phoenix, Socma's annual meeting in Palm Beach, or the New Jersey Chemical Industry Council's annual get-together in Absecon, one topic dominated industry meetings this April: the Chemical Scorecard Web site unveiled by the Environmental Defense Fund (EDF; New York) that month (CW, April 22, p. 14).

The first look at Scorecard's power to transform Toxics Release Inventory (TRI) data into images of toxic doom sent executives into spasms. "This is the worst thing that could possibly happen to the industry," said one executive.

At the same time, EDF was fighting a problem of its own: popularity well beyond expectations, bringing almost 40 demands/second for database searches to a site designed for 20/second. "We actually melted down the first day," says Scorecard designer William Pease, an EDF senior scientist who serves as an adjunct toxicology professor at the University of California at Berkeley's School of Public Health.

EPA, information access services such as RTK-NET (Washington), and state regulators have been expanding on-line access to environmental data for several years (CW, April 2, 1997, p. 43), but Scorecard has outdesigned the competition. It is user friendly, offers context to help users understand the data, and flexes the Internet's interactive muscle to give users a quick way to act on their newfound knowledge. "It's very powerful," says Mark Greenwood, a former EPA pollution prevention official who represents the Coalition for Effective Environmental Information (Washington) as a partner with law firm Ropes & Gray (Washington).

Scorecard has begun to achieve something Greenwood has been trying to do for several years: awaken industry--particularly top executives--to the Internet's power to shape public perceptions of environmental issues and companies' environmental performance. "They're beginning to see how the information can be manipulated," says Greenwood. As the initial scare passes, some are beginning to perceive opportunity in Scorecard.

"I think it's great to see our numbers out there--we're proud of them," says John V. Van Hulle, president of CasChem (Bayonne, NJ). Scorecard shows that releases from CasChem's Bayonne, NJ plant decreased 95% between 1990 and 1995--a message that Van Hulle says takes on added credibility coming from EDF, a group perceived as an industry critic.

C.P. Hall (Chicago) CEO and president George Vincent says that one product emission at one of his facilities has a high-visibility ranking on the EDF site. As a result, he is challenging his staff to reduce those emissions, "so that C.P. Hall will receive less attention."

Nevertheless, interviews by CW reveal that most companies are at the earliest stages of formulating a response to Scorecard and the Internet. Some have yet to examine sites such as EPA's and EDF's to check the accuracy of the data provided for their plants. Only a handful of companies, including Ashland Chemical, Dow Chemical, and DuPont, have begun linking their Web sites to EPA's and others to get their side of the story told.

Pease says EDF plans to provide links to company Web sites, as well as dedicated feedback pages, because generating a dialogue about chemical hazard is what Scorecard is all about. "This is really just the beginning of a long-term interactive campaign, involving people in plant communities and the companies," he says.

The interactive nature of the Internet and the interconnectivity of modern digital communication channels makes the campaign possible. Scorecard gets the dialogue going by giving users an easy way to e-mail EPA, hyperlink to a local environmental group's Web site, or fax a message to a plant after viewing its profile. Pease says Scorecard users have already contacted 1,000 of the system's 4,000 faxable facilities, some more than once.

To keep the dialogue going, EDF plans to follow up with Scorecard users to learn whether a facility responded and with companies to challenge them to post a response on-line. Pease says EDF will post company responses, whether or not it agrees with the answer, as long as companies evaluate the threat posed by the plant's emissions. "We're going to allow companies to post their side of the story, but it has to be supported." Plants could post a risk assessment, for example, or more recent data than the two-year-old TRI release numbers.

The potential impact on community relations is staggering. The site has cooled off from its first day, when about 500,000 data requests were filed, but it is still receiving about 200,000 data requests/day from about 50,000 users. In comparison, RTK-NET, which provides the complete TRI database, received 230,000 data requests in 1997. EPA's site logs millions of hits/day and is growing exponentially (chart, p. 26), but the measure exaggerates the number of requests for data, and only a fraction of the requests are for EPA's Envirofacts database on facility performance.

