Regulatory Flexibility Act: Implementation in EPA Program Offices and
Proposed Lead Rule (Letter Report, 09/20/2000, GAO/GGD-00-193).

Pursuant to a congressional request, GAO reviewed the Office of
Prevention, Pesticides and Toxic Substances' (OPPTS) implementation of
the Regulatory Flexibility Act (RFA), focusing on the: (1) guidance that
OPPTS and the Environmental Protection Agency's (EPA) three other major
program offices have used during the past 10 years to determine whether
their proposed rules could be certified as not having a significant
economic impact on a substantial number of small entities (SEISNSE); (2)
comparing rate at which OPPTS certified that its substantive proposed
rules published during calendar years 1994 through 1999 would not have a
SEISNSE with the rates in EPA's other major program offices; and (3)
methodology that OPPTS used in the economic analysis for the proposed
lead rule and: (a) key aspects of that methodology that may have
contributed to the Office's conclusion that the rule would not have a
SEISNSE; (b) if so, how OPPTS has changed its economic analysis since
publication of the rule; and (c) whether additional data or analysis
could have yielded a different conclusion about the rule's impact on
small entities.

GAO noted that: (1) OPPTS and EPA's other major program offices have
adopted three very different sets of guidance documents during the past
10 years to determine if their proposed rules could be certified as not
having a SEISNSE; (2) the current guidance was implemented because of
the enactment of the Small Business Regulatory Enforcement Fairness Act
(SBREFA) in 1996; (3) the guidance includes specific numerical
guidelines to assist EPA program offices in determining whether a rule
should be certified as not having a SEISNSE; (4) those guidelines focus
on the magnitude of a rule's impact and the number and percent of
affected small entities that are expected to experience that impact; (5)
OPPTS certified 86 percent of the substantive proposed rules that it
published during calendar years 1994 through 1999--about the same rate
of rule certification as EPA's three other major program offices; (6)
the RFA certification rate in OPPTS and in each of the three other major
program offices increased after the implementation of SBREFA in 1996;
(7) by the end of 1999, OPPTS and one other program office had certified
all of their post-SBREFA proposed rules as not having a SEISNSE; (8) two
other EPA program offices used the discretion afforded them in the
guidance and did not certify at least four proposed rules; (9) OPPTS
estimated that the smallest manufacturers affected by the proposed lead
rule had annual revenues of $4 million; (10) OPPTS concluded that the
rule would affect about 5,600 small businesses, but none of them would
experience first-year compliance costs of 1 percent or more of their
annual revenues; (11) therefore, OPPTS certified that the proposed lead
rule would not have a SEISNSE; (12) according to the summary of the
draft revised economic analysis, OPPTS expected that the proposed lead
rule would affect more than 8,600 small companies, and as many as 464 of
them would experience first-year compliance costs of at least 1 percent
of their annual revenues; (13) nevertheless, OPPTS again concluded that
the rule would not have a SEISNSE; (14) GAO estimated that as many as
1,098 additional small manufacturing companies could experience
compliance costs of at least 1 percent of their annual revenues, and as
many as 78 small companies could experience a 3 percent economic impact;
(15) therefore, if OPPTS had used this analytic approach and the
discretion permitted in EPA's RFA guidance, it could have chosen not to
certify the rule; and (16) the Office's initial and draft revised
analyses and the conclusions that it based on those analyses were within
the discretion permitted under the RFA and the EPA guidance.

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

 REPORTNUM:  GGD-00-193
     TITLE:  Regulatory Flexibility Act: Implementation in EPA Program
	     Offices and Proposed Lead Rule
      DATE:  09/20/2000
   SUBJECT:  Proposed legislation
	     Agency proceedings
	     Economic analysis
	     Small business assistance
	     Federal regulations
	     Regulatory agencies
	     Toxic substances
	     Environmental policies
IDENTIFIER:  EPA Toxic Release Inventory

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

GAO/GGD-00-193

REGULATORY FLEXIBILITY ACT

Implementation in EPA Program Offices and Proposed Lead Rule

United States General Accounting Office

GAO Report to the Chairman Committee on Small Business

U. S. Senate

September 2000 GAO/ GGD- 00- 193

United States General Accounting Office General Government Division
Washington, D. C. 20548

Page 1 GAO/ GGD- 00- 193 Regulatory Flexibility Act

B- 284265 September 20, 2000 The Honorable Christopher S. Bond Chairman,
Committee on Small Business United States Senate

Dear Mr. Chairman: On August 3, 1999, the Environmental Protection Agency's
(EPA) Office of Prevention, Pesticides and Toxic Substances (OPPTS)
published a proposed rule in the Federal Register that would, upon
implementation, lower certain reporting thresholds for lead and lead
compounds under the Toxics Release Inventory (TRI) program from 25,000
pounds to 10 pounds. 1 OPPTS said the proposed lead rule would have positive
effects on health, safety, and the natural environment, but it did not
assign a dollar value to those benefits due to the lack of adequate
methodologies. OPPTS estimated that the rule would cost businesses $116
million during the first year of implementation and that approximately 5,600
small businesses would be affected by the rule. However, OPPTS said that
none of these small businesses would experience annual compliance costs
above 1 percent of annual sales and that the proposed rule would, therefore,
not impose significant compliance costs on any of these small businesses. As
a result, OPPTS certified that the rule would not have a significant
economic impact on a substantial number of small entities (SEISNSE) and did
not trigger certain analytical and procedural requirements of the Regulatory
Flexibility Act (RFA), as amended. 2 Concerns were subsequently raised
regarding the methodologies that OPPTS used in its analysis and its
conclusions about the impact of the rule on small businesses.

