Toxic Chemical Releases: EPA Actions Could Reduce Environmental  
Information Available to Many Communities (30-NOV-07,		 
GAO-08-128).							 
                                                                 
Federal law requires certain facilities that manufacture,	 
process, or use any of 581 toxic chemicals to report annually to 
the Environmental Protection Agency (EPA) and their state on the 
amount of those chemicals released into the air, water, or soil. 
It also requires EPA to make this information available to the	 
public electronically through the Toxics Release Inventory (TRI) 
database. Facilities must either (1) submit a detailed TRI Form R
for each designated chemical that they use in excess of certain  
thresholds or (2) file a simpler Form A certifying that they need
not do so. To reduce companies' burden, EPA issued a rule in	 
December 2006 intended to expand Form A eligibility for certain  
facilities and chemicals. GAO was asked to analyze (1) how EPA	 
and others use TRI data, (2) whether EPA followed internal	 
guidelines in developing its rule, (3) the rule's impact on	 
information available to the public, and (4) the extent of burden
reduction that is likely to result from EPA's changes.		 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-08-128 					        
    ACCNO:   A77668						        
  TITLE:     Toxic Chemical Releases: EPA Actions Could Reduce	      
Environmental Information Available to Many Communities 	 
     DATE:   11/30/2007 
  SUBJECT:   Agency proceedings 				 
	     Data collection					 
	     Environmental law					 
	     Environmental monitoring				 
	     Environmental policies				 
	     Federal facilities 				 
	     Fees						 
	     Government information dissemination		 
	     Internal controls					 
	     Pollution						 
	     Pollution control					 
	     Pollution prevention				 
	     Program management 				 
	     Regulatory agencies				 
	     Reporting requirements				 
	     Toxic substances					 
	     Waste management					 
	     Benefit-cost tracking				 
	     Cost estimates					 
	     EPA Toxics Release Inventory Program		 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-08-128

   

     * [1]Results in Brief
     * [2]Background
     * [3]The TRI Is Widely Used by Federal Agencies, States, and the

          * [4]EPA Offices and Other Federal Agencies Use the TRI for Multi
          * [5]Most States Have TRI Programs and Use TRI Data for a Range o

               * [6]Chemical Identification or Pollution Prevention
                 Activities
               * [7]Compliance Assistance or Program Enforcement
               * [8]Public Information or Other Data Services

          * [9]The Public Uses the TRI for a Variety of Purposes

     * [10]EPA Did Not Follow Key Steps for Developing the TRI Burden R

          * [11]Internal Stakeholders Had No Opportunity to Comment on Effec
          * [12]Public  on EPA's Proposed Rule Were Overwhelmingly N

     * [13]EPA Changes Could Significantly Limit the TRI Data Available

          * [14]Thousands of Detailed Form R Reports May No Longer Be Submit
          * [15]Data about Toxic Chemicals May Be Reduced or Eliminated for
          * [16]Some Facilities Will No Longer Have to Report Detailed Infor
          * [17]Many Commenters Have Stated That EPA's Changes Will Signific

     * [18]Changes to the TRI Will Likely Do Little to Reduce Industry'

          * [19]EPA Relied on Outdated Data and Questionable Assumptions to
          * [20]Other Analyses Indicate Actual Savings Could Be Much Less Th
          * [21]Other EPA Efforts Have Provided More Burden Relief without A

     * [22]Congressional Interest and Actions to Reverse EPA's Changes
     * [23]Conclusions
     * [24]Agency  and Our Evaluation
     * [25]Matter for Congressional Consideration
     * [26]GAO 
     * [27]GAO 
     * [28]GAO Contact
     * [29]Staff Acknowledgments
     * [30]GAO's Mission
     * [31]Obtaining Copies of GAO Reports and Testimony

          * [32]Order by Mail or Phone

     * [33]To Report Fraud, Waste, and Abuse in Federal Programs
     * [34]Congressional Relations
     * [35]Public Affairs

Report to Congressional Requesters

United States Government Accountability Office

GAO

November 2007

TOXIC CHEMICAL RELEASES

EPA Actions Could Reduce Environmental Information Available to Many
Communities

GAO-08-128

Contents

Letter 1

Results in Brief 4
Background 7
The TRI Is Widely Used by Federal Agencies, States, and the Public 13
EPA Did Not Follow Key Steps for Developing the TRI Burden Reduction Rule
25
EPA Changes Could Significantly Limit the TRI Data Available to Many
Communities 31
Changes to the TRI Will Likely Do Little to Reduce Industry's Reporting
Burden 41
Congressional Interest and Actions to Reverse EPA's Changes 50
Conclusions 51
Agency  and Our Evaluation 52
Matter for Congressional Consideration 53
Appendix I Scope and Methodology 55
Appendix II GAO Estimates of the Impact of Reporting Changes on the TRI 58
Appendix III Comparison of Information Collected on the Form R and the
Form A Certification Statement 60
Appendix IV  from the Environmental Protection Agency 64
GAO  68
Appendix V  from the Office of Management and Budget 72
GAO  80
Appendix VI GAO Contact and Staff Acknowledgments 84

Tables

Table 1: Key Characteristics of State TRI Programs 16
Table 2: Comparison of Average Annual TRI Burden Estimates by Activity:
OMB-approved Versus Engineering Analysis 46
Table 3: Estimated Impact of TRI Reporting Changes on Number of Form Rs,
Chemicals, and Facilities, by State 58
Table 4: Information Collected on the TRI Form R and Form A Certification
Statement 60

Figures

Figure 1: Summary of Information Reported on TRI Form R 9
Figure 2: TRI Activities Reported by States 20
Figure 3: Estimate of Impact Allowed by EPA's Changes on Number of Form
Rs, by State 33
Figure 4: Estimate of Percentage of Chemicals for Which Facilities Could
Report on Form A, by State 35
Figure 5: Estimate of Percentage of Chemicals for Which Facilities Could
Report on Form A, by County 36
Figure 6: Estimate of Percentage of Facilities That Could Convert All Form
Rs to Form A, by State 38
Figure 7: Result of GAO Survey of State TRI Coordinators' Views about
Impact of EPA's Changes on Various State Activities 41
Figure 8: Burden Savings for Each PBT Chemical and Non-PBT Chemical Report
43

Abbreviations

ADP Action Development Process
CAS Chemical Abstracts Service
e-FDR Electronic Facility Data Release
EPA Environmental Protection Agency
EPCRA Emergency Planning and Community Right-to-Know Act of 1986
Form A Form A Certification Statement
Form R Form R report
GIS geographic information system
ICR Information Collection Request
IRIS Integrated Risk Information System
IRS Internal Revenue Service
NAICS North American Industry Classification System
NGO non-governmental organization
NSC No Significant Change
OEI Office of Environmental Information
OMB Office of Management and Budget
OPEI Office of Policy, Economics, and Innovation
PBT Persistent Bioaccumulative Toxic
POTW Publicly Owned Treatment Works
PPA Pollution Prevention Act of 1990
PRA Paperwork Reduction Act
RCRA Resource Conservation and Recovery Act
RSEI Risk-Screening Environmental Indicators
RTK Net Right-to-Know Network
SAB Science Advisory Board
TRI Toxics Release Inventory
TRI-ME Toxics Release Inventory-Made Easy

This is a work of the U.S. government and is not subject to copyright
protection in the United States. The published product may be reproduced
and distributed in its entirety without further permission from GAO.
However, because this work may contain copyrighted images or other
material, permission from the copyright holder may be necessary if you
wish to reproduce this material separately.

United States Government Accountability Office
Washington, DC 20548

November 30, 2007

The Honorable Barbara Boxer
Chairman, Committee on Environment and Public Works
United States Senate

The Honorable Frank R. Lautenberg
United States Senate

The Honorable Olympia J. Snowe
United States Senate

Each year, U.S. industry uses billions of pounds of toxic chemicals to
produce the nation's goods and services. The release of these chemicals
during transport, storage, use, or disposal as waste, can potentially harm
human health and the environment. In 1984, a catastrophic accident caused
the release of methyl isocyanate--a toxic chemical used to make
pesticides--at a Union Carbide plant in Bhopal, India, killing thousands
of people, injuring many others, and displacing many more from their homes
and businesses. One month later, it was disclosed that the same chemical
had leaked at least 28 times from a similar Union Carbide facility in
Institute, West Virginia. Eight months later, 3,800 pounds of chemicals
again leaked from the West Virginia facility, sending dozens of injured
people to local hospitals. In the wake of these events, the Congress
passed the Emergency Planning and Community Right-to-Know Act of 1986
(EPCRA) to inform citizens about releases of toxic chemicals to the
environment in their communities. Each year, U.S. industry uses billions
of pounds of toxic chemicals to produce the nation's goods and services.
The release of these chemicals during transport, storage, use, or disposal
as waste, can potentially harm human health and the environment. In 1984,
a catastrophic accident caused the release of methyl isocyanate--a toxic
chemical used to make pesticides--at a Union Carbide plant in Bhopal,
India, killing thousands of people, injuring many others, and displacing
many more from their homes and businesses. One month later, it was
disclosed that the same chemical had leaked at least 28 times from a
similar Union Carbide facility in Institute, West Virginia. Eight months
later, 3,800 pounds of chemicals again leaked from the West Virginia
facility, sending dozens of injured people to local hospitals. In the wake
of these events, the Congress passed the Emergency Planning and Community
Right-to-Know Act of 1986 (EPCRA) to inform citizens about releases of
toxic chemicals to the environment in their communities.

Certain facilities that manufacture, process, or otherwise use any of 581
individual chemicals and 30 chemical categories must report annually to
the Environmental Protection Agency (EPA) and their respective state, the
amount of those chemicals that they released to air, soil, or water. EPCRA
further requires EPA to make this information available to the public,
which the agency does electronically through the Toxics Release Inventory
(TRI) database. The Pollution Prevention Act of 1990 requires facilities
that report to the TRI to also provide certain information about their
waste management practices, including amounts of chemicals recycled or
treated. The purpose of making this information available is to inform
citizens about releases of toxic chemicals to the environment; to assist
governmental agencies, researchers, and other persons in the conduct of
research and data gathering; and to aid in the development of Certain
facilities that manufacture, process, or otherwise use any of 581
individual chemicals and 30 chemical categories must report annually to
the Environmental Protection Agency (EPA) and their respective state, the
amount of those chemicals that they released to air, soil, or water. EPCRA
further requires EPA to make this information available to the public,
which the agency does electronically through the Toxics Release Inventory
(TRI) database. The Pollution Prevention Act of 1990 requires facilities
that report to the TRI to also provide certain information about their
waste management practices, including amounts of chemicals recycled or
treated. The purpose of making this information available is to inform
citizens about releases of toxic chemicals to the environment; to assist
governmental agencies, researchers, and other persons in the conduct of
research and data gathering; and to aid in the development of appropriate
regulations, guidelines, and standards. Using EPA's Web site, the public
and others can search the TRI for information by state, county, zip code,
chemical, and type of facility, among other options.

Facilities must submit a detailed Form R report (Form R) for each
designated chemical that they manufactured, processed, and/or otherwise
used in excess of certain thresholds, or certify that they are not subject
to the reporting requirement by submitting a brief Form A Certification
Statement (Form A). Form A captures general information about the
facility, such as address, parent company, industry type, and basic
information about the chemical or chemicals it released. Form R includes
the same information, but also requires facilities to provide details
about the quantity of the chemical they disposed or released on-site to
the air, water, land, and injected underground, or transferred for
disposal or release off-site. Since 1995, EPA has allowed certain
facilities to submit a Form A in lieu of Form R if they release or manage
no more than 500 pounds of any chemical that is not considered to be a
persistent bioaccumulative toxic (PBT) chemical. PBTs are toxic chemicals
such as lead, mercury, and dioxins that remain in the environment for long
periods of time, are not readily destroyed, and accumulate in body tissue.

During the past several years, the Office of Management and Budget (OMB)
has encouraged EPA to engage in a multiphased effort to reduce the
reporting burden on industry, particularly small business, by revising TRI
reporting regulations to eliminate redundant items on Form A and Form R
and to expand Form A eligibility. On December 22, 2006, EPA issued the TRI
Burden Reduction Final Rule, an action that expanded Form A eligibility
for certain facilities by raising the release threshold from 500 pounds to
2,000 pounds for any non-PBT chemical.^1 The rule also allows, for the
first time, certain facilities to use Form A to report nondioxin PBT
chemicals, provided that the facilities do not release the chemicals into
the environment.^2 The burden reduction rule goes into effect for
reporting 2006 releases.

^1Specifically, the rule expanded non-PBT chemical eligibility for Form A
by raising the eligibility threshold to 5,000 pounds of total annual waste
management (i.e., releases, recycling, energy recovery, and treatment for
destruction) provided total annual releases of the non-PBT chemical
comprise no more than 2,000 pounds of the 5,000-pound total waste
management limit and provided the facility does not exceed a
one-million-pound manufacture, process, or otherwise use activity
threshold for the specific non-PBT chemical.

EPA's Action Development Process is internal guidance that outlines a
series of steps that the agency is to follow when it develops actions such
as policy statements, risk assessments, and regulations--including the TRI
Burden Reduction Rule. The purpose of the process is to ensure that
scientific, economic, and policy issues are adequately addressed at the
appropriate stages of action development and to ensure adequate
stakeholder participation across EPA's offices (e.g., Office of Air,
Office of Water) until the final action is completed. Steps in the process
include, among others, (1) chartering a workgroup comprised of
representatives from various EPA headquarters and regional offices to
develop the action; (2) preparing and executing an analytic blueprint,
which identified the analyses needed to support the action; and (3)
conducting final agency review by senior EPA management.

In February 2007, we testified on our preliminary assessment of the
process EPA followed in developing the TRI Burden Reduction Rule and its
impact on TRI data.^3 As you requested, this report provides the final
results of our work on (1) how EPA and other federal agencies, the states,
and the public use the TRI; (2) the extent to which EPA followed its
action development guidelines in developing the burden reduction rule; (3)
the impact of the changed reporting requirements on the amount of
environmental information available to the public; and (4) the burden
reduction that is likely to result from EPA's changes.

To address these four objectives, we analyzed documents pertaining to the
development of the TRI Burden Reduction Rule (including stakeholder
 in the public docket for the proposed rule), EPA's report on uses
of TRI data, annual TRI public data release reports, and EPA guidance for
facilities reporting to TRI. For the first objective, we reviewed
documentation from EPA, other federal agencies, the states,
nongovernmental organizations, and businesses about their uses of TRI
data. We obtained further information from states through a Web-based
survey of state TRI coordinators that achieved 100 percent response from
the 50 states and the District of Columbia. This report does not contain
all the results from the survey. The survey and a more complete tabulation
of the results can be viewed at [36]GAO-08-129SP . For the second
objective, we interviewed EPA officials who served on the TRI workgroup
that developed the TRI rule and reviewed internal EPA documents that
detailed the workgroup's process and EPA's decisions. For the third
objective, we used 2005 TRI data to estimate the impact of changes to the
TRI reporting requirements on the information provided by facilities about
their chemical releases. We performed a reliability assessment of the data
we analyzed and determined that the data were sufficiently reliable for
the purposes of this report. For the fourth objective, we reviewed EPA's
economic analysis of the costs and impacts of expanding eligibility for
Form A and other relevant documents, and we interviewed EPA officials
about their burden reduction analyses. A more detailed discussion of our
scope and methodology is included in appendix I. We conducted our work
from August 2006 to September 2007 in accordance with generally accepted
government auditing standards.

^2Specifically, this rule allows the use of Form A for PBT chemicals,
except dioxin and dioxin-like compounds, when total annual releases of a
PBT chemical are zero and the total annual amount of the PBT chemical
recycled, combusted for energy, and treated for destruction does not
exceed 500 pounds provided the facility does not exceed a 1-million-pound
manufacture, process, or otherwise use activity threshold for the specific
PBT chemical.

^3GAO, Environmental Information: EPA Actions Could Reduce the
Availability of Environmental Information to the Public, GAO-07-464T
(Washington, D.C.: Feb. 6, 2007).

Results in Brief

The TRI is used widely by EPA and other federal agencies, the states, and
the public for a variety of purposes. Nearly all of EPA's program offices
depend on TRI data to inform their decision making about toxic releases.
For example, EPA's Office of Prevention, Pesticides and Toxic Substances
used TRI data to develop its Risk-Screening Environmental Indicators
(RSEI) model, which incorporates detailed facility data from the TRI,
toxicity information from EPA's Integrated Risk Information System (also
known as IRIS), population data from the U.S. Census, and other EPA data
to develop human health hazard and risk-related perspectives on long-term
(chronic) exposures to TRI chemicals. Many other federal agencies also use
the TRI. For example, the Internal Revenue Service (IRS) uses TRI data to
identify companies that release chlorofluorocarbons (chemicals that
deplete the earth's ozone layer) to enforce a tax to help phase out their
use. In addition, state TRI coordinators reported that their states use
the TRI to carry out a variety of pollution prevention initiatives, assess
fees on facilities, and assist with emergency preparedness functions,
among other uses. The public also uses the TRI for a variety of
activities. For example, researchers use TRI data to assess environmental
policies and strategies for pollution reduction, and investment companies
use the data to determine socially-responsible investment options.
Individual citizens and local advocacy organizations also use TRI data to
learn about the type and quantity of toxic chemicals used and released in
their communities.

When developing the TRI Burden Reduction Rule, EPA did not follow several
important steps from its Action Development Process--guidelines designed
to ensure that the agency conducts appropriate analyses and receives
adequate input before the rule is finalized. According to EPA documents
that we reviewed and program officials who we interviewed, in mid-2006,
senior EPA management directed inclusion of a burden reduction option--a
10-fold increase in eligibility for reporting non-PBT chemicals on Form
A--in response to direction from OMB. The TRI workgroup that was charged
with developing the rule had previously dropped the non-PBT option from
consideration because of its impact on the TRI and, consequently, had not
conducted an economic analysis to assess its costs and benefits. Those
documents also showed that the EPA Administrator expedited the process in
order to meet a commitment to OMB to provide burden reduction by the end
of December 2006. The expedited schedule did not provide time for EPA to
complete the economic analyses necessary to evaluate the costs and
benefits of the option or request input from the EPA program offices that
rely on the TRI data for decision making. Specifically, although EPA held
a Final Agency Review for program offices to state their position on the
proposed rule, the review package that we viewed did not include the
non-PBT option for their review and comment. These deviations from EPA's
rule-development process resulted in inadequate justification for the
proposed rule. As a result, a substantial majority of public 
expressed opposition to EPA's rule, whereas few commenters supported the
rule. EPA finalized the rule in December 2006 with a fourfold increase in
Form A eligibility threshold--allowing facilities to use it for releases
up to 2,000 pounds of a non-PBT toxic chemical.

