Toxic Substances Control Act: Legislative Changes Could Make the Act More
Effective (Chapter Report, 09/26/94, GAO/RCED-94-103).

Thousands of toxic chemicals are in commercial use. Although these
chemicals are important in producing goods and services, they are often
toxic and can harm human health and the environment. Congress passed the
Toxic Substances Control Act in 1976 to enable the Environmental
Protection Agency (EPA) to obtain more information on chemicals and to
control those that pose an unreasonable risk. In response to
congressional concerns that EPA has been slow to implement the act, this
report discusses EPA's efforts to (1) assess chemicals under the act,
(2) control those found to be harmful, and (3) make the act's
information on chemical risks publicly available by reducing the amount
of information that the chemical industry claims as confidential.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-94-103
     TITLE:  Toxic Substances Control Act: Legislative Changes Could 
             Make the Act More Effective
      DATE:  09/26/94
   SUBJECT:  Toxic substances
             Health hazards
             Information dissemination operations
             Data collection operations
             Chemical research
             Environmental law
             Environmental monitoring
             Manufacturing industry
             Information disclosure
             Proprietary data
IDENTIFIER:  European Union
             EPA Toxics Release Inventory Program
             EPA National Occupational Exposure Survey
             EPA Confidential Business Information Review Challenge 
             Program
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed 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.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

September 1994

TOXIC SUBSTANCES CONTROL ACT -
LEGISLATIVE CHANGES COULD MAKE THE
ACT MORE EFFECTIVE

GAO/RCED-94-103

Toxic Substances Control Act


Abbreviations
=============================================================== ABBREV

  CFC - chlorofluorocarbon
  EPA - Environmental Protection Agency
  GAO - General Accounting Office
  OSHA - Occupational Health and Safety Administration
  PCB - polychlorinated biphenyl
  RCRA - Resource Conservation and Recovery Act
  SAR - structure activity relationships
  TCDD - 2,3,7,8-tetrachlorodibenzo-p-dioxin
  TSCA - Tosic Substances Control Act

Letter
=============================================================== LETTER


B-256039

September 26, 1994

The Honorable Harry M.  Reid
Chairman, Subcommittee on Toxic
 Substances, Research and Development
Committee on Environment and Public Works
United States Senate

The Honorable Mike Synar
Chairman, Environment, Energy,
and Natural Resources Subcommittee
Committee on Government Operations
House of Representatives

As requested, this report discusses the Environmental Protection
Agency's (EPA) efforts under the Toxic Substances Control Act to
assess the risks of chemicals, control those found to be harmful, and
make information on chemical risks publicly available.  The report
also contains options for revising the act to improve EPA's efforts
in each of these areas. 

As arranged with your offices, unless you publicly announce its
contents earlier, we will make no further distribution of this report
until 30 days after the date of this letter.  At that time, we will
send copies to the appropriate congressional committees; the
Administrator, EPA; and the Director, Office of Management and
Budget.  We will also make copies available to other interested
parties on request. 

This work was performed under the direction of Peter F.  Guerrero,
Director, Environmental Protection Issues, who can be reached at
(202) 512-6111 if you or your staff have any questions.  Other major
contributors to this report are listed in appendix IV. 

Keith O.  Fultz
Assistant Comptroller General


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

Thousands of chemicals are in commercial use.  Although these
chemicals are important in producing goods and services, they are
often toxic and can have adverse health and environmental effects. 
The Congress passed the Toxic Substances Control Act (TSCA) in 1976
to enable the Environmental Protection Agency (EPA) to obtain more
information on chemicals and to control those that pose an
unreasonable risk. 

Concerned that EPA has been slow to implement TSCA, Senate and House
Subcommittee Chairmen with responsibilities for overseeing the act
asked GAO to review the agency's progress and identify changes to
make the act more effective.  Specifically, the Chairmen asked GAO to
review EPA's efforts to (1) assess chemicals under TSCA, (2) control
those found to be harmful, and (3) make TSCA's information on
chemical risks publicly available by reducing the amount of
information that the industry claims as confidential. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

TSCA authorizes EPA to review chemicals before and after they enter
commerce.  To assess risks, EPA examines a chemical's toxicity or
potential adverse effects and the amount of human and environmental
exposures.  If EPA finds that a chemical's risks are unreasonable, it
can prohibit or limit its production, distribution, use, and disposal
or take other action, such as requiring warning labels on the
substance. 

TSCA requires the industry to notify EPA at least 90 days before
producing or importing a new chemical.  These notices contain
information, such as the chemical's molecular structure and
anticipated uses, that EPA uses to evaluate the chemical's potential
risks.  TSCA also authorizes EPA to require manufacturers to perform
tests or provide other data, such as production volumes, on existing
chemicals.  In addition, TSCA requires the industry to report to EPA
any data that reasonably support a conclusion that a chemical
presents a substantial risk. 

Of about 72,000 substances in EPA's inventory of TSCA chemicals,
62,000 were already in commerce when EPA began to review new
chemicals in 1979.  EPA reviewed the remaining 10,000 substances as
new chemicals and added them to the inventory when their manufacture
began. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

TSCA's unique authorities to limit the manufacture, distribution, and
use of toxic chemicals could be important tools in a comprehensive
program for these chemicals.  However, the act's legal standards are
so high that they have usually discouraged EPA from using these
authorities.  In addition, EPA has generally interpreted TSCA as
giving preference to dealing with chemical risks under other laws. 
As a result, EPA has issued regulations to control only nine
chemicals in almost 18 years.\1

Although EPA has reviewed new chemicals in a timely manner, its
process does not ensure that their potential risks are fully assessed
before they enter commerce.  EPA usually has few if any test data,
and it predicts chemicals' potential effects with mixed results.  In
addition, the data that EPA uses to assess exposure may change
substantially after manufacture begins.  For existing chemicals, the
burden is essentially on EPA to compile the data, which is
time-consuming and costly.  As a result, EPA has reviewed the risks
of about 2 percent of the 62,000 chemicals that were already in
commerce when the agency began to review new chemicals. 

TSCA's provisions on confidential business information are difficult
for EPA to implement.  The chemical information collected under TSCA
can be useful to others, such as state environmental officials. 
However, EPA cannot disseminate much of the information because
industry claims that it is confidential.  EPA believes that many
claims are not necessary to protect trade secrets.  The agency has
successfully challenged the validity of some claims, but it does not
have the resources to challenge a significant portion. 


--------------------
\1 EPA can impose temporary controls on new chemicals pending the
development of sufficient information on their effects and has done
so for a small percentage of chemicals. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4


      EPA REGULATES FEW CHEMICALS
      UNDER TSCA
-------------------------------------------------------- Chapter 0:4.1

EPA has issued regulations to control four new and five existing
chemicals determined to present an unreasonable risk.  TSCA does not
define what risk is unreasonable.  However, according to EPA
officials, the threshold is very high in that the agency must present
substantial evidence that the benefits to society of implementing the
controls outweigh the costs.  This standard is especially difficult
for major controls or restrictions because the costs can be extensive
and the full range of benefits may be difficult to document. 
Although EPA had considerable evidence of serious health problems and
spent several years developing a rule to phase out the use of nearly
all products containing asbestos, the Fifth Circuit Court of Appeals
decided in 1991 that the agency had issued the rule on the basis of
insufficient evidence. 

EPA's interpretation that TSCA gives preference to dealing with
chemical risks under other laws, such as the Clean Air and
Occupational Safety and Health acts, has been controversial within
and outside the agency.  While these laws can limit environmental
releases and certain exposures, they do not offer TSCA's flexibility
to ban or restrict chemicals' production, distribution, use, and
disposal.  Some EPA staff, Members of Congress, and environmental
groups believe that EPA should pursue more chemical regulations under
TSCA. 


      EPA HAS NOT FULLY ASSESSED
      CHEMICAL RISKS
-------------------------------------------------------- Chapter 0:4.2

TSCA does not require routine testing of chemicals, and industry
performs limited tests on new chemicals.  Because test data are
generally insufficient, EPA uses a method known as structure activity
relationships analysis to predict health and environmental effects by
comparing new chemicals with chemicals of similar molecular
structures that have been tested.  However, a 1993 study comparing
EPA's predictions with the results of testing required for new
chemicals in the European Union showed that EPA's predictions were
not always accurate, especially for such aspects as the chemicals'
physical properties.  These properties, such as vapor pressure, are
important factors in how much exposure occurs during a chemical's
manufacture and use. 

The information in premanufacture notices that EPA uses to assess
potential exposures to new chemicals, such as production volume and
anticipated uses, are estimates that can change substantially once
EPA completes its review and manufacture begins.  Although TSCA
authorizes EPA to require a manufacturer to submit a new notice under
these conditions, the agency must promulgate a rule in which it
identifies the new uses or activities that may pose health or
environmental hazards. 

EPA has reviewed the risks of about 1,200 existing chemicals.  At the
current rate of about 100 chemicals per year, EPA will need many
years to review just the higher-priority chemicals.  According to
EPA, over 16,000 chemicals are potentially of some concern because of
their production volume and chemical structure.  In addition, the
reviews that have been conducted may already be or may become
outdated as production expands and new uses develop for the chemicals
reviewed. 

TSCA authorizes EPA to issue rules to require testing if the agency
finds that chemicals may present an unreasonable risk or may result
in significant human or environmental exposure.  However, EPA must
expend considerable time and resources to obtain chemical information
from the industry.  According to EPA, promulgating a test rule for a
chemical can take as long as 24 to 30 months and cost about $68,500
to $234,000. 


      LARGE AMOUNTS OF TSCA'S DATA
      ARE CLAIMED AS CONFIDENTIAL
-------------------------------------------------------- Chapter 0:4.3

A 1992 study by a consulting firm found that more than 90 percent of
the premanufacture notices that the firm reviewed contained
information claimed as confidential.  Although EPA officials
recognize that some of these claims are needed to protect trade
secrets, they believe, on the basis of the 1992 study and their
experience with the data, that the claims are excessive.  Under
certain conditions, federal officials can obtain access to
confidential information, but state health and environmental
officials cannot. 

To discourage excessive confidentiality claims, EPA is considering
various actions, including revisions to its regulations to require
substantiation of claims.  Currently, the burden is on EPA to perform
a series of labor-intensive steps to declassify data that it believes
should not be treated as confidential. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
---------------------------------------------------------- Chapter 0:5

The Congress could strengthen EPA's ability to regulate chemicals by
allowing TSCA to be used in preference to other environmental laws,
when appropriate, and establishing a framework for taking action that
is less burdensome for EPA.  The Congress could also improve EPA's
ability to conduct chemical reviews by requiring industry to submit
additional data on new chemicals and shifting to industry some of the
burden for compiling data on existing chemicals.  To increase the
dissemination of TSCA data, the Congress could give the states access
to confidential data and limit confidentiality claims.  GAO also
presents options for the Congress to consider in reauthorizing TSCA
in these areas. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:6

GAO discussed the facts in this report with EPA officials, including
the Director of the Chemical Control Division, Office of Pollution
Prevention and Toxics, who generally agreed with the report's
accuracy.  These officials acknowledged that TSCA's effectiveness has
been limited and that changes to the act's provisions could
strengthen EPA's ability to control harmful chemicals, assess
chemical risks, and increase the dissemination of TSCA data on
chemical hazards.  The EPA officials also suggested changes to
clarify and update information on the agency's implementation of
TSCA, and GAO made changes as appropriate.  As requested, GAO did not
obtain written comments on the draft report. 


INTRODUCTION
============================================================ Chapter 1

Thousands of chemicals are in commercial use in the United States,
and chemical manufacturers introduce many new chemicals into the
marketplace each year.  Although these chemicals are important in
producing many American goods and services, they are often toxic and
can have adverse effects on human health and the environment if
significant levels of exposure to them occur. 

The Toxic Substances Control Act (TSCA) was enacted in October 1976
to provide a safeguard against the introduction of additional
contaminants into the environment and to address the risks posed by
existing chemicals.  Under TSCA, the Environmental Protection Agency
(EPA) may require chemical manufacturers and processors to test
potentially harmful chemicals for the purpose of assessing their
health and environmental effects.  If a chemical poses an
unreasonable risk to health or the environment, EPA can take action
to control its manufacturing, processing, distribution, use, and/or
disposal.  These controls can range from banning the chemical to
requiring warning labels on the chemical or products containing the
chemical when they are sold. 


   CHEMICALS ARE IMPORTANT BUT
   PRESENT POTENTIAL DANGERS
---------------------------------------------------------- Chapter 1:1

More than 7 million recognized chemicals are in existence, and
approximately 80,000 of them are in common use worldwide.  Over
72,000 chemicals have been produced for commercial use in the United
States and are listed in the TSCA inventory of chemicals.  In
addition, about a thousand new chemicals are developed and added to
the inventory each year. 

Chemicals play an important role in people's lives.  In performing
common household activities, consumers use products containing
chemicals, such as cleaning detergents, soaps, and paints.  In
producing a wide variety of other products and industrial processes,
industries use chemicals as solvents, resins, and additives.  Also,
the production of chemicals makes a significant contribution to the
national economy.  The chemical industry, one of the largest
industries in the United States, has a work force of about 850,000,
almost 5 percent of all U.S.  manufacturing workers.  Ten percent of
U.S.  scientists and engineers work for chemical companies, a
reflection of the industry's high technology products and
capital-intensive, complex manufacturing processes.  In 1991, more
than 178 billion pounds of synthetic organic chemicals were produced,
representing $85 billion in chemical sales. 

While the production of chemicals generates many benefits to
consumers and the national economy, the health and environmental
risks associated with chemicals are not always known.  For many
chemicals, there is little knowledge of the ill-effects they might
cause to people and the environment exposed to them.  Human exposure
to some chemicals can contribute directly to health problems, such as
cancers, birth defects, respiratory disorders, and other acute and
chronic diseases.  Also, chemicals released into the environment have
been responsible for problems such as contaminated drinking water
supplies, contaminated fish, air pollution, hazardous waste dumps,
and other adverse impacts on environmental quality. 


   THE FEDERAL GOVERNMENT'S
   IMPORTANT ROLE IN PROTECTING
   AGAINST HARMFUL CHEMICALS
---------------------------------------------------------- Chapter 1:2

Federal laws have been enacted over the years to determine the health
and environmental hazards associated with toxic chemicals and to
address these problems.  These laws, such as the Clean Air Act, the
Clean Water Act, the Resource Conservation and Recovery Act, the
Federal Food, Drug and Cosmetic Act, and the Federal Insecticide,
Fungicide, and Rodenticide Act, are designed to control hazardous
chemicals that may be present in food, drugs, and pesticides and in
the air, water, and soil.  Other laws, such as the Occupational
Safety and Health Act and the Consumer Product Safety Act, can be
used to protect workers and consumers from unsafe exposures to
chemicals in the workplace and at home. 

These laws were generally enacted or substantially amended in the
early 1970s.  Despite their existence, problems with toxic chemicals
continued to occur.  In addition, the Congress became increasingly
concerned about the long-term effects of substantial amounts of
chemicals entering the environment.  TSCA was enacted to authorize
EPA to collect information about the hazards posed by chemical
substances and to take action to control unreasonable risks by either
preventing dangerous chemicals from making their way into use or
placing restrictions on those already in commerce.  Under the act,
EPA can control the entire life cycle of chemicals from their
production, distribution in commerce, and use to their disposal. 
Other environmental and occupational health laws generally control
only disposal or release to the environment, or exposures in the
workplace. 

TSCA applies to new (not yet in commerce) and existing (currently in
the marketplace) "industrial" chemicals.  The act does not apply to
eight categories of chemical products--pesticides, tobacco, nuclear
material, firearms and ammunition, food, food additives, drugs, and
cosmetics--that are regulated under other laws. 


   TSCA'S MAJOR PROVISIONS
---------------------------------------------------------- Chapter 1:3

TSCA's primary purpose of ensuring that chemicals in commerce do not
present an unreasonable risk of injury to health or the environment
is carried out through six major sections of the act, as shown in
table 1.1. 



                          Table 1.1
           
           TSCA's Major Sections for Chemical Data
                    Collection and Control

Section                 Purpose
--------------  ------  ------------------------------------
4                       Chemical testing
5                       New chemical review and control
6                       Chemical regulation
8                       Industry reporting of chemical data
9                       TSCA's relationship to other laws
14                      Disclosure of chemical data
------------------------------------------------------------
Under section 4, EPA can promulgate rules to require chemical
manufacturers and processors to test potentially harmful chemicals
for their health and environmental effects.  To require testing, EPA
must determine that the chemical may present an unreasonable risk or
that substantial exposure may exist. 

