| EW
Home
See additional articles on |
Reform or Revolution in the Mother of Environmental Databases
SPECIAL REPORT Reports in previous years ran more than 400 pages — but this years totaled only about six pages in length. Did it make any difference? Most of the data is still accessible to reporters and the public. The real questions revolve around how EPA presents the data and how (or whether) journalists are using it effectively. TRI Background TRI was born in the “Bhopal Bill” of 1986—an effort by Congress to reduce the chances that the Union Carbide chemical disaster that killed thousands in Bhopal, India in December 1984 could happen in the United States. The “Emergency Planning & Community Right to Know Act” (EPCRA), a title of the 1986 Superfund reauthorization bill, was revolutionary in that it proposed to fight pollution with information. TRI was one of four important things EPCRA accomplished. First, it set up mechanisms for local and state officials to plan chemical disasters. Second, it required chemical-handling facilities to report immediately to local and state emergency officials when an unusually large chemical release occurred. Third, it required plants to tell local and state emergency responders what chemicals they were handling on site. And fourth, it required chemical plants to report annually to the state and EPA an inventory of all their chemical releases to the environment (routine and catastrophic). That fourth item was the TRI. One more revolutionary thing that the law did was to
require EPA to put that data into a computer database, and to “make
these data accessible by computer telecommunication and other means
to any person on a cost reimbursable basis.” This was a decade
before E-FOIA, at the dawn of “Computer-Assisted Reporting,”
in the era of the 2400 bps modem, long before the World Wide Web—possibly
the first instance where reporters could It was also the heyday of the Reagan administration—when
“regulation” was a dirty word to many. The bargain made between
environmentalists and industry in getting EPCRA through Congress was
that the government would forego regulating toxic releases and settle
for informing the public about them. The hope was that information would
allow the public to better protect itself—and It was, many thought, lack of information that had contributed to the deaths of those thousands of victims at Bhopal. Lack of information about hazards had allowed hundreds of Bhopals poorest to live as squatters close to the Carbide plants fence line. Lack of any warning siren when the leak occurred. Lack of information about whether to stay indoors or flee. Even the hospital emergency rooms lacked information about what was causing the flood of casualties or how to treat them. Almost 20 years after Bhopal, information is still almost all the U.S. public has for protection from routine toxic pollution of air and water, much less catastrophic toxic releases. For various reasons, there are even today few mandatory or blanket federal limits on routine discharges of specific toxic pollutants into air or water. TRI and Environmental Journalism TRI has been a key tool in some of the best environmental
journalism during the last 15 years. Some journalists for the New
York Times saw early its potential as an investigative tool. Phil
Shabecoff wrote a 1,500-word piece emphasizing TRIs importance when
data collection began in 1988. Prize-winning science writer Cristine
Russell, the Washington Post, was one of those heralding TRIs
arrival as early as 1989. She envisioned it as a way for anyone with
a PC to find environmental threats in their own back yard. And the Post
published TRI-based local stories over the years. Russell herself Rae Tyson, USA Today, found the TRI ideal for producing the kind of story his paper specialized in—a national story that could be parsed into 50 state stories. Tyson did a huge (and very visual) take-out on TRI back in May 1992, including the inevitable “state-by-state look.” But USA Today's take was far from routine. Only a handful of national newspapers wanted to devote that kind of space to the TRI story, and none of the national broadcast or cable networks addressed these stories. In the second year of TRI data release (1990 for 1988 data), the real news involved a drop of about nine percent in total releases, seeming to prove the argument that publicity would push companies to clean up. Even while it pushed this story, EPA cautioned against over-interpreting the data. This added a ho-hum factor, and the second-year stories tended to be shorter and less grandiose than those in year one. Within a few years, after the novelty began wearing off, TRI-based coverage settled into a much more mundane groove, both in national and local stories. Enterprise stories by PC-pounding investigative reporters were notable for their rarity. Usually, a wave of stories was set off by each years Public Data Release—fresh meat, yes, but also a predictable and reliable story. The national stories tended to recite which states (or chemicals) were highest or lowest in the rankings, and which states had moved up or down on the list, and what the trends were. People did not read these lists as passionately as they did the Major League standings. TRI had become a statistical story. State and local stories, especially those built on the annual data release, tended to follow suit. They reported how