Most significantly, Scorecard is drawing people who might not even know what the letters EPA stand for, says Pease. "These aren't environmental activists," he says. He credits the draw to coverage by major news media and scores of regional newspapers, as well as EDF banner ads running on major sites on the Web, such as ESPN SportsZone and Yahoo. The ads invite browsers to type in their zip code and find out what toxic chemicals "polluters" are releasing into their community's air and water.

"We paid attention to what makes public disclosure effective," says Pease. "The key was interpreting raw data to answer common sense questions such as 'What does this stuff do to me?,' 'Which is the worst?,' and 'Have things gotten better?'" Providing context, particularly ranking chemical releases by risk rather than by pounds, is something that the chemical industry has been clamoring for EPA to do since TRI was created. However, industry officials are expressing a mix of appreciation for Scorecard's ability to share information on their plants and concern about the accuracy of the picture it paints.

DuPont manager/pollution prevention Ed Mongan says Scorecard is providing a legitimate public service by publishing TRI data but adds that EDF's system for ranking the relative threat posed by different TRI chemicals contains questionable data. "It's not clear what science they've used," he says. Similarly, Terry Cox, manager of Dow's EH&S management center, shares concerns about how information is "packaged" on Scorecard but says EDF is providing a useful service. "If it makes it easier for someone in our company to access information and rankings, we would support that," Cox says.

Michael Desmond, v.p./technology at BP Chemicals, says Scorecard is "extremely professional," but he finds its hazard indexing system misleading. "There's some merit to indexing, since some chemicals are more toxic than others," he says. But Desmond says EDF's system errs in being too specific. "I'd prefer a much simpler system than assigning specific numbers--something like a scale of one to five on which five is the highest carcinogenicity." EDF's ranking system--designed by Pease and colleagues at Berkeley and peer reviewed and published by Environmental Science & Technology in March--calculates each TRI chemical's toxicity and exposure potential relative to benzene, a known carcinogen, and toluene, a developmental toxin. The system is similar in concept to the approach ICI has taken in recent years in its environmental reports, but Desmond says EDF's methodology exaggerates the threat posed by most chemicals.

For example, Scorecard lists DuPont's site at El Paso, IL as the top polluter in the U.S., with carcinogenic emissions equivalent to 4.2 billion lbs of benzene in 1995. "If that were true, their hospitals would be chock-a-block [with cancer patients]," says Desmond. Scorecard's relative rankings for acetonitrile and acrylonitrile are similarly misleading, he says. "Knowledgeable people would prefer acetonitrile exposure [over] acrylonitrile exposure. Acetonitrile is not particularly toxic. Acrylonitrile we're very careful with," but Scorecard lists acetonitrile as more toxic. "From a public right-to-know perspective, it's very misleading."

Pease says complaints about the ranking system are "inevitable" because there is much at stake for chemical producers and because there is no universally accepted ranking system. In addition, he says, many industry toxicologists find the rankings counterintuitive because they know chemicals primarily by acute risk to workers, whereas EDF's rankings also take into account the potential for community exposure--for example, a chemical with low toxicity can rate high if it is persistent. He says Desmond's judgment on acetonitrile and acrylonitrile is a clear case of intuitive scoring based on potential for occupational exposure.

Still, Pease says, he recognizes one mistaken ranking in the system--the carcinogenic hazard posed by releases of acrylamide to water--which Dow toxicologists brought to his attention. Scorecard ranked acrylamide as a high risk because it used government data showing acrylamide to be highly persistent in sterile water, whereas published data presented by Dow shows that it readily biodegrades in aquatic environments.

Pease says EDF also sympathizes with criticism from Desmond and others in the industry that counting millions of pounds of deepwelled wastes as environmental releases--in keeping with EPA practice--overestimates risk to neighboring communities. He says EDF plans to address this when it updates the system with 1996 TRI data.