Section 5 requires industry to notify EPA at least 90 days before
beginning to manufacture or process a new chemical.  EPA generally
has these 90 days to review the chemical information in the
notification and identify the chemical's potential risks.  If the
chemical will present an unreasonable risk, EPA must act to protect
against the risk.  If insufficient data exist and an unreasonable
risk may be present, EPA can impose temporary controls or
restrictions until sufficient data are developed. 

Section 6 requires EPA to take actions against chemicals for which a
reasonable basis exists to conclude that the chemicals present or
will present an unreasonable risk to health or the environment.  To
adequately protect against a chemical's risk, EPA can take a range of
actions that include banning or restricting the chemical's
production, processing, distribution, or use and requiring warning
labels on the chemical. 

Section 8 directs chemical manufacturers and processors to maintain
records and to submit such information as the EPA Administrator
reasonably requires.  This information can include, among other
things, chemical identity, categories of use, production levels,
by-products, existing data on adverse health and environmental
effects, and the number of workers exposed to the chemical.  In
addition, chemical manufacturers, processors, and distributors are
required to submit lists or copies of certain health and safety
studies to EPA and to report to EPA information which indicates that
a chemical presents a substantial risk. 

Section 9 sets out TSCA's relationship to other laws.  If EPA
determines, in its discretion, that an unreasonable chemical risk may
be prevented or sufficiently reduced by action under a federal law
not administered by EPA, it must refer information on the chemical's
risk to the agency administering the other law.  That agency must
initiate action to regulate the chemical or publish in the Federal
Register why no action is needed.  The section also directs EPA to
use other laws it administers to protect against unreasonable risks,
unless EPA determines that it is in the public interest to protect
against such risks under TSCA. 

Section 14 authorizes EPA to release chemical information obtained by
the agency under the act.  Certain information, such as data
disclosing chemical processes, can be claimed as confidential
business information by chemical manufacturers and processors.  EPA
must protect such information against disclosure unless public
disclosure is necessary to protect against an unreasonable risk. 


   EPA'S ORGANIZATION AND
   RESOURCES FOR IMPLEMENTING TSCA
---------------------------------------------------------- Chapter 1:4

The primary responsibility for implementing TSCA resides with EPA's
Office of Pollution Prevention and Toxics, formerly the Office of
Toxic Substances.  Nearly all program activities are carried out by
EPA headquarters staff.  EPA's regional offices have some enforcement
responsibilities. 

The Office's TSCA activities center around two principal programs: 
new chemicals and existing chemicals.  The new chemicals program
implements TSCA section 5, whereas the existing chemicals program
implements the section 4 testing, section 6 chemical control, and
section 8 information-gathering provisions.  Section 14 governs the
disclosure of information collected from industry under both
programs. 

One of EPA's first task's under the act was to compile an inventory
of chemicals already in U.S.  commerce.  This initial inventory,
which EPA published in July 1979 and then in revised form in July
1980, contained about 62,000 chemicals.  EPA's new chemical program
began with publishing the initial inventory.  Any chemicals not on
the inventory were to be considered new substances under TSCA and
subject to the section 5 premanufacture notification requirements. 
EPA adds new chemicals to the inventory when it completes its review
of the premanufacture notices and is informed that manufacture or
import of the chemicals has begun.  About 10,000 chemicals have been
added to the inventory since 1979, bringing the total to over 72,000
chemicals.  Chemicals on the inventory are referred to as existing
chemicals. 

The Office of Pollution Prevention and Toxics' fiscal year 1994
budget for implementing TSCA is about $67.9 million.  Of this amount,
about $14.3 million is for new chemicals and $53.6 million is for
existing chemicals.  Although the amount for existing chemicals
appears substantially larger than that for new chemicals, about $5.7
million is for chemical testing and $31.4 million is to carry out
congressionally mandated activities related to the control of three
chemicals:  asbestos, lead, and polychlorinated biphenyls (PCBs). 

EPA's regulatory control actions for new and existing chemicals are
discussed in chapter 2.  EPA's programs for the review of the risks
of new and existing chemicals are discussed in chapters 3 and 4,
respectively.  Chapter 5 discusses EPA's implementation of TSCA's
confidential business information provisions. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:5

Citing concerns about EPA's implementation of TSCA, the Chairman,
Subcommittee on Toxic Substances, Research and Development, Senate
Committee on Environment and Public Works, and the Chairman,
Subcommittee on Environment, Energy, and Natural Resources, House
Committee on Government Operations, requested that we review EPA's
efforts to (1) control chemicals under TSCA, (2) assess the risks of
chemicals before and after they enter commerce, and (3) reduce the
amount of information collected under TSCA that cannot be
disseminated because industry claims that it is confidential.  The
Chairmen also requested that we identify ways to make TSCA more
effective in these areas. 

To review EPA's progress in evaluating the risks of chemicals and
controlling those harmful to human health or the environment, we
determined the results of EPA's reviews for about 1,500 new chemicals
from January 1990 to May 1993.  In addition, we determined the number
of existing chemicals reviewed by EPA for the various types and
levels of review conducted by the agency and examined EPA's files for
selected chemicals.  We also identified EPA's requirements, policies,
and guidelines on how the new and existing chemical review and
control programs work and determined the extent of actions taken by
EPA to control new and existing chemicals.  These efforts were
augmented by interviews with officials of EPA and the Occupational
Safety and Health Administration and representatives of the Chemical
Manufacturers Association and the Environmental Defense Fund. 

To identify potential regulatory changes to make TSCA more effective,
we (1) interviewed officials from EPA's Office of Pollution
Prevention and Toxics, the Interagency Testing Committee, the
Chemical Manufacturers Association--which represents companies that
account for over 90 percent of the U.S.  bulk chemical
production--and the Environmental Defense Fund; (2) reviewed
literature and congressional hearings on TSCA and attended various
public meetings and conferences sponsored by EPA and others, such as
the American Chemical Society; (3) reviewed written comments
submitted to us by representatives of the Chemical Manufacturers
Association; and (4) compared EPA's new and existing chemical
programs with programs implemented in three other countries--Canada,
Germany, and Sweden. 

These countries were chosen to help us identify potential changes to
TSCA because each country has recently revised its chemical control
laws and taken other actions to improve its chemical review and
control programs.  These countries were also selected because they
have important characteristics that are similar to those of the
United States:  All are industrialized nations and have extensive
experience with the review and control of chemical substances.  In
addition, Canada and Germany produce a considerable amount of
chemicals. 

For each of the foreign countries we reviewed, we obtained national
laws, technical literature, and government documents that describe
the country's chemical control programs.  We also interviewed the
foreign officials responsible for implementing the chemical
substances control laws in the three countries.  We limited our data
collection efforts in these countries to the national level,
recognizing that there may be significant differences across
provinces.  We further interviewed representatives from the Canadian
Chemical Producers' Association, the German Institute for Applied
Ecology (Oko-Institut), the German Chemical Industry Association, the
Association of Swedish Chemical Industries, and the Swedish Society
for the Conservation of Nature to obtain views on their respective
countries' chemical control programs.  Finally, we interviewed
officials of the Commission of the European Union and of the European
Center for Ecotoxicology and Toxicology of Chemicals in Brussels,
Belgium, to obtain their views on reviewing and controlling new and
existing chemicals.  Our descriptions of these countries' laws are
based on interviews with government officials and written materials
they provided. 

This report does not discuss all possible options for revising TSCA. 
Those options that are discussed were selected because they address
major constraints to EPA's implementation of the act.  Our selection
of these options reflects (1) our knowledge of EPA's implementation
of TSCA obtained during this and previous reviews of the agency's
toxics programs, (2) foreign countries' approaches to reviewing and
controlling harmful chemicals, and (3) views provided by U.S. 
government officials and representatives of the chemical industry and
environmental groups. 

Our review was performed between September 1992 and July 1994 in
accordance with generally accepted government auditing standards. 


TSCA'S REGULATORY CONTROL
AUTHORITIES ARE SELDOM USED
============================================================ Chapter 2

Because issuing regulations under TSCA is so difficult, this course
of action is generally not a viable alternative when EPA is
considering how best to deal with toxic chemical concerns.  In the
almost 18 years since TSCA was enacted, EPA has issued regulations
under the act to control only nine chemicals--five existing chemicals
and four new ones.  The agency has been more successful in entering
into individual agreements with manufacturers of new chemicals to
take certain actions, such as to implement workplace practices,
pending the development of additional data.  New- chemical
manufacturers have also voluntarily conducted certain toxicity tests
or withdrawn their plans to manufacture the chemicals when EPA has
indicated plans to require these controls.  EPA does not have similar
authority for existing chemicals but has begun to encourage
industry's voluntary actions to reduce the risks of these chemicals. 

EPA has issued only a few regulations under TSCA because the act's
legal standards are very high, and the burden of proof is essentially
on EPA.  In addition, consensus does not exist in the Congress and
the environmental community on whether the act is intended to have a
greater role in addressing toxic chemicals.  Some in the Congress and
in the environmental community believe that TSCA is a comprehensive
or umbrella law; their belief is based on the fact that the act deals
exclusively with industrial chemicals and provides EPA with
authorities to control chemicals throughout their life cycle, from
production to use and disposal.  However, some EPA officials and
industry representatives believe that the act's purpose is generally
limited to filling gaps in other health and environmental laws.  The
act does not clearly articulate what EPA is to achieve through the
use of its regulatory authorities. 


   EPA HAS CONTROLLED FEW
   CHEMICALS UNDER TSCA
---------------------------------------------------------- Chapter 2:1

EPA seldom uses TSCA to regulate existing chemicals.  In addition,
nearly all of the actions for new chemicals have been temporary
controls or voluntary actions by individual chemical manufacturers,
primarily to address occupational or worker exposure. 


      EPA HAS PLACED CONTROLS ON
      FIVE EXISTING CHEMICALS
-------------------------------------------------------- Chapter 2:1.1

Since the enactment of TSCA in 1976, EPA has issued regulations under
the act to control five existing chemicals:  (1) polychlorinated
biphenyls (PCB), (2) chlorofluorocarbons (CFC), (3) dioxin, (4)
asbestos, and (5) hexavalent chromium.  Even this small number of
chemicals does not fully indicate EPA's reluctance to use TSCA's
regulatory authorities because the act itself required EPA to
regulate PCBs.  In addition, EPA attempted to refer asbestos to the
Occupational Safety and Health Administration for action until the
Congress and public interest groups objected.  Furthermore, EPA's
Office of Air and Radiation requested that the Office of Pollution
Prevention and Toxics regulate hexavalent chromium under TSCA because
substantially more resources would be needed to enforce a Clean Air
Act regulation.  Officials in the Office of Pollution Prevention and
Toxics told us that a ban under TSCA was a better approach than EPA's
having to issue and monitor compliance with an emissions standard for
hexavalent chromium under the Clean Air Act. 

The regulations for two of the chemicals were comprehensive.  The PCB
rules, as required in the act, address manufacture, distribution,
use, and disposal.  A 1989 asbestos rule would have phased out most
uses, but the rule was overturned by a 1991 court decision.\1 On the
other hand, the CFC rule banned the substance's use in aerosol spray
cans--other uses were phased out later under the Clean Air Act. 
Under the dioxin rule, EPA prevented land disposal of one kind of
dioxin by one manufacturer (other companies were required to notify
EPA if they intend to dispose of this substance).  In addition, the
hexavalent chromium rule covered only its use in commercial cooling
towers, not industrial ones.  (The TSCA regulations for these five
chemicals are discussed in more detail in app.  I.)


--------------------
\1 Corrosion Proof Fittings v.  EPA, 947 F.2d 1201 (5th Cir.  1991). 


      A SMALL PERCENTAGE OF NEW
      CHEMICALS HAS BEEN
      CONTROLLED
-------------------------------------------------------- Chapter 2:1.2

Of the 23,971 new chemicals reviewed, some action to reduce risks
were taken on 2,431, or about 10 percent of the chemicals.  In
addition to issuing regulations to impose certain controls on 4
chemicals, EPA entered into agreements or consent orders with
manufacturers of 626 chemicals to implement temporary workplace
practices or controls during manufacture and/or to perform toxicity
testing when the chemicals' production volumes reached certain
levels.  For 827 chemicals, the manufacturers voluntarily agreed to
perform toxicity testing before EPA completed its reviews.  For the
remaining 974 chemicals, the manufacturers or processors withdrew the
premanufacture notices after EPA had indicated its plans to require
testing or controls. 

In the four cases involving rule-making, EPA determined that the
chemicals presented an unreasonable risk of injury to health or the
environment and promulgated rules to protect against the risks. 
These chemicals--
(1) mixed mono and diamides of an organic acid, (2) triethanolamine
salts, (3) triethanolamine salt of tricarboxylic acid, and (4)
tricarboxylic acid--are ingredients in metalworking fluid.  The
rules, promulgated in 1984, prohibit adding any nitrosating agent,
including nitrites, to metalworking fluid that contains these
substances.  According to EPA, the addition of nitrites or other
nitrosating agents to the substances leads to the formation of a
substance known to cause cancer in laboratory animals. 

In the cases of the consent orders for 626 chemicals, EPA determined
that insufficient data existed to assess the chemicals' health and
environmental effects and that an unreasonable risk may be present. 
Under its section 5 authority, EPA cannot require that these data be
developed, but it is authorized to control human and environmental
exposures to the substances until sufficient data are available. 
Consent orders allow limited manufacture, processing, distribution in
commerce, use, and disposal of chemicals until the data are
developed.  In return, the firms agree in writing to implement
workplace practices or controls to eliminate or reduce exposures to
the chemicals.  Examples of these controls are requirements for
employees' use of protective equipment, such as gloves when skin
contact is a concern, and implementation of worker training programs
on how to handle the chemicals.  Consent orders may also involve
certain restrictions, such as a prohibition or limitation on a
chemical's release to water, or certain tests that must be performed
when a chemical reaches a designated production volume.  According to
EPA, this production volume level is set at the point at which the
agency estimates that profits from a chemical will support the cost
of testing. 

In the 827 cases in which manufacturers voluntarily agreed to develop
certain toxicity data before EPA completed its review, EPA determined
that sufficient toxicity information was not available to evaluate
the chemicals' effects and anticipated that health or environmental
exposures could not be controlled through the routine workplace
practices or controls normally incorporated into consent orders.  EPA
recommended additional testing and evaluation of the results before
allowing the chemicals to be produced or imported.  According to EPA,
voluntary testing may be the best option available to the
manufacturer if releases or exposures cannot be controlled pending
testing or if the requested testing is relatively cheap and not very
time-consuming.  The only other alternative is to withdraw the
premanufacture notice. 

In the 974 cases in which premanufacture notices were withdrawn from
the review process, the manufacturers decided not to proceed with
plans to market the chemicals in the face of EPA's plans to require
testing, impose controls, or prohibit production or use.  EPA
officials told us that manufacturers often drop plans to market a new
chemical when the chemical's niche in the marketplace is uncertain
and EPA informs them that toxicity data will be needed for the agency
to complete its evaluation of the chemical's risks.  This testing can
cost from a few hundred to thousands of dollars to perform. 


   TSCA'S REGULATORY FRAMEWORK
   ENSURES THAT EPA WILL CONTROL
   FEW CHEMICALS
---------------------------------------------------------- Chapter 2:2

Although TSCA appears to give EPA broad authority to control harmful
chemicals, several key provisions and their interpretation or
implementation by the agency and the courts have worked together to
discourage chemical control actions under the act.  The legal
standards for taking action and the burden of proof placed on EPA by
the act make it extremely difficult for the agency to use this
authority.  In addition, the act encourages EPA to use other
environmental laws to control chemical risks or to refer concerns to
other federal agencies with health and safety responsibilities. 


      EPA MUST MEET HIGH LEGAL
      STANDARDS TO TAKE CONTROL
      ACTIONS
-------------------------------------------------------- Chapter 2:2.1

Under section 6 of TSCA, the existence of an unreasonable risk of
injury to human health or the environment is the trigger for
controlling an existing chemical.  To make an unreasonable risk
determination, the act requires EPA to consider more than whether the
chemical is toxic or harmful to humans, animals, plants, and other
organisms.  The agency is to also determine the magnitude of human
and environmental exposures to the chemical.  Once it determines the
extent of the risks presented by the chemical, EPA must determine
whether these risks are unreasonable.  According to EPA officials,
the agency must, in effect, perform a cost-benefit analysis,
considering the economic and societal costs of placing controls on
the chemical.  EPA must take into account the benefits provided by
the various uses of the chemical, the availability of substitutes,
and the reasonably ascertainable economic consequences of regulating
the chemical after considering the effects of such regulation on the
national economy, small business, technological innovation, the
environment, and public health. 