Our analysis of the latest TRI data shows that, by increasing the number
of facilities that may use Form A, the TRI Burden Reduction Rule could
significantly reduce environmental information available to the public on
dozens of toxic chemicals released from thousands of facilities across the
country. EPA estimated that expanding Form A eligibility would eliminate
detailed reports for less than 1 percent of the total release pounds
reported annually to the TRI. Therefore, the agency concluded that the
changes will not compromise the usefulness of the TRI to the public.
However, EPA's estimate of the impact in terms of national-level aggregate
pounds masks the impact on important toxic chemical information available
to many individual communities and states. We analyzed the impact of EPA's
new Form A thresholds at the local level and found they would allow more
than 3,500 facilities currently submitting Form R to submit Form A
instead. As a result, detailed information about toxic chemical releases
and waste management practices from more than 22,000 of the nearly 79,000
Form Rs could no longer be available to communities throughout the
country. We found that Alaska, California, Connecticut, Hawaii,
Massachusetts, New Jersey, and Rhode Island could receive up to one-third
fewer detailed reports if eligible facilities opt to begin using Form A
instead of Form R. As a result, those states and others may no longer
receive detailed information for a number of chemicals, ranging from 3
chemicals in South Dakota to 60 in Georgia. Finally, we estimated that as
many as 3,565 facilities would no longer have to report any specific
quantitative information about their chemical releases and other waste
management practices to the TRI.

EPA's estimated savings from the reduced reporting burden associated with
the TRI rule--3 percent of total annual burden hours, worth about $6
million annually--are likely overstated. EPA's estimated savings from the
new rule are based on OMB-approved estimates of total burden hours
associated with completing Form R and Form A that are based on outdated
and unreliable data. EPA's more recent engineering estimates--developed
from a systematic examination of the amount of time needed to collect and
report each data element on Form R and Form A--suggest a significantly
lower overall burden associated with current TRI reporting and,
consequently, about 25 percent lower burden savings from the new rule. In
addition, the OMB-approved burden estimates do not give EPA credit for
burden reduction efforts already completed or under way--efforts that,
taken together, go further to reduce reporting burden than the TRI Burden
Reduction Rule, and which do not reduce toxic release information to the
public. For example, EPA estimated that its Toxics Release Inventory-Made
Easy (TRI-ME) software, which helps facilities determine TRI requirements
and submit forms electronically, has reduced the burden associated with
reporting by 15 percent without reducing the amount of information
available to the public. Similarly, EPA's Central Data Exchange, a
computer system that receives electronic submissions from facilities and
disseminates them to participating states automatically, virtually
eliminates time-consuming handling, coding, and reconciling of TRI forms
at the federal and state levels, and thereby reduces the number of errors
that facilities later have to correct.

In a draft of this report, we recommended that EPA thoroughly evaluate the
costs and benefits anticipated to communities and reporting industries
from increased use of TRI Form A and report the agency's findings to
relevant congressional committees within 30 days, along with its
determination as to whether it will reconsider the TRI rulemaking. In
commenting on the draft, EPA's Assistant Administrator for Environmental
Information and Chief Information Officer disagreed with our
recommendation regarding the need for such an analysis and determination,
noting that EPA believes that all appropriate and necessary analyses were
conducted in the rulemaking process. Because EPA did not agree to
implement the recommendation, and in light of the significant problems
with the TRI rule that we identified in this report, we believe the
Congress should consider legislation specifically addressing EPA's
expansion of Form A eligibility. EPA's letter and our detailed response to
it are contained in appendix IV. EPA also provided technical ,
which we have incorporated into this report, as appropriate.

We also provided excerpts of the draft report that discuss OMB's role in
EPA's development of the TRI Burden Reduction Rule. In commenting on the
excerpts, OMB's Deputy Administrator of Information and Regulatory Affairs
raised concerns related to our characterization of OMB's role in reviewing
EPA's burden estimates for the TRI information collections and OMB's role
in EPA's decision to include a burden reduction option to raise the Form A
eligibility threshold in the rule. We acknowledge OMB's concerns, and have
made minor clarifications in the report as appropriate. However, no new
information was provided by OMB that changes the findings, conclusions, or
recommendations. OMB's letter and our detailed response to it are
contained in appendix V.

Background

In 1984, a catastrophic accident caused the release of methyl
isocyanate--a toxic chemical used to make pesticides--at a Union Carbide
plant in Bhopal, India, killing approximately 4,000 people, injuring
thousands of others, and displacing many more from their homes and
businesses. One month later, it was disclosed that the same chemical had
leaked at least 28 times from a similar Union Carbide facility in
Institute, West Virginia. Eight months later, 3,800 pounds of chemicals
again leaked from the West Virginia facility, sending dozens of injured
people to local hospitals. In the wake of these events, the Congress
passed EPCRA.

Among other things, EPCRA provides individuals and communities with access
to information regarding chemical releases in their communities.
Specifically, Section 313 generally requires certain facilities that
manufacture, process, or otherwise use any of 581 individual chemicals,
and 30 additional chemical categories, to report annually the amount of
those chemicals that they released to the environment and whether they
were released into the air, water, or soil.

Owners of facilities that are subject to EPCRA comply with its reporting
requirements by submitting an annual Form R to EPA and the state in which
they are located for each TRI-listed chemical that they manufactured,
processed, and/or otherwise used in excess of certain thresholds during
the previous calendar year. These reports must be submitted on or before
July 1 of the following year. Form R captures information about facility
identity, such as address, parent company, industry type, and detailed
information about the toxic chemical, such as quantity of the chemical
disposed or released on-site to air, water, and land or injected
underground, or transferred off-site for release or disposal. This
information is labeled as "Disposal or Other Releases" on the left half of
figure 1. In 2005, facilities reported a total of 4.5 billion pounds of
disposal or other releases on Form R.

In the Pollution Prevention Act of 1990 (PPA), the Congress declared that
(1) pollution should be prevented or reduced at the source whenever
feasible, (2) pollution that cannot be prevented should be recycled in an
environmentally safe manner whenever feasible, (3) pollution that cannot
be prevented or recycled should be treated in an environmentally safe
manner whenever feasible, and (4) disposal or other release into the
environment should be employed only as a last resort and should be
conducted in an environmentally safe manner. Consequently, beginning with
reports for calendar year 1991, EPA expanded TRI to require facilities to
report about their efforts to reduce pollution at its source, including
the quantities of TRI chemicals they manage through other waste management
activities such as recycling, energy recovery, or treatment. In 2005,
facilities reported 20.1 billion pounds of "Other Waste Management," as
shown on the right half of figure 1.

Figure 1: Summary of Information Reported on TRI Form R

Beginning in 1995, EPA allowed facilities to use a two-page Form A
Certification Statement (Form A) to certify that they are not subject to
Form R reporting for a given chemical provided that they (1) did not
release more than 500 total pounds that year and (2) did not manufacture,
process, or otherwise use more than 1 million total pounds of the
chemical.^4 Form A includes general information identifying the facility
and the chemical(s) being reported but does not contain details about the
specific quantities of chemicals released or otherwise managed as waste.
EPA considers Form A to be a range report, indicating that a facility's
chemical release was between 0 pounds and the maximum allowable amount.
For the purpose of representing range report amounts in the TRI database,
EPA uses the midpoint of the allowable range. Therefore, until reporting
year 2006, the midpoint of the Form A range was 250 pounds. Under EPA's
TRI Burden Reduction Rule, the midpoint became 1,000 pounds for non-PBT
chemicals.

EPCRA requires EPA to make toxic chemical release information available to
the public, and EPA compiles the reports and stores them in a national
database that can be accessed on the agency's Web site.^5 In addition,
about 3 months after facility TRI reports are due, EPA publishes
individual facility submissions in its publicly available Electronic
Facility Data Release (e-FDR). In spring of the following year, after EPA
has completed its data quality checks and analysis, the agency issues the
official Public Data Release. The public may access TRI data on EPA's Web
site and aggregate it by zip code, county, state, industry, and chemical.
EPA also publishes an annual report that summarizes national, state, and
industry data. In 2005, the latest year for which data are publicly
available, 23,461 facilities filed a total of nearly 90,000 forms,
including nearly 11,000 Form As.

In September 2002, EPA initiated a stakeholder dialog process to identify
improvements to the TRI and to develop opportunities to reduce the burden
on reporting facilities. As part of the process, EPA issued a white paper
with five specific options and one general category of other burden
reduction options. At the time, EPA stated that it was looking to more
fully explore these broadly outlined options with the intention of
identifying a specific burden reduction initiative that effectively
lessens the burden on facilities but at the same time ensures that TRI
continues to provide communities with the same high level of significant
chemical release and other waste management information. Each option
included in the white paper was, according to EPA "intended to encourage
thoughtful comment to help the agency develop a meaningful burden
reduction initiative that is technically, practically, and legally
feasible."

^4EPA reporting guidance states that the information contained in the Form
R constitutes a "report," and the submission of a report to the
appropriate authorities constitutes "reporting."

^5 [37]http://www.epa.gov/tri .

At OMB's urging, EPA embarked on a three-phase regulatory effort to
streamline TRI reporting requirements and reduce the reporting burden on
industry. During the first phase, EPA removed some data elements from the
Form R and Form A that could be obtained from other EPA information
collection databases, streamlined other TRI data elements, and eliminated
a few data elements from the Form R. As part of the second phase--and the
focus of this report--EPA proposed a rule that would have allowed a
reporting facility to use Form A for (1) non-PBT chemicals, so long as its
total annual waste management (i.e., releases, recycling, energy recovery,
and treatment for destruction) was not greater than 5,000 pounds, and (2)
for PBT chemicals if there were no releases or other disposal of the
chemical and the annual reportable amount was not greater than 500
pounds.^6 The phase III changes that EPA had considered proposing would
have allowed alternate-year reporting, rather than yearly reporting.

The phase II and III proposals generated considerable public concern about
the negative impact the changes would have on federal and state
governments' and the public's access to important public health
information. EPA received well over 100,000 commenters, representing about
5,000 distinct . The vast majority expressed opposition to EPA's
proposed changes, whereas about 30 commenters expressed support for the
proposals. EPA decided not to pursue phase III in the wake of
overwhelmingly negative reaction from stakeholders--including members of
Congress, various state attorneys general, and researchers--but the agency
continued to pursue the phase II burden reduction rule.^7 On December
2006, EPA finalized the rule, extending Form A to PBT chemicals as it had
proposed. However, in the final rule, the agency moderated its proposed
increase in the Form A non-PBT threshold from 500 to 5,000 pounds, thereby
allowing Form A for a non-PBT chemical provided the release comprises no
more than 2,000 pounds of the overall 5,000 pound total waste management
limit.

^6The annual reportable amount is the combined total quantity released at
the facility, treated at the facility, recovered at the facility as a
result of recycling operations, combusted for the purpose of energy
recovery at the facility, and amounts transferred from the facility to
off-site locations for the purpose of recycling, energy recovery,
treatment, and/or disposal.

^7In a November 28, 2006, letter to Senator Lautenberg, the EPA
Administrator announced that the agency had decided against moving forward
with any changes to TRI reporting frequency. The letter did not specify
EPA's reasons for abandoning the phase III initiative.

EPA's Action Development Process is a 70-page guidance document designed
to improve how the agency develops actions such as the TRI Burden
Reduction Rule.^8 According to EPA, the process contains the following
five goals:

           o planning sound scientific and economic analysis to support the
           action;
           o developing and selecting regulatory and nonregulatory options
           based on relevant scientific, economic, and policy analysis;
           o involving affected headquarters and regional managers early in
           development and until the final action is completed;
           o ensuring active and appropriate cross-agency participation; and
           o encouraging appropriate and meaningful consultation with
           stakeholders through substantive consultative procedures.

           To accomplish these goals, the Action Development Process outlines
           five major stages, each with multiple steps that are to be
           followed. A critical step early in the process is chartering a
           workgroup comprised of representatives from interested EPA
           headquarters offices and regions to develop the action. The
           various steps provided in the Action Development Process are
           directed to the workgroup and their managers who provide policy
           direction and ensure the integrity of the process. Throughout the
           rule development process, senior EPA management generally has the
           discretion to depart from the guidelines, including by
           accelerating the development of the proposed regulations.

           Under the Paperwork Reduction Act (PRA), a federal agency may not
           conduct or sponsor the collection of information unless OMB has
           approved the agency's information collection request.^9
           Information collection requests lacking current OMB approval are
           invalid under the PRA.^10 Accordingly, EPA submits an information
           collection request for the TRI program to OMB every 1 to 2 years.
           Before approving information collections, OMB is required to
           determine that the agency's collection of information is necessary
           for the proper performance of the functions of the agency,
           including whether the information will have practical utility.^11
           As part of EPA's information collection requests for the TRI, OMB
           approves the agency's estimates of the burden associated with
           completing the Form R and Form A.

^8Other types of actions that are covered by the guidance include policy
statements, risk assessments, guidance documents, or models that may be
used in future rulemakings.

           The TRI Is Widely Used by Federal Agencies, States, and the Public

           Numerous EPA offices, other federal agencies, the states, and the
           public use the TRI for a wide range of activities, from assessing
           fees on regulated facilities to identifying the location of toxic
           releases in local communities. One of the TRI's most notable
           attributes is the applicability of its data across air, water,
           solid waste, and other media, making it useful to a wide range of
           federal and state agencies and the public.
			  
			  EPA Offices and Other Federal Agencies Use the TRI for Multiple Purposes

           EPA and other federal agencies use TRI data for a variety of
           purposes, including developing guidelines, standards, and programs
           for controlling pollution from regulated entities, assessing the
           effectiveness of environmental regulations, and planning office
           priorities. Some agencies may rely solely on TRI data while
           others, particularly the agency's media program offices, have used
           its data to supplement their own data for a variety of purposes.

           EPA's Office of Prevention, Pesticides, and Toxic Substances, for
           example, used TRI data to create its RSEI model that estimates the
           impacts associated with air and water releases from TRI
           facilities, taking into account the human health hazard and risk
           associated with different chemical substances. RSEI's models use
           TRI data on the location and amount of chemical releases,
           adjusting for key factors such as the toxicity of the chemicals,
           their destination and transport through the environment, the route
           and potential extent of human exposure, and the number of people
           potentially affected. RSEI then creates numerical values that can
           be added and compared in different ways to assess the human health
           hazards and related risks of chemicals, facilities, regions,
           industries, and other parameters.

^944 U.S.C. S 3507(a).

^10The Paperwork Reduction Act provides that no person shall be subject to
any penalty for failing to maintain or provide information to any agency
if the information collection request is invalid. 44 U.S.C. S 3512.
However, this provision is inapplicable to information explicitly required
by statute, and accordingly does not relieve a facility of its reporting
duty under EPCRA even in the absence of a valid TRI information collection
request. Gossner Foods, Inc. v. Environmental Protection Agency, 918
F.Supp. 359, 362-63 (D. Utah 1996).

^1144 U.S.C. S 3508.

           Similarly, EPA's Office of Policy, Economics, and Innovation
           (OPEI) uses the TRI to, among other things, measure trends in the
           environmental performance of industries participating in its
           Sector Strategies program, an industry-EPA partnership
           administered by OPEI's Office of Business and Community Innovation
           that seeks innovative ways to improve environmental performance
           among participating industry sectors. That office then includes
           the data in its Sector Strategies Performance Report, which
           informs the public about changes in environmental performance of
           these sectors.

           In addition, EPA's key media programs, including its Office of Air
           and Radiation, Office of Water, and Office of Solid Waste and
           Emergency Response, use TRI data to supplement their own data to
           fulfill key core responsibilities. Specifically:

           o Office of Air and Radiation uses TRI data as a quality assurance
           tool when filling in missing data in its triennial National
           Emissions Inventory, which provides data on air emissions from
           about 85,000 sources and is used to assess risks from hazardous
           air pollutants at the local, regional, and national levels. The
           program has also used TRI data to assist in the development of
           emission standards required by the Clean Air Act.
           o Office of Water compares TRI facilities with those in its Permit
           Compliance System, a database that tracks facilities permitted to
           discharge releases into water. Among other things, the office also
           uses TRI data as an input in its watershed analysis software,
           allowing it and other stakeholders to analyze water quality at a
           select stream site or throughout an entire watershed.
           o Office of Solid Waste and Emergency Response uses TRI data in
           its National Priority Chemicals Trends Report, which analyzes and
           evaluates 24 highly toxic chemicals found in industrial wastes in
           the United States. This report tracks progress toward achieving
           EPA's national goal to reduce the amount of highly toxic chemicals
           in waste and to identify these chemicals through a variety of
           methods.
           o Other key EPA offices making wide use of the TRI include the
           Office of Enforcement and Compliance Assurance, which has used TRI
           data to locate pollution sources that may be out of compliance
           (among other purposes) and the agency's regional offices, which
           use the data extensively in their interactions with both their
           regulated entities and the public at large.

           Finally, other federal agencies that make use of the TRI include
           (1) the IRS, which has used TRI data to identify companies that
           release chlorofluorocarbons (known to deplete the earth's ozone
           layer), and to enforce a tax on those releases as part of a plan
           to phase out these environmentally hazardous chemicals; (2) the
           Census Bureau's Center for Economic Studies, which has included
           TRI data in reports that highlight changes in the U.S. economy;
           (3) the Centers for Disease Control and Prevention's Agency for
           Toxic Substances and Disease Registry, which has used TRI data on
           air and water releases in its local-level health consultations and
           assessments; and (4) the Small Business Administration, which uses
           TRI data for evaluating other environmental regulations.
			  
			  Most States Have TRI Programs and Use TRI Data for a Range of Activities

           EPCRA requires that facilities submit their TRI forms to EPA and
           their state. We surveyed TRI coordinators in the 50 states and the
           District of Columbia to better understand how states use the TRI
           data they receive.^12 We found that the states have TRI programs
           that carry out a range of TRI-related activities. The state
           programs vary on several key parameters, including whether they
           have statutory requirements for chemical release reporting, how
           they fund TRI-related activities, how much they spend annually on
           TRI-related activities, and whether they assess fees on facilities
           that report to the TRI. Table 1 shows these key characteristics
           for each state. We provide a complete tabulation of the results of
           our survey in an electronic supplement.^13
			  
^12Throughout the remainder of our report, we refer to our survey of state
TRI coordinators simply as the survey of states. Unless otherwise
specified, our discussion includes responses from 51 coordinators from the
states and the District of Columbia.

^13GAO, Toxic Chemical Releases: Survey of State Toxics Release Inventory
Coordinators, [38]GAO-08-129SP (Washington, D.C.: Oct. 26, 2007).