Similar debates over contextualizing of data have been simmering in Washington as EPA has increased its use of the Internet. EPA's enforcement office recently launched its Sector Facility Indexing Project (SFIP), which allows users to compare the environmental performance of facilities in six industrial sectors, but it dropped a planned hazard ranking system following intense criticism from industry and some state regulators (CW, March 25, p. 58). "There's scientific uncertainty about this and a mixed desire for it in the business community," says Charles Fox, EPA's associate administrator/reinvention. EPA is expected to use a hazard index with its 1996 TRI data release this month to rank chemicals and industry sectors, but it has decided the system is not robust enough for facility ranking.

EPA took eight years to develop its hazard ranking system and is still cautious about using it, but EDF can charge ahead with facility ranking since it is not a government agency. "I don't think EPA could get a benzene equivalence index like EDF's through its Science Advisory Board, whereas EDF just does it," says Greenwood.

But even though the Internet allows groups such as EDF to bypass EPA, Greenwood says there are two reasons industry should continue pressing federal and state agencies on their use of the Internet: Government resources carry added credibility, and the government collects the hard data used by most systems. Greenwood says industry efforts are showing results, particularly in persuading EPA to correct quality gaps in its databases. EPA was shrugging off the problem last year, but Greenwood says the agency now recognizes that it bears some responsibility to ensure that online data is accurate.

Data quality is not a major issue for Scorecard, which uses the thoroughly checked TRI, but it is for resources such as SFIP and Envirofacts, which use older databases never intended for public distribution. "The TRI database is the only one checked carefully to ensure that what goes in is correct--that's what's frightening about Envirofacts," says Greenwood. Companies looking through SFIP and Envirofacts have identified many mistakes in their own reports to state agencies, as well as systematic errors introduced when EPA's systems pick up state data.

Renee Bobal, senior manager/corporate environment and safety affairs at Hoffmann-La Roche (Nutley, NJ), says data accuracy is a serious issue, given that Scorecard and SFIP are the beginning of what will likely be a barrage of information available on the Web. "Right now, there's no way of controlling anything," she says. "It's a free and open system. Right now the burden is on the company whose data is out there to make sure the information is valid." Despite EPA's recent commitment to assure data quality, the problem looks likely to grow as the agency continues to post additional information on-line. Envirofacts is expected to add data on plant compliance, which is already available for the sectors covered by SFIP. And EPA is expected to unveil an Internet site for its new Center for Environmental Information and Statistics (CEIS) within two months. Rather than profiling facilities, CEIS will provide risk assessments and environmental quality modeling at the community level.

EDF also plans to add resources to Scorecard. Pease says the group is hoping to add an EPA database that projects levels of hazardous air pollutants for each census tract in the U.S. Rankings of corporations and information on chemicals in consumer products may be added.

Environmental activists say access for those who cannot afford a computer--a problem facing the Internet as a whole--must be solved before systems such as Scorecard replace other modes of communication. "Communities hit hardest by toxic pollution are nowhere near having access to the Internet," says Carolyn Hartmann, director/environmental programs for the U.S. Public Interest Research Group (Washington).

Hartmann says EDF's site does a great job of highlighting the lack of toxicity information for many chemicals, but other unknowns are harder to advertise. In particular, no release data is available for toxic chemicals not yet listed on the TRI, and many polluters, such as municipal waste incinerators and airports, do not report their emissions, Hartmann adds.

Industry associations are beginning to assist members in formulating a response to Scorecard. The National Paint and Coatings Association (Washington), for example, is advising members to review the Scorecard for accuracy and to be prepared to explain why their plants are releasing a particular chemical. It recommends that members compile a list of improvements at plants, such as reductions in toxic releases and shifts to less toxic chemicals, as well as jobs provided.

To alert its members to Scorecard, The New Jersey Chemical Industry Council issued printouts of data on member companies to attendees at its annual meeting. Socma is telling members that EDF's use of 1995 TRI data is a ready-made opportunity for firms to show off improvements by offering Scorecard users more recent data.



1998 Copyright Chemicalweek Associates