Section 5 of TSCA contains standards for imposing controls on the
production, distribution, use, and disposal of new chemicals.  If EPA
determines that a new chemical presents an unreasonable risk, the
agency must impose controls to protect against the risk.  In the
event that EPA does not have sufficient information to make a
determination of risk, the agency has the authority under section 5
to impose controls to limit exposures to the chemical until
sufficient data are developed.  In the latter case, EPA can impose
these controls if it can demonstrate that the chemical may present an
unreasonable risk or that substantial human exposure may occur. 
According to EPA officials, it is less difficult to demonstrate that
a chemical may present rather than actually presents an unreasonable
risk. 

The reason for these requirements lies in congressional concern about
the cost to industry and the economic impact of EPA's actions.  TSCA
provides EPA with such sweeping authority to impose controls,
including bans or limits on production and restrictions on the use of
harmful chemicals, that a single action could have substantial
economic consequences.  The requirements also reflect an underlying
philosophy of the statute that manufacturers and processors have the
right to produce and market chemicals and that before EPA can take
any legal action to restrict this right, it must demonstrate that the
risks outweigh both the costs to industry and the lost benefits of
the unrestricted use of the chemical. 

While concerns about the potential impact of EPA's regulations on
industry are legitimate, the requirement for a finding of
unreasonable risk has proven difficult for EPA to implement.  TSCA
does not define "unreasonable risk" and provides little guidance on
what level of risk should be considered unreasonable under the act. 
In a June 1991 report, we cited three cases in which EPA had
determined, on the basis of test results, that the
chemicals--cyclohexanone, ethylhexanoic acid, and octylphenol--were
dangerous but took no regulatory action.\2 We reported, for example,
that over 839,000 workers were exposed to cyclohexanone and that test
data had shown that the chemical adversely affected the development
of embryos and fetuses in laboratory test animals.  According to EPA
officials, these chemicals' risks were not such that they constituted
a significant risk for priority review under TSCA section 4(f) or an
unreasonable risk for control under section 6.\3

EPA officials told us the agency uses a "high threshold of risk" to
assess whether a chemical poses significant or unreasonable risks. 

Given the recognized dangers of cyclohexanone, ethylhexanoic acid,
and octylphenol and the fact that EPA had no criteria and methodology
to guide its managers in determining when chemicals present a
significant or unreasonable risk, our 1991 report questioned the
basis for EPA's failure to take regulatory action on the chemicals. 
We recommended that EPA establish criteria and a methodology for
determining when chemicals present risks that would trigger
implementation of TSCA's regulatory provisions.  We further
recommended that the criteria and methodology include definitions of
significant and unreasonable risk and quantitative and qualitative
measures to determine when such risks are present.  EPA officials
told us that they do not believe that the agency can develop overall
criteria and must determine significant or unreasonable risk for each
individual case because each circumstance is so different.  We
continue to believe that some criteria or guidelines are needed to
provide for the agency's consistent and systematic implementation of
TSCA's regulatory provisions. 

A major difficulty in making an unreasonable risk finding is the
level of evidence that the act requires.  EPA must develop
"substantial evidence" through rule-making sufficient to withstand
judicial review.  This requirement, together with the unreasonable
risk requirement, imposes a burden of legal proof that is greater
than in the Administrative Procedure Act, which requires only that an
agency's actions not be arbitrary and capricious.\4 The burden is on
EPA to obtain the necessary evidence to (1) prove that the use of a
particular chemical will cause great harm that cannot be alleviated
in a less costly manner and (2) demonstrate that the agency has
considered whether less harmful, economically viable substitutes are
readily available. 

Even if EPA had the substantial resources needed to develop this
evidence, the outcome would not be assured.  In current
state-of-the-art risk assessments, some uncertainty and some basis
for a legal challenge almost always exist.  Furthermore, the costs to
the economy of regulating a chemical are usually much more easily
documented than the risks of the chemical or the benefits associated
with controlling it, according to EPA officials. 

The court decision on EPA's 1989 asbestos rule illustrates the
substantial burden that TSCA places on EPA.  As discussed above, the
rule prohibited the future manufacture, importation, processing, and
distribution of asbestos in almost all products.  Some of the
manufacturers of these asbestos products filed suit against EPA,
arguing that the rule was not promulgated on the basis of substantial
evidence.  In October 1991, the U.S.  Court of Appeals for the Fifth
Circuit agreed with the manufacturers and sent the rule back to EPA
for further consideration.\5

In its ruling, the court concluded that the burden is on EPA to
justify that the products it bans present an unreasonable risk and
that EPA did not present sufficient evidence to justify the ban on
asbestos.  In reaching this conclusion, the court found that EPA did
not consider all necessary evidence and failed to show that the
control action it chose was the least burdensome reasonable
regulation.  As articulated by the court, the proper course of action
for EPA would have been to consider the costs and benefits of each
regulatory option available under section 6, starting with the less
restrictive options, such as product labeling, and working up through
a partial ban to a complete ban.  The court further criticized EPA's
ban of products for which no substitutes are currently available
because TSCA explicitly requires the agency to consider the benefits
of the substance for various uses and the availability of various
substitutes for those uses.  Thus, EPA would need to analyze each
product or use of a chemical, which can number up to a hundred or
more. 

The court's decision on the asbestos rule is especially revealing
about section 6 because EPA spent 10 years preparing the rule.  In
addition, asbestos is generally regarded as one of the substances for
which EPA has the most scientific evidence or documentation of
substantial adverse health effects.  As a result of the court
decision, EPA finds itself faced with the need to commit even more
resources to any effort to regulate under section 6 and the
possibility that the effort may not end in regulation.  Officials of
EPA's Office of Pollution Prevention and Toxics told us that with the
court decision in the asbestos case, EPA most likely will not attempt
to issue regulations under section 6 for comprehensive bans or
restrictions on chemicals. 


--------------------
\2 Toxic Substances:  EPA's Chemical Testing Program Has Not Resolved
Safety Concerns (GAO/RCED-91-136, June 19, 1991). 

\3 Section 4(f) of TSCA is a priority review provision that requires
EPA to initiate appropriate control action or publish a Federal
Register notice that the risk is not unreasonable within 180 days
(plus a 90-day extension) after receipt of test data or any other
information which indicates that there may be a reasonable basis to
conclude that a chemical presents, or will present, a significant
risk of serious or widespread harm to human beings from cancer, gene
mutations, or birth defects. 

\4 The Administrative Procedure Act, among other things, sets
procedures and standards for federal agencies' rule-makings. 
Agencies are to adhere to these procedures and standards unless the
law under which a particular rule-making is taking place sets
different procedures or standards. 

\5 Corrosion Proof Fittings v.  EPA, 947 F.2d 1201 (5th Cir.  1991). 


   TSCA'S RELATIONSHIP TO OTHER
   LAWS LIMITS ITS USE
---------------------------------------------------------- Chapter 2:3

Section 9 of TSCA generally requires that other environmental laws be
used to address the risk posed by a chemical, if the EPA
Administrator determines that such laws can eliminate or sufficiently
reduce the chemical's risk.  EPA has usually interpreted this section
to mean that TSCA should be used primarily to fill gaps in the
authorities of other laws.  As a result, the Office of Pollution
Prevention and Toxics has referred essentially all risks identified
for existing chemicals to other EPA offices or agencies for action
under these other laws.  (New chemicals are dealt with under TSCA
because the act is the only federal legislation that can address
concerns about industrial-use chemicals before they are manufactured
and enter commerce.)

TSCA contains little guidance on section 9 and does not clearly
articulate what EPA is expected to achieve through the use of the
act's regulatory authorities.  Some controversy has occurred when EPA
has referred chemicals to other agencies rather than take control
action, and the Congress has at times expressed concern about how
little EPA has achieved under TSCA.  The center of this debate is
whether TSCA is a comprehensive toxics law or is a gap-filling act,
as it has usually been viewed by EPA. 


      EPA HAS GENERALLY USED TSCA
      TO FILL GAPS IN OTHER LAWS
-------------------------------------------------------- Chapter 2:3.1

According to EPA officials, TSCA's relationship to other laws has
been controversial within the agency.  EPA does not have a written
policy or guidance on when it should exercise its discretion to use
TSCA or to refer a chemical for action under these other laws. 
However, a 1985 legal opinion by the agency concluded that section 9
of the act states a preference for the other laws to be used to
control chemical risks.  In addition, various staff members in the
Office of Pollution Prevention and Toxics told us that under section
9, they must refer chemical concerns to other offices or agencies for
action if either an existing regulation could be revised or a new
regulation could be promulgated under the authorities of the other
acts to address the concerns.  On the other hand, other office
representatives told us that they believe that TSCA is more than a
gap-filler.  These officials agreed, however, that TSCA's role and
relationship to other environmental laws are unsettled issues. 

Because these other laws collectively cover the major sources of
toxic chemical exposure, Office of Pollution Prevention and Toxics
officials have identified very few chemical risks that could not be
addressed, at least to some extent, under the authorities of other
laws.  (As shown in app.  II, virtually all risk concerns--human
health, environmental, worker, and consumer--associated with the
industrial use of chemicals are covered by other laws.) As a result,
nearly all chemical risks that EPA officials believe should be
controlled are referred to the EPA offices or other agencies
responsible for implementing those laws. 

The five existing chemicals that EPA regulated under TSCA also could
have been addressed under other legislation.  EPA's Office of Air
requested that hexavalent chromium in cooling towers be controlled
under TSCA because a TSCA regulation could be enforced with fewer
resources than a Clean Air Act regulation.  Likewise, the dioxin
chemical could have been regulated under the Resource Conservation
and Recovery Act, but EPA decided to use TSCA because a rule under
the act would become effective immediately and the reporting
requirement would cover all facilities. 

By far the largest number of chemicals that EPA has controlled under
TSCA are new chemicals, and EPA usually takes these actions to
address concerns about workers' exposure.  The Director of the Office
of Pollution Prevention and Toxics told us that it is understandable
that most control actions under TSCA are directed at new chemicals
and worker protection because no other statute or agency program
addresses worker exposure concerns before the chemicals are on the
market and in the workplace. 


      EPA FORMALLY REFERS FEW
      CHEMICAL CONCERNS TO OTHER
      AGENCIES
-------------------------------------------------------- Chapter 2:3.2

For chemicals referred to other agencies, section 9 of TSCA requires
that these agencies respond to EPA's concerns.  EPA is to submit to
the other agency a report describing the risk, the activity or
activities that present the risk, and the information on which the
report is based.  EPA is to publish the report in the Federal
Register.  The other agency is to determine if action under the law
it administers can prevent or sufficiently reduce the risk and to
issue an order declaring whether the activity or activities cited in
the report present such risk.  The agency is to respond to EPA within
the time specified in EPA's report (the specified time cannot be less
than 90 days).  The agency must also submit a detailed statement of
its findings and conclusions and publish them in the Federal
Register. 

EPA has referred four chemicals to other agencies under the formal
process established by section 9.  Three referrals were to the
Occupational Safety and Health Administration (OSHA) for
4,4'-Methylene dianiline in 1985, 1,3,-Butadiene in 1985, and glycol
ethers in 1986, all because of potential adverse health effects from
exposure in the workplace.  EPA also referred dioxin in bleached wood
pulp and paper products to the Food and Drug Administration in 1990
because of concerns about exposure to dioxin in paper materials used
to package food.  According to EPA officials, the process of formally
referring chemicals to other agencies has not been very effective for
two reasons.  First, EPA has difficulty meeting the unreasonable risk
standard contained in the section, and the term "unreasonable risk"
is not relevant to agencies, such as OSHA, whose legislation does not
contain this standard.  Second, EPA does not find it easy to
influence the agendas of other agencies that have their own
priorities. 

EPA has also informed OSHA through Federal Register notices of its
concerns about four other chemicals under a provision of section 9
that requires the Administrator to consult and coordinate with other
agencies to obtain maximum enforcement of the act with the least
burden of duplicative requirements.  For these chemicals, EPA did not
determine that the risks were unreasonable, and OSHA was not required
to either formally rebut the finding or to report on actions to
control the exposure.  EPA believed that these
chemicals--toluenediamine, p-Dichlorobenzene, 4,4'-Methylene bis
chloroaniline, and formaldehyde--presented a significant risk to
workers but not an unreasonable risk considering other factors, such
as the cost of regulation and the availability of suitable
substitutes.  In 16 additional cases, EPA officials informally
referred or made known their concerns about chemicals to other
agencies.  According to EPA officials, it is also general practice to
share risk assessments related to occupational issues with OSHA and
the National Institute for Occupational Safety and Health. 

TSCA does not establish a formal mechanism for the Office of
Pollution Prevention and Toxics to use when it refers chemical risks
to other offices within EPA.  The act also does not require the other
offices to formally report on the actions they plan to take to
address the risks.  Although EPA did not have information on the
number of these referrals, our discussions with EPA officials and
review of selected chemical files indicate that these referrals are
more numerous than formal referrals. 


      SOME BELIEVE EPA SHOULD USE
      TSCA AS COMPREHENSIVE
      LEGISLATION
-------------------------------------------------------- Chapter 2:3.3

The intent of Section 9 and its effect on TSCA's role in
environmental protection has been a source of contention.  The first
attempt to use section 9 referrals to other agencies was in February
1985, when EPA announced its intention to abandon its efforts to
regulate asbestos and to refer the substance to the Occupational
Safety and Health Administration.  This announcement caused a public
controversy in which the Congress and environmental groups charged
that EPA and the Office of Management and Budget were subverting the
regulatory process.  In March 1985, EPA suspended the referral
process and continued with its efforts to issue the 1989 asbestos
rule. 

Before and during the debate on the attempted asbestos referral,
various offices within EPA, the Office of Management and Budget, and
the Congress expressed various positions on the proper use of section
9 for coordinating interagency regulatory action.  The major point of
contention was whether TSCA should be an "umbrella" statute aimed at
regulating all unreasonable chemical risks or a gap-filler
essentially addressing only those risks that cannot be controlled
under other laws. 

This contention over TSCA's role has continued as congressional
oversight and authorization subcommittees have recently criticized
EPA's lack of progress in controlling potentially harmful chemicals. 
For example, the Chairman of the Subcommittee on Toxic Substances,
Research and Development, Senate Committee on Environment and Public
Works, introduced a bill, the Lead Exposure Reduction Act of 1993, in
April 1993 because EPA had not taken action to reduce the levels of
lead in the environment, including restrictions on the continuing
uses of certain lead-containing products. 

A central aspect of this debate is that TSCA contains authorities to
eliminate or restrict the production or use of toxic chemicals,
whereas other health and environmental laws generally accept the
production and use of the chemicals and provide for limits on
exposures or discharges to the environment.  For example,
environmental laws generally provide for standards governing the
amount of pollution that can be emitted or discharged by a single
source (performance standards) or standards governing pollution
abatement technology and practices that companies must adopt
(technology standards). 

In addition, these other laws have been slow to address toxics.  Few
of the chemicals used in commerce have been regulated under the laws. 
(See app.  III for a description of the status of regulation of
toxics under four of these laws:  the Clean Water, Clean Air, Safe
Drinking Water, and Occupational Safety and Health Acts.) Billions of
pounds of toxic chemicals are released to the environment each year,
and workers and consumers are exposed to chemicals that may put them
at risk.\6

Recognizing the need for a more comprehensive approach to addressing
toxics, the Office of Pollution Prevention and Toxics is working to
establish a program that focuses on pollution prevention and other
strategies, such as promoting the design, development, and
application of safer chemicals, processes, and technologies, to
reduce toxic exposures, releases, and risks.\7 Although TSCA's
regulatory control authorities can approach toxic chemical concerns
at the front-end by providing for controls or restrictions on
chemical production, processing, distribution, and use, the office's
approach is to seek voluntary actions by industry to reduce toxic
releases or take other risk reduction actions.  EPA officials believe
that voluntary actions will take less time and fewer office resources
than promulgating regulations under TSCA.  In addition, they believe
that the approach has the advantage that industry, which should have
the most knowledge about the operations or circumstances generating
the risk, identifies and plans the specific actions that it will
take.  According to EPA, industry has taken voluntary actions, such
as labeling, testing, and undertaking an emissions reduction program,
on 12 chemicals or groups of chemicals. 