           Table 1: Key Characteristics of State TRI Programs
			  
                               Source of                    Annual   Assess   
State               Statute funds                expenditures^a    fees    
Alabama                     general                         $$$            
Alaska                      dedicated                         $            
Arizona                o    general                           $            
Arkansas                    general                         $$$     o      
California                  dedicated                         $            
Colorado                                                      $     o      
Connecticut            o    general                         n/a            
Delaware               o    general                         $$$            
District of                 general                           $            
Columbia                                                                   
Florida                                                      $$     o      
Georgia                     dedicated                         $     o      
Hawaii                 o                                      $            
Idaho                                                       n/a            
Illinois               o    dedicated                        $$            
Indiana                     general                          $$            
Iowa                        general                           $            
Kansas                      general                          $$     o      
Kentucky                    general                           $            
Louisiana                                                   n/a            
Maine                  o                                      $     o      
Maryland                    general                           $     o      
Massachusetts         o^b   general                       $$$$$     o      
Michigan                    dedicated                       n/a            
Minnesota             o^b   general                         $$$     o      
Mississippi                 general                         $$$            
Missouri               o                                      $            
Montana                                                       $            
Nebraska                    general                          $$            
Nevada                      general                           $     o      
New Hampshire               general                         n/a            
New Jersey            o^b                                    $$            
New Mexico                                                    $            
New York              o^b                                   n/a            
North Carolina                                              n/a            
North Dakota                general                         n/a            
Ohio                   o    dedicated                      $$$$     o      
Oklahoma                                                    $$$            
Oregon                      dedicated                         $            
Pennsylvania           o    general                         $$$     o      
Rhode Island                                                n/a            
South Carolina                                                $            
South Dakota                                                n/a     o      
Tennessee                   general                         $$$            
Texas                  o    general                           $     o      
Utah                        general                         $$$            
Vermont                     general                           $            
Virginia                    general                          $$            
Washington                  dedicated                       $$$            
West Virginia          o                                      $            
Wisconsin              o                                      $            
Wyoming                     general                           $      			  

           Source: GAO survey of state TRI coordinators.

           ^a$=Less than $25,000; $$=$25,000-$50,000; $$$=$50,001-$150,000;
           $$$$=$150,001-$250,000; $$$$$=more than $250,000; n/a=no answer
           provided.

           ^bState has statute that requires facilities to report additional
           chemicals or provide information not required by EPCRA.

           As table 1 also shows, officials in 17 states reported that their
           states have statutes that have similar requirements to EPCRA for
           toxic release reporting elements. Of note are four states with
           statutes that require facilities to report additional chemicals or
           provide information not required by EPCRA--Massachusetts,
           Minnesota, New Jersey, and New York. For example, Massachusetts
           officials stated that facilities must report information, such as
           the risk of chemicals used at the facility or present in products,
           and the number of employees at the facility. In addition, New
           Jersey officials stated that facilities must report on their
           chemical use, including information about how chemicals travel
           through the facility's processes. The states also vary in how they
           fund TRI programs, with 25 states using general funds and 8 states
           using dedicated funds for TRI. With regard to total amount spent
           on TRI-related activities, most states (29) reported spending
           $50,000 or less in fiscal year 2006, and 12 states spent more than
           $50,000. Fourteen states reported assessing fees on facilities
           that submit to the TRI, 11 of which dedicated those funds to their
           TRI programs. States base their fees on one or more of the
           following criteria: which TRI form is submitted (Form R or Form
           A); the number of employees at the facility; the amount of a
           chemical reported; the type of chemical reported; and when the
           form is submitted. As a consequence, changes to federal TRI
           reporting requirements that affect any of those criteria may
           consequently impact a source of state funding for TRI-related
           activities.

           We also asked the TRI coordinators to describe how their state
           uses the information from the TRI forms that it receives. Although
           a few states reported doing little or nothing with the reports,
           many states make wide use of the data contained therein. Figure 2
           provides their responses, which fall into one of three categories:
           (1) chemical identification or pollution prevention, (2)
           compliance assistance or program enforcement, and (3) public
           information or other data services.

           This page intentionally left blank

Figure 2: TRI Activities Reported by States

  Chemical Identification or Pollution Prevention Activities

States reported using TRI data for chemical identification or pollution
prevention activities such as identifying which facilities release large
volumes of chemicals or the location of chemical releases posing the
greatest risk to public health. States also reported using TRI data to
monitor facility pollution prevention efforts, establish state permit
limits, monitor facility performance in a pollution reduction program, and
publicly recognize facilities that reduce pollution. State coordinators
provided us with specific examples of pollution prevention programs that
used TRI data. Following are examples:

           o Oklahoma used TRI data to identify companies that used hazardous
           chemicals as inputs in their manufacturing processes and work with
           them to substitute less hazardous chemicals.
           o New Jersey used TRI data to identify facilities that reported
           releasing high volumes of reproductive toxics (i.e., chemicals
           that can damage human reproductive systems) for participation in a
           pollution reduction program.
           o Colorado used TRI data to identify the state's 40 largest toxic
           chemical releasers to participate in the Governor's Pollution
           Reduction Challenge. The program encourages facilities to optimize
           their production processes to reduce emissions, improve their
           recycling capabilities, and add pollution control technology, and
           it has resulted in reductions in overall emissions.
			  
			    Compliance Assistance or Program Enforcement

           The second most common category of state TRI-related activities
           relates to compliance assistance and program enforcement. Most
           states helped facilities comply with TRI reporting requirements in
           some way, including answering facilities' questions via phone or
           e-mail (48 states) or referring facilities to EPA for assistance
           (46 states). Over half the states informed facilities about
           reporting requirements (37 states), distributed TRI forms or
           materials (31 states), or participated in EPA-sponsored
           information or training sessions (30 states). About one-third of
           states conducted state-sponsored information or training sessions
           (16 states) or visited reporting facilities (14 states). In
           addition to compliance assistance for TRI reporting, states used
           TRI data to assist with enforcement of other state environmental
           regulations by, for example, comparing TRI data with other
           databases and identifying facilities that must comply with state
           environmental regulations. For example, Ohio has cross-checked
           facilities' TRI submissions with their Resource Conservation and
           Recovery Act (RCRA) reports, National Pollutant Discharge
           Elimination System permits, and air permits to check for possible
           noncompliance violations and to target facilities for inspection.
           Forty-one states reported using TRI data for at least one of the
           enforcement activities listed in figure 2.
			  
			    Public Information or Other Data Services

           States most commonly reported using TRI data to provide
           information about toxic chemical releases to the general public
           and for providing data services to other state programs. Forty-one
           states reported providing TRI-related information or referrals to
           the public through annual TRI reports, TRI Web pages, state TRI
           data files, data analyses conducted upon request, copies of Form
           Rs from individual facilities, public reading rooms with TRI
           information, EPA TRI documents, referrals to EPA regional TRI
           contacts, and other means (not listed as a response option on the
           survey) such as printed fact sheets, audio for radio interviews,
           and TRI-related CDs. For example, Ohio prepares a report listing
           the facilities with the greatest releases, the most common TRI
           chemicals, and the counties with the largest releases. Indiana
           correlates TRI releases and waste management practices with its
           gross state product, a measure of the state's economic output, to
           provide better context for the public to understand TRI data. The
           state TRI coordinator for Colorado reported using the TRI as part
           of a state mercury program that sought, in part, to increase
           public awareness of mercury releases to inform the public about
           possible health hazards of mercury, a PBT. In addition, more than
           two-thirds of states reported responding to data requests from
           other state programs (36 states) or comparing state TRI data with
           other databases (31 states). States also integrated TRI data with
           a geographic information system (GIS) or other state mapping
           capabilities (14 states), or used TRI data to make environmental
           justice assessments (7 states).
			  
			  The Public Uses the TRI for a Variety of Purposes

           Academic researchers, environmental groups, educational
           organizations, businesses, and public interest groups use the TRI
           for a variety of purposes. The public can generally access TRI
           data from EPA's Web site, as well as from nonprofit organizations'
           Web sites. EPA officials told us that the public accessed the TRI
           database through its Web site 829,682 times during the 12-month
           period ending March 2007.

           Academic researchers. According to an EPA summary of TRI users,
           universities and research institutions used TRI data as a means
           for "examining environmental policies and strategies, and
           clarifying risks associated with toxic chemicals at the state and
           local level." Through an EPA program called Science to Achieve
           Results that awards grants to scientific researchers, 25 grantees
           have used the TRI to study
           environmental performance. For example, one grantee used TRI data
           to study areas with cancer risks and found that high-risk areas
           are concentrated around certain aluminum and cement industries. In
           addition, a public policy researcher used TRI data to study the
           accuracy of self-reported regulatory programs. This researcher
           found a strong correlation between estimated cancer cases in a
           region, rates of voter turnout, and the likelihood that facilities
           in that region reduce emissions.

           Educational organizations. According to EPA, various education
           institutions have integrated the TRI into curricula so students
           can use it in the classroom. For example, the National Science
           Teachers Association included the TRI in an instructional resource
           guide for high school teachers as an example of how science can be
           used in an everyday context. In addition, the Delaware Department
           of Natural Resources designed lessons for high school and middle
           school students to learn about the effects of emissions on air
           quality and how to locate facilities that report air emissions to
           the TRI.

           Private business. Businesses also use the TRI to assist with
           decision making and chemical release monitoring. Businesses within
           the regulated community that report to the TRI have used their own
           reports to achieve gains in cost reduction and performance
           management. For example, Dupont lists its TRI data on its Web site
           and uses its progress in emissions reductions as a marketing tool.
           Boeing also tracks its progress at reducing TRI emissions and
           invests in pollution prevention technology that has resulted in
           more than 81 percent reductions in emissions since 1991. For some
           managers and operators at these businesses, TRI reporting has
           increased their awareness of the quantity of chemicals released
           from their facilities, and they have used TRI to set goals for
           reducing their chemical releases, sometimes resulting in increased
           efficiency, greater profits, identification of pollution
           prevention opportunities, and evidence of a public commitment to
           reduce those releases. Labor unions that represent employees who
           work at TRI facilities have used TRI data to support demands for
           safer working conditions and have trained their members to access
           and interpret TRI data. Outside of the regulated community,
           investment companies have used TRI data to advise clients who want
           to invest in companies with a record of reducing environmental
           releases. An adviser at one investment firm told us that TRI data
           have been useful in measuring companies' overall environmental
           performance, which includes their compliance with regulations and
           their overall emissions.

           Public interest groups. Many organizations use the TRI to educate
           citizens about toxic releases in their communities and to empower
           citizens to take actions that can reduce those releases. At the
           local and state levels, for example, the Citizen's Environmental
           Coalition in New York maintains a Web site with an interactive
           mapping tool for citizens to view areas that may be cause for
           environmental concern, including the location of TRI facilities.
           In Wisconsin, the Oneida Environmental Resources Board used TRI
           data to convince leaders in the Oneida tribe to find cleaner ways
           to manufacture pulp and paper. At the national level,
           Environmental Defense uses TRI data in its Web-based "pollution
           locator," which allows users to compare states and communities by
           criteria such as the presence of lead concentrations, chemicals
           known to cause birth defects, and chemicals known to cause cancer.
           OMB Watch, an organization that seeks to increase government
           transparency, maintains the Right-to-Know Network (RTK Net), a Web
           site that provides links to 11 environmental databases, including
           the TRI. The National Environmental Trust, an organization that
           seeks to inform citizens about environmental problems and the
           effects of those problems on human health, conducts TRI data
           analyses and tracks TRI program developments, information it makes
           publicly available on its Web site. Physicians for Social
           Responsibility, a nonprofit health and environmental advocacy
           organization, has used TRI data in reports that describe the
           threats that children face from different types of pollution.
			  
			  EPA Did Not Follow Key Steps for Developing the TRI Burden Reduction Rule

           EPA did not follow key steps from its Action Development Process
           (ADP) when developing the proposed TRI Burden Reduction Rule. This
           process guides the internal development of proposed EPA
           regulations through a series of milestones to ensure that the
           agency uses sound information to support its actions, and that it
           adequately addresses scientific, economic, and policy issues.
           Throughout the rule development process, senior EPA management
           generally has the discretion to depart from the guidelines,
           including by accelerating the development of the proposed
           regulations. However, in reviewing the process EPA followed when
           developing the rule, we found several significant differences that
           resulted in inadequate input from internal stakeholders and
           insufficient analytical support for the proposed rule's burden
           reduction options.
			  
			  Internal Stakeholders Had No Opportunity to Comment on Effects of
			  Selected Rulemaking Option

           As part of the first stage in ADP, the agency assigns an action to
           one of three tiers that determine the process the agency uses when
           developing the rule. On March 30, 2004, EPA approved the TRI rule
           as a tier 2 action--targeting it for extensive cross media or
           cross-agency involvement and resting primary decision authority
           with the Assistant Administrator for Environmental Information.
           Next, EPA charters a workgroup to develop the specific action. The
           workgroup that was assembled to shepherd the TRI Burden Reduction
           Rule through the ADP was chartered in late May 2004 and led by a
           representative of EPA's Office of Environmental Information (OEI),
           which manages the TRI program. Other workgroup members included at
           least one representative from EPA's Office of Solid Waste and
           Emergency Response, 3 of the 10 Regional Offices, the Office of
           General Counsel, and several additional representatives from OEI.

           In December 2004, pursuant to ADP, the TRI workgroup completed a
           Preliminary Analytic Blueprint, a document that outlined the five
           burden reduction options they planned to consider. These options
           were the following:

           o a higher reporting threshold for small businesses;
           o a higher reporting threshold for categories of facilities or
           classes of chemicals with small reportable amounts;
           o expanded eligibility for Form A (by raising the maximum release
           threshold for non-PBTs from 500 to 5,000 pounds);
           o a new No Significant Change (NSC) Certification Statement for
           facilities with releases that changed less than a specified
           amount; and
           o use of range reporting on Form R for releases and waste
           management practices (e.g., indicating that releases were between
           100 and 499 pounds).

           In April 2005, pursuant to the ADP, workgroup members submitted a
           Detailed Analytic Blueprint, which narrowed the list to three
           options for which they would prepare further analyses, seek
           internal stakeholder input, and forward to senior management for
           final consideration. Importantly, none of these three options
           would have expanded Form A eligibility for non-PBT chemicals.

           The first two options would have allowed facilities to use Form A
           in lieu of Form R for PBT chemicals, provided the facility had no
           releases to the environment. The only difference between the two
           options was whether the facility could have a limited quantity of
           other waste management activities (e.g., recycling). Specifically,
           these two options were to allow facilities to

           o report PBT chemicals using Form A if they have zero releases and
           zero total other waste management activities, or
           o report PBT chemicals using Form A if they have zero releases and
           no more than 500 pounds of other waste management activities.

           These PBT options received general approval from the EPA offices
           involved in the workgroup because they would have little effect on
           the TRI. That is, the public would not have less information about
           releases because neither option allowed facilities to use Form A
           if they released any of the chemical. However, the PBT options did
           not provide the bulk of the anticipated burden reduction.

           The third option--which provided significantly more burden
           reduction--would have created a new NSC Certification Statement,
           in lieu of Form R, for facilities whose releases changed little
           from the previous year. Facilities eligible to file the NSC
           Certification Statement could do so during alternate years if
           their releases and waste management practices did not change by
           more than 10 percent or if less than 40 percent of their combined
           releases and waste management practices were releases. According
           to EPA officials on the TRI workgroup, NSC was the most discussed
           option, and the one that the workgroup had the highest
           expectations for significant burden reduction.

           In accordance with the ADP, the TRI workgroup presented their
           three options to the EPA Administrator during an "Options
           Selection Briefing." However, based on internal documents that we
           reviewed, senior EPA management not involved with the workgroup or
           its analyses had begun considering a different option than the
           ones the workgroup had presented. Specifically, the briefing
           slides for the Administrator stated that OMB's preferred burden
           reduction option was to increase the Form A eligibility threshold
           for non-PBT chemicals from 500 to 5,000 pounds. The TRI workgroup
           had dropped that option from consideration, at an earlier step in
           the ADP, because of its impact on the TRI. An internal memorandum
           that we reviewed stated that OMB preferred a 10-fold increase in
           the non-PBT threshold to show a "demonstrable threshold increase"
           as a way of achieving a "sizable reduction" in reporting burden
           beyond expansion of Form A reporting for PBT chemicals. Those
           documents also showed that, after the Options Selection Briefing,
           the Administrator directed that the process be expedited through
           Final Agency Review in order to meet EPA's commitment to OMB to
           provide burden reduction by the end of December 2006.

           The next milestone in EPA's ADP is Final Agency Review, a meeting
           for internal and regional EPA offices and senior management (i.e.,
           representatives of EPA's Assistant Administrators) to review the
           draft proposed rule and discuss whether they concur, concur with
           comment, or do not concur with it. To meet the EPA Administrator's
           deadline, the final agency review package--including the draft
           text of the proposed rule--was distributed to internal
           stakeholders to obtain their  on June 17, 2005. At the
           June 29, 2005, Final Agency Review meeting, all Assistant
           Administrators (or their representatives) concurred with EPA's
           proposed rule. However, the proposed rule that was circulated in
           the final agency review package contained only the options that
           the workgroup had developed and advanced. Consequently, the Final
           Agency Review discussion and concurrence pertained to the PBT and
           NSC options that the TRI workgroup had developed, rather than the
           increased non-PBT threshold option that OMB favored.

           By the time EPA held the Final Agency Review meeting, the TRI
           workgroup was working to create the NSC form and establish
           eligibility criteria to determine which facilities could use it.
           This option was based on the premise that the information on many
           facilities' Form Rs does not change much from year to year,
           especially as a percentage of reported releases. However, the
           workgroup's internal analysis also showed that Form Rs do change
           from year to year, especially for facilities with large releases
           and large quantities of managed waste. That analysis showed that,
           for facilities with large quantities of releases or other waste
           management practices, a 5 or 10 percent increase would represent a
           significant change in the absolute quantity of release or other
           waste management. Moreover, Form R captures the location of
           facilities' off-site transfers for disposal or recycling to a
           different location. Therefore, changes in waste management
           practices during the "no significant change year" would no longer
           be available to the public. The workgroup also was considering
           incentives to allow facilities participating in EPA's pollution
           prevention/reduction program to file two consecutive NSC
           Certification Statements before having to file a new Form R.

           At the senior management level, discussions about increasing the
           Form A threshold from 500 to 5,000 pounds continued through July
           with little, if any, input from the TRI workgroup, according to
           internal documents we reviewed and interviews with staff from
           EPA's Office of Environmental Information. Consequently, the
           economic analysis that the workgroup finalized in late July 2005
           did not examine burden reduction option to raise the non-PBT
           threshold to 5,000 pounds. In early August, the final package for
           the proposed rule was circulated to the workgroup, but the rule
           did not include the non-PBT option or an analysis of the option.
           As a result, TRI workgroup members, and other internal
           stakeholders from the program offices, did not have the
           opportunity for input because the senior EPA management had
           already decided which options would and would not be in the
           proposed rule that went to OMB for review on August 26, 2005.