a state was doing in the national standings (high or low, up or down), and what major chemicals were being released by whom in what part of the state. The one journalistic use of TRI that generated some voltage involved stories that ranked or listed the companies or facilities (locally or nationally) that released the highest amounts of toxics. Nobody wanted to be top-ranked on those lists, and companies wanted to stay out of the spotlight in the worst possible way. Journalists doing these stories quickly learned that they needed to wear figurative rubber boots. Upon seeing the reporter, companies would immediately and loudly complain about the accuracy, meaningfulness, and interpretation of the data, and they sometimes had a point. Such stories tended more and more often to be done by environmentalist research groups—and less by the media. Yet throughout the 1990s, a few tough and talented investigative reporters or teams used TRI as a starting point for prize-winning journalism, combining it with ground truthing, other kinds of reporting, and insightful understanding of local ecosystems, economics, and politics. One contemporary example of such work is ongoing coverage by the Louisville Courier-Journal. An extensive July 13, 2003, story by James Bruggers analyzing TRI data in the Louisville area was an important piece in a larger mosaic of coverage of the areas toxic air pollution issues by Bruggers and other reporters. ( http://www.courier-journal.com/cjextra/2003projects/toxicair/0713/2wirfront-air0622-26963.html) TRI 2002–2003 Stakeholder Dialogue On October 15, 2002, EPA launched a major reform project, a TRI “Stakeholder Dialogue.” Its purpose, as stated in various documents at the time, was to “improve,” “modernize,” “re-engineer,” or “streamline” the program. But the effort came at a critical time. For one thing, it marked the beginning of a new administrations efforts to get a grip on the reins of EPAs information programs—after getting administrators con.rmed and installed and weathering the upsets of 9/11. But for another, it came at a time when the TRI was being attacked with unprecedented intensity. The mining industry in 2002 was hopping mad. Metals and coal mining operations were among six new categories added to TRI by Clinton administration regulatory action in 1997. They didnt actually start reporting data to EPA until 1998, and EPA didn't report that data to the public until 2000, the closing year of the Clinton administration. TRI releases are counted by weight, not by toxicity, and because mining wastes are heavy and voluminous, the mining industry in 2000 suddenly leaped to the top of the chart, accounting for almost half of the nations “toxic” releases. The mining industry looked bad, and it didnt think the picture was fair. The National Mining Association had gone to court in May 1998 to challenge EPA's action, but the court upheld EPA's authority to add mining in January and March of 2001. But while the Clinton administration had been progressing in one direction (more reporting on toxics), the Congress had been moving in quite another direction. Western land interests had long dominated the House Interior Committee, but when the GOP took over the committee in 1995 as a result of the Gingrich electoral revolution (renaming it the Resources Committee), the ascendancy of “Wise Use” and “Sagebrush Rebellion” voices was full speed ahead. The committee became a firebase from which the mining industry artillery could pound EPA and TRI. Nor was mining the only industry unhappy with TRI. Electric utilities hardly rejoiced when they, too, were added to the TRI list in 1997. Until then, most attention had been focused on the high-volume “conventional” pollutants from coal-burning power plants, such as particulates, SO2 and NOx (which were not labeled “toxic”). The addition of utilities to TRI opened the curtain for public attention to a whole new set of toxic utility emissions like mercury and dioxins. EPAs 2002 Stakeholder Dialogue on TRI actually was not the first. There had been another back in 1997–99 as EPA was adapting to the “new” industries on the TRI list (and vice versa). The 2002 Stakeholder Dialogue was unsurpassed in its openness and deliberativeness—in fact, it was a landmark initiative in the new field of electronic government. Even as many federal agencies were erecting barriers in the post-9/11 era that cut them off from the public, EPA was throwing its doors open wide—at least in a cybernetic way, even as many in the federal government viewed the Internet with dark suspicion as a potential terrorist tool. The 2002 Stakeholder Dialogue consisted of two phases, which took place over a year and a half. Phase I focused on data collection and processing; the analysis and release of the TRI data; and EPAs help to companies submitting data. Phase II, which began in November 2003, focused on “reducing the burden” on companies submitting data. EPA started each phase by publishing “white papers” on the subjects under discussion. Then it threw open successive online “virtual public meetings” in which anyone could offer and respond to comments. Those virtual public meetings were some of the earliest manifestations of EPAs “EDOCKET” system, which held out the tantalizing promise of opening up most aspects of the agencys regulatory process to much broader forms of participatory democracy. EPA remains a leader among