--------------------
\6 For example, the Toxics Release Inventory reported that
manufacturing facilities in 1992 (the latest available data) released
about 3.2 billion pounds of toxics into the environment and sent an
additional 4.4 billion pounds off-site for treatment, disposal,
energy recovery, and recycling.  The inventory reports on releases of
about 300 chemicals and 20 chemical categories. 

\7 EPA, on an agencywide basis, is also placing an increasing
emphasis on pollution prevention.  Citing the difficulty of achieving
substantially greater reductions in environmental releases through
the technology-based, end-of-the-pipe type of controls on which EPA's
major programs are based, the agency's pollution prevention efforts
focus on actions taken up-front, or during manufacturing and
processing, to avoid or reduce the generation of toxic wastes. 


   OTHER COUNTRIES HAVE DIFFERENT
   APPROACHES TO REGULATING
   CHEMICALS
---------------------------------------------------------- Chapter 2:4

Although the other countries that we visited during our
review--Canada, Germany, and Sweden--consider economic impacts in
their regulatory decision-making process, none of them used the
unreasonable risk standard.  For example, the Canadian Environmental
Protection Act of 1988 authorizes the government to control chemicals
that are "toxic." The act defines a substance as toxic if it is
entering or may enter the environment in a quantity or concentration
or under a condition having or that may have an immediate or
long-term harmful effect on the environment, health, or human life. 
A chemical is also defined as toxic if it is entering or may enter
the environment in a quantity or concentration or under conditions
constituting or that may constitute a danger to the environment on
which human life depends.  Canadian officials, who are familiar with
TSCA, told us that they believe it is easier to control chemicals
under their toxic standard than under the United States' unreasonable
risk standard.  According to Canadian officials, some type of control
is likely for chemicals found toxic.  The type of control will depend
on such factors as the extent and type of risks and economic impacts. 

In Germany, the major focus of the chemical control law is to
classify and label chemical products on the basis of their toxicity. 
In addition to determining the labeling of a chemical, classification
is the starting point for risk assessment.  The classifications also
influence legislation concerned with aspects of risk management, such
as worker protection.  The risk assessments can result in additional
testing or the imposition of certain controls on the chemical, such
as use restrictions.  Bans or major restrictions on chemicals are
rare because of the complex process established for taking these
actions. 

In Sweden, the major focus is also on the classification and labeling
of chemicals on the basis of their toxicity.  Certain mandatory
controls are established for each classification category.  Use
restrictions may also apply, depending on the chemical's
classification.  Sweden's chemical control law also requires industry
to substitute less hazardous chemicals for more hazardous ones. 

Although the Swedish government has banned or severely restricted
only a few chemicals, it has established a list of 13 undesirable
chemicals that it wants to eliminate or significantly reduce by the
year 2000.\8 These chemicals have been identified for complete bans
or severe restrictions on use because they are considered to have
sufficiently hazardous properties in combination with widespread use
to warrant such action.  According to Swedish officials, the phaseout
of chemicals by the government cannot be avoided when the chemical
industry does not follow through with its responsibility under the
act to substitute for hazardous chemicals. 


--------------------
\8 The 13 chemical substances that Sweden plans to phase out are
lead, mercury, cadmium, organic tin or organotin compounds, methylene
chloride, trichloroethylene, tetrachlorethylene, creosote, arsenic,
brominated flame retardants, phthalates, nonylphenolethoxylates, and
chlorinated paraffins. 


   OPTIONS FOR STRENGTHENING
   TSCA'S CHEMICAL CONTROL
   AUTHORITIES
---------------------------------------------------------- Chapter 2:5

TSCA's regulatory provisions could be strengthened by revising the
act to clarify its role and relationship to other laws and to reduce
EPA's regulatory burden.  Additional legislative changes would be
needed if TSCA is to be a comprehensive toxics statute. 


      CLARIFYING TSCA'S ROLE AND
      RELATIONSHIP TO OTHER LAWS
-------------------------------------------------------- Chapter 2:5.1

The overlap among TSCA and EPA's other environmental laws
demonstrates the need for the Congress to establish clear
expectations for EPA's use of TSCA's regulatory authorities.  If EPA
is to limit use of the act to filling gaps in other legislation, the
act requires little change.  The Congress could revise the act to
define what gaps the act should fill or what circumstances are in the
public interest and would invoke TSCA's authorities. 

If TSCA is to be an umbrella act, the Congress could simply revise
the act to eliminate its provisions on its relationship to other
laws.  EPA would then have the option to pursue risk reductions
through restrictions on chemical production, distribution, and use or
through limits on releases or exposures, depending on which would be
the most cost-effective.  EPA's efforts to address toxic chemical
concerns or risks could be carried out either through a strategy of
actions under other environmental and health and safety laws, through
its own authorities, or both.  TSCA, with its authorities to control
chemical production, distribution, and use, would be on an even
footing with other environmental laws that prescribe performance and
technology standards limiting the amount of environmental releases or
exposures.  An alternative would be to revise these provisions to
explicitly provide for EPA to select the most cost-effective course
of action using TSCA and/or one or more of the other laws. 


      REDUCING EPA'S REGULATORY
      BURDEN
-------------------------------------------------------- Chapter 2:5.2

To ensure that EPA is not discouraged from taking chemical control
actions under TSCA, the Congress could revise the act in two ways. 
First, it could change the unreasonable risk standard and the
requirement that EPA use the least burdensome regulation adequate to
regulate a substance.  Second, it could revise the act's requirement
that EPA develop substantial evidence to support a regulation. 

One way to change the risk standard would be to authorize EPA to take
control actions when it identifies significant rather than
unreasonable risks.  EPA could then make a judgment on whether the
chemical's toxicity and exposures constitute significant risks of
harm to health or the environment.  EPA officials view the term
"significant risk," as they use it in section 4(f) of TSCA to
identify chemicals for priority review, as a very high threshold for
action.  However, they expect that demonstrating significant risk
would be less demanding than demonstrating unreasonable risk.  While
"significant" risk implies a finding that the risks are substantial
or serious, EPA believes that a finding of "unreasonable" risk
requires an extensive cost-benefit analysis. 

For those chemicals that pose a significant risk, EPA would determine
the most cost-effective actions to adequately reduce the risks.  The
costs and benefits of the contemplated regulations would not be
factors in deciding whether to reduce risks; they would be
considerations in selecting a specific course of action to deal with
the risks.  This two-step process for taking regulatory action would
be similar to the Canadian process. 

EPA's burden of proof could be reduced by changing the requirement in
TSCA that EPA have "substantial evidence" to a requirement similar to
the "arbitrary and capricious" standard in the Administrative
Procedure Act.  Alternatively, requiring that EPA demonstrate that a
chemical may present an unreasonable risk could require less
documentation than requiring the agency to demonstrate that the
chemical presents or will present an unreasonable risk. 


      BROADENING TSCA'S IMPACT
-------------------------------------------------------- Chapter 2:5.3

Given the thousands of chemicals in use and the many ways that
exposures and releases to the environment can occur, TSCA's
chemical-by-chemical and risk-based approach means that the act is
unlikely to address more than the most serious chemical risks, even
if EPA's regulatory burden is reduced.  The process of collecting
information on chemical effects and exposures to support TSCA's
regulations is a resource-intensive and time-consuming process. 

A different approach would be to set goals for reducing the use of
toxic chemicals overall.  Under this approach, legislation could
establish national goals for reductions in the use of toxic chemicals
and provide EPA with various tools, such as pollution taxes and other
economic incentives, to achieve these goals.  In a February 1993
report, we concluded that because of their inherently greater
flexibility, market-based incentives can be both a less costly and
more effective means of controlling pollution.\9

More chemicals could also be addressed under TSCA, if the act's goal
or purpose were to achieve reductions in exposures to and
environmental releases of toxic chemicals.  EPA then would not have
to document the risks associated with each use of each chemical and
show that they are unreasonable or significant; rather, it could show
that a chemical is toxic to humans and/or the environment, exposures
or releases to the environment occur or are likely to occur in
substantial amounts, and opportunities exist to reasonably reduce
exposures and releases.  To implement this goal, changes could be
made in the act to expand the types of circumstances under which EPA
could take action under TSCA to specifically include situations in
which (1) it identifies pollution prevention opportunities, such as
when safer chemical substitutes can be shown to exist at a reasonable
cost, or (2) the use of a toxic chemical cannot be shown to pose a
current problem, but its continued use could be a long-term problem
because it persists in the environment or accumulates in plant or
animal tissue. 

To better support EPA's pollution prevention initiatives, TSCA could
also be revised to expand the range of regulatory control options
available to EPA to reduce chemical risks.  These additional options
could include the authority to require the use of safer chemical
substitutes or manufacturing processes that result in less exposure
or fewer environmental releases. 


--------------------
\9 Environmental Protection:  Implications of Using Pollution Taxes
to Supplement Regulation (GAO/RCED-93-13, Feb.  17, 1993). 


   CONCLUSIONS
---------------------------------------------------------- Chapter 2:6

Although TSCA contains some unique regulatory authorities, EPA has
seldom used them because the act's legal standards are so high.  In
addition, EPA officials responsible for implementing TSCA do not
believe that the act gives them a clear mandate or directive to
control more than a few chemicals.  Consequently, TSCA has not had a
major role in EPA's efforts to protect human health and the
environment from the potential adverse effects of toxic chemicals. 

EPA instead plans to continue to rely primarily on standards under
other laws aimed at limiting exposures to or releases of toxic
chemicals, on voluntary risk reduction, and on pollution prevention
initiatives.  Although we recognize that EPA has made progress in
these areas, we believe that controls or restrictions on production,
distribution, and use should be viable alternatives for EPA when it
considers how to effectively deal with the adverse effects of toxic
chemicals. 

Despite their achievements, the various environmental, health, and
safety programs have been slow to deal with toxic chemicals, and
EPA's emphasis on voluntary industry actions is relatively new. 
Voluntary efforts may provide faster results than rule-making under
TSCA, and industry, when presented with concerns, may be willing to
take action even before EPA has the evidence or documentation
required for rule-making under TSCA.  However, the chemical industry
is large and diverse, and it is uncertain that all firms will always
act responsibly or agree with the concerns raised by EPA and its
efforts to obtain voluntary action, especially when it may involve
more expensive controls or reduced production of an important
chemical product.  In addition, strong regulatory authorities can
ensure a "fair playing field" when all companies are not willing to
cooperate in voluntary action and by their presence encourage
industry to develop solutions on their own.  For these reasons,
effective and viable regulatory authorities are essential whether
TSCA is to be a supplemental or umbrella act. 

Even if TSCA's regulatory authorities are made easier to use, the act
is unlikely to address more than the most serious chemical risks. 
TSCA's approach, as well as that of the other environmental laws, is
to address concerns on a chemical-by-chemical basis.  Given the large
number of chemicals in use, this approach is resource-intensive and
time-consuming, and major reductions in toxic releases to the
environment are difficult to achieve. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
---------------------------------------------------------- Chapter 2:7

In some cases, banning, restricting, or placing other controls on
production, distribution in commerce, use, or disposal may be the
most cost-effective way to deal with toxic chemicals.  TSCA contains
authorities to take these actions, but because of high legal
standards and uncertainty about the act's role and relationship to
other laws, these authorities are too difficult to use. 
Consequently, these types of actions are not viable alternatives when
EPA is considering how to address chemical risks. 

In its deliberations on reauthorizing TSCA, the Congress may wish to
consider changes to allow the act to be used in preference to other
environmental laws, when appropriate, and reduce EPA's burden of
proof in using the act's regulatory authority, as discussed earlier
in this chapter.  In addition, to supplement TSCA's
chemical-by-chemical and risk-based approach, the Congress may wish
to consider establishing overall goals for reductions in the use of
toxic chemicals and provide EPA with tools, such as market-based
incentives, to achieve these goals. 


EPA'S NEW CHEMICAL REVIEW PROCESS
DOES NOT FULLY ASSESS HEALTH AND
ENVIRONMENTAL RISKS
============================================================ Chapter 3

EPA's new chemical review process has enabled the agency to review
over 20,000 substances in a timely manner.  However, the reviews do
not ensure that the potential human health and environmental risks of
new chemicals are fully identified because EPA has limited data on
their toxic effects and exposures. 

TSCA does not require industry to test new chemicals for their
toxicity, and industry generally does not voluntarily perform this
testing.  Thus, EPA must predict toxicity by comparing new chemicals
to other chemicals with similar molecular structures and for which
toxic effects are known.  However, its prediction method has been
questioned in the scientific community, and a recent EPA study has
confirmed that the method does not always produce accurate results. 

EPA must also predict exposure to new chemicals on the basis of the
limited information contained in premanufacture notifications. 
Notifications are not binding, and once EPA completes its reviews and
production begins, manufacturers are not required under TSCA to limit
chemical production levels or uses to their estimates in the
notifications.  Increased production or use of the chemicals in
different ways can significantly increase exposure and risks. 
Although TSCA authorizes EPA to require the industry to submit
additional notifications for significant new uses or production
increases, the agency has made limited use of these authorities
because applying them on an individual chemical basis is a
resource-intensive process. 


   EPA'S NEW CHEMICAL REVIEW
   PROCESS
---------------------------------------------------------- Chapter 3:1

From the start of the new chemical review program in 1979 through
September 1993, EPA reviewed a total of about 24,000 new chemical
submissions, an annual average of about 1,700 notifications over this
period.\1 In more recent years, the number of notices has
increased--EPA estimates that it will review 2,500 notifications in
1994.  In nearly all cases, the agency has performed its review
within the 90 days TSCA requires.\2

Despite the increase in the number of notifications, EPA's budget for
new chemical review has not substantially changed.  For example, the
new chemical review budget was $13.7 million in fiscal year 1984 to
review fewer than 1,300 notifications and $15 million in fiscal year
1993 to review an estimated 2,500 notifications.  In more recent
years, the agency's resources for new chemical review have declined. 
For example, much of the detailed work involved in developing
toxicity and exposure assessments is performed under contracts, for
which funding decreased from $8.6 million in fiscal year 1988 to
about $5 million in fiscal year 1993. 

According to EPA officials, the agency has dealt with an increasing
workload by making the program more efficient and focusing its review
process on the substances of greatest concern.  These substances are
those about which little is known other than that they are
structurally related to known harmful chemicals.  According to EPA,
about 5 percent of the premanufacture notices received annually go
through the agency's more detailed or full review process.  In the
process's earlier stages, those chemicals are screened out that EPA
(1) anticipates will have a limited amount of exposure or (2) using
preliminary structure activity relationships analyses, estimates will
have few, if any, adverse human health or environmental effects.  In
addition, substances in certain chemical categories are identified
for possible regulatory action without undergoing a full review. 

During a full review, EPA evaluates the chemical's risks by
conducting a chemistry analysis, searching the scientific literature
and agency files for and analyzing toxicity data on structurally
similar chemicals, calculating potential releases of and exposures to
the chemical, and identifying potential new uses of the chemical.  On
the basis of this review, EPA makes a decision to either take no
action, require controls on the use, manufacture, or disposal of the
chemical, or ban the chemical pending the receipt and evaluation of
test studies performed by the chemical's manufacturer or processor. 


--------------------
\1 Included in this total is EPA's review and approval of about 5,000
applications for exemption from premanufacture notification reporting
requirements.  Under the authority contained in section 5 of TSCA,
EPA has limited the reporting of relatively low risk substances, such
as chemicals that are to be produced in volumes less than 1,000
kilograms per year or manufactured in small quantities solely for
research and development. 

\2 EPA, for good cause, can extend the premanufacture notice review
period for an additional 90 days. 


   EPA USUALLY DOES NOT HAVE TEST
   DATA WHEN REVIEWING NEW
   CHEMICALS
---------------------------------------------------------- Chapter 3:2

To assess chemicals' risks, EPA needs to know the chemicals' toxic or
adverse human health and environmental effects.  Potential health
effects that may be considered include skin and eye irritation, blood
effects, cancer, birth defects, and harm to the central nervous
system.  Consideration of potential environmental effects include
those on aquatic life, such as fish and algae.  Another major
consideration is chemical fate, that is, the characteristics of a
chemical, such as its ability to be absorbed in water, and its
ultimate disposition in the environment. 