           EPA program staff ultimately revised the economic analysis so as
           to consider the impact of raising the Form A reporting threshold.
           However, their analysis was neither completed before EPA sent the
           proposed rule to OMB for review nor was it circulated for review.
           Instead, the analysis was completed just prior to the September
           21, 2005, date when the EPA Administrator signed the proposed rule
           for increasing Form A reporting eligibility from 500 to 5,000
           pounds. EPA published this proposal in the Federal Register for
           public comment on October 4, 2005. The accompanying economic
           analysis did not adequately support EPA's stated benefits or
           costs. Specifically, although EPA stated in the proposed rule that
           it provided significant incentives for facilities to reduce or
           eliminate releases (of PBT chemicals, especially), the agency's
           economic analysis did not attempt to estimate those incentives
           quantitatively, citing lack of data. Moreover, EPA's analysis did
           not estimate the value of the information that would no longer be
           reported to the TRI.
			  
			  Public  on EPA's Proposed Rule Were Overwhelmingly Negative

           The agency received well over 100,000  in response to the
           proposed rule, most of which were overwhelmingly negative. Some
           commenters supported EPA's proposed option to extend Form A to PBT
           chemicals because it would provide burden relief, but no actual
           release data would be lost. Some commenters also stated that the
           proposal would not compromise public health or reduce the ability
           to plan for emergency responses and that most people are
           interested solely in releases to the environment. Other commenters
           suggested that EPA's proposal would encourage pollution
           prevention, as facilities would work to eliminate releases and
           minimize waste generation of PBT chemicals in order to quality for
           Form A. However, as EPA pointed out in its summary of the
           , many more commenters expressed opposition to the
           proposed option for allowing Form A for PBT chemicals because the
           proposal provided minimal burden reduction while losing important
           publicly available data. As an example, EPA highlighted a
           commenter that estimated that the average cost savings per
           facility would be only $1,035, which the commenter argued does not
           justify the expected loss of information from the rule. Another
           commenter estimated that 77 percent of facilities eligible to use
           Form A for PBTs currently report zero for both releases and other
           waste management and, therefore, would not save burden by
           switching to Form A. Other  disagreed and asked that EPA
           instead expand Form A for reporting small, nonzero releases of PBT
           chemicals. In the final rule, EPA responded to these  by
           reiterating its belief that the rule would still result in
           significant burden reduction without losing crucial information.

           Commenters who supported EPA's proposed expansion of Form A
           eligibility for non-PBT chemicals asserted that the proposed rule
           would provide significant burden relief from TRI
           reporting--especially for small facilities. These proponents argue
           that this relief would be significant despite the need to
           calculate releases and other waste management amounts to determine
           if they quality for Form A. However, many more commenters
           expressed opposition to the proposed option to expand Form A
           eligibility for non-PBT chemicals by raising the threshold from
           500 to 5,000 pounds. These commenters, including 12 state
           attorneys general, focused on the local-level impacts from the
           detailed chemical release and waste management information that
           would no longer be reported on Form R. These and other commenters
           recognized that the potential nonreporting represented less than 1
           percent of total release and waste management quantities reported
           nationwide on Form R but argued that a 5,000-pound Form A would
           adversely affect the ability of data users to perform local trend
           analyses, monitor the performance of individual facilities and,
           more generally, meet the intended purpose of the data collection
           to inform the public, government, and other data users about
           releases of toxic chemicals to the environment.^14 Many commenters
           gave specific examples of the local data use that could be
           affected by the proposed rule, such as identifying
           pollution-prevention opportunities, conducting risk analyses,
           identifying trends in toxic exposure, conducting spatial analyses
           of toxic hazards, setting environmental and public health policy,
           and evaluating trends in the performance of individual companies.

           In response to these public commenters, EPA conducted an
           additional analysis of the impact of its changes at the local
           level and made several modifications before finalizing the TRI
           Burden Reduction Rule in December 2006. Specifically, EPA
           finalized the rule to raising the Form A eligibility threshold for
           non-PBT chemicals to 5,000 pounds of total annual waste management
           (i.e., releases, recycling, energy recovery, and treatment for
           destruction) provided total annual releases of the non-PBT
           chemical comprise no more than 2,000 pounds of the 5,000-pound
           total waste management limit. The agency also included its
           proposed option to allow use of Form A for PBT chemicals when
           total annual releases of a PBT chemical are zero, and the total
           annual amount of the PBT chemical recycled, combusted for energy,
           and treated for destruction does not exceed 500 pounds.
			  
^14Form A essentially serves as a range report, revealing to the public
that the facility released between 0 and 5,000 pounds of a non-PBT
chemical.

           EPA Changes Could Significantly Limit the TRI Data Available to Many
			  Communities

           Our analysis shows that EPA's TRI Burden Reduction Rule could, by
           increasing the number of facilities that may use Form A,
           significantly reduce the amount of information currently available
           to many communities about toxic chemicals used, transported, or
           released in their environment. EPA estimated that the impact of
           its change to TRI reporting requirements would be minimal;
           amounting to 5.7 million pounds of releases (0.14% of total
           release pounds) and 10.5 million pounds of waste management
           activities (0.06% of total waste management pounds) not being
           reported to the TRI if all eligible facilities switch from Form R
           to Form A. However, our analysis shows that EPA's changes could
           have far more significant impacts on information available to
           communities than EPA's national aggregate totals would appear to
           indicate. In addition, our survey of state TRI coordinators
           indicates that EPA's changes will have, on balance, a negative
           impact on state TRI programs and other users of the TRI. We
           acknowledge that not all eligible facilities will take advantage
           of the ability to file Form A, based on historical rates of Form A
           filing. However, because of the many assumptions necessary to
           calculate a "likely" impact, we present the total possible impact
           on the TRI under EPA's new Form A eligibility rules.
			  
			  Thousands of Detailed Form R Reports May No Longer Be Submitted to the TRI

           We estimated that 22,200 Form R reports (28 percent) could convert
           to Form A under EPA's new Form A thresholds.^15 EPA has observed
           that facilities used Form A for only 54 percent of the Form R
           reports potentially eligible under the previous threshold.
           According to EPA, eligible facilities may choose not to submit a
           Form A for a number of reasons. First, an unknown number of
           facilities may exceed the 1 million pound alternative threshold
           (e.g., facilities that use large quantities of feedstock chemicals
           to produce pesticides or pharmaceuticals) and, therefore, are
           ineligible for Form A.^16 EPA does not know how many facilities
           exceed the alternative threshold because that information is not
           reported on current TRI forms. For this reason, EPA's utilization
           rate is likely an underestimate. Second, some facilities report on
           Form R out of a desire to showcase their pollution prevention
           efforts. Third, a facility, having collected all this information,
           may also submit a Form R to demonstrate good environmental
           stewardship. Fourth, some facilities find the Form R to be an
           efficient mechanism for tracking their material balances. Last,
           EPA and industry officials also told us that facilities are less
           likely to submit a Form A in lieu of a Form R for a given chemical
           if they must also submit Form Rs for other chemicals. We
           understand that not all eligible facilities will take advantage of
           EPA's new thresholds for one or more of these reasons. Although
           the agency stated that the utilization rate will not likely be
           significantly higher under the new threshold, additional
           facilities that were formerly eligible to file Form A may choose
           to file Form R now that they are eligible to do so for more of the
           reports. Given the uncertainties in projecting a "likely"
           utilization rate, we present our results in terms of the total
           number of Form R reports that are currently eligible to be filed
           on a Form A under the thresholds provided for in EPA's TRI rule.

^15We estimated that approximately 11,700 new and 10,500 formerly-eligible
Form R reports could convert to Form A under EPA's increased Form A
thresholds.

^16EPA cannot determine with certainty whether a facility exceeded the 1
million pound threshold, because facilities are not required to report
quantities of a chemical that they manufactured, processed, or otherwise
used.

           According to our analysis, the number of Form Rs that may no
           longer be submitted ranges by state from 25 forms in Vermont (27.2
           percent of Form Rs in state) to 2,196 forms in Texas (30.6 percent
           of Form Rs in state). As figure 3 shows, Arkansas, Idaho, and
           Nevada, North Dakota, and South Dakota could have up to 20 percent
           fewer of the detailed forms, while Alaska, California,
           Connecticut, Georgia, Hawaii, Illinois, Maryland, Massachusetts,
           New Jersey, New York, North Carolina, Rhode Island, and Texas
           could have at least 30 percent fewer Form Rs. We provide estimates
           of these impacts, by state, in appendix II.

Figure 3: Estimate of Impact Allowed by EPA's Changes on Number of Form
Rs, by State

For each facility that chooses to file a Form A instead of Form R, the
public would no longer have available quantitative information about a
facility's releases and waste management practices for a specific chemical
manufactured, processed, or otherwise used at the facility.^17 Form R and
Form A both capture information about a facility's identity, such as
mailing address, parent company, and basic information about a chemical's
identity, such its generic name. However, only Form R provides detailed
information about the chemical, such as quantity disposed or released
on-site to air, water, and land, or injected underground, or transferred
for disposal or release off-site. Form R also provides information about
the facility's efforts to reduce pollution at its source, including the
quantities managed in waste, both on- and off-site, such as amounts
recycled, burned for energy recovery, or treated. We provide a detailed
comparison of the data captured on Form R versus Form A in appendix III.

^17According to EPA, Form A serves as a range report, informing the public
that a facility filing a Form A for a specific non-PBT chemical has total
annual releases of that chemical in the range of zero to 2,000 pounds and
total waste management (which includes releases) in the range of zero to
5,000 pounds.

Data about Toxic Chemicals May Be Reduced or Eliminated for Many Communities

One way to capture the impact of Form Rs converting to Form A is to
examine what currently available public data could no longer be reported
about specific chemicals at the state level. The number of chemicals for
which only Form A information could be reported under the new rule ranges
from 3 chemicals in South Dakota to 60 chemicals in Georgia. That means
that the specific quantitative information currently reported about those
chemicals on Form R may no longer be included in the TRI. Figure 4 shows
that 13 states--Delaware, Georgia, Hawaii, Iowa, Maryland, Massachusetts,
Missouri, North Carolina, Oklahoma, Tennessee, Vermont, Wisconsin, and
West Virginia--may no longer receive specific quantitative information
about at least 20 percent of TRI-reported chemicals in the state.

Figure 4: Estimate of Percentage of Chemicals for Which Facilities Could
Report on Form A, by State

The impact of EPA's change on toxic chemical information available to many
local communities may be more significant than indicated by national or
state estimates. As figure 5 shows, citizens in more than 1,700 counties
may no longer receive detailed information about at least one toxic
chemical currently reported on Form R in their county. We estimated that,
as a result of the allowable reduction in number of reports, citizens in
64 counties across 28 states--Texas (10); Virginia (7); Colorado,
Kentucky, and Mississippi (4 each); California, Georgia, and Missouri (3
each); Florida, Iowa, Illinois, Louisiana, North Carolina, New Mexico (2
each); and Alaska, Kansas, Maryland, Montana, Nebraska, Nevada, New York,
Oklahoma, Oregon, South Dakota, Tennessee, Utah, Vermont, and West
Virginia (1 each)--may no longer receive detailed information about any
toxic chemical releases from facilities in their county.

Figure 5: Estimate of Percentage of Chemicals for Which Facilities Could
Report on Form A, by County

Some Facilities Will No Longer Have to Report Detailed Information to the TRI

Another way to present the impact of EPA's changes to TRI reporting
requirements is to examine how many facilities would no longer be required
to submit a Form R. We estimated that 6,620 facilities nationwide could
choose to convert at least one Form R to a Form A, and about 54 percent of
those are eligible to convert all their Form Rs to Form A. That means
3,565 facilities would no longer have to report any specific quantitative
information about their chemical releases and other waste management
practices to the TRI, according to our estimates. The number of facilities
ranges from 5 in Alaska to 302 in California.^18 For example, ATSC Marine
Terminal--a bulk petroleum storage facility in Los Angeles County,
California--reported releasing 13 different chemicals to the air,
including xylenes, toluene, and highly toxic benzene. Because the facility
released less than 2,000 pounds of each chemical, it could use Form A to
report each chemical it released. As figure 6 shows, more than 10 percent
of facilities in every state except Idaho would no longer have to report
any quantitative information to the TRI. The most affected states are
Colorado, Connecticut, Hawaii, Massachusetts, and Rhode Island, where more
than 20 percent of facilities could choose to not disclose the details of
their chemical releases and other waste management practices.

^18Appendix II provides the number of affected facilities for each state.

Figure 6: Estimate of Percentage of Facilities That Could Convert All Form
Rs to Form A, by State

Many Commenters Have Stated That EPA's Changes Will Significantly Limit TRI Data

Many commenters, including attorneys general of 12 States, EPA's Science
Advisory Board, and state TRI coordinators, have expressed concern about
EPA's changes to the TRI reporting requirements will significantly reduce
the amount of useful information reported to the TRI. In commenting on the
rule in a jointly signed January 12, 2006, letter, the Attorneys General
from of New York, California, Connecticut, Illinois, Iowa, Maryland,
Massachusetts, New Hampshire, New Jersey, New Mexico, Vermont, and
Wisconsin stated that, rather than repairing any problems with the TRI
program, EPA's changes will harm it by raising the reporting thresholds
for nearly all chemicals current subject to TRI requirements. The
Attorneys General added that the changes would significantly reduce the
amount of information about releases of toxic chemicals available to the
public and, as a result, would impair efforts by federal, state, and local
governments; workers; firefighters; and citizens to protect Americans and
their environment from the harm caused by discharges of toxic chemicals to
the air, water, and land. They added that "because the changes work
contrary to the purpose of the TRI--providing comprehensive information
about toxic releases across the United States--[they are arbitrary and
capricious, an abuse of discretion, and otherwise contrary to law." For
those, and other reasons, the states' Attorneys General concluded that "in
addition to being contrary to the public interest and sound policy, the
proposed changes would violate EPCRA,^19 PPA,^20 and the Administrative
Procedure Act."^21 However, EPA contends that the rule complies with all
applicable laws.

EPA's Science Advisory Board's (SAB) Environmental Economics Advisory
Committee also expressed concerns about the proposal in a July 12, 2006,
letter to the EPA Administrator. Noting that TRI data are widely used to
evaluate changes in facility and firm environmental performance and for
other purposes, and that TRI data often provide the only reliable source
of longitudinal data for this type of research, the committee said that
its primary concern was that increased eligibility for Form A reporting
will obscure the extent of facilities' releases of toxic chemicals.
According to the committee, the changes in reported toxic chemical release
levels will make the data incomparable over time and across facilities. It
further stated that they will impair researchers' ability to use TRI data
to assess spatial health impacts of toxic chemical releases and may also
reduce variation in the data that are useful in identifying
epidemiological and other relationships. The committee suggested that
these impairments on research could significantly limit the national
picture of the effect of toxic chemicals in the environment.

In addition, we surveyed the TRI coordinators in the 50 states and the
District of Columbia about their states' views on the two changes to Form
A eligibility that EPA finalized in December 2006--(1) raising the non-PBT
eligibility threshold from 500 pounds to 2,000 pounds and (2) allowing PBT
chemical reporting on Form A. Most states reported that EPA's changes
would have either a negative impact on various aspects of TRI. States most
frequently reported that the first change would have a negative impact on
information available to the public, efforts to protect the environment,
efforts to inform citizens about toxic releases, and community
right-to-know. For example, 23 of the 51 states responded that the non-PBT
change would have a negative impact on information available to the
public, while 15 indicated that the change would have no impact, as shown
in figure 7. Fewer states reported a negative impact as a result of the
PBT change, and more states reported no impact, most likely because the
change is limited to facilities that do not release any of the PBT
chemical. States most commonly reported that the PBT change would have a
negative impact on efforts to community right-to-know. Specifically, 19 of
the states said that the change would negatively impact community
right-to-know, while 2 said that it would have a positive impact. Although
as many as one-third of the states responded that they were uncertain
about the impact of these changes, few states reported that the changes
would have a positive impact on any aspect of TRI.

^1942 U.S.C. SS 11001-11050.

^2042 U.S.C. SS 13101-13109.

^215 U.S.C. S 706.

Figure 7: Result of GAO Survey of State TRI Coordinators' Views about
Impact of EPA's Changes on Various State Activities

Note: We provide complete wording of the survey questions and the
responses we received in a related electronic supplement (
[39]GAO-08-129SP ).

Changes to the TRI Will Likely Do Little to Reduce Industry's Reporting Burden

EPA's rule is intended to reduce the burden on reporting facilities by
allowing more facilities to use Form A, but the agency's estimated savings
are likely overstated. EPA's projected savings are based on calculations
that use OMB-approved estimates of burden hours associated with completing
Form R and Form A, but these estimates are based on outdated data. EPA's
more recent "engineering estimates" suggest a lower overall burden
associated with current TRI reporting and, consequently, lower burden
savings from the new rule.

EPA Relied on Outdated Data and Questionable Assumptions to Calculate Reporting
Burden and Cost Savings

In order to estimate the reduction in burden resulting from EPA's changes
to TRI reporting requirements, we started with the baseline burden
associated with current TRI reporting--that is, the amount of time that
facilities need to complete Form R and Form A. Using these baseline
estimates, cost savings are calculated by multiplying three factors: (1)
the difference in time needed to complete the two forms, (2) the number of
potentially eligible forms, and (3) the cost of labor. However, EPA's
baseline estimates rely on outdated, incomplete, or uncertain data
concerning the amount of time facilities need to complete the forms.
Therefore, the agency's derivative estimates of burden reduction and cost
savings are also unreliable.

Under the Paperwork Reduction Act, EPA must submit its Form R and Form A
to OMB for review and approval as part of an Information Collection
Request (ICR). As part of the review process, EPA provided OMB with its
re-estimates of the amount of time that facilities need to complete the
two forms, including reductions in the estimated number of reports that
would be filed and the amount of time needed to complete each report.
Those estimates also reflected differences in the amount of time
facilities need to report PBT chemicals versus non-PBT chemicals and
differences in time needed for first-time filers versus subsequent-year
filers. The agency used these baseline burden estimates from its current
ICRs to determine cost savings from the TRI Burden Reduction Rule.^22 As
shown in figure 8, EPA calculated that facilities would save 15.5 hours
for each PBT chemical submitted on Form A in lieu of Form R and 9.1 hours
for each non-PBT chemical.

^22Toxic Chemical Release Inventory Toxic Chemical Release Reporting
Information Collection Request Supporting Statement; OMB Control No.
2070-0093; EPA ICR #1363.14; October 2005 and Toxic Chemical Release
Inventory Alternative Threshold For Low Annual Reportable Amounts; Toxic
Chemical Release Reporting Information Collection Request Supporting
Statement; OMB Control No. 2070-0143; EPA ICR #1704.08; October 2005.