federal agencies trying to implement egovernment (see http://www.regulations.gov/). But the experiment also demonstrated the drawbacks of electronic government. For one thing, the lack of limits on participation allowed highly motivated commenters to swamp the docket with their views (although this happens in paper rulemakings and regular public meetings also). For another, the EDOCKET system allowed (because it was incapable of preventing) completely anonymous comments, which sometimes encouraged frivolity or abuse. Such cyber-pioneering was entirely appropriate to TRI. Almost since its beginning, the program had tried to offer companies ways to submit their data electronically, and the agency actually encouraged them to do it. The TRI program began in the decade when the Paperwork Reduction Act of 1980 was cutting a serious swath through the information requests coming from the federal government. Instead of requiring companies to submit their “Form R” on paper, EPA provided software to ease compilation of emissions estimates and allowed companies to submit their data on disk and eventually online. The Debate Perhaps the most noteworthy thing about the 2002 Stakeholder Dialogue was how few groups or individuals took part in it. It was, of course, only the latest chapter in a dialogue that had been going on for 15 years. The organized industry groups that participated were the American Chemistry Council, the American Petroleum Institute, Chlorine Chemistry Council, Consumer Specialty Products Association, and the Edison Electric Institute. The Coors Brewing Co., Ashland Co., Xcel Energy, John Deere, and Weyerhaeuser. About a dozen lesser-known companies made mostly brief comments, as did the Varnum law firm, which represents some companies who have to file TRI reports. That was about it from industry. The mining industry made no comments at all. For all its openness, the dialogue was notably lopsided. Missing from the dialogue was the environmental movement. Virtually the only voice from that side was Paul Orum, who heads the Working Group on Community Right-to-Know and has single-mindedly and persistently watchdogged chemical right-to-know issues for more than a decade. Also missing from the dialogue were the news media and journalism organizations. Companies complained bitterly about the reporting burden—electronic or otherwise—in the 2002 Stakeholder Dialogue, just as they had whenever opportunities arose in the past. Michael P. Walls of the American Chemistry Council (ACC) said ACC came to the dialogue with “a sense of pessimism about the prospects for significant change,” and slammed “the lack of any meaningful response by EPA to the years of effort and comment that the chemical industry has provided the Agency in an attempt to resolve even the most basic problems.” Industry interests used the opportunity to repeat most of the basic objections they had to the TRIs very existence: The quality of the data was bad. The data were meaningless or misleading because they did not account for human exposure levels, much less potential or actual harm to human health. By ranking and comparing poundage of different toxics (which not only had different kinds of toxicity, but different toxicity-per-unit-weight ), the TRI distorted the comparative toxic impacts of different facilities and different companies. The term “release,” industry had long argued, was a definitional nightmare including permitted emissions, fugitive emissions, the use of materials as manufacturing feedstocks, recycling, transporting materials from one plant to another, and proper legal disposal in licensed hazardous waste landfills. It was misleading to imply all these “releases” were going into the environment, when in many cases the truth was the exact opposite. Robert F. Van Voorhees, representing ACCs Underground Injection Control Task Group, repeatedly argued that well-injection of toxics into deep underground formations should not be considered a “release” into the environment. But in the end, such arguments were as much questions of perceptual engineering as hydraulic engineering. Many of these problems were unfixable, because they grew from the essential nature of TRI dictated by its legislative mandate. Indeed, industry groups and their Congressional allies had seriously proposed terminating TRI in 1995 shortly after the GOP takeover. But it was a two-edged sword. The almost-steady annual downward trend in total U.S. toxic releases was not only evidence that the regulation-by-information approach was working, it was also evidence that U.S. industry was making progress in cleaning up. Data quality had actually long been a complaint from both sides— environmentalists and industry. The beef from environmentalists and right-to-know advocates was that TRI data was not based on actual measurements or monitoring, but on “best estimates” by companies themselves. It was an “honor system.” Companies had every motive to minimize their releases and emissions in order to avoid appearing atop the “worst polluters” lists. EPA limited its datapolicing largely to catching omissions, inconsistencies, arithmetic errors, and the like, and had hardly ever been known to prosecute a company for deliberately falsifying data. Glen J. Barrett of the American Petroleum Institute
(API) argued against shifting toward greater reliance on electronic
submission of data. API preferred paper. He argued that many of the