EPA officials told us, however, that the agency frequently lacks
comprehensive toxicity data when reviewing new chemicals.  Although
TSCA does not require manufacturers to test new chemicals, they are
to submit with premanufacture notices the results of the testing that
has been done.  According to EPA, less than half of all
premanufacture notices contain any toxicity data. 


   EPA'S PERFORMANCE IN PREDICTING
   TOXICITY FOR NEW CHEMICALS IS
   MIXED
---------------------------------------------------------- Chapter 3:3

Because comprehensive test data are generally not available when
reviewing new chemicals, EPA uses structure activity relationships
(SAR) analysis to identify potential chemical hazards.  In SAR
analysis, EPA scientists' predictions of the characteristics and
potential effects of a new chemical are based on the known
characteristics and effects of chemicals with structures whose key
parts are similar to those of the new chemical.  According to EPA
officials, the SAR technique is used for each new chemical under
review. 


      THE U.S.-EUROPEAN UNION'S
      PROJECT TO COMPARE SAR
      PREDICTIONS WITH TEST
      RESULTS
-------------------------------------------------------- Chapter 3:3.1

In 1993, EPA completed a study with the European Union (formerly the
European Community), whose member countries require certain toxicity
testing before new chemicals are reviewed and go on the market.  For
144 chemicals, the European Commission (the executive arm of the
European Union) sent EPA information similar to that which is
generally contained in premanufacture notices.  EPA then used this
information in SAR analysis to predict the toxic properties of the
chemicals and carry out preliminary hazard assessments.  The physical
properties and hazards that EPA identified for the individual
chemicals were then compared with those identified by the European
Union using test data.  EPA wanted to determine how well its SAR
approach worked in identifying chemicals' toxic effects, and the
European Union wanted to determine whether SAR could be used as a
substitute for testing to avoid costs and the destruction of
laboratory animals. 

The comparisons of EPA's predictions of individual chemical
properties or effects with the actual test results showed that SAR's
accuracy varied, depending on the effect or property being compared. 
For example, the SAR predictions were correct for only 50 percent and
63 percent of the chemicals when predicting their boiling point and
vapor pressure, respectively.  On the other hand, the accuracy rate
for biodegradation\3 was 93 percent. 


--------------------
\3 Biodegradation refers to the capability of chemicals, under normal
conditions, to be broken down in the environment by the action of
living things, especially microorganisms. 


      THE U.S.  PERSPECTIVE ON THE
      STUDY RESULTS
-------------------------------------------------------- Chapter 3:3.2

The EPA officials' overall conclusion from the study was that the SAR
approach to screening new chemicals is useful and effective in
identifying chemicals that may be toxic and in need of further
scrutiny to determine whether to regulate them.  However, according
to the officials, the SAR approach appears to have limitations in
predicting physical chemical properties under some circumstances and
in predicting the exact type and level of toxicity of the chemical,
especially in connection with general systemic (health) effects. 

EPA officials further concluded that although the SAR approach has
largely been successful in identifying chemicals of concern, the
process could be improved by selectively incorporating specific
testing schemes into the process.  The officials said that results of
this testing would provide insight into chemical toxicities and
improve SAR's predictive capabilities.  According to the officials,
the testing would provide a richer data base upon which to base
predictions. 


      THE EUROPEAN UNION'S
      PERSPECTIVE ON THE STUDY
      RESULTS
-------------------------------------------------------- Chapter 3:3.3

The European Union officials concluded that the study identified a
number of possibilities for making greater use of SAR as part of the
testing package for new chemicals.  According to the officials, the
SAR approach performed extremely well in predicting biodegradation
and acute toxicity to fish and daphnia.  On the other hand, the
officials said that given the relatively low cost of tests for
physical chemical properties, the results of this study did not
constitute a persuasive argument for introducing SAR as an
alternative for testing.  The officials stated that of the physical
chemical properties, SAR performed best in predicting water
solubility, but even with this property SAR could not be used with
confidence for all chemical groups. 

For health effects, the European Union officials did not believe that
the SAR approach was sufficiently developed for estimating eye and
skin irritation or sensitization.  On the other hand, the officials
believed that SAR was relatively successful in assessing acute lethal
toxicity.  The officials also noted that although SAR tended to
underestimate the severity of the effects, it provides an excellent
additional tool to decide about further testing for subchronic,
repeated dose toxicity.  In addition, the officials stated that the
study results suggest that SAR could usefully be incorporated into a
battery of approaches for evaluating the mutagenic potential of a new
chemical. 


      EXAMPLE OF POTENTIAL IMPACT
      OF INACCURATE SAR
      PREDICTIONS
-------------------------------------------------------- Chapter 3:3.4

In 1990, EPA received a premanufacture notification for a new
chemical generically described as dialkyldialkoxysilane.  On the
basis of SAR analysis, EPA identified potential adverse health
effects (skin and eye irritation and liver and lung toxicity) and
environmental effects (toxicity to aquatic organisms at low
concentrations) for the chemical.  EPA, however, did not make an
unreasonable risk finding and did not impose regulatory controls
because exposures and releases of the chemical were expected to be
low. 

In 1991, EPA received another premanufacture notification for the
same chemical from a different manufacturer.  Because the submitter
of the earlier premanufacture notification had not yet commenced
production and the chemical was not yet on the TSCA inventory, EPA
decided to evaluate the second notification as a new chemical.  The
second notification contained the results of several toxicity tests
for the chemical; EPA found, on the basis of these tests, that
potential inhalation exposure considerably increased because the
actual vapor pressure of the chemical was 100 times greater than had
been previously predicted using SAR analysis.  On the basis of the
new data, EPA found that the chemical posed an unreasonable risk and
proposed regulations to control new uses of the chemical. 

An official in EPA's new chemical review program told us that in the
vast majority of cases, a second notification would not be sent
before production of the chemical had begun and that the second
notification would not result in a new chemical review of the
chemical.  The official also pointed out that most notifications do
not contain test data and that had such data not been available for
dialkyldialkoxysilane, EPA would not have identified and acted upon
the unreasonable risk presented by exposure to the chemical. 


      SAR DEPENDS ON THE
      AVAILABILITY OF TEST DATA ON
      SIMILAR CHEMICALS
-------------------------------------------------------- Chapter 3:3.5

Without adequate test data on chemicals with similar molecular
structures, SAR analysis does not have a reliable basis for
predicting the toxicity of new chemicals.  For example, an EPA
official who participated in the project with the European Commission
told us that the SAR approach had some difficulty in predicting
systemic toxicity because of a lack of test data on similar
chemicals.  Situations also arise in which EPA must review new
chemicals with molecular structures not similar to any chemical for
which it has health and environmental effects data.  An official in
EPA's new chemical review program told us that he was not comfortable
using SAR analysis to predict the hazards posed by such chemicals but
that no other option exists since EPA cannot require chemical
manufacturers to conduct tests of new chemicals before the agency's
review. 


   UNCERTAINTY IN EPA'S EXPOSURE
   ESTIMATES MAY LIMIT REGULATION
---------------------------------------------------------- Chapter 3:4

EPA's exposure assessments for new chemicals are based on information
contained in premanufacture notices.  This information generally
includes data on how the manufacturer anticipates that the chemical
will be used and estimates of production volumes, chemical releases,
and number of employees in contact with the chemical at the company's
plants.  These estimates are not binding and do not have to be
amended once manufacturing begins, unless EPA promulgates a rule
specifically to require the reporting of new uses or significant
increases in production or releases. 

Chemical manufacturers can also modify the information in
premanufacture notices during EPA's 90-day new-chemical review
process.  EPA officials told us that manufacturers have, when
informed that EPA had targeted one of their chemicals for regulation,
amended exposure information.  The officials believe that the
manufacturers made some of these changes to avoid regulatory control
of their chemicals. 

At our request, EPA officials provided us with examples of exposure
data changes, including reductions in production volumes and
environmental releases and revisions in predicted uses of the
chemicals resulting in lower levels of expected exposure.  For
example, when informed of EPA's determination that uncontrolled
release of a new chemical to surface waters could present an
unreasonable risk to aquatic organisms, the company informed the
agency of a revised manufacturing process that would result in a
tenfold reduction in predicted water releases.  In another case, EPA
determined that production of the new chemical in substantial
quantities could potentially cause significant human exposure to the
substance.  The manufacturer then submitted new information
indicating that the chemical would have a narrower use than
originally identified and would be processed using a system that
would prevent releases of the chemical outside of the production
process.  As a result of the manufacturers' reported changes in these
cases, EPA could not support its initial concerns about the chemicals
and could take no regulatory action. 


   EPA MAKES LIMITED USE OF TSCA'S
   AUTHORITIES TO OBTAIN
   ADDITIONAL DATA
---------------------------------------------------------- Chapter 3:5

Although TSCA authorizes EPA to control chemicals pending development
of needed test data and to require chemical manufacturers to notify
EPA of significant new uses that develop after manufacture begins,\4
the agency uses these authorities for only a small percentage of new
chemicals.  Using these authorities is difficult because EPA must
take these actions on a chemical-by-chemical basis, which can require
considerable resources. 


--------------------
\4 The term "new uses" refers not only to additional applications of
a chemical, but also to significant increases in the projected volume
of production or a significant change in the type, magnitude, or
duration of human or environmental exposure. 


      EPA REQUIRES ADDITIONAL
      TESTING FOR A SMALL
      PERCENTAGE OF CHEMICALS
-------------------------------------------------------- Chapter 3:5.1

If EPA determines that a new chemical may present an unreasonable
risk or may result in significant exposure but that insufficient data
exist to adequately assess the chemical's health and environmental
effects, TSCA section 5(e) authorizes EPA to control exposure to the
substance until sufficient data are available.  If EPA believes that
the manufacturer can control exposure to the chemical through routine
workplace practices or controls, it will enter into an agreement or
consent order with the manufacturer to allow production under these
controls until test data are developed.  The manufacturer can perform
the testing and seek to have the controls removed, or it can choose
to forgo the testing and continue the controls.  In some consent
orders, manufacturers have agreed to perform certain testing when the
production volume reaches a designated level.  If EPA anticipates
that routine workplace practices or controls will not adequately
reduce exposure, the manufacturer cannot begin production until it
performs the additional testing and EPA evaluates the results. 

For about 6 percent of the new chemicals reviewed, the manufacturers
or processors either agreed to manufacture the chemicals with
controls on exposure pending the development of test data or provided
additional information during EPA's review process.\5 In about
one-third of these cases, EPA would not allow manufacture to begin
until the necessary testing had been performed.  According to EPA,
this 2 percent of chemicals reviewed accounted for 80 percent of the
test data obtained through its new-chemical review process. 

In 1988, EPA implemented a policy aimed at increasing the amount of
testing performed on high-volume new chemicals by making use of its
section 5(e) authority.  This policy calls for basic toxicity testing
for new chemicals with planned annual production volumes of 100,000
kilograms or more, if EPA believes that substantial or significant
human exposures or substantial environmental releases will occur. 
From October 1989 through March 1993, EPA issued 54 orders under the
policy, about 1 percent of about 7,200 premanufacture notices it
received during this period. 


--------------------
\5 EPA made section 5(e) findings for these chemicals and for an
additional 4 percent of premanufacture notices that the submitters
later withdrew. 


      EPA HAS ISSUED SIGNIFICANT
      NEW USE RULES FOR A SMALL
      PORTION OF CHEMICALS
-------------------------------------------------------- Chapter 3:5.2

TSCA section 5(a)(2) rules requiring that manufacturers or processors
submit premanufacture notices for significant new uses of chemicals
provide EPA with the opportunity to obtain additional exposure data
and reassess the risks posed by new chemicals.  However, according to
EPA officials, to issue one of these rules, the agency generally must
show that (1) a reason exists to be concerned about the chemical's
toxicity and (2) the potential exists for uses, other than those
contained in the premanufacture notice, that could lead to higher
exposure. 

In July 1989, EPA issued a rule to establish an expedited process for
issuing significant new use rules.  The process enables EPA to more
quickly issue these rules for chemicals for which the agency has
entered into section 5(e) consent orders and for chemicals that may
present human health and environmental hazards, if exposures or
releases are substantially different from those in the premanufacture
notice.  However, the burden is still on EPA to identify future uses
and to support the need for the additional reporting on a
chemical-by-chemical basis.  From October 1989 through March 1993,
EPA issued significant new use rules for 382 of the approximately
7,200 new chemicals reviewed.  EPA issued a total of 12 rules before
fiscal year 1990. 


   OTHER COUNTRIES PLACE MORE OF
   THE BURDEN ON INDUSTRY
---------------------------------------------------------- Chapter 3:6

Two of the countries that we visited, Canada and Germany, also review
new chemicals before they enter commerce--at the premanufacture stage
in Canada and at the premarketing stage in Germany.  However, unlike
the U.S.  practice, these countries require manufacturers to perform
certain tests and submit the results to the government at the
beginning of the review process.  Sweden's Act on Chemical Products
places the main responsibility on manufacturers to assess the risks
of both new and existing chemicals and provide adequate information
on their effects to chemical users.  Sweden plans to implement
requirements similar to those of Germany in 1995, when it joins the
European Union. 

Germany and other European Union countries maintain a separate
inventory for new chemicals, and chemicals on this inventory are
subject to additional testing and review when the annual amount
marketed reaches certain levels.  In addition, a new manufacturer has
to notify the government before marketing the chemical, even though
the chemical may have been marketed by the original manufacturer for
several years. 

In Canada, requirements for additional testing are also triggered as
production or import volumes increase.  A new chemical is generally
added to the existing chemicals inventory only after a certain level
of production or import has been reached and specified testing for
that level has been performed without conditions being placed on the
chemical's manufacture or import.  Until the chemical is placed on
the inventory, anyone planning to produce the chemical must notify
the government.  In Sweden, a new manufacturer of a chemical product
generally has to notify the government that it is manufacturing the
product. 


   OPTIONS FOR REVISING TSCA
---------------------------------------------------------- Chapter 3:7

Various options exist for revising TSCA to enhance EPA's ability to
obtain the information it needs to assess the potential risks of new
chemicals and to take control actions to protect against those risks. 
We discuss below some options to address the problems that we noted
in TSCA. 


      INCREASING THE AVAILABILITY
      OF TOXICITY DATA FOR NEW
      CHEMICALS
-------------------------------------------------------- Chapter 3:7.1

To help ensure that EPA has sufficient information on toxic effects
when reviewing new chemicals, manufacturers could be required to test
their chemicals and submit the results to EPA with their
premanufacture notifications.  The major drawback to testing is its
cost to the chemical industry, which representatives of the Chemical
Manufacturers Association believe could reduce chemical research and
innovation.  For example, Canadian officials told us that testing a
chemical in that country can cost up to $130,000, and a
representative of the Swedish chemical industry told us that the
government's planned testing requirement will cost about $188,000 per
chemical. 

To reduce these costs or to delay them until the chemicals are
produced in larger quantities, Canada and Germany require testing
based on production volume.  The testing requirements for low volume
chemicals are less extensive and complex than those for high volume
chemicals. 

As an option, testing could be targeted to those areas in which EPA's
SAR analysis does not accurately predict toxicity.  TSCA could be
amended to authorize EPA to establish a set of tests for new chemical
reviews.  Some tests could be required with the premanufacture
notifications; more complex and costlier tests could be required on
the basis of a chemical's production or environmental release levels. 
According to EPA officials, the tests needed would not be as
comprehensive as those in Germany and Canada because SAR analysis
works well for characterizing some toxic effects.  They also said
that testing could result in more accurate SAR analysis because the
testing would provide more toxicity information on which to base
future SAR analysis. 

On the basis of the results of the EPA/European Union study, SAR
analysis is generally not reliable for predicting physical chemical
properties.  Testing for these properties is not very costly.  For
example, according to EPA, testing to determine a chemical's boiling
point can range from about $420 to $560.  The cost to determine vapor
pressure can range from about $1,790 to $2,520. 

Revising TSCA to require EPA to review new chemicals before they are
marketed rather than before they are manufactured would reduce the
costs of initial testing.  According to EPA, about half of the
premanufacture notices that the agency reviews are for chemicals that
never enter the marketplace.  Thus, a requirement that certain
initial test results be submitted with notifications would apply to
far fewer new chemicals. 

Another option for increasing the availability of toxicity data for
new chemicals would be to enable EPA to require certain testing
without going through rule-making.  For example, EPA could be
authorized to require such testing if it finds that it cannot be
confident of the results of its SAR analysis because it does not have
sufficient toxicity data on chemicals with structures similar to the
new chemicals submitted by manufacturers. 