Figure 8: Burden Savings for Each PBT Chemical and Non-PBT Chemical Report

Note: PBT chemical savings does not equal the difference between Form R
and Form A burden savings because of rounding in the constituent estimates
for recordkeeping and mailing or form completion.

During OMB's review of EPA's ICR, it revised three of EPA's original
estimates without conducting an independent analysis of reporting burdens.
These revisions had the effect of significantly increasing EPA's proposed
re-estimated burden associated with TRI reporting, and thus increasing the
estimated burden reduction and cost savings from the new rule. OMB
revisions to EPA's re-estimates were the following:

           o increasing the non-PBT chemical, Form R burden from 14.5 hours
           to 25.2 hours for subsequent-year filers,
           o increasing the PBT chemical Form R burden from 14.5 hours to
           47.1 hours for subsequent year filers, and
           o increasing Form A burden from 9.3 hours to 16.2 hours for
           subsequent-year filers.

           Taken together, OMB's revisions resulted in the estimated burden
           savings being about 97,000 hours more than what they would have
           been if the EPA re-estimates had been used.^23 OMB's revisions to
           these three estimates constituted approximately 78 percent of
           total burden reduction from the new rule. In an agency memorandum,
           EPA raised concerns prior to accepting the revisions and stated
           that EPA had only agreed to them to obtain ICR clearance, which is
           required before the agency can request facilities to complete
           their TRI forms.^24 OMB's revisions appeared to favor the use of
           older data, and assumptions based on older data, despite the fact
           that more recent data were available.

           For the first revision, OMB increased EPA's proposed burden
           re-estimate of the time needed to calculate and complete a Form R
           non-PBT chemical for subsequent year filing from 14.5 hours to
           25.2 hours. This revision was based on a 1998 facility survey of
           18 respondents that OMB believed was the most recent data.
           However, EPA pointed out that more recent data were available,
           including 17 observations from 2000 and 2001 and 99 additional
           observations from 2001. Furthermore, EPA advised OMB that if the
           burden estimate had been based on all observations from 1998
           onward, the average per form burden would decrease to 12.5
           hours.^25 Notwithstanding, the EPA memorandum stated, "OMB did not
           like this result. Apparently, as with other recent initiatives,
           OMB is outcome-oriented."

           For the second revision, OMB increased EPA's proposed re-estimate
           of the time needed to calculate and complete a Form R PBT chemical
           for subsequent-year filing from 14.5 hours to 47.1 hours.
           According to internal memorandum, EPA had requested approval for a
           lower burden on the basis that PBT and non-PBT reporting are
           similar enough that the same burden estimate should be used for
           both. OMB disagreed, according to the memorandum, citing 
           from trade associations that special conditions of PBT reporting,
           namely not having an exemption for reporting de minimis (very
           small or insignificant) amounts and not being able to use range
           reporting, created a higher burden for PBT reporting. EPA agreed
           to leave this estimate unchanged because available data on
           reporting burdens was incomplete, (i.e., the data did not
           specifically address whether reporting was for PBT or non-PBT
           chemicals). Nevertheless, EPA expressed concern that using the
           OMB-approved burden estimates would create the appearance of a
           rather large relative difference in the reporting burden of a PBT
           chemical versus a non-PBT chemical (47.1 hours vs. 25.2 hours),
           which the agency said was not supported by the data. In addition,
           a team of experts disagreed with OMB's position and stated that if
           overall differences do exist in the reporting burden for PBT and
           non-PBT forms, the difference would stem largely from compliance
           determination activities and not from form completion.^26
			  
^23The original EPA burden amounts submitted to OMB for approval were
expected to reduce total burden by approximately 62,000 hours, and the
OMB-approved revisions increased burden reduction by about 97,000 hours to
approximately 158,000 hours.

^24USEPA/OEI, Terms of Clearance for TRI ICR Renewal, January 20, 2004.

^25Research Triangle Institute surveyed 18 facilities and collected 1998
reporting year data indicating an average burden of 25.2 hours per form.
The American Petroleum Institute survey 99 facilities and collected 2001
reporting year data, and EPA surveyed a total of 17 facilities and
collected 2000 and 2001 reporting year data from TRI-ME users.

           For the third revision, OMB increased EPA's proposed re-estimate
           of the time needed to complete a Form A for subsequent year filing
           from 9.3 hours to 16.2 hours. The OMB revision was based on the
           historical assumption used since the form was created in 1994 that
           Form A calculations take approximately 64 percent of the time of
           Form R calculations. EPA accepted this change even though the
           agency stated in its memorandum that the change was not supported
           by more recent data from a 2002 survey of nine facilities that
           showed a much lower Form A burden.^27
			  
			  Other Analyses Indicate Actual Savings Could Be Much Less Than EPA Estimated

           During the last TRI ICR renewal, EPA cited industry data
           indicating the burden of current TRI reporting was lower than
           previously estimated. Furthermore, while the total time for each
           major form-completion activity was estimated in the ICR, it was
           not broken down by the individual tasks (i.e., data elements) that
           comprise each activity. To help develop more reliable baseline
           estimates of TRI burden, EPA contracted for an engineering
           analysis of the time required to complete a Form R, and the agency
           requested public  on the results of this analysis as part
           of the TRI Burden Reduction Proposed Rule.^28
			  
^26TRI Reporting Burden Estimates, Memorandum, from Abt Associates, dated
July 16, 2004.

^27In April 2002, EPA contacted nine facilities that file Form As to
request information on the typical facility level burden associated with
using the Form A. EPA found that the average facility level burden per
chemical certification ranged from 11.2 to 15.5 hours depending on whether
the midpoint or maximum range was used. However, one facility reported a
much higher per chemical burden than the other eight facilities. Without
this outlier, the average of facility-level burden hours per chemical
certification would be 3.8 to 4.9 hours per chemical certified.

           To get at the burden associated with each activity, the
           engineering analysis divided the Form R into item-specific tasks.
           Then, the analysis calculated the total realistic burden for the
           specific activity under consideration by adjusting the total time
           for each activity by combining the time required to complete each
           task with the percentage of time individual tasks are typically
           completed.^29 As shown in table 2, the burden estimates from the
           engineering analyses are substantially lower than the current
           OMB-approved estimates.

Table 2: Comparison of Average Annual TRI Burden Estimates by Activity:
OMB-approved Versus Engineering Analysis

                                                                EPA              
                                                        engineering              
                                           OMB-approved    analysis   Difference 
Category  Activity                              (hours)     (hours) (percentage) 
Facility  Compliance determination--all                                          
level     facilities                                  4         2.5        -37.5 
          Rule familiarization--first-time                                       
          filers                                   34.5        23.5        -31.9 
          Rule                                                                   
          familiarization--subsequent-year                                       
          filers                                    n/a         5.6          n/a 
          Supplier notification                      24          24            0 
Per Form  Calculations and report                                                
R         completion--first-time                                                 
          filers--PBTs                             66.8         7.5        -88.8 
          Calculations and report                                                
          completion--first-time                                                 
          filers--non-PBTs                         67.6         9.5        -85.9 
          Calculations and report                                                
          completion-- subsequent year                                           
          filers--PBTs                             46.3         5.9        -87.3 
          Calculations and report                                                
          completion-- subsequent year                                           
          filers--non-PBTs                         24.6         7.0        -71.5 
          Recordkeeping/submission--all                                          
          filers                                      5           5          0.0 

Source: GAO analysis of EPA & Abt Associates, Inc. Burden Estimates.

Note: Both sets of estimates have been adjusted to account for burden
savings associated with the TRI Reporting Forms Modification Rule.

^28TRI Reporting Burden Estimates, Memorandum from Hilary Eustace, David
Cooper, and Susan Day of Abt Associates to Paul Borst, EPA dated July 16,
2004.

^29The engineering analysis derived estimates are based on the TRI
reporting experiences of a typical facility. The Abt team of experts
defined a typical facility as, among other things, (1) reasonably modern
and well-organized; (2) having Internet access with reasonable connection
speed; (3) normally, having no difference in completing a data element for
a non-PBT versus PBT chemicals; and (4) having no significant changes to
facility operations or waste management practices for subsequent-year
reports.

For example, the OMB-approved Form R burden for calculations and report
completion of subsequent year PBT and non-PBT chemicals are 46.3 and 24.6
hours, respectively. Under the engineering estimates, Form R estimates for
PBT chemicals are reduced to 5.9 hours (a reduction of about 87 percent)
and for non-PBT chemicals to 7.0 hours (a reduction of about 72 percent).
According to the proposed rule, if these estimates had been used, burden
reduction would have been about three-fourths (75 percent) of what was
estimated using the OMB-approved reporting burden estimates.

In addition to its engineering analysis, EPA also sought public 
on purported burden savings. Overall, EPA received thousands of public
 on the TRI Burden Reduction Proposed Rule, some of which
commented on burden reduction or cost savings estimates.^30 In general,
 were quite diverse: some stated that the savings were meaningful,
while others said that they were either too high or not needed. Some
expressed concern with the accuracy of the savings estimate or said that
it did not take into account other burden reduction actions while others
mentioned that the new rule offered little or no savings. For example,
eight commenters supported EPA's decision to extend Form A reporting for
PBT chemicals because of the helpful burden reduction for facilities that
have zero chemical releases, but five expressed general opposition to it
because it provided minimum burden reduction and did not justify the loss
in publicly available data. An official from the South Carolina Department
of Health and Environmental Control, for example, pointed out that
increases in the number of electronic filings are achieving real burden
reduction and that EPA's burden reduction estimates do not take this into
account. Four commenters stated that the burden reduction cost savings
estimates were too high. Fifteen commented that the current requirements
to complete a Form R in lieu of a Form A were not a significant burden on
industry. Three commenters stated that they work for, or had worked for,
facilities that submit Form Rs; did not find the current requirements
burdensome; and found that reporting helped facilities keep track of plant
operations. Others commented that the reporting burden is insignificant if
compared with the value of the TRI data to a wide range of stakeholders,
and the pollution reductions that have resulted. However, the National
Mining Association,^31 whose members have consistently reported the
largest numbers of TRI releases, commented that the proposed rule affords
little or no benefit for the metal and coal mining sectors, which tend to
release chemicals in excess of the Form A thresholds.

^30Response To  Toxics Release Inventory Phase 2 Burden Reduction
Rule Office of Information Analysis and Access, Office of Environmental
Information, U.S. Environmental Protection Agency, December 18, 2006.

In response to these , EPA agreed that savings may not represent a
significant amount for all eligible facilities but disagreed that expanded
Form A eligibility would not provide burden relief. The agency believes
that the proposed rule might provide meaningful burden relief for some
reporters, such as small facilities. EPA also agreed that increases in
electronic filing are achieving burden reduction and that the agency's
burden reduction estimates do not take these savings into account and thus
the agency's estimates of the cost savings associated with the rule could
be overstated.

Other EPA Efforts Have Provided More Burden Relief without Affecting the TRI

While the wide range of  submitted to EPA suggests that there may
be some uncertainty about the precise extent of burden reduction offered
by EPA's changes, there is considerably more certainty that the burden
reduction benefits of the changes are relatively small compared with other
initiatives. Throughout the history of the TRI program, EPA has
implemented measures to reduce the regulated community's reporting burden
and still maintain the public's access to information consistent with the
purpose of the TRI program. Through a range of compliance assistance
activities, such as the Toxic Chemical Release Inventory Reporting Forms
and Instructions (which is updated every year), industry training
workshops, chemical-specific and industry-specific guidance documents, and
the TRI Information Center (with a telephone hotline), we believe the
agency has shown a commitment to enhancing the quality and consistency of
reporting and assisting those facilities that must comply with EPCRA
section 313. In addition, EPA has made considerable progress in reducing
the TRI reporting burden through technology-based processes.

Beginning with Reporting Year 2001, EPA provided TRI-ME to help facilities
determine their TRI obligations and complete their TRI forms. TRI-ME leads
prospective reporters interactively through a series of questions that
eliminate a good portion of the analysis required to determine whether a
facility must comply with the TRI reporting requirements, and it includes
threshold calculations needed to determine Form A eligibility. If TRI-ME
determines that a facility is required to report, the software provides
guidance for each of the data elements on the reporting forms. The
software also provides an integrated assistance library with detailed
guidance for each step. Prior to submission, TRI-ME performs a series of
validation checks before the facility prints the forms for mailing,
transfers the data to CD-ROM, or submits the information electronically
over the Internet. According to EPA, since the release of TRI-ME, there
has been an estimated 15 percent reduction in burden-hours for facilities
certain activities associated with completing TRI forms^32--approximately
twice the annual cost savings resulting from the new Burden Reduction
Rule.^33

^31The National Mining Association is the industry association representing
the producers of most of the nation's coal, metals, and industrial and
agricultural minerals; the manufacturers of mining and mineral processing
machinery, equipment and supplies; and the engineering and consulting
firms, financial institutions and other firms serving the coal and
hardrock mining industry.

Moreover, TRI-ME has no adverse impact on the amount of information
submitted to the TRI and has likely improved the overall quality and
timeliness of the data, according to TRI program officials. Similarly,
other technology-based processes, including (1) EPA's Central Data
Exchange for form submission and (2) the population of data fields with
data submitted through other EPA programs have reduced the time, cost, and
complexity of existing environmental reporting requirements, while
enhancing reporting effectiveness and efficiency and continuing to provide
useful information to the public that fulfills the purposes of the TRI
program. At the same time, these efficiencies were not accounted for in
EPA's official estimates of TRI reporting burden as approved by OMB. EPA
has stated that the availability of TRI-ME reporting software is likely to
assist and streamline the reporting process--which could mean that the
estimated burden of current Form R reporting that EPA used in developing
the new rule were overstated. If this is the case, then the agency's
estimated cost savings associated with the rule would likewise be
overstated.

^32These activities are (1) Form R calculations and report completion, (2)
Form R recordkeeping/submission, (3) Form A calculations/certification,
and (4) Form A recordkeeping/submission.

^33Based on EPA's estimated TRI-ME savings of $11,737,699 from EPA's
October 2003 ICR supporting statements.

Congressional Interest and Actions to Reverse EPA's Changes

Members of Congress have expressed interest in EPA's changes to the TRI by
writing letters to the EPA Administrator, holding hearings, and
introducing legislation. In November 2005, a bipartisan group of Senators
wrote to the Administrator with serious concerns about EPA's actions and
analyses and to request additional data analysis from the agency. In
January 2006, the same day that public  closed for the rule,
members of the House of Representatives wrote to the Administrator
expressing similar, serious concerns that EPA's changes would undermine
the ability of local communities to take actions to protect themselves
from exposure to toxic chemicals. Meanwhile, EPA continued to develop its
final TRI Burden Reduction Rule.

Responding to these congressional concerns, on December 22, 2006, the
Administrator announced the decision to maintain annual reporting for the
TRI and not pursue any changes in reporting frequency. However, EPA also
finalized the TRI Burden Reduction Rule the same day. The first reports
using the revised reporting requirements were due on or before July 1,
2007, for reporting year (i.e., calendar year) 2006.

The Senate Committee on Environment and Public Works subsequently held a
hearing to consider, among other things, the impact of the TRI Burden
Reduction Rule and other EPA decisions affecting public right-to-know.^34
Soon thereafter, the Toxic Right-To-Know Protection Act was introduced to
effectively repeal EPA's Burden Reduction Rule and prevent future
consideration of alternate year reporting.^35 Specifically, the bill would
amend EPCRA (1) requiring the Administrator of EPA to establish the
eligibility threshold for use of Form A at not greater than 500 pounds for
non-PBT chemicals, (2) prohibit use of Form A for PBT chemicals,^36 and
(3) release provisions allowing the Administrator of EPA to modify the
frequency of toxic chemical release reporting. Similar legislation has
been introduced in the House of Representatives,^37 which recently held a
hearing to consider it.^38 At that hearing, we testified that based on our
analysis, EPA's recent changes to TRI reporting requirements will reduce
the amount and specificity of toxic chemical information that facilities
have to report to the TRI and that would, in turn, impact communities'
ability to assess environmental justice and other issues.^39

^34Oversight of Recent EPA Decisions, Hearing Before the Senate Committee
on Environment and Public Works, 110th Cong. (2007).

^35S. 595, introduced February 14, 2007.

^36The bill specifically prohibits the use of Form A with respect to any
chemical identified by the Administrator as a chemical of special concern
under 40 C.F.R. section 372.28 (or a successor regulation).

^37H.R. 1055.

Conclusions

To improve the prospects for successful regulations, EPA's Action
Development Process seeks to ensure that scientific, economic, and policy
issues are adequately addressed at the appropriate stages of regulations
development and to ensure adequate stakeholder participation across EPA's
offices. However, EPA did not follow the process in key respects when
developing the TRI Burden Reduction Rule, leading to a proposed rule whose
projected costs and benefits were not adequately analyzed or reviewed in
accordance with that process. EPA deviated from its process, in part,
because of pressure from OMB to provide significant burden reduction by
the end of 2006. In response to overwhelmingly negative public ,
EPA modified the proposed expansion of Form A eligibility by capping
allowable releases of non-PBT chemicals at 2,000 pounds.

However, we believe that EPA did not adequately address the analytical
concerns we raised with its proposed rule in the supporting analyses the
agency completed for the final rule. In particular, despite the TRI rule's
stated purpose--"to reduce burden while continuing to provide valuable
information to the public"--EPA did not adequately weigh the benefits
provided to facilities against the reduction in information available
about toxic chemical releases to affected communities. As a result, the
final rule has been widely criticized by TRI users for curtailing key
information about the release of toxic chemicals, and by the regulated
community as providing insufficient burden relief. Hence, the rule
provides neither meaningful burden reduction nor sufficient information to
the public. This outcome contrasts sharply with previous, openly conceived
TRI burden reduction efforts that have achieved substantial burden
reduction without reducing information to TRI users.

^38Environmental Justice and the Toxics Release Inventory Reporting
Program: Communities Have a Right to Know, Hearing Before the Subcommittee
on Environment and Hazardous Materials, Committee on Energy and Commerce,
House of Representatives, 110th Cong. (2007).

^39GAO, Environmental Right-to-Know: EPA's Recent Rule Could Reduce
Availability of Toxic Chemical Information Used to Assess Environmental
Justice, GAO-08-115T (Washington, D.C.: Oct. 4, 2007).

Accordingly, we believe reexamination of the benefits and costs to ensure
that the changes to TRI fully reflect the considered judgment of EPA staff
as provided for in the Action Development Process could help to avoid
similarly problematic outcomes in the future, while ensuring the
credibility and effectiveness of future TRI rulemakings. In addition,
congressional committees of jurisdiction currently have pending bills that
would effectively repeal EPA's rule by capping the eligibility threshold
for Form A at 500 pounds for non-PBT chemicals and prohibiting use of Form
A for PBT chemicals.