multi-billiondollar Few things drew more complaints from companies than EPA's data-checking procedures, which required companies to re-do their work when they submitted bad data. “Data quality” had for some years been the slogan emblazoned on the banners owned by industry lobbyists in various anti-regulatory campaigns. Companies blamed EPA for publishing data that was inaccurate. Companies blamed EPA for hassling them about submitting inaccurate data. But in many cases the companies themselves had submitted inaccurate data in the .rst place. Many of EPA's delays in publishing the data resulted from its efforts to clean up and correct industry data. Companies, of course, also blamed EPA for those delays. Bad as it was, TRI was not only the best publicly available data that described the quantity and fate of toxic chemicals handled by U.S. companies, it was the only such data set. It set a precedent that a number of foreign countries rushed to follow. Right-to-know advocate Orum argued that what was important was the quality of the data. Orum argued along with EPA that the way to improve data quality was to shift toward electronic submission, rather than away from it. The technology of computer databases, largely perfected a decade ago, allowed data to be validated and checked even as it is being entered. EPA wanted to improve further the data-entry client software, known as TRI-ME that it distributed to companies. Orum argued that requiring electronic submission “would save money for both EPA and industry, reduce errors, speed the process, and enable more advanced forms of error checking BEFORE companies ever submit information to the agency.” The Outcome On May 19, 2004, EPA .nally announced what changes
it would make in response to the 2002 Stakeholder Dialogue (http://www.epa.gov/tri/TRI%20Re-Engineering%20Memo.pdf)
in a memo from an EPA assistant administrator, Kim Nelson. There were
a number of major changes: Reform or Revolution?—2002 Public Data Release On June 23, 2004, a month after Nelsons May memo, the Public Data Release for 2002 went public. On paper, it consisted only of six very general overview pages (including graphs and charts), between a front and back cover. That was a far cry from the reports of previous years, which had typically run more than 400 pages. It was distributed online, along with a Web-based slide show briefing and a downloadable Powerpoint. The press kit included links to many tables online. Perhaps the essence of Nelsons new strategy was to let the data speak for themselves. What was missing was interpretation, comparison, conclusions, and hand-holding of readers. There was a strong and direct statement about all the things the data werent—e.g. a measure of health risk. Over the years, as reporting rules changed and toxics were subtracted from or added to the list, charting year-by-year trends and comparisons had become more complex. EPA had to do more massaging of the data to normalize it. That was largely gone. But users could still download the entire years database all at once—faster and easier than ever in the era of broadband Internet—and massage and interpret it to their hearts content. And EPA said it was allowing users to directly execute larger and more complex explorations of the data via its online query engine, TRI Explorer. One irony in the story came a day or two before EPAs announced date for releasing this years public data. Two environmental groups—the Environmental Integrity Project (EIP) and the Galveston-Houston Association for Smog Prevention (GHASP)—launched a preemptive press conference, releasing a report charging that “EPA knowingly underreports toxic air emissions from refineries, chemical plants by 330 million pounds per year” (about 16 percent). Kim Nelson and EPA press officials scrambled to move up the time of their own press conference call-in response. Nelson could only say she had not read the report, and that the law did not actually require EPA to verify by monitoring the TRI estimates companies submitted. (http://www.courier-journal.com/localnews/2004/06/23ky/A1-toxic0623-8968.html and http://www.environmentalintegrity.org/pub205.cfm) It was a coup. Not only had the initiative by EIP (headed by EPA's former enforcement chief Eric Schaeffer) and GHASP seized the momentum of the news cycle away from EPA—they had seized the “data quality” issue away from industry. And it seemed they had also taken the job of skeptical fact-checking and muck-raking from media who had long since settled into the comfortable role of conveyor-belt for whatever EPA gave them in its annual Public Data Release. Was EPA starving a data-hungry press and public, or had it just stopped hand feeding them? At the end of the day, the question remained unanswered whether EPA's decision to truncate the Public Data Release was going to deprive the press and public of meaningful information about toxics going into their environment. One thing seemed likely, though, in the future, only the enterprise and investigative stories requiring far more of journalists would be likely to give the public meaningful information about toxic releases. Many fear that institutional pressures and .nancial and resource constraints affecting the news media are making those in-depth reporting efforts more and more difficult and unusual. —EPA TRI Page: http://www.epa.gov/tri/index.htm September 2004
|