      IMPROVING EPA'S ASSESSMENTS
      OF EXPOSURES TO NEW
      CHEMICALS
-------------------------------------------------------- Chapter 3:7.2

To provide additional assurances that new chemicals do not result in
harmful use, TSCA could be revised to require manufacturers to submit
additional premanufacture notices or comparable data when production
volumes or uses change significantly from the estimates in the
previous premanufacture notices.  This requirement could be
implemented in various ways.  One way would be to make statements in
premanufacture notices concerning projected production levels and
uses binding on manufacturers.  Under this approach, the manufacturer
or processor submitting the notice would be required to submit
additional notices for significant increases in production volume or
new uses.  In addition, other firms that decide to manufacture or
process the chemical would also have to submit premanufacture notices
of their plans.  Currently, once EPA's review of a premanufacture
notice has been completed and the agency is notified that manufacture
has begun, the chemical is added to the TSCA inventory of chemicals. 
At that point, any manufacturer can produce the chemical in any
amount and for any use without notifying EPA, unless EPA has
promulgated a significant new use rule. 

Making premanufacture notices binding would provide EPA with
up-to-date exposure information without the burden of issuing a large
number of significant new use rules.  This approach would also
encourage chemical manufacturers and processors to pay more attention
to their production volume and use estimates and would provide an
opportunity for EPA to reassess the risks of new chemicals as their
production increases and uses expand. 

Under this approach, manufacturers would be required to submit new
notices to EPA, if production volumes, releases, and exposures
increased significantly or if the manufacturing process or the use of
the chemical changed significantly from those in the original
premanufacture notice.  The burden of identifying when new notices
are needed would be shifted to the industry, and EPA would not have
to expend the time and resources to issue significant new use rules. 
On the other hand, industry would have some additional costs to
prepare new notices.  The number of new notices would largely depend
on how EPA defines a "significant" change and how many new chemicals
become commercial successes.  Of the over 20,000 premanufacture
notices reviewed, about 10,000 chemicals entered commerce.  According
to EPA officials, only about 2,000 of these chemicals attained any
significance in the marketplace. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 3:8

Although TSCA recognizes that the best time to assess the risks posed
by chemicals is before the production and use of the substances
begin, the act's authorities for obtaining the information on which
EPA can base these assessments are ineffective.  TSCA does not
require routine testing of new chemicals, and the chemical industry
provides little test data with its premanufacture notifications.  In
addition, the EPA/European Union study showed that EPA's use of SAR
analysis to determine potential toxicity posed by new chemicals often
does not accurately determine the chemicals' properties and the full
extent of their adverse effects. 

Concerns also exist about whether the data submitted in
premanufacture notifications accurately reflect actual exposures once
production and use of new chemicals begin.  The production and
exposure information submitted in premanufacture notices is not
binding under TSCA and may not reflect actual conditions once the
chemical commences production and its marketing expands.  Given these
limitations in TSCA's new-chemical review program, we do not believe
that EPA can be assured that it identifies the potential risks posed
by the chemicals. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
---------------------------------------------------------- Chapter 3:9

In its deliberations on reauthorizing TSCA, the Congress may wish to
consider revising the act to place more of the burden on industry to
demonstrate that new chemicals are safe.  Some of the burden could be
shifted by requiring industry to test new chemicals and to notify EPA
of significant increases in production, releases, and exposures or of
significant changes in manufacturing processes and uses after new
chemicals enter commerce. 


WITHOUT CHANGES, EPA WILL NOT
REVIEW THE RISKS OF MANY EXISTING
CHEMICALS
============================================================ Chapter 4

EPA has made little progress in reviewing the risks of existing
chemicals.  EPA's information on chemical effects and exposures is
often scarce, incomplete, or outdated, and TSCA's data-gathering
authorities are difficult to use and not very effective in supporting
EPA's review process.  As a result, the reviews are generally lengthy
and resource-intensive.  Because of competing priorities, EPA has
made limited resources available for review of existing chemicals. 


   EPA HAS BEEN SLOW TO ASSESS
   EXISTING CHEMICALS' RISKS
---------------------------------------------------------- Chapter 4:1

Under its existing chemicals program, EPA has reviewed the risks of
about 1,200 substances.  This amount represents only about 2 percent
of the approximately 62,000 chemicals in the TSCA inventory that were
already in commerce when EPA began its new- chemical review program
in 1979 and were not reviewed as new substances.  Although EPA
reviewed the other 10,000 chemicals in the inventory as new
chemicals, it generally has not reviewed them after their production
began and they were placed on the inventory.  In total, EPA has
reviewed about 16 percent of the inventory, either as new chemicals
or existing ones. 

However, even this small percentage overstates EPA's progress in
fully identifying chemical risks.  First, EPA's reviews often do not
include all uses of individual existing chemicals.  For example, EPA
is clustering some chemicals by use, such as chemicals used in paint
stripping.  This approach enables EPA to review more chemicals and
compare their relative risks and trade-offs.  On the other hand, EPA
does not examine other uses for the chemicals that may pose risks. 

EPA's progress may also be overstated because many of the completed
reviews may be outdated.  Production volumes may have increased and
new uses may have developed since EPA conducted the reviews.  In
addition, the toxicity information that EPA had available at the time
of the reviews may not have included all the potential toxic effects
of the chemicals. 

According to EPA, the existing chemical review process typically
takes from 12 to 16 weeks and requires about 100 staff hours.  During
this process, nearly all the burden is on EPA to compile and analyze
the available data on a chemical.  The agency must search its files
and public data bases for information on the chemical's effects,
physical properties, production volumes, manufacturing processes,
uses, releases to the environment, and other data, such as the number
of workers exposed to the chemical.  Because limited information is
generally available, EPA uses various computer models to estimate or
project certain data, such as the amounts and types of environmental
releases.  A chemical can have various potential health and
environmental effects that EPA needs to consider in evaluating its
risks.  In addition, chemicals often have several uses--as many as
100, or even more, if they have been on the market for a long time. 
The agency has to also consider a chemical's individual uses because
use largely determines the amount of exposure. 

Of about $53.6 million for existing chemicals in fiscal year 1994,
about $11.2 million was allocated for chemical review and associated
risk management activities.  The remainder went primarily to chemical
testing and carrying out congressionally mandated activities related
to the control of asbestos, lead, and PCBs. 

With its current level of resources and review process, EPA is
reviewing about 100 existing chemicals per year.  At this rate, it
would take over a century and a half just to cover the approximately
16,000 chemicals in the inventory that EPA believes are of higher
priority because of their production levels and chemical structures. 


   TSCA PROVIDES LIMITED
   INFORMATION TO SUPPORT EPA'S
   CHEMICAL REVIEWS
---------------------------------------------------------- Chapter 4:2

The completeness and quality of EPA's reviews are dependent on the
information that the agency has on the chemicals' potential toxic
effects and the amounts and types of exposures that occur.  TSCA
authorizes EPA to require manufacturers and processors to conduct
toxicity testing and to provide exposure-related information on their
chemicals.  However, EPA rarely uses these authorities because they
are costly and time-consuming to use.  As a result, EPA generally has
to rely on (1) limited and sometimes outdated information in its
files and in publicly available data bases and (2) computer modeling
to project or estimate key information, such as environmental
releases. 


      LITTLE CHEMICAL TESTING IS
      DONE UNDER TSCA
-------------------------------------------------------- Chapter 4:2.1

Industry voluntarily conducts some chemical tests.  In addition,
other federal agencies, such as the Department of Health and Human
Services' National Toxicology Program, have chemical testing
programs.  Universities and other organizations may also test
chemicals for their effects. 

However, few test data are available for many chemicals.  A 1984
study by the National Research Council is the most comprehensive
analysis of the availability of test data.  The Council sampled the
entire TSCA inventory and then determined what test data on health
effects (the analysis did not include environmental effects) were
available for the sample.  According to the analysis, no health
toxicity information was available for almost 80 percent of the
chemicals.  A 1989 analysis by EPA staff of about 1,300 substances on
the Organization for Economic Cooperation and Development's list of
international high volume chemicals showed that many of them did not
have the major types of health and environmental toxicity
information.  Only acute (effects of a single exposure) or general
toxicity data were found for more than half of the chemicals.  Less
than one-fourth of them had any reproductive or developmental
toxicity data.  Aquatic toxicity data were found for 41 percent. 
(According to EPA officials, the analysis did not determine the
data's validity or reliability, only whether data were available.) A
more recent analysis performed by the Organization for Economic
Cooperation and Development identified about 500 of over 1,300 high
volume international chemicals as having potential health or
environmental concerns but having few test data publicly available to
assist in evaluating their risks.  According to the Director of EPA's
Office of Pollution Prevention and Toxics, little is known about the
effects of many chemicals used in commerce. 

As of May 1994, EPA had issued only 30 test rules covering 121
chemicals.  In addition, EPA has entered into negotiated test
agreements or consent agreements to test 59 more chemicals, and
industry has voluntarily agreed to test 230 chemicals.  According to
EPA officials, the agency has not used its authority to require more
testing, largely because it must undergo a lengthy and costly
rule-making process.  EPA must demonstrate that it needs the test
results, issue a proposed rule for public comment, consider the
comments it receives, and issue the final rule.  According to EPA
officials, this process can take as long as 24 to 30 months and cost
between about $68,500 and $234,000. 


      EPA HAS LIMITED EXPOSURE
      DATA
-------------------------------------------------------- Chapter 4:2.2

To fully assess human exposure to a chemical, EPA needs to know how
many workers, consumers, and others are exposed; whether the exposure
occurs through inhalation or other means, such as skin absorption;
and the amount and duration of the exposure.  For environmental
exposure, EPA needs to know whether the chemical is being released to
the air, water, or land; how much is being released; and the extent
of the area affected.  Another important factor in environmental
exposure is chemical fate, that is, how it acts and its ultimate
disposition in the environment. 

Actual measurements of exposure in the environment, workplace, and
home, on an extensive basis, for the thousands of chemicals in use
are not practicable because of the monitoring equipment and staff
resources required.  Consequently, EPA estimates the types and
amounts of exposure on the basis of a chemical's physical properties,
how it is used, the industrial processes for producing and processing
it, production volumes, and the type and amount of releases to the
environment. 

However, the basic data that EPA needs to develop its estimates are
often not available, and EPA must rely on models or other analytical
techniques.  According to EPA officials, sufficient data on exposures
rarely exist to permit full analysis of a chemical, and the agency
has little assurance that its exposure assessments are accurate and
complete. 


      CHEMICAL USE AND RELEASE
      DATA
-------------------------------------------------------- Chapter 4:2.3

The amount of exposure to a chemical can vary substantially depending
on its use.  For example, only workers are potentially exposed to
chemicals that are used solely for industrial purposes, whereas both
workers and consumers can be exposed to chemicals used in household
products.  However, TSCA does not require routine reporting of
chemical use information, and EPA receives very little of this
information under its current industry reporting rules. 

EPA is planning to revise its TSCA Inventory Update Rule to add
reporting requirements for information on chemical use.\1

According to EPA officials, the current rule collects some key
information, such as production volumes, needed to identify chemicals
of concern, but additional information on use is essential to
determine possible exposure routes and scenarios and identify
potentially safer substitute chemicals.  EPA is also considering
broadening the scope of industries that have to report and increasing
the reporting frequency from every 4 years to every 2 years.  The
chemical industry has expressed concern about the additional
reporting burden and cost of these changes; its specific concerns are
how much information it must report on each chemical and whether EPA
will require it to report on all chemicals. 

Even with additional information provided under an expanded inventory
update rule, EPA will lack key data on "downstream" chemical uses. 
Manufacturers produce certain chemicals and often sell to processors,
who prepare these chemicals or mixtures for distribution in commerce. 
The processors' customers may then buy the chemicals for use as
solvents, lubricants, or various other purposes. 

According to EPA officials, manufacturers generally know how the
processors use chemicals, and processors generally know their
customers' major uses.  However, a processor's customer may use a
chemical for a purpose that the processor is not aware of or even
sell some of the chemical to another firm for a different purpose. 
EPA officials said that some chemicals have many uses, some of which
may surprise the manufacturer or processor.  According to EPA
officials, these downstream uses can be important factors in
assessing exposure, but TSCA does not provide EPA with the authority
to collect information from these users. 

Another important factor in assessing exposure is the release of a
chemical to the environment during its manufacture, processing,
distribution, use, and disposal.  Such releases are a potential
source of exposure for plants, animals, and people living near
industrial and disposal sites.  In assessing the extent of this
exposure, EPA considers the quantity released; whether the releases
were to the air, water, or land; the chemical's migration through the
air, water, and soil; the time required for it to biodegrade or break
down in the environment; and whether it accumulates in the tissue of
animals or plants. 

However, according to the head of the EPA office responsible for
developing exposure estimates, few release data are available for
chemicals not included in the Toxics Release Inventory, which
contains estimates of annual releases to the air, water, and land for
about 300 chemicals.  Many other potentially harmful chemicals are
produced in large quantities.  For example, over 16,000 nonpolymer
existing chemicals are produced in amounts of 10,000 pounds or more
annually.  To obtain some indication of the types and amounts of
releases for chemicals not in the inventory, EPA develops estimates
on the basis of where and how the chemicals are manufactured and how
much is produced, but it acknowledges that the accuracy of these
estimates is uncertain. 


--------------------
\1 The rule requires chemical manufacturers and importers to provide
updated information for the TSCA inventory every 4 years.  This
information includes chemical identity, plant sites, and production
volumes.  Two updates have taken place, the first in 1986 and the
other in 1990.  The next scheduled update is 1994. 


      WORKER AND CONSUMER EXPOSURE
      DATA
-------------------------------------------------------- Chapter 4:2.4

The extent of worker exposure is especially important in assessing a
chemical's risks because workers are often in contact with the
substance.  The key information needed is the number of exposed
workers and the concentration and duration of the exposure. 

EPA's major source of data on the number of workers exposed is the
National Occupational Exposure Survey.  This survey, which the
National Institute for Occupational Safety and Health conducted in
the early 1980s, contains estimates of the number of workers exposed
nationally to over 10,000 chemicals.  It also contains data such as
the number of sites at which a chemical is manufactured or used. 
Although EPA officials recognize that the survey is old and probably
outdated, it is often the only available data on the number of
workers exposed to a particular chemical.  According to National
Institute for Occupational Safety and Health and EPA officials, funds
are not available to update worker exposure information. 

Because the survey does not provide the needed information and little
is available elsewhere, EPA generally estimates how exposure occurs
and its duration.  These estimates are made on the basis of
information that the agency can compile on the chemical's physical
properties, its uses, and the industrial operations or procedures
involved in manufacturing, processing, and using the chemical. 

EPA's files on 2-Mercaptobensothiazole, a chemical primarily used to
produce rubber, inhibit corrosion, and act as a fungicide, illustrate
some of the agency's difficulty in obtaining worker exposure data. 
EPA's late 1991 and early 1992 review of this chemical found only
outdated information on the number of workers exposed and no data on
how long exposure occurs.  An occupational hazard survey conducted by
the National Institute for Occupational Safety and Health from 1972
to 1974 resulted in an estimate that as many as 558,893 workers may
be exposed to the chemical.  On the other hand, the National
Occupational Exposure Survey estimated that only 2,398 workers were
exposed.  EPA officials did not know the reason for the difference in
the two estimates but speculated that (1) the hazard survey may have
included all of the workers in the industries using the chemical and
(2) the exposure survey may have included only those workers in
direct contact with the chemical.  The number of exposed workers was
important because test results on the chemical indicated possible
adverse effects on the fetuses of pregnant women who are exposed to
the chemical. 

Consumers may also be exposed to potentially toxic chemicals through
skin contact or inhalation of the fumes of products containing these
substances.  When reviewing a chemical, EPA often does not have
information on all of its uses, including as an ingredient in
household products.  Even if EPA is aware of consumers' use of a
chemical, it does not have good information on the number of people
using the products or the frequency and duration of their use. 
According to an EPA official, the agency needs much better
information on consumers' exposure to assess the risks of chemicals
but does not believe that extensive monitoring of consumers' exposure
to products is practicable. 

To obtain the information that it needs, EPA is considering
estimating human exposures to chemicals in consumer products on the
basis of measurements of the concentrations of product residue on
indoor and outdoor sources.  However, industry representatives have
expressed concerns about the cost and complexity of such an effort. 
For example, assessing consumers' exposure to a single chemical may
involve making measurements for its use in various consumer products. 