Agency  and Our Evaluation

In commenting on a draft of this report, EPA's Assistant Administrator for
Environmental Information and Chief Information Officer disagreed with our
recommendation that EPA perform sufficient analyses to support the rule.
Specifically, we had recommended that the Administrator of EPA thoroughly
evaluate the costs and benefits anticipated to communities and reporting
industries from increased use of TRI Form A and, based on this evaluation,
make a determination as to whether it would reconsider the TRI rulemaking.
We further recommended that EPA submit a report of its findings and
determination to relevant congressional committees within 30 days so as to
inform congressional deliberation on proposed legislation.

EPA's letter stated that the agency "believes fully that all appropriate
and necessary analyses were conducted ... in the context of the full
rulemaking process." EPA also noted that the December 2006 TRI rule put
into place important incentives to reduce chemical releases by permitting
additional facilities to use Form A. However, we continue to believe that
EPA did not adequately substantiate its assertion that the rule provides
such incentives, among other assertions that the agency made in the
proposed and final rules. Moreover, if TRI is, as EPA noted, the
cornerstone of its environmental information programs--allowing local
citizens and governments to hold facilities accountable for how they
manage toxic chemicals--then reducing the amount of information that those
facilities must disclose would provide less accountability for facilities
to reduce emissions resulting from manufacturing, processing, and using
toxic chemicals. Hence, it is unclear how EPA's course of action would
improve, rather than hinder, facilities' overall environmental
performance. Because EPA did not agree with the need to implement our
recommendations, and given ongoing congressional interest and pending
actions to address EPA's TRI Burden Reduction Rule, we have included a
matter for consideration by the Congress. EPA's letter and our detailed
response to it are contained in appendix IV. EPA also provided technical
, which we have incorporated into this report as appropriate.

In commenting on excerpts from a draft of this report, OMB's Deputy
Administrator of Information and Regulatory Affairs raised three concerns
regarding our characterization of OMB's activities under the Paperwork
Reduction Act (PRA) and Executive Order 12866. The first two,
closely-related concerns pertained to OMB's activities in reviewing the
burden estimates for the TRI information collection (i.e., Form R and Form
A) under the PRA. Specifically OMB felt that the excerpts did not provide
the necessary context, or a complete and balanced presentation, that would
enable the reader to understand OMB's successive interactions with EPA on
its approval of EPA's collection of information for the TRI. The
OMB-related excerpts that we provided to OMB were taken from the draft
report, as a whole. Given that we did not evaluate an OMB program or make
a recommendation to OMB, we believe that our report provides the necessary
context for the reader to understand EPA's actions, and OMB's related
role, with regard to recent changes to the TRI reporting requirements. In
its third concern, OMB states that the excerpts do not provide a complete
and balanced presentation of the facts with regard to EPA's decision to
include an option for raising the Form A eligibility threshold in the TRI
Burden Reduction Rule. Specifically, OMB provided additional details about
the development of the TRI Burden Reduction Rule not included in the
OMB-related excerpts that we provided for its review. We acknowledge OMB's
concerns and have made minor changes in the report as appropriate;
however, OMB did not provide new information that changes our findings,
conclusions, or recommendations. We believe that the report, taken in its
entirety, provides a fair, balanced, and complete understanding of EPA's
development of the TRI rule. OMB's letter and our detailed response to it
are contained in appendix V.

Matter for Congressional Consideration

Because EPA did not agree to our recommendation that it sufficiently
analyze the costs and benefits of increased use of TRI Form A, we suggest
that the Congress consider taking appropriate actions to address concerns
about reduced environmental information available to many communities.
Specifically, unless EPA provides the Congress with such an analysis
within 30 days of the public release of this report, the Congress may wish
to consider enacting legislation, including bills already introduced, that
would reverse EPA's expansion of TRI Form A eligibility for certain
facilities and chemicals.

As agreed with your offices, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days from
the report date. At that time, we will send copies of this report to the
appropriate congressional committees. We are also sending this report to
the Administrator of the Environmental Protection Agency. We will also
make copies available to others on request. In addition, this report will
be available at no charge on the GAO Web site at [40]http://www.gao.gov .

If you or your staffs have any questions about this report, please contact
me at (202) 512-3841 or [41][email protected] . Contact points for our
Offices of Congressional Relations and Public Affairs may be found on the
last page of this report. GAO staff who made major contributions to this
report are listed in appendix VI.

John B. Stephenson
Director, Natural Resources and Environment

Appendix I: Scope and Methodology

We assessed (1) how federal users, the states, and the public use Toxics
Release Inventory (TRI) data, (2) the extent to which the U.S.
Environmental Protection Agency (EPA) considered the views of internal and
external stakeholders in developing its burden reduction proposal, (3) the
impact of reporting changes on information available to the public, and
(4) the likely burden reduction that reporting facilities could receive
from the reporting changes.

To address the four main objectives, we analyzed documents, including
stakeholder  in the Federal Register about the proposed rule
change, those pertaining to the development of rule change, EPA's report
on stakeholder uses of the TRI, annual TRI reports, and EPA guidance for
facilities reporting to the TRI. In addition, we interviewed EPA
officials, industry representatives, officials from nongovernmental
organizations (NGO), and state TRI contacts. We also attended the annual
TRI Data Users' Conference in February 2007.

To respond to the first objective, how federal users, the states, and the
public use the TRI, we reviewed documentation from EPA, states,
nongovernmental agencies, and industries/businesses about their uses of
TRI data. We also interviewed EPA officials from program offices that use
TRI data, and officials from certain NGOs, industries, and states. We
selected officials from the NGOs and industries based on their previous
involvement with conducting TRI data analyses, attending EPA-sponsored TRI
stakeholder meetings, and  they submitted to the Federal Register
regarding the TRI. We selected state officials from New Jersey and
Massachusetts for interviews because those states have laws that require
facilities to submit additional data about their toxic chemical usage.

We obtained further information from states through our state survey that
we administered to TRI contacts in all states and the District of
Columbia. While developing the survey, we conducted pretests over the
phone with state contacts from five states. Based on these pretests, we
made revisions to the survey and found that not all individuals listed by
EPA as TRI contacts had sufficient expertise to complete each survey
section. We sent an introductory presurvey to all respondents so that they
could indicate whether they were the most knowledgeable person in their
state to answer the main survey sections. In 10 states, two or more
individuals were identified as most knowledgeable for a given survey
section. We administered this survey with a self-administered electronic
questionnaire sent in e-mails on January 4, 2006.

The survey asked respondents to indicate, in the first section, the types
of activities for which the state uses the TRI, whether facilities are
required to pay fees when submitting TRI reports and, if so, factors used
to calculate those fees. In the second section, the survey asked
respondents to indicate specific actions the state took to help facilities
comply with TRI reporting requirements and actions they took to enforce
reporting requirements. In the third section, the survey asked respondents
to indicate information about number of state employees who work on
TRI-related activities, the amount of the state TRI program budget, and
sources of funding for the budget. In the fourth section, the survey asked
respondents to indicate whether the state requires facilities to submit
additional data not required under national TRI reporting requirements,
the status of the state's involvement with EPA's Central Data Exchange,
and ways that the state makes TRI data available to the public. And in the
fifth, and final, section, the survey asked respondents to indicate the
likely impact of the proposed rule, their satisfaction with TRI-related
communication, and actions EPA could take to improve the TRI. In some
cases, we asked survey respondents additional questions via e-mail and
telephone. For the 10 states that had multiple contacts, we sent a
follow-up e-mail with a copy of their surveys to ensure that they
concurred with each other's answers. We closed the survey on February 12,
2007, after the 51st state responded, thus making our response rate 100
percent. This report does not contain all the results from the survey. The
survey and a more complete tabulation of the results can be viewed at
[42]GAO-08-129SP .

To respond to the second objective, the extent to which EPA considered the
views of internal and external stakeholders in developing its burden
reduction proposal, we reviewed documents related to the stakeholder
process EPA used to help identify possible burden reduction options. We
also interviewed EPA officials who were on the TRI workgroup that
developed the final rule and reviewed internal EPA documents that detailed
the workgroup's processes and decisions. In addition, we interviewed
knowledgeable Office of Management and Budget (OMB) officials about their
office's role in the TRI rule-making process and Information Collection
Request proposals. Finally, we reviewed the public  submitted in
response to EPA's proposed rule.

To respond to the third objective, the impact of reporting changes on
information available to the public, we conducted our analyses using 2005
TRI data, the most current available data. We performed a reliability
assessment of the data we obtained from EPA and determined that the data
were sufficiently reliable for the purposes of this report. To understand
the potential impact of EPA's changes to TRI reporting requirements at the
local level, we used 2005 TRI data to estimate the number of Form Rs that
would no longer have to be submitted in each state and the impact this
would have on data about specific chemicals and facilities. We provide
estimates of these impacts, by state, in appendix III.

To respond to the fourth objective, the likely burden reduction that
reporting facilities could receive from reporting changes, we reviewed the
proposed and final TRI Burden Reduction Rules that expanded eligibility
for using Form A Certification Statement in lieu of the more detailed Form
R by TRI facilities submitting required annual reports on releases and
other waste management. We also reviewed EPA's economic analysis of the
costs and impacts of the expanded eligibility for Form A, as well as other
relevant documents, and interviewed EPA officials about the burden
reduction savings analysis.

We conducted our work from August 2006 to September 2007 in accordance
with generally accepted government auditing standards.

Appendix II: GAO Estimates of the Impact of Reporting Changes on the TRI

We analyzed 2005 TRI data provided by EPA to estimate the number of Form
Rs that could no longer be reported in each state and determine the
possible impacts that this could have on data about specific chemicals and
facilities. Table 3 provides our estimates of the total number of Form Rs
eligible to convert to Form A, including the percentage of total Form Rs
submitted by facilities in each state.^1 The table also provides our
estimates of the number of unique chemicals for which no quantitative
information would have to be reported in each state, including the
percentage of total chemicals reported in each state. The last two columns
provide our estimates for the number of facilities that would no longer
have to provide quantitative information about their chemical releases and
waste management practices, including the percentage of total facilities
reporting in each state.

Table 3: Estimated Impact of TRI Reporting Changes on Number of Form Rs,
Chemicals, and Facilities, by State

               Form Rs              Chemicals                  Facilities
                Percentage of         Percentage of        Percentage of      
State Number         total  Number         total Number         total      
AK        59          36.6       8          17.0      5          15.6      
AL       456          22.0      34          17.1     69          12.9      
AR       247          17.7      18           5.8     39          11.0      
AZ       221          27.7      12          10.8     50          15.0      
CA     1,533          37.5      36          18.2    302          19.9      
CO       162          25.8      11          11.1     51          21.8      
CT       299          33.5      16          15.4     73          20.6      
DC         4          28.6       2          18.2      2          28.6      
DE        80          27.7      24          23.3     10          14.1      
FL       479          27.4      19          13.2    119          17.2      
GA       678          30.9      60          29.1    132          16.7      
HI        67          37.9      12          26.1      9          23.1      
IA       371          27.7      34          22.2     46          10.6      
ID        41          14.4       8          10.4      8           7.3      
IL     1,155          30.0      37          16.4    171          14.3      
IN       900          25.6      29          14.6    143          14.4      
KS       291          28.3      23          16.0     41          14.0      
KY       490          25.7      28          15.3     63          13.4      
LA       665          25.6      34          13.1                   46 12.4 
MA       574          38.0      23          20.4                  119 20.1 
MD       221          32.6      24          22.6                   34 16.6 
ME       105          26.1       8          11.3                   14 13.7 
MI       965          29.7      36          19.0                  145 16.1 
MN       263          21.0      20          15.4                   55 11.5 
MO       498          27.3      43          21.7                   80 14.2 
MS       265          25.0      29          18.7                   37 11.8 
MT        61          21.8      10          13.5                    7 15.2 
NC       705          30.1      43          24.9                  148 17.8 
ND        29          13.8       7          11.5                    6 12.5 
NE       116          20.3      11           7.9                   24 12.9 
NH        98          29.1      13          17.3                   23 16.1 
NJ       582          35.1      34          16.0                  101 19.3 
NM        96          29.2      11          15.3                   15 19.2 
NV        96          21.2      14          18.9                   19 14.3 
NY       663          31.8      33          19.1                  122 17.2 
OH     1,557          28.5      38          12.6                  218 13.8 
OK       273          26.1      30          23.3                   50 15.2 
OR       236          28.6      16          15.5                   47 15.5 
PA     1,253          29.9      30          15.2                  192 14.9 
RI       112          39.3      12          17.4                   30 23.4 
SC       596          29.0      36          17.6                   78 15.0 
SD        44          19.6       3           5.8                   10 10.5 
TN       569          27.6      40          20.9                  105 16.2 
TX      2196          30.6      29           9.3                  210 14.1 
UT       146          19.9      11           9.9                   25 12.6 
VA       401          25.2      23          14.8                   70 14.3 
VT        25          27.2       9          23.7                    6 14.6 
WA       276          26.4      22          19.8                   43 12.5 
WI       692          25.4      31          21.2                  113 12.5 
WV       222          22.8      40          24.1                   35 17.4 
WY        60          23.6       9          14.5                    5 10.9 
Total 22,193                                                    3,565      

Source: GAO analysis of EPA 2005 TRI data.

^1We estimated that approximately 11,700 new and 10,500 formerly-eligible
Form R reports could convert to Form A under EPA's increased Form A
thresholds.

Appendix III: Comparison of Information Collected on the Form R and the
Form A Certification Statement

Facilities must submit a detailed Form R report for each designated
chemical that they manufactured, processed, and/or otherwise used in
excess of certain thresholds, or certify that they are not subject to the
reporting requirement by submitting a brief Form A certification
statement. Form A captures general information about the facility, such as
address, parent company, industry type, and basic information about the
chemical or chemicals it released. Form R includes the same information
but also requires facilities to provide details about the quantity of the
chemical they disposed or released on-site to the air, water, land, and
injected underground, or transferred for disposal or release off-site.
According to EPA, Form A can be used by the public as a "range report"
because it indicates that the facility managed between 0 and 500 pounds of
a PBT chemical as waste and had no releases or other disposal quantities.
For a non-PBT chemical, the Form A indicates that a facility managed
between 0 and 5,000 pounds of a chemical as waste, of which no more than
2,000 pounds was released or otherwise disposed. Table 4 provides details
about the specific information that facilities provide on the Form R and
Form A.

Table 4: Information Collected on the TRI Form R and Form A Certification
Statement

Form RR                                         Form Am A                  
Facility Identification Information             Facility Identification    
                                                   Information                
      o TRI Facility ID Number                                                
      o Reporting year                                o TRI Facility ID       
      o Trade secret information (if claiming that    Number                  
      toxic chemical is trade secret)                 o Reporting year        
      o Certification by facility owner/operator      o Trade secret          
      or senior management official                   information (if         
      o Facility name, mailing address                claiming that toxic     
      o Whether form is for entire facility, part     chemical is trade       
      of facility, federal facility, or contractor    secret)                 
      at federal facility                             o Certification by      
      o Technical contact name, telephone number,     facility owner/operator 
      Email address                                   or senior management    
      o Public contact name, telephone number         official                
      o North American Industry Classification        o Facility name,        
      System (NAICS) codes                            mailing address         
      o Dun & Bradstreet number                       o Whether form is for   
      o Parent company information (name, Dun &       entire facility, part   
      Bradstreet number)                              of facility, federal    
                                                      facility, or contractor 
                                                      at federal facility     
                                                      o Technical contact     
                                                      name, telephone number, 
                                                      Email address           
                                                      o North American        
                                                      Industry Classification 
                                                      System (NAICS) codes    
                                                      o Dun & Bradstreet      
                                                      number                  
                                                      o Parent company        
                                                      information (name, Dun  
                                                      & Bradstreet number)    
Chemical Specific Information                   Chemical Specific          
                                                   Information                
      o Chemical Abstracts Service (CAS) registry                             
      number                                          o Chemical Abstracts    
      o EPCRA Section 313 chemical or chemical        Service (CAS) registry  
      category name                                   number                  
      o Generic name                                  o EPCRA Section 313     
      o Distribution of each member of the dioxin     chemical or chemical    
      or dioxin-like compound category                category name           
      o Generic name provided by supplier if          o Generic name          
      chemical is component of a mixture                                      
      o Activities and uses of the chemical at                                
      facility, whether chemical is:                                          
                                                                              
              o produced or imported for on-site                              
              use/processing, for                                             
              sale/distribution, as a byproduct,                              
              or as an impurity                                               
              o processed as a reactant, a                                    
              formation component, article                                    
              component, repackaging, or as an                                
              impurity                                                        
              o otherwise used as a chemical                                  
              processing aid, manufacturing aid,                              
              or as an ancillary or other use                                 
                                                                              
      o Maximum amount onsite at any time during                              
      the year                                                                
On-site Chemical Release Data                   On-site Chemical Release   
                                                   Data                       
      o Quantities released on-site to:                                       
                                                   Not reported on Form A     
              o air as fugitive or non-point                                  
              emissions                                                       
              o air as stack or point emissions                               
              o surface water as discharges to                                
              receiving streams or water bodies                               
              (including names of streams or water                            
              bodies)                                                         
              o underground injection                                         
              o land, including RCRA Subtitle C                               
              landfills, other landfills, land                                
              treatment/application farming, RCRA                             
              Subtitle C surface impoundments,                                
              other surface impoundments, other                               
              land disposal                                                   
                                                                              
      o Basis for estimates of releases (i.e.,                                
      monitoring data or measurements, mass                                   
      balance calculations, emissions factors,                                
      other approaches)                                                       
      o Quantity released as a result of remedial                             
      actions, catastrophic events, or one-time                               
      events not associated with production                                   
      processes                                                               
On-site Chemical Waste Management Data          On-site Chemical Waste     
                                                   Management Data            
      o Quantities managed on-site through:                                   
                                                   Not reported on Form A     
              o recycling                                                     
              o energy recovery                                               
              o treatment                                                     
                                                                              
      o Recycling processes (e.g., metal recovery                             
      by smelting, solvent recovery by                                        
      distillation)                                                           
      o Energy recovery methods (e.g., kiln,                                  
      furnace, boiler)                                                        
      o Waste treatment methods (e.g., scrubber,                              
      electrostatic precipitator) for each waste                              
      stream (e.g., gaseous, aqueous, liquid                                  
      non-aqueous, solids)                                                    
      o On-site waste treatment efficiency                                    
Off-site Transfers for Release or Other Waste   Off-site Transfers for     
Management                                      Release or Other Waste     
                                                   Management                 
      o Quantities transferred to any Publicly                                
      Owned Treatment Works (POTW)                 Not reported on Form A     
                                                                              
              o POTW name(s), address(es)                                     
                                                                              
      o Quantities transferred to other location                              
      for disposal or other release                                           
                                                                              
              o underground injection                                         
              o other land release                                            
                                                                              
      o Quantities transferred to other location                              
      for waste management                                                    
                                                                              
              o treatment                                                     
              o recycling                                                     
              o energy recovery                                               
                                                                              
      o Quantity transferred off-site for release,                            
      treatment, recycling, or energy recovery                                
      that resulted from remedial actions,                                    
      catastrophic events, or one-time events not                             
      associated with production processes                                    
      o Off-site location(s) name and address                                 
      o Basis for estimates for amounts                                       
      transferred                                                             
      o Whether receiving location(s) is/are under                            
      control of reporting facility/parent company                            
Source Reduction and Recycling Activities       Source Reduction and       
                                                   Recycling Activities       
      o Total quantities, for (1) the prior and    Not reported on Form A     
      (2) current reporting years and estimated                               
      totals for (3) the following and (4) second                             
      following years for:                                                    
                                                                              
              o on-site disposal to underground                               
              injection wells, RCRA Subtitle C                                
              landfills, and other landfills                                  
              o other on-site disposal or other                               
              releases                                                        
              o off-site transfer to underground                              
              injection wells, RCRA Subtitle C                                
              landfills, and other landfills                                  
              o other off-site disposal or other                              
              releases                                                        
              o on-site treatment                                             
              o on-site recycling                                             
              o on-site energy recovery                                       
              o off-site treatment                                            
              o off-site recycling                                            
              o off-site energy recovery                                      
                                                                              
      o Production ratio or activity index                                    
      o Source reduction activities the facility                              
      engaged in during the reporting year (e.g.,                             
      inventory control, spill/leak prevention,                               
      product modifications)                                                  
      o Option to submit additional information on                            
      source reduction, recycling, or pollution                               
      control activities                                                      

Sources: EPA TRI Form R and Form A Certification Statement.

Appendix IV:  from the Environmental Protection Agency

Note: GAO  supplementing those in the report text appear at the
end of this appendix.

See comment 2.

See comment 1.

See comment 6.

See comment 5.

See comment 4.

See comment 3.

See comment 9.

See comment 8.

See comment 7.

The following are GAO's  on the Environmental Protection Agency's
letter dated October 4, 2007.

GAO 

           1. We disagree with EPA's assertion that the TRI rule put into
           place important incentives to reduce chemical emissions and
           increase the use of alternatives to disposal and other releases.
           To the contrary, we concluded that the rule may actually reduce
           the incentives for a facility to prevent pollution by allowing up
           to 2,000 pounds of releases for those chemicals--a quadrupling of
           the threshold. Furthermore, as we show in appendix III, Form A
           provides no details about a facility's efforts to increase use of
           alternatives such as source reduction, recycling, or treatment of
           chemicals. We agree with EPA's assertion that the only change in
           requirements is that facilities are permitted to use the short
           form (Form A) if they maintain releases and total wastes below
           levels set in the rule. Indeed, we described the old and new Form
           A requirements in the first pages of this report. However, we did
           not state or imply that any facilities would be excused from
           submitting a TRI Form R or Form A under EPA's rule. Instead, our
           analysis showed that thousands of facilities that previously filed
           Form R may file Form A under the new levels. Given differences in
           the amount and specificity of information on Form R and Form A,
           which we presented in appendix III, we concluded that EPA's change
           could result in significantly less information about toxic
           chemicals being reported to communities around the country. For
           example, our analysis shows that EPA's levels do not maintain any
           Form R reporting for certain chemicals. Also, as we recently
           testified, the change appears likely to disproportionately impact
           minority and low-income communities.^1 Consequently, we believe
           that EPA's assertion about the change in requirements, while
           technically correct, misrepresents the impact of that change on
           the intended recipients of toxic chemical information provided by
           the TRI.
           2. We disagree with EPA's contention that it completed a thorough
           evaluation of the uses of TRI data and the associated economic and
           policy issues. Although EPA does have discretion to accelerate
           regulatory development, we continue to believe that, in this
           instance, the agency's acceleration of the rule relatively late in
           the development process, coupled with pressure from OMB to provide
           burden reduction at levels advocated by the Small Business
           Administration (e.g., a 5,000 pound Form A threshold), resulted in
           a poorly analyzed proposed rule that drew criticism from thousands
           of commenters. In response to EPA's comment, however, we have
           added a sentence to clarify that we considered in our review the
           results of EPA's stakeholder process, which started well before
           the proposed rulemaking commenced. In fact, that process began in
           November 2003 and concluded in February 2004, even earlier than
           EPA indicated in its letter. We also added a sentence to make
           clear that EPA officially began its rule-making process by
           approving the Action Initiation Form on March 30, 2004.
           3. We agree with EPA's statement that the full rulemaking record
           is best reviewed to determine whether adequate analyses were
           conducted before EPA issued the final rule. Indeed, we reviewed
           all relevant documentation in the record, and none of that
           material--including the studies that EPA inserted into the record
           between the proposed and final rules--addresses the shortcomings
           that we have identified in our report. For example, EPA did not
           fully consider the impacts of the rule on users and recipients of
           TRI data, including the states, which under the law, receive TRI
           data directly from facilities. It is for this reason that many
           commenters, including states and EPA's own Science Advisory Board,
           objected to any increase in the Form A threshold provided for in
           the proposed or final rules. EPA also did not substantiate its
           assertion that the final TRI rule puts into place important
           incentives to reduce chemical emissions and increase recycling and
           treatment as alternatives to disposal and other releases. If that
           claim is true, as EPA contends, then some analysis is warranted to
           support it. Prior to the rule, a facility could use Form A if its
           total waste management did not exceed 500 pounds. For example, a
           facility could use Form A if it released 100 pounds of a non-PBT
           chemical and treated an additional 400 pounds of the chemical for
           disposal (i.e., a total annual reportable amount of no more than
           500 pounds). Now, instead of treating the 400 pounds of chemical,
           that same facility could simply release all 500 pounds and qualify
           for Form A. EPA has not demonstrated in any analysis that the
           incentives to decrease releases outweigh the incentives to
           increase releases of chemicals.
           4. We disagree that all appropriate and necessary analyses were
           conducted for either the proposed or final rule. Contrary to EPA's
           assertion, we considered the entire economic analysis that EPA
           included in the publicly accessible docket for the final rule--in
           addition to reviewing extensive internal documentation that is not
           publicly available and conducting interviews with knowledgeable
           EPA staff to clarify our understanding--regarding the rulemaking
           process. We do not believe that any further discussion is
           necessary without analyses or  from its program offices
           that were not available in the publicly-accessible docket or made
           available for our review. We also considered the 122,000 public
           --99 percent of which did not support the TRI rule--and
           EPA's analysis of them. GAO continues to believe that EPA has not
           demonstrated that the costs of the TRI rule exceed its benefits.
           5. We considered the supporting analyses of the final rulemaking
           in our evaluation. We clearly stated in our draft report that EPA
           decided to modify the proposed expansion of Form A eligibility
           from 500 to 5,000 pounds of releases so that the final rule only
           allowed a fourfold increase in releases (from 500 to 2,000) pounds
           within a 10-fold increase in total waste management practices
           (from 500 to 5,000 pounds).
           6. We have included additional explanation in the report to
           clarify that EPA considers the Form A to be a range report
           indicating to the public that a facility released between 0 and
           2,000 pounds of a chemical. At the same time, we note EPA's own
           statement in the final rule that some facilities eligible for Form
           A continue to report on Form R out of a desire to showcase their
           pollution prevention efforts or to demonstrate good environmental
           stewardship. That is, some facilities choose not to use Form A to
           avoid giving the impression that their releases may have been as
           large as the range allows. For example, a facility that released
           100 pounds of a chemical may file Form R because it does not want
           the public to assume its releases may have been as high as 2,000
           pounds.
           7. We disagree that EPA has demonstrated that the rule provides
           incentives to reduce chemical emissions while minimizing the loss
           of information to communities about toxic releases and pollution
           prevention, especially for non-PBT chemicals. In fact, we
           concluded that the rule may actually reduce the incentives for a
           facility to prevent pollution by allowing up to 2,000 pounds of
           releases for those chemicals. Appendix III of this report shows
           that Form A provides no details about a facility's efforts to
           increase source reduction, recycling, or treatment of chemicals.
           Form A leaves the public to assume that the facility is releasing
           between 0 and 2,000 pounds of a chemical and is managing (e.g.,
           recycling, treating) between 0 and 5,000 pounds of that chemical.
           With regard to EPA's comment that the rule promotes national
           policy under the Pollution Prevention Act of 1990--it is unclear
           whether expanded use of Form A serves that policy. The Congress
           declared in the Pollution Prevention Act that the national policy
           of the United States is that (1) pollution should be prevented or
           reduced at the source whenever feasible; (2) pollution that cannot
           be prevented should be recycled in an environmentally safe manner,
           whenever feasible; (3) pollution that cannot be prevented or
           recycled should be treated in an environmentally safe manner
           whenever feasible; and (4) disposal or other release into the
           environment should be employed only as a last resort and should be
           conducted in an environmentally safe manner. EPA apparently
           assumes that some facilities will reduce the amount of toxic
           chemicals they release in order to qualify for the new, higher
           Form A thresholds. However, facilities that are presently below
           these thresholds could now increase their releases of toxic
           chemicals without triggering the requirement to file Form R. EPA
           provides no evidence that the higher Form A thresholds will result
           in net source reduction. Therefore, we disagree that increasing
           the amount of releases and other waste management allowed on Form
           A supports national policy to prevent or reduce pollution at its
           source.
           8. We disagree that EPA has sufficiently supported its assertion
           that the final rule promotes improvements in environmental
           performance. In fact, only Form R allows facilities to showcase to
           the public their improvements. In drafting our report, we reviewed
           all supporting materials that EPA provided in the docket for the
           proposed and final rules. Although an analysis of actual results
           of the rule would be meaningful, we agree with EPA that it is
           premature. However, EPA could have lent credibility to this
           assertion by providing an analysis to determine whether, for
           example, whether introduction of Form A in 1995 led to
           improvements in facilities' environmental performance.
           9. We disagree that requiring less information from many
           facilities will encourage them to reduce emissions and improve
           their environmental performance. As a general matter, EPA failed
           to establish that reducing the amount of information that those
           facilities must disclose would improve accountability for
           facilities to reduce emissions resulting from their use of toxic
           chemicals. Also, as discussed above, EPA did not explain the basis
           for its apparent belief that the incentives provided by the new
           Form A threshold will result in a net reduction in toxic chemical
           releases. Hence, it is unclear how EPA's course of action would
           improve, rather than hinder, facilities' overall environmental
           performance.

^1 [43]GAO-08-115T.

Appendix V: Comments from the Office of Management and Budget:

Note: GAO comments supplementing those in the report text appear at the 
end of this appendix. 

Executive Office Of The President: 
Office Of Management And Budget: 
Washington, D. C. 20503: 

October 30, 2007: 

Mr. John B. Stephenson: 
Director: 
Natural Resources and Environment: 
U.S. Government Accountability Office: 
441 G Street, NW: 
Washington, D.C. 20548: 

Dear Mr. Stephenson:

This provides the comments of the Office of Management and Budget (OMB) 
on the excerpts that you have provided to us from the General 
Accounting Office's (GAO) draft report entitled, Toxic Chemical 
Releases, EPA Actions Could Reduce Environmental Information Available 
to Many Communities.

As explained below, we have serious concerns regarding how GAO has 
characterized, in these excerpts, OMB's activities under the Paperwork 
Reduction Act (PRA) and Executive Order (EO) 12866. Our concerns fall 
into three areas.

I. The GAO excerpts do not provide the necessary context for properly 
understanding OMB's role in reviewing the burden estimates for the 
Toxics Release Inventory (TRI) information collection.

Our first concern with the GAO excerpts is that they do not provide the 
necessary context (including an accurate and coherent chronology) that 
would enable the reader to understand properly OMB's successive 
interactions with the Environmental Protection Agency (EPA) on, 
initially, EPA's request that OMB renew our approval under the 
Paperwork Reduction Act of EPA's collection of information for the TRI 
and, subsequently, OMB's review under Executive Order 12866 of EPA's 
draft proposed and final rules to amend EPA's TRI regulations. [See 
comment 1] 

As a result, the GAO excerpts suggest, erroneously, that OMB 
"increased" the estimated paperwork burden for the TRI collection of 
information (allegedly due to OMB's interest in influencing the burden 
estimates in connection with EPA's subsequent proposed rule on TRI 
burden reduction), when the fact is that OMB approved ï¿½ in response to 
EPA's request ï¿½ a reduction of 59% in the estimated PRA burden for the 
TRI collection. What the GAO excerpts describe as OMB's "increase" in 
the burden estimate is OMB's decision to approve a 59% reduction in the 
burden estimate, rather than the 73% reduction that EPA had requested. 
We do not understand how GAO can describe OMB's decision to approve a 
59% reduction in the PRA-approved burden estimate as OMB having 
"increased" the burden estimate. 

Here is the chronology. OMB's review of EPA's proposed downward re-
estimates of the paperwork burden took place during OMB's PRA reviews 
of two successive EPA requests for renewal of OMB's approval for the 
TRI collection of information. These PRA reviews took place in 2002 and 
2003, which was a year and a half before there was discussion involving 
OMB of any specific burden reductions proposals or estimates associated 
with EPA's proposed rule on TRI burden reduction (which was published 
in October 2005).

In the first of the two PRA reviews, EPA requested that the approved 
burden estimate for the TRI collection be reduced by 73% (from the 
previously-approved estimate of 10.257 million hours downward to 2.737 
million hours). Not only would this be a very large reduction in the 
previously-approved burden estimate, but it was not associated with any 
change in the paperwork requirements of the TRI collection. In other 
words, the requested burden-estimate reduction of 73% did not reflect 
any actual real-world reduction in paperwork burden. Nevertheless, 
based on OMB's review of EPA's burden re-estimate, OMB approved a 
reduction of 41% in the burden estimate (a reduction of 4.227 million 
hours). A primary reason OMB did not approve all of EPA's requested 
reduction is that OMB was concerned that EPA's estimates did not appear 
to account for all categories of burden, including time for data 
tracking and assembly; creation, operation and maintenance of data 
tracking systems; training; and compliance determinations. OMB 
explained the basis for this determination in the public PRA file, 
[Footnote 42] and OMB provided EPA with a 10-month approval so that EPA 
and OMB could revisit again whether additional reductions in the burden 
estimate were warranted. [See comment 2] 

During that second PRA review, EPA again requested a further reduction 
in the burden estimate, again with no associated reduction in the TRI 
paperwork requirements. OMB ultimately agreed to a further reduction of 
1.859 million hours.[Footnote 43] When combined with OMB's prior 
approval of the 41% reduction, this represented OMB's approval of a 
total reduction of 59% in the approved burden estimate, from the 
original 2002 baseline of 10.257 million hours. Moreover, in approving 
a 59% reduction, OMB had approved the lion's share (80%) of EPA's 
original request for a 73% reduction in the burden estimate.

In addition, with respect to the engineering analysis that is discussed 
in the excerpts, EPA did not present this information to OMB as part of 
our PRA review, but instead later on ï¿½ after OMB had renewed the PRA 
approval for the TRI collection (and prior to the discussions and 
review of EPA's proposed TRI rule). Therefore, when OMB ï¿½ during its 
PRA review ï¿½ decided to accept most, but not all, of EPA's proposed 
downward re-estimate of paperwork burden (as discussed above), OMB was 
not rejecting the engineering analysis, for the obvious reason that EPA 
had not yet presented that analysis to OMB for our consideration. [See 
comment 3] 

EPA presented the engineering analysis to OMB in the summer of 2004, 
outside of the regular PRA approval process (in other words, EPA was 
not requesting the downward re-estimate as part of a request for OMB to 
renew the approval of the TRI collection). In response, OMB expressed 
concerns about the merits of the engineering analysis and suggested to 
EPA that it discuss the analysis in the preamble of the proposed TRI 
burden-reduction rule the following year. EPA agreed with this 
approach. [See comment 4] 

In the preamble of the proposed rule, EPA described the engineering 
analysis. In addition, EPA also indicated that, if EPA had used this 
methodology as the baseline for estimating the burden reduction under 
the proposed rule, the total burden reduction resulting from the 
proposed rule would have been about three-fourths of the reduction as 
estimated using the currently approved PRA-burden estimate as the 
baseline.

Thus, neither OMB nor EPA concealed either the engineering analysis or 
what the results would be if its methodology were used to estimate the 
burden reductions that would occur under the proposed rule. EPA placed 
in the public docket for the proposed rule both estimates of burden 
reduction (the one using the current PRA-approved burden estimate and 
the other using the engineering analysis), and EPA requested public 
comment on both estimates. EPA also summarized, in the preamble, the 
results of the independent peer review that EPA had arranged of the 
engineering analysis:

The peer review panel was generally favorable to both the general 
methodology used in the engineering analysis (summing across Form R 
elements to derive total burden) and the specific form completion steps 
described. However, the panel felt that the time allocated for many of 
the tasks should be increased. The panel disagreed with the assumption 
in the Agency's engineering analysis that a typical TRI reporting 
facility was reasonably modern and well-organized. A majority of the 
panel thought that EPA overestimated the experience and knowledge that 
a typical TRI reporting facility would have in completing its Form R 
and thus underestimated the time it would take to complete the Form. 
[Footnote 44] 

In response to the proposed rule, EPA received a number of comments on 
the engineering analysis. The comments were mixed, with several 
comments questioning the validity of its methodology and/or the lower 
baseline burden estimates that would result from it. [Footnote 45] 
After reviewing the public comments, EPA decided not to use the 
methodology of the engineering analysis in estimating the burden 
reduction that would result from the final rule.

II. The GAO excerpts do not provide a complete and balanced 
presentation of OMB's review of the revised burden estimates that EPA 
proposed.

Our second concern, closely related to the first, is that the GAO 
excerpts do not provide a complete and balanced presentation of OMB's 
actions in approving most, but not all, of the proposed downward re-
estimates that EPA had put forward for the paperwork burden that is 
imposed by the TRI collection of information. [See comment 5] 

As part of our responsibilities under the Paperwork Reduction Act, OMB 
regularly reviews the re-estimates that agencies develop of the burdens 
imposed by their paperwork requirements. Such re-estimates and reviews 
are a normal part of the PRA process. That is because, under the PRA, 
OMB's approval for a collection of information must be renewed at least 
once every three years.