   OTHER COUNTRIES HAVE MAJOR
   INITIATIVES UNDER WAY
---------------------------------------------------------- Chapter 4:3

The other countries that we visited--Canada, Germany, and
Sweden--have recently established systematic processes for reviewing
existing chemicals.  Recognizing that it was not feasible to
immediately review all of the chemicals in commerce, the Canadian
Environmental Protection Act of 1988 required the development of
Priorities Substances Lists.  The government is to periodically
develop a list of high priority chemicals for assessment.  Each
chemical on the list is to be assessed and the results made available
to the public within 5 years.  The government's goal is to
comprehensively assess and report on 100 priority substances of
greatest concern by the year 2000. 

Germany is implementing a 1993 European Union regulation that
requires the member countries to participate in a systematic review
process for existing chemicals.  On the basis of information
submitted by chemical manufacturers and importers, the European Union
plans to periodically develop priority lists of chemicals for member
countries to review, with the initial focus on high volume chemicals. 
The Union plans for its members to assess up to 50 chemicals--and
possibly more--each year. 

In addition to a requirement that manufacturers and importers assess
the risks of their own chemicals, the Swedish government has
implemented a project to develop a systematic process for identifying
chemicals with properties and patterns of use that make them
potential candidates for general restrictions on their use.  During
the pilot project, the government started with a list of about 2,000
high volume or environmentally hazardous chemicals.  This list was
narrowed down to about 200 chemicals on the basis of their
environmental hazards and production volumes.  After eliminating
low-volume, regulated, and unregistered substances, the government
identified 41 chemicals for review. 

These countries place more of the burden on industry for the review
of chemical risks.  Sweden's Act on Chemical Products makes
manufacturers and importers primarily responsible for determining
chemical risks.  In Canada and Germany, the government is responsible
for assessing chemical risks.  However, it is easier for the
government to obtain chemical information from industry.  In Germany,
the European Union regulation being implemented requires industry to
compile and report the data needed to review the risks of existing
chemicals.  On the other hand, in Canada, the government is
responsible for compiling available chemical information.  However,
under the Canadian Environmental Protection Act, the government can
collect additional information on a chemical by informing
manufacturers and importers of the needed data through a notice in
the Canada Gazette, which is similar to the Federal Register in the
United States.  The government can also write letters to the
manufacturers or importers to demand that they submit the needed data
by a certain date. 


   OPTIONS TO STRENGTHEN TSCA'S
   PROVISIONS ON REVIEWING
   EXISTING CHEMICALS
---------------------------------------------------------- Chapter 4:4

TSCA's chemical review provisions can be strengthened by requiring
the systematic review of existing chemicals and expanding EPA's
information-gathering authorities.  A requirement for systematic
review would ensure that EPA's risk assessments of existing chemicals
receive sufficient agency priority and resources.  The expanded
information-gathering authorities would enable the agency to obtain
more of the data it needs to perform these reviews. 


      SYSTEMATIC REVIEW OF
      EXISTING CHEMICALS
-------------------------------------------------------- Chapter 4:4.1

In providing for EPA to review premanufacture notices within 90 days,
TSCA established a firm requirement for reviewing new chemicals, but
the act contains no similar requirement for existing chemicals. 
Accordingly, EPA's priority has been to comply with the statutory
requirement for new chemicals.  EPA's Office of Pollution Prevention
and Toxics has increased its emphasis on existing chemicals but has
not been able to obtain additional resources for this purpose because
of competing needs for limited resources. 

TSCA could be revised to establish a goal for the review of existing
chemicals.  Such a goal could focus EPA's and the chemical industry's
attention on completing these reviews and put existing chemicals on
more of an equal footing with new chemicals.  Because of limited
resources and the large and growing number of existing chemicals, a
goal to review all existing chemicals within the near future would be
impracticable.  Even the 16,000 chemicals that EPA considers of
potential concern because of their production volumes and chemical
structures could take many years to review.  EPA would need the
flexibility to identify those chemicals most in need of review and
give them priority.  TSCA could be revised to set out a systematic
review process for EPA to implement or to require EPA to establish
and carry out such a process. 


      TSCA'S INFORMATION-GATHERING
      AUTHORITIES
-------------------------------------------------------- Chapter 4:4.2

Approaches similar to those of Germany and Canada would be two basic
options for improving TSCA's information-gathering powers.  Under one
option, EPA would still rely largely on its files and publicly
available data bases as the major sources of information for its
reviews.  EPA would have authority to demand, without undergoing
costly and time-consuming rule-making, that industry provide certain
additional information that EPA needs to complete its reviews.  To
limit the potential reporting burden on industry, this authority
could be limited to filling specific data needs on those chemicals
scheduled for review during a specific period. 

The other option would be to place on industry more of the burden for
existing chemical reviews.  The industry could be required to compile
and submit essential chemical data to EPA, as is done for new
chemicals in submitting premanufacture notices.  EPA would specify
the reporting form and the information to be contained in these
"manufacture notices." Manufacturers or importers could individually
submit the required data or they could collectively submit it for an
individual chemical. 

The industry's costs to prepare the notices could be reduced by
requiring this reporting for only those chemicals scheduled to
undergo EPA's review in the short term.  The agency could notify the
industry of chemicals that it will assess during a certain period and
require that the notices be submitted by a particular date.  Because
more data are available to review and EPA generally has to consider
more uses and potential pathways of exposure for existing chemicals,
the review period probably would need to be longer than the 90 days
for new chemicals. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 4:5

Very little is known about the risks of many of the chemicals to
which workers, consumers, the general public, and plant and animal
life are potentially exposed.  EPA has been slow to assess the risks
of existing chemicals, and for those that have been assessed,
frequent gaps in the data raise questions about the completeness of
the reviews. 

The burden is essentially on EPA in reviewing existing chemicals, and
several factors have contributed to EPA's slow progress.  These
include limited resources, a greater emphasis on new chemicals, and
the difficulty in using TSCA's data-gathering authorities.  Although
these authorities would appear to provide EPA with the tools to
obtain the data it needs, the act's restrictions on EPA, such as the
requirement that EPA promulgate a rule to obtain data on a chemical,
have made using the authorities time-consuming and costly.  EPA's
experience in implementing the act has shown that the restrictions
place a heavy burden on the agency, given available resources. 

The countries that we visited have initiatives under way to
systematically review existing chemicals.  These initiatives
illustrate potential options for revising TSCA to enable EPA to make
better progress in the agency's efforts to review existing chemicals. 
We believe that a substantial increase in the number of these reviews
is unlikely unless TSCA is changed to give such reviews more emphasis
and to place more responsibility on industry to compile the
information needed to assess chemical risks. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
---------------------------------------------------------- Chapter 4:6

To put existing chemicals on a more equal footing with new chemicals,
the Congress could consider revising TSCA to set specific deadlines
or targets for the review of existing chemicals.  These deadlines or
targets would provide for EPA to establish priorities to review those
chemicals that, on the basis of their toxicity, production volumes,
and potential exposure, present the highest risk to health and the
environment. 

The Congress could also consider revising TSCA to shift to the
industry more of the burden for the review of existing chemicals.  If
more of this responsibility were shared by the industry, EPA could
review more chemicals with its current level of resources.  In
deciding how much burden to shift, the major consideration for the
Congress is to what extent is providing the data to show that
chemicals are safe a cost of doing business for the chemical
industry.  As discussed in this chapter, the burden could be shifted
in various ways and to varying extents. 


TSCA'S PROVISIONS ON CONFIDENTIAL
BUSINESS INFORMATION NEED TO BE
REVISED
============================================================ Chapter 5

A key component of EPA's strategy to revitalize its toxics program is
to disseminate information on toxic chemicals to educate the public
on chemical risks and involve them in environmental decision-making. 
Nonetheless, recognizing the need to protect against the disclosure
of trade secrets or other proprietary data to competitors, TSCA
allows companies to designate the information submitted to EPA as
confidential if it meets certain criteria.  Other federal agencies
and federal contractors can obtain access to this confidential
business information if they can demonstrate that they need it to
carry out their responsibilities and they can protect the information
against unauthorized release. 

According to the industry, a large amount of the TSCA data it submits
to EPA is confidential business information.  While EPA considers
many of the confidentiality claims to be legitimate, it does not
believe that others meet the criteria established in the act.  EPA
further believes that the resources necessary to protect the
information designated as confidential discourages other federal
agencies with health and safety responsibilities from obtaining it. 
In addition, the confidentiality designation limits the dissemination
and usefulness of the data because many interested groups are not
allowed access to the data, including private organizations and
public, state, and local groups responsible for health and
environmental issues. 

EPA is currently reviewing confidentiality claims and thus far has
been successful when it has challenged their appropriateness. 
However, this effort is resource-intensive, and EPA has challenged
only a small percentage of the claims.  EPA is also implementing
various voluntary and regulatory measures to reduce claims. 


   CONFIDENTIALITY CLAIMS LIMIT
   DATA'S USEFULNESS AND ARE
   COSTLY
---------------------------------------------------------- Chapter 5:1

EPA believes that public release of important environmental data
gives everyone the ability to participate in the broader national
effort to address chemical issues.  EPA's Toxic Release Inventory
illustrates the ability of information to dramatically promote and
empower initiatives by the toxics community.  Toxic Release Inventory
data have helped industry to identify problems and target actions and
have given the public an opportunity to learn about problems and
become involved in their solutions. 

Information on toxic substances is available in numerous trade
publications and scientific literature, and nonconfidential TSCA data
are available through several sources, such as the National Technical
Information Service, the National Library of Medicine, and the TSCA
hotline.  The confidential TSCA data, however, are available only
through the companies that submit them.  TSCA's data also provide the
only available comprehensive view of what is known about the flow and
environmental effects of commercial chemicals. 

Consequently, the data are of great value to various government
officials and to scientists and researchers in general.  Providing
state government officials with access to TSCA's confidential data
could enable them to control potential risks from chemicals subject
to TSCA reporting, using their authorities under state laws.  The
scientific community could review risk assessment decisions made by
EPA or test EPA's hazard and risk predictions if it had the
confidential TSCA information to perform the tasks.  In addition,
federal agencies, such as the Occupational Safety and Health
Administration, could also use the information for promulgating
worker protection standards. 

TSCA's confidential data is available upon request to staff in
various EPA offices and to other federal agencies.  However, most EPA
offices are generally not familiar with the data available under
TSCA, and even those that are aware of the data do not attempt to use
them because of the difficulties associated with obtaining security
clearances and handling confidential data.  Federal officials outside
of EPA also limit their requests for confidential data because they
are required to give the information the same level of protection
afforded by EPA.  In addition, confidentiality claims make it
difficult for federal officials to know that the data exist within
EPA. 

State and local governments and public groups, such as environmental
organizations and unions, would like to obtain access to the
confidential business information submitted to EPA to protect the
health and safety of the public.  However, TSCA does not provide for
disclosure of such information to them. 

Safeguarding confidential information also imposes significant costs
on EPA.  Staff discussions on chemicals must be held in secure areas,
documents can be reviewed only in secure environments, meeting notes
themselves become confidential documents and must be logged and
guarded under lock and key, and computers must have their memories
and permanent storage media erased after processing confidential
data.  In addition, all staff working with or accessing confidential
documents must have appropriate security clearances.  The protection
that EPA provides for TSCA's confidential data is equivalent to that
provided to information deemed "secret" for national security
purposes.  Since EPA does not account for many of these costs
separately, it cannot quantify how much these requirements cost the
agency. 


   MANY CLAIMS ARE INAPPROPRIATE
---------------------------------------------------------- Chapter 5:2

EPA initiated its Confidential Business Information Review and
Challenge Program in August 1990 to review confidentiality claims in
TSCA filings and, when appropriate, to challenge the claims.  By
September 1992, industry had voluntarily amended and withdrawn over
600 claims after EPA's inquiries.  On the basis of their experience
with TSCA data, EPA officials believe that the problem with
inappropriate claims is extensive.  They plan to continue to review
claims but do not expect to increase their efforts because of limited
resources. 

Concerned about how often its challenge program identified
inappropriate confidentiality claims, EPA retained the Hampshire
Research Associates, Inc., in 1992 to study how EPA's policies on
confidential information have affected the implementation of TSCA. 
The study examined the confidential business information claims made,
the validity of the claims, and the impact of inappropriate claims on
the usefulness to the public of TSCA data. 

The Hampshire Study found that many of the claims submitted under
TSCA were not appropriate, particularly for health and safety data. 
For example, the study noted that between September 1990 and May
1991, EPA reviewed 351 health and safety studies submitted by
chemical companies and challenged 77, or 22 percent.  In each of
these cases, the submitter amended the claim when challenged. 

EPA is concerned about the industry's confidentiality claims over
information on the identity of chemicals in health and safety
studies.  TSCA mandates the availability of health and safety data,
with protection given for proprietary information.  Although EPA
takes a broad view of what is a health and safety study and believes
the legislative history supports this position, it is difficult for
the agency to prevent the chemical industry from making
confidentiality claims for data that relate to health and safety,
given that the statute does not actually define the term. 
Consequently, EPA believes that it would be helpful if TSCA were
revised to clarify that data submitted under the statute be made
available to the public if the data relate to a chemical's effects on
public health and safety. 

The Hampshire study also pointed out that information already
publicly available was submitted to EPA as confidential business
information.  For example, information contained elsewhere in
newspaper articles and corporate annual reports was submitted as was
publicly available information from EPA's Toxics Release Inventory, a
system that contains nationwide information on toxic chemicals
emitted into the air, ground, and water by manufacturing facilities. 
In one case, a submitter provided EPA with a final draft study of the
effects of working with particular chemicals for prolonged periods, a
study that had already been provided to union representatives of the
submitter's employees.  Despite the fact that all of the relevant
information had already been made public, the submitter claimed the
company's name, the union's name, plant sites, and chemical
identities as confidential information.  Following discussions with
EPA, the submitter agreed to drop all confidentiality claims for the
information submitted. 

The Hampshire study also compared data collected under TSCA with
similar data collected under the Emergency Planning and Community
Right-to-Know Act of 1986.  This act required thousands of industrial
facilities, starting in 1988, to report annually to EPA and to states
the estimated quantities of hundreds of toxic chemicals emitted
directly into the environment or transported to waste treatment,
storage, or disposal locations.The study shows that under the
Emergency Planning and Community Right-to-Know Act, confidentiality
claims are much more restrictive.  Under this act, industry can make
claims only for chemical identity, a top corporate official must sign
the claims, a company must submit information to substantiate a trade
secret claim, and EPA can levy penalties on corporate officials
making false trade secret claims under the act.  None of these
conditions applies for claims under TSCA.  The Hampshire study
pointed out that industry is far more likely, because of the more
liberal requirements of TSCA, to claim as confidential information
the data collected under TSCA than nearly identical data reported
under the Emergency Planning and Community Right-to-Know Act. 

EPA invited all interested parties to discuss the points raised in
the Hampshire study and to identify ways to improve EPA's existing
confidential business information policies under TSCA.  EPA held its
first public meeting in October 1992, and subsequent meetings have
been held among a variety of states, environmental groups, and the
regulated community. 

All 12 chemical industry commentators were skeptical of the points
and conclusions raised in the study.  They said that the purpose of
TSCA information is to provide EPA with a factual basis for chemical
regulation, not to provide a basis for disseminating data on the
chemicals to other interested organizations. 

The industry commentators were also concerned about the protection of
confidential information on new chemicals submitted for EPA's review
before they are manufactured.  One commentator observed that
requiring the disclosure of identity for these chemical substances
could provide a company's competitors with critical information on
research direction and on the timing and developmental progress of
new products.  Such information can affect the marketability of new
chemical products.  EPA responded by saying that it is aware of the
sensitivity of information on the new-chemicals process and that for
this reason, the focus of EPA's effort to disseminate toxics
information has been on those chemicals that have already gone
through the new-chemicals review process and may be legally used in
commerce. 

Although the chemical industry was critical of the report, most of
the chemical industry commentators accepted the study's basic finding
that the chemical industry does make improper confidentiality claims
and needs to address such claims.  The industry commentators stated
that EPA could use its workshops, newsletters, and
question-and-answer bulletins to inform the industry of EPA's
confidential business information policy.  In addition, some
commentators were amenable to a change in the TSCA process to provide
for holding information confidential for only a specific period,
provided that the chemical industry is given an opportunity to
justify extending the period, as circumstances warrant.  For example,
a chemical company could furnish comments in response to an EPA
request that the company resubstantiate a claim by a specific date. 
If the company did not submit such comments, EPA would automatically
declassify the information previously submitted on the chemical. 


   EPA IS CONSIDERING ALTERNATIVES
   TO REDUCE INAPPROPRIATE CLAIMS
---------------------------------------------------------- Chapter 5:3

Using study recommendations and meeting results, EPA published in the
Federal Register, dated June 4, 1993, the notice of a public meeting
to obtain comments on confidential business information policies and
certain proposed EPA actions that would change confidential business
information requirements.  After considering comments on the proposed
actions, EPA announced in a July 6, 1994, Federal Register notice the
availability of its final action plan. 

EPA's plan includes various voluntary and regulatory measures to
reduce confidentiality claims.  One of the voluntary measures is the
development of educational programs to apprise industry of EPA's
policy on confidentiality claims and procedures.  In addition, the
Chemical Manufacturers Association has offered to implement
educational programs for its members and the regulated chemical
community.  The Chemical Manufacturers Association presented a course
on the subject in December 1993 and in February and June 1994.  EPA
is also planning to prepare instructional papers clearly identifying
the types of confidentiality claims that it will not accept. 

Other measures of EPA's reform program include: 

  Exploring how to make TSCA's confidential information available to
     the states.  This project is under way and is supported by the
     Chemical Manufacturers Association and the Forum on State and
     Tribal Toxics Action. 

  Having the regulated community voluntarily review old submissions
     of claims on a periodic basis to weed out unnecessary ones. 

  Continuing to review claims and challenging those that appear to be
     inappropriate. 

EPA is also considering changes to its regulations to require (1)
industry to substantiate claims at the time of submission, (2) senior
corporate officials to sign claims to ensure that the information
claimed as confidential meets statutory and regulatory requirements,
and (3) submitters to resubstantiate claims at a future date when
confidentiality may no longer be required.  According to EPA, the
agency has already begun to implement many of the measures in its
plan and expects to achieve significant progress by January 1995. 


   AUTHORITIES TO CLAIM
   CONFIDENTIALITY ARE MORE
   RESTRICTIVE IN COUNTRIES
   VISITED
---------------------------------------------------------- Chapter 5:4

The other countries that we visited also allow industry to make
confidentiality claims.  However, these countries generally specify
more types of data that cannot be claimed as confidential.  In
addition, they generally have more requirements for substantiating or
justifying confidentiality claims. 

While health and safety studies are the only type of information on
which TSCA restricts confidentiality claims, Canada generally does
not allow claims on data such as chemical uses and safe handling
procedures.  Exposure data are confidential in Germany, but it
generally does not allow claims for information such as the
chemical's trade name, physical chemical properties, precautionary
and emergency measures, and evaluations of toxicological tests. 
Sweden is more restrictive in that it generally limits claims to
chemical identity and some business aspects, such as the volume of
production.  (Although these types of data would appear to provide
the public with information on the potential dangers of chemicals
without revealing proprietary information, some U.S.  industry
representatives have expressed concern that the information could, in
some cases, reveal trade secrets.)

Upon making confidentiality claims in Sweden, chemical manufacturers
must submit evidence substantiating that disclosure of the data would
adversely affect business.  In Germany, chemical manufacturers must
indicate the information that is commercially sensitive and provide
evidence that its disclosure might cause the company industrial or
commercial harm.  Canada also requires some substantiation of claims
but not the extent of justification required in Sweden and Germany. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 5:5

EPA needs a process that will both protect legitimate confidential
business information claims under TSCA and discourage the chemical
industry from filing inappropriate claims.  EPA is concerned that the
health and safety data that should be available to the public is now
being claimed by industry as confidential business information. 
While EPA currently reviews a limited number of confidentiality
claims for appropriateness, this process is very resource-intensive
and would be difficult to expand, given EPA's limited resources. 

TSCA does not permit the sharing of confidential business information
with states.  If state governments could be provided with access to
this information, it would enable them to control potential risks
from chemicals subject to TSCA reporting, using their state
authorities.  This would provide the public with another line of
defense to protect health and the environment. 

Canada, Germany, and Sweden also recognize the need to protect trade
secrets.  However, they place more restrictions than the United
States does on the types of information that can be claimed as
confidential. 

Administrative initiatives planned by EPA include educating the
chemical industry on confidential business information policies and
procedures and having the chemical industry periodically review old
claims to weed out those that have expired.  Although EPA plans to
pursue these and other changes administratively, we believe that
legislation may be needed, especially for some of them, such as
making confidential business information available to states. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
---------------------------------------------------------- Chapter 5:6

To ensure that EPA can implement its initiatives without having to
face legal challenges and delays, the Congress may wish to consider
revising TSCA to

  clarify that health and safety data cannot be claimed as
     confidential business information,

  require substantiation of confidentiality claims at the time that
     the claims are submitted to EPA,

  limit the length of time for which information may be claimed as
     confidential without resubstantiation of the need for
     confidentiality,

  establish penalties for the false filing of confidentiality claims,
     and

  authorize states to have access to confidential business
     information when they can demonstrate to EPA that they have a
     legitimate need for the information and can adequately protect
     it against unauthorized disclosure. 


REGULATIONS PROMULGATED UNDER
SECTION 6 OF THE TOXIC SUBSTANCES
CONTROL ACT
=========================================================== Appendix I


      POLYCHLORINATED BIPHENYLS
      (PCB)
------------------------------------------------------- Appendix I:0.1

Because the Congress believed that PCBs posed a significant risk to
public health and the environment, TSCA specifically prohibits the
manufacture, processing, distribution in commerce, and use of all
PCBs in other than a totally enclosed system, unless authorized by
EPA, and requires their proper disposal.  PCBs, which have been used
primarily in electrical equipment, are toxic and very persistent in
the environment.  When released into the environment, they decompose
very slowly and can accumulate in plants, animals, and human tissue. 
Laboratory tests show that they cause cancer in rats and mice and
that they have adverse effects on fish and wildlife.  EPA has issued
various rules to implement these requirements. 


      CHLOROFLUOROCARBONS (CFC)
------------------------------------------------------- Appendix I:0.2

In 1978, EPA banned nonessential uses of CFCs as propellants in
aerosol spray containers.  EPA took this action because of a concern
that these chemicals were destroying the upper atmosphere's ozone
layer, which shields the earth from ultraviolet radiation.  Increased
exposure to ultraviolet radiation has been linked to increased skin
cancer.  Depletion of the ozone layer is also thought to lead to
climate changes and other adverse effects.  CFCs have numerous other
uses, ranging from solvents to air conditioning.  Litigation for EPA
to act on these other uses and international concern about ozone
depletion led to an international agreement in late 1988 to freeze
production of five major CFCs at 1986 levels and cut production
levels up to 50 percent by 1999.  EPA issued a final rule under the
Clean Air Act in August 1988 that allocates production quotas to
current producers on the basis of their 1986 production levels. 


      DIOXIN
------------------------------------------------------- Appendix I:0.3

In 1980, EPA promulgated a rule prohibiting Vertac Chemical Company
from disposing of its wastes containing
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) stored at its
Jacksonville, Arkansas, facility.  The rule also required that any
other persons planning to dispose of TCDD-containing wastes notify
EPA 60 days before their intended disposal.  TCDD, the most toxic of
the about 75 dioxins in existence and an animal carcinogen, is a
contaminant or waste product formed during the manufacture of certain
substances.  EPA concluded that it is likely to result in adverse
human health effects. 


      ASBESTOS
------------------------------------------------------- Appendix I:0.4

In 1982, EPA issued a rule requiring all public and private
elementary and secondary schools to inspect for friable (easily
crumbled into powder) asbestos-containing materials.  Asbestos, which
refers to several minerals that typically separate into very tiny
fibers, is a known human carcinogen that can cause lung cancer and
other diseases if inhaled.  Asbestos-containing materials were used
widely for fireproofing, thermal and acoustical insulation, and
decoration in building construction and renovation before the adverse
effects of asbestos were known.  Asbestos also has numerous other
applications, for example, in friction products such as brake
linings.  In July 1989, EPA issued a final rule to ban the
manufacturing, importing, and processing of nearly all asbestos
products.  The rule was to begin phasing out asbestos-containing
products in August 1990 and complete the phaseout by 1997.  EPA was
challenged in federal court by asbestos manufacturers, and in October
1991 the Fifth Circuit Court of Appeals vacated most of the rule (the
rule continued to apply to asbestos products no longer in commerce)
and remanded it to the agency for further consideration.  EPA first
regulated asbestos in the early 1970s as a hazardous air pollutant
under the Clean Air Act by prescribing, among other things, work
practices to prevent or minimize the release of asbestos into the air
during the demolition or renovation of buildings containing asbestos. 


      HEXAVALENT CHROMIUM
------------------------------------------------------- Appendix I:0.5

In 1990, EPA banned the use of hexavalent-chromium-based water
treatment chemicals in commercial cooling towers on the basis of
health risks associated with human exposure to air emissions. 
Hexavalent chromium is a known human carcinogen that is also widely
used in industrial cooling towers.  EPA could have issued an
emissions standard under the Clean Air Act.  However, the agency
believed that regulation under TSCA would be more efficient and
effective because the act could be used to regulate use and
distribution of hexavalent-chromium-based water treatment chemicals. 
According to EPA, hexavalent chromium was being released from a large
number of unidentified cooling towers. 


OTHER MAJOR LAWS THAT CAN ADDRESS
INDUSTRIAL CHEMICAL CONCERNS
========================================================== Appendix II

Law                            Regulated action
-----------------------------  -----------------------------
Clean Air Act of 1963, as      Protect and enhance air
amended                        quality to promote public
                               health and welfare.

Federal Water Pollution        Restore and maintain the
Control Act of 1956, as        chemical, physical, and
amended (Clean Water Act)      biological integrity of the
                               nation's waters.

Safe Drinking Water Act of     Protect the quality of all
1974, as amended               sources of drinking water.

Resource Conservation and      Regulate the generation,
Recovery Act of 1976, as       transportation, treatment,
amended (RCRA)                 shortage, and disposal of
                               hazardous wastes.

Comprehensive Environmental    Finance cleanup measures for
Response, Compensation, and    releases of hazardous
Liability Act of 1980, as      substances and leaking
amended (Superfund)            hazardous waste dumps.

Marine Protection, Research,   Regulate the dumping of
and Sanctuaries Act of 1972,   materials into oceans and
as amended                     prevent or strictly limit the
                               dumping of material that
                               adversely affects human
                               health or the marine
                               environment.

Occupational Safety and        Develop and enforce mandatory
Health Act of 1970, as         job safety and health
amended                        standards to ensure as far as
                               possible that employees have
                               safe and healthful working
                               conditions.

Consumer Product Safety Act    Protect the public against
of 1972, as amended            unreasonable risks of injury
                               associated with consumer
                               products.
------------------------------------------------------------

STATUS OF REGULATION OF TOXIC
CHEMICALS UNDER SELECTED LAWS
========================================================= Appendix III

The following narratives summarize the status of toxic chemical
regulation under the Clean Water, Clean Air, Safe Drinking Water, and
Occupational Safety and Health Acts. 


      CLEAN WATER ACT
----------------------------------------------------- Appendix III:0.1

In July 1991, we reported that excessive levels of toxic pollutants
getting into the nation's rivers, lakes, and streams is a serious
water quality problem.\1 Under the Clean Water Act, the principal
tools for controlling toxic water pollution are numeric discharge
limits in permits issued to facilities discharging wastes to surface
waters.  These permit limits are largely based on national effluent
guidelines and criteria documents developed by EPA. 

We also reported that these national standards control only a limited
number of toxic discharges.  EPA was slow to revise existing national
effluent guidelines and to develop new ones.  Specifically, some
existing guidelines did not reflect the latest advances in treatment
technologies available to eliminate toxic and nonconventional
discharges, and many categories of industries discharging toxics were
not covered by such guidelines.  EPA was also slow to develop and
revise criteria documents.  According to EPA officials, the lack of
resources hampered their efforts to issue more timely criteria
documents. 

Our recent update of EPA's progress in developing criteria documents
showed that only a few of them had been developed or updated since
our 1991 report.  For a list of 126 priority toxic pollutants, EPA,
as of September 30, 1993, had issued human health criteria documents
for 91 of the chemicals and aquatic life criteria documents for only
26 of them.  Aquatic life criteria documents for 12 of the priority
pollutants had been revised, whereas none of the human health
criteria documents had been revised.  Human health and aquatic life
criteria documents had also been issued for eight nonpriority
pollutants.  EPA was in the process of developing criteria documents
for 12 additional priority pollutants and 11 nonpriority ones and
revising the criteria documents for 1 priority and 2 nonpriority
pollutants. 


--------------------
\1 Water Pollution:  Stronger Efforts Needed by EPA to Control Toxic
Water Pollution (GAO/RCED-91-154, July 19, 1991). 


      CLEAN AIR ACT
----------------------------------------------------- Appendix III:0.2

Section 112 of the Clean Air Act authorized EPA to establish
standards to regulate emissions of hazardous air pollutants. 
However, by 1990 EPA had promulgated standards for only seven
substances.  Dissatisfied with this slow pace, the Congress
established a list of 189 toxic pollutants in the Clean Air Act
Amendments of 1990 and directed EPA to impose standards with a new
two-phase strategy.  For the first phase, the amendments set out a
schedule for promulgation of standards for all 189 of the pollutants
by November 2000.  As the second phase of control, after these
standards have been met, certain facilities may be subject to further
regulation in situations in which EPA determines that additional
standards are required to protect health and the environment. 

EPA is already behind in meeting the schedule set by the 1990
amendments.  The agency's first deadline was to promulgate 40
standards by November 15, 1992.  It did not promulgate its first
standard until September 1993 and did not plan to issue the remaining
standards until February 1994, nearly 15 months behind schedule. 


      SAFE DRINKING WATER ACT
----------------------------------------------------- Appendix III:0.3

Chemicals can enter drinking water supplies as a consequence of
surface or ground water contamination in conjunction with chemical
production, use, or disposal.  The Safe Drinking Water Act, enacted
in 1974, required EPA to, among other things, establish standards or
treatment techniques for drinking water contaminants that could
adversely affect human health.  By the mid-1980s, EPA had regulated
only 23 contaminants.  Concerned about drinking water quality and
frustrated with the pace at which EPA was developing regulations, the
Congress amended the act in 1986 to require EPA to regulate 83
specific contaminants by 1989, plus an additional 25 contaminants
every 3 years thereafter. 

A September 1993 EPA report to the Congress stated that 76 of the 83
contaminants had been regulated and the remaining 7 were in process. 
According to the report, EPA will regulate an estimated total of 112
contaminants by 1995.  Although the report noted that other
contaminants could be a problem in some locations and at some wells
(over 77,000 industrial, pesticide, food, and drug chemicals are
released to the environment to some extent), EPA said that a
continuing stream of these regulations would add considerably to the
regulatory burden on states and drinking water systems and detract
from implementation of the priority contaminants among the first 112
standards.  EPA estimated that the annual state funding shortfall for
implementing federal drinking water requirements is approximately
$162 million. 


      OCCUPATIONAL SAFETY AND
      HEALTH ACT
----------------------------------------------------- Appendix III:0.4

A key responsibility of the Occupational Safety and Health
Administration (OSHA) under the Occupational Safety and Health Act is
to issue safety and health standards for the workplace.  However,
employees continue to be exposed to many hazardous work practices,
conditions, and substances because of delays by OSHA in issuing these
standards.  Since 1971, OSHA has promulgated fewer than 30 health and
40 safety standards, and it routinely takes up to 10 years from the
time the agency recognizes the need to regulate until the regulation
becomes final.  One attempt by OSHA to speed up the process was
successfully challenged in court.  OSHA had updated the permissible
exposure limits for over 400 substances in one rule-making effort
that took less than 2 years.  However, the court ruled that OSHA
could not change permissible exposure levels for multiple substances
without providing substantial evidence in support of each change. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix IV


   RESOURCES, COMMUNITY, AND
   ECONOMIC DEVELOPMENT DIVISION,
   WASHINGTON, D.C. 
-------------------------------------------------------- Appendix IV:1

Bernice Steinhardt, Associate Director, (202) 512-6501
Edward A.  Kratzer, Project Director
Raymond H.  Smith, Jr., Project Manager
Donald E.  Pless, Senior Evaluator
Frank J.  Gross, Senior Evaluator


   PHILADELPHIA REGIONAL OFFICE
-------------------------------------------------------- Appendix IV:2

Denise R.  Dias, Senior Evaluator
Victoria A.  Snyder, Evaluator