In this case, the downward re-estimate that EPA put forward, as part of 
EPA's request for OMB's renewal of the PRA approval for the TRI 
collection, were not associated with any actual change (i.e., 
reduction) in the paperwork requirements that are imposed by the TRI 
collection. In other words, in PRA terminology, the re-estimate was not 
associated with what we refer to as a "program change" (e.g., the 
elimination of certain questions, or a reduction in the frequency of 
reporting). Instead, EPA's re-estimate was just that ï¿½ a change in 
EPA's estimate of the burden that was imposed by the (unchanged) TRI 
collection. In PRA terminology, we refer to such re-estimates as 
"adjustments." As OMB has explained in its annual PRA report to 
Congress (the "Information Collection Budget"), such adjustments can 
occur due to demographic and economic factors (e.g., during a 
recession, more people submit applications for unemployment benefits) 
or due to an agency's reconsideration of the prior burden estimates 
(e.g., because the agency has made a revision to its methodology for 
estimating paperwork burden). The downward re-estimates that EPA put 
forward for the TRI program included both types of "adjustments." That 
is, EPA reduced its estimates of both the number of TRI reports that 
would be filed and the time needed to complete each report.

Under the PRA, the preparation, review, and issuance of burden 
estimates is not an end in itself. Instead, burden estimates are an 
important part of the evaluation, which the PRA requires both the 
agency and OMB to conduct, of whether the "practical utility" of a 
proposed collection of information justifies the burden that the 
collection would impose on the public. For this reason, the burden 
estimates are most important when those estimates are associated with 
an agency's proposal to adopt a new collection or to make changes in a 
collection, such as by increasing ï¿½ or decreasing ï¿½ the reporting 
requirements (i.e., when there is a proposed "program change"). In such 
cases, the burden estimates are a key component in the evaluation, by 
the agency and by OMB, of whether the new collection (or the collection 
changes) should be proposed and approved. As noted above, EPA's request 
for a downward re-estimate of 73% did not result from a "program 
change" to the TRI collection, but instead was an "adjustment" based on 
EPA's reconsideration of the prior PRA-approved burden estimate.

As we have explained above, OMB approved 80% of EPA's proposed downward 
re-estimate (i.e., OMB approved a 59% reduction in the burden estimate, 
in response to EPA's request for a 73% reduction). This fact is not at 
all clear in the GAO excerpts. The excerpts describe the situation as 
OMB having increased the burden estimate for the TRI collection, and 
this is because the GAO excerpts use ï¿½ as the operative "baseline" ï¿½ 
EPA's burden estimate under its proposed 73% downward re-estimate. 
However, EPA's proposal is not the appropriate baseline to evaluate 
OMB's actions, because that is not the baseline that OMB, properly, 
applies under the PRA. Instead, the proper baseline is the estimate 
that had been previously approved for the TRI collection (which 
estimate EPA was proposing to reduce).

OMB approved most, but not all, of EPA's proposed burden re-estimate 
for the following reason. In the case of an information collection that 
is not being changed (as was the case with the TRI collection during 
OMB's PRA review), the starting point for the analysis is, as it should 
be, the collection's prior approved burden estimate. Re-estimates of 
burden ï¿½ upward or downward ï¿½ are permitted, and do occur, but they 
need to be explained and justified by reference to the prior approved 
estimate. In the absence of such explanation and justification, 
periodic re-estimates of burden could result in these estimates 
becoming meaningless. Accordingly, OMB decided to accept most but not 
all (i.e., 80%) of EPA's proposed downward re-estimate, because OMB had 
concluded that EPA sufficiently demonstrated that the original 2002 PRA-
approved burden estimate should be reduced by 59%, but OMB was not 
persuaded that the estimate should be reduced by 73%. OMB therefore 
proceeded to approve EPA's request that OMB renew its approval for the 
TRI collection, with the reduced burden estimate (reflecting the 59% 
reduction in the burden estimate from the original 2002 PRA-approved 
baseline). It should be noted that this did not mean that this would be 
the "final" burden estimate. That is because EPA could request a 
further re-estimate the next time that EPA submitted the proposed 
collection to OMB for review and approval. In fact, that is what had 
already happened in this case; OMB initially approved a 41% reduction 
in the burden estimate, and OMB subsequently approved a further 
reduction to the estimate, resulting in OMB having approved a total 
reduction of 59% in the burden estimate.

III. The GAO excerpts do not provide the necessary context for properly 
understanding EPA's decision to include an option for raising the Form 
A eligibility threshold in the proposed TRI Burden Reduction Rule.

Our third concern with the GAO excerpts relates to their discussion of 
OMB's role with respect to the inclusion, in EPA's proposed TRI rule, 
of an option to raise the Form A eligibility threshold from 500 to 
5,000 pounds. Again, we do not believe that the GAO excerpts provide a 
complete and balanced presentation of the facts. [See comment 6] 

As an initial matter, based on the GAO excerpts, one might conclude 
(erroneously) that the option of raising the Form A eligibility 
threshold was unheard of, and unanalyzed, before it was included in 
EPA's draft of the proposed rule. That is definitely not the case. The 
idea that one way to reduce TRI reporting burden would be by raising 
the Form A eligibility threshold had been a topic of discussion ï¿½ both 
within the Federal Government and between the Federal Government and 
stakeholders ï¿½ for many years, in part because several prior EPA rule 
makings had substantially increased the program's reporting burden.

In September 2002, EPA initiated a Stakeholder Dialogue process to 
identify improvements to the TRI and to develop opportunities to reduce 
the burden on reporting facilities. As part of that process, EPA issued 
a White Paper with options for TRI burden reduction intended to 
stimulate public comment. Option 3 in the White Paper was "Expanded 
Eligibility for the Form A Certification Statement," which included a 
discussion of raising the Form A eligibility threshold from 500 to 
either 1,000, 2,000 or 5,000 pounds,[Footnote 46] exactly the same 
options discussed in the proposed burden reduction rule several years 
later.

Moreover, the idea of raising the eligibility threshold had arisen in 
another public forum, which was OMB's request in February 2004 for the 
public to submit nominations for reducing regulatory burdens on the 
manufacturing sector.[Footnote 47] In response to this request, OMB 
received nominations for a wide variety of regulatory reforms, several 
of which ï¿½ including one from the Small Business Administration's (SBA) 
Office of Advocacy ï¿½ focused on the TRI program. [Footnote 48] (This 
was not the first time that the TRI program had been the subject of a 
reform nomination; the TRI program had also been nominated for reform 
in response to each of OMB's requests for nominations going back to 
2001. [Footnote 49]) One reform that the SBA Office of Advocacy 
specifically suggested was "expanding the number of filers eligible to 
use a `short form' annual form (Form A)." [Footnote 50] OMB and the 
relevant regulatory agencies then reviewed the 189 different sets of 
regulatory nominations. Based on that review, OMB and the regulatory 
agencies mutually agreed to pursue 76 of these nominations, and the 
agreed-upon course of action took various forms, such as the regulatory 
agency conducting a review and reporting back to OMB or the agency 
initiating a rulemaking.

OMB announced these 76 reforms in a report that OMB issued in March 
2005, and one of these reforms (#52) involved "Reporting and Paperwork 
Burden in the Toxic Release Inventory (TRI) Program." As the report did 
for each of the 76 nominations, it summarized the comments that OMB had 
received in support of reform of the TRI program: "The required TRI 
database contains thousands of reports that show little or no release 
of toxic chemicals, an indication that expensive and time-consuming 
reports are required with little environmental benefit. Burden-
reduction reforms are needed such as raising the reporting thresholds 
on the amount of material that can be used without triggering a report. 
"[Footnote 51] The report then went on to outline the course of action 
to which OMB and EPA had mutually agreed, which included EPA issuing a 
proposed rule on TRI burden reduction in August 2005 and issuing the 
final rule in December 2006. 

Although it is not clear from the GAO excerpts, the March 2005 report 
must be what the excerpts are discussing when the excerpts refer, 
obliquely, to "EPA's commitment to OMB to provide burden reduction by 
the end of December 2006." Yes, this was a commitment that EPA had made 
to OMB, but this was also a commitment that EPA had made to itself and 
to the public.

Thus, the idea of increasing the Form A eligibility threshold, and 
EPA's commitment to issue a final burden-reduction rule by the end of 
2006, were a matter of public knowledge. By failing to provide this 
context (at least in the excerpts that we have received for review), 
GAO does not present a complete and balanced picture of EPA's 
development of the draft proposed rule.

Another context point, to which the GAO excerpts give insufficient 
weight, is that the discussion in the excerpts focuses on EPA's 
inclusion in the proposed rule of the option to increase the Form A 
eligibility threshold. In other words, this is a discussion of an 
option on which EPA was inviting public comment. Thus, by including 
this option in the proposed rule, EPA was not making a final decision 
to increase the eligibility threshold. Instead, EPA had decided that 
the long-discussed idea of increasing the eligibility threshold was 
serious enough to warrant inclusion in the proposed rule, so that this 
option would be subject to public scrutiny and comment (both in favor 
and against) that EPA would subsequently consider as it reviewed the 
public comments and developed the final rule. [See comment 7] 

We see nothing inappropriate about EPA including, in the proposed-rule, 
an option ï¿½ namely, to increase the Form A eligibility threshold ï¿½ that 
had long been a matter of policy discussion and that would be one 
possible and logical way of reducing the reporting burden of the TRI 
program. At the end of the day, what ultimately matters is the 
decisions that a rulemaking agency makes in its final rules. In this 
regard, courts review final rules; they do not review proposed rules. 
And, when the courts review a final rule, they evaluate whether the 
rulemaking agency has sufficiently justified the decisions that the 
agency has made in the final rule (not whether the agency had 
sufficiently justified, in the proposed rule, the agency's decision to 
include the particular options on which the agency was requesting 
public comment). Finally, we should note that, in its final rule on TRI 
burden reduction, EPA did not adopt the threshold increase as it had 
been proposed. Instead, in the final rule, EPA responded to the public 
comments on the proposal by adopting a significantly different (and 
smaller) increase in the eligibility threshold for releases to the 
environment (2000 pounds), while keeping the threshold for chemicals 
managed as waste at the proposed level of 5,000 pounds. 

In conclusion, OMB takes seriously its responsibility under the 
Paperwork Reduction Act to ensure the integrity of burden estimates. 
OMB reviewed a series of requests by EPA to reduce its estimates of TRI 
reporting burden, well before the development of specific options for 
the proposed burden reduction rule, and OMB approved most, but not all, 
of those estimated reductions. Contrary to the GAO excerpts, OMB's 
approval of a very substantial (59%) reduction in the burden estimate 
does not constitute OMB increasing the burden estimate. Moreover, with 
respect to EPA's subsequent rulemaking on reducing TRI burden, there 
was nothing inappropriate about EPA including in the proposed rule for 
which EPA was requesting public comment the long-discussed concept of 
increasing the Form A eligibility threshold. [See comment 8] 

Thank you for the opportunity to comment on the excerpts from the draft 
GAO report. In order to correct the report, I request that your staff 
incorporate these comments.

Sincerely, 

Signed by: 

Kevin F. Neyland
Deputy Administrator: 
Office of Information and Regulatory Affairs: 

Enclosures: 

The following are GAO's comments on the Office of Management and 
Budget's letter dated October 30, 2007.

GAO Comments:

1. We believe that the reader would be able to clearly understand OMB's 
successive interactions with EPA on the agency's request that OMB renew 
its approval under the Paperwork Reduction Act (PRA) for EPA's 
collection of information for the TRI. The report provides considerable 
detail about the process used for estimating the baseline burden 
associated with collecting TRI information on Form R versus Form A. We 
provided that level of detail to help the reader understand that the 
net burden savings associated with EPA's changes is derived from the 
difference in burden estimates associated with each form. Nonetheless, 
we reviewed the relevant sections of the report and clarified that the 
reductions in TRI burden ultimately approved by OMB were smaller than 
the EPA requested but not necessarily "increased" relative to prior OMB 
information collection approvals.

2. We agree with OMB's assertion that EPA's request for a reduction in 
the approved burden estimate for the TRI collection was not associated 
with any change in the paperwork requirements of the TRI collection. 
However, we disagree with OMB's assertion that the revision did not 
reflect any "actual real-world reduction" in paperwork burden. In fact, 
the request for reduction was based on EPA's evolving understanding of 
the actual burden associated with reporting to the TRI based on best 
available information that the agency had at the time. We note that 
despite EPA's requests, OMB has not allowed EPA to contact more than 10 
TRI facilities to develop its burden reduction estimates.

3. OMB's comments implied that the report stated that the office 
rejected EPA's engineering analysis as part of a PRA review. The report 
simply compares the lower burden estimates derived from EPA's 
engineering analysis with the OMB-approved burden estimates for the TRI 
information collection. The report neither states nor implies that OMB 
rejected EPA's engineering analysis as part of an ICR. EPA staff 
knowledgeable about the ICR process stated that OMB had expressed 
concerns about the merits of the engineering analysis. When we met with 
OMB staff who were knowledgeable about the engineering analysis, they 
declined to provide details about the office's rationale on the grounds 
that such information was related to internal deliberations and OMB 
policy was to not discuss internal deliberations.

4. OMB stated that, as a result of concerns it raised, EPA decided to 
discuss the engineering analysis in the proposed TRI Burden Reduction 
Rule's preamble. OMB's comments also provided additional details about 
the comments that EPA received from the public on its engineering 
analysis. The excerpts that we provided to OMB did not include details 
of EPA's engineering analysis, or the public comments pertaining to it, 
because those details did not pertain to OMB's involvement in the TRI 
rule. Therefore, OMB had no reason to know that our draft report 
already included many of the details that OMB provided in its letter. 
OMB further stated that "neither OMB nor EPA concealed either the 
engineering analysis or what the results would be if its methodology 
were used to estimate the burden reductions that would occur under the 
proposed rule." Our report neither stated nor implied that OMB 
concealed any information. However, as noted earlier, OMB chose to not 
provide us with details about the office's concerns about the merits of 
EPA's engineering analysis.

5. OMB's comments raised a concern that the report excerpts did not 
provide a complete and balanced presentation of OMB's actions in 
approving most, but not all, of the proposed downward re-estimates that 
EPA requested in its TRI information collection. As we acknowledged 
previously, OMB did not have the benefit of the complete draft report, 
which provides context and balance. Specifically, none of the report's 
three objectives focused on OMB programs or even OMB's role in the TRI 
program. Nonetheless, to provide the reader with necessary context to 
understand the complexity in estimating the baseline burden associated 
with the TRI information collection requests (i.e., TRI Form R and Form 
A), we discuss OMB's actions under the PRA in approving EPA's recent 
TRI information collection requests. As we stated in an earlier 
comment, we have reviewed the report to ensure that our presentation of 
the facts clearly states that OMB did not approve all the reduction in 
estimated burden that EPA had requested and to ensure that the 
relatively smaller reduction is not construed as an increase in the 
baseline. Nonetheless, we do not believe that the level of detail 
provided by OMB's comments is proportional to the relative importance 
of OMB's role in the context of our report's scope and objectives.

6. OMB stated its belief that the excerpts provided by GAO do not 
provide a complete and balanced presentation of the facts associated 
with OMB's role in the development of EPA's TRI Burden Reduction Rule. 
OMB stated that one might erroneously conclude that the option of 
raising the Form A eligibility threshold was unheard of, and 
unanalyzed, before it was included in EPA's proposed rule. OMB stated 
that was "definitely not the case." We agree, and the draft report 
clearly states that we considered in our review the results of EPA's 
stakeholder process, which started well before the proposed rulemaking 
commenced. In fact, we state in the report that the TRI workgroup had 
carefully analyzed and decided to eliminate the option to raise the 
Form A eligibility threshold from its consideration, at an early step 
in the rule development process, because of the option's potential 
adverse impact on the TRI. Therefore, we do not agree that the reader 
will erroneously conclude that the option was unheard of or unanalyzed. 
Instead, we believe that the report as a whole provides the reader with 
the context to understand when in the rule development process OMB's 
preferred burden reduction option was identified, considered, 
evaluated, dropped from further consideration, and ultimately picked up 
again for inclusion in EPA's TRI Burden Reduction Rule.

7. OMB commented that it sees nothing inappropriate about EPA 
including, in the proposed rule, an option to increase the Form A 
eligibility threshold--that had long been a matter of policy 
discussion. Although our report explains that the option was identified 
well before the rulemaking process commenced, we do not believe that 
the length of time the option was discussed has much bearing on its 
appropriateness. Instead, our report focuses on the extent to which EPA 
followed its internal rule development process, which is intended to 
ensure that scientific, economic, and policy issues are adequately 
addressed at the appropriate stages of rule development and to ensure 
adequate stakeholder participation across EPA's offices until the final 
action is completed. OMB further stated that "at the end of the day, 
what ultimately matters is the decisions that a rulemaking agency makes 
in its final rules." We agree. Accordingly, our report discusses that 
EPA responded to the overwhelming negative public comments to its 
proposed rule by raising the Form A threshold only 4-fold, rather than 
10-fold, as EPA had proposed. Furthermore, our report evaluated the 
impact that the final rule may have on the TRI, and we concluded that 
EPA did not adequately address the analytical concerns we raised with 
its proposed rule in the supporting analysis the agency completed for 
the final rule [emphasis added]. Specifically, we found that despite 
the EPA's statement of the rule's purpose--to reduce burden while 
continuing to provide valuable information to the public--EPA did not 
adequately weight the benefits provided to facilities against the 
reduction in information available about toxic chemical releases to 
affected communities.

8. OMB concludes its comments by stating its belief that there was 
nothing inappropriate about EPA including in the proposed rule the long-
discussed concept of increasing the Form A eligibility threshold. Our 
report does not assert that inclusion of that option was inappropriate. 
Instead, we believe that EPA did not fully consider the true impacts 
its TRI Burden Reduction Rule would have on environmental information 
available to, and used by, many communities. In addition, we continue 
to believe that burden reduction can be achieved in ways that do not 
simultaneously reduce publicly-available information about use and 
management of toxic chemicals in many communities across the United 
States.

Appendix VI: GAO Contact and Staff Acknowledgments:

GAO Contact:

John Stephenson, (202) 512-3841, or [email protected]:

Staff Acknowledgments:

In addition to the contact named above, Steven Elstein, Assistant 
Director; Mark Braza; John Delicath; Karen Febey; Timothy Guinane; 
Terrance Horner, Jr.; Richard Johnson; Alison O'Neill; Jennifer 
Popovic; Steven Putansu; Kim Raheb; Michael Sagalow; and Jena Sinkfield 
also made key contributions.

References

Visible links
  36. http://www.gao.gov/cgi-bin/getrpt?GAO-08-129SP
  37. http://www.epa.gov/tri
  38. http://www.gao.gov/cgi-bin/getrpt?GAO-08-129SP
  39. http://www.gao.gov/cgi-bin/getrpt?GAO-08-129SP
  40. http://www.gao.gov/
  41. mailto:[email protected]
  42. http://www.gao.gov/cgi-bin/getrpt?GAO-08-129SP
  43. http://www.gao.gov/cgi-bin/getrpt?GAO-08-115T
*** End of document. ***