This report responds to your request that we review OPPTS' implementation of
the RFA and certain aspects of OPPTS' economic analysis of its proposed lead
rule. Specifically, you asked us to

1 64 Fed. Reg. 42222 (1999). In this report, the rule will be referred to as
“the proposed lead rule.” 2 The RFA is codified at 5 U. S. C.
601- 612. Section 601 of the RFA defines a “small entity” as
including small businesses, small governmental jurisdictions, and other
small organizations. The RFA incorporates the generally accepted meanings of
small business as established by the Small Business Administration through
size standards that define whether a business is small. For example, the
size standard is 500 employees for about 75 percent of the manufacturing
industries, and either 750, 1, 000, or 1,500 for the remaining manufacturing
industries.

B- 284265 Page 2 GAO/ GGD- 00- 193 Regulatory Flexibility Act

ï¿½ describe the guidance that OPPTS and EPA's three other major program
offices have used during the past 10 years to determine whether their
proposed rules could be certified as not having a SEISNSE;

ï¿½ compare the rate at which OPPTS certified that its substantive proposed
rules published during calendar years 1994 through 1999 would not have a
SEISNSE with the rates in EPA's other major program offices; and

ï¿½ review the methodology that OPPTS used in the economic analysis for the
proposed lead rule and (1) describe key aspects of that methodology that may
have contributed to the Office's conclusion that the rule would not have a
SEISNSE, (2) determine whether and, if so, how OPPTS has changed its
economic analysis since publication of the rule, and (3) determine whether
additional data or analysis could have yielded a different conclusion about
the rule's impact on small entities.

Under the RFA, federal agencies have broad discretion regarding how key
terms in the act should be defined and how RFA certification decisions
should be made. Using that discretion, OPPTS and EPA's other major program
offices have adopted three very different sets of guidance documents during
the past 10 years to determine if their proposed rules could be certified as
not having a SEISNSE. The current guidance was implemented because of the
enactment of the Small Business Regulatory Enforcement Fairness Act (SBREFA)
in 1996. The guidance includes specific numerical guidelines to assist EPA
program offices in determining whether a rule should be certified as not
having a SEISNSE. Those guidelines focus on the magnitude of a rule's impact
(e. g., estimated annual compliance costs as a percentage of estimated
annual revenues) and the number and percent of affected small entities that
are expected to experience that impact. For example, the guidance indicates
that a rule should be presumed eligible for certification if it does not
impose a 1 percent economic impact on any number of small entities. However,
this guidance and the two preceding sets of guidance also give agency
personnel broad discretion in how they should be applied.

OPPTS certified 86 percent of the substantive proposed rules that it
published during calendar years 1994 through 1999- about the same rate of
rule certification as EPA's three other major program offices. The RFA
certification rate in OPPTS and in each of the three other major program
offices increased after the implementation of SBREFA in 1996. By the end of
1999, OPPTS and one other program office had certified all of their
postSBREFA proposed rules as not having a SEISNSE. Two other EPA program
offices used the discretion afforded them in the guidance and did not
certify at least four proposed rules published during this period that the
numerical guidelines would have allowed them to certify. Results In Brief

B- 284265 Page 3 GAO/ GGD- 00- 193 Regulatory Flexibility Act

OPPTS made a number of assumptions in the original economic analysis for the
proposed lead rule that contributed to its determination that no small
entities would experience significant economic effects. For example, to
estimate the annual revenues of companies expected to file new reports for
lead, OPPTS assumed that (1) the new filers would have employment and
economic characteristics similar to current TRI filers, (2) different types
of manufacturers would experience similar economic effects, and (3) the
revenues of the smallest manufacturers covered by the proposed rule could be
exemplified by the firm at the 25 th percentile of the Office's projected
revenue distribution for small manufacturers. As a result of these and other
assumptions, OPPTS estimated that the smallest manufacturers affected by the
proposed lead rule had annual revenues of $4 million. Using that and other
information, OPPTS concluded that the rule would affect about 5,600 small
businesses, but none of them would experience first- year compliance costs
of 1 percent or more of their annual revenues. Therefore, OPPTS certified
that the proposed lead rule would not have a SEISNSE.

OPPTS revised these and other parts of the economic analysis for the
proposed lead rule before submitting it to the Office of Management and
Budget (OMB) for final review in July 2000. 3 According to a summary of the
draft revised economic analysis, OPPTS expected that the proposed lead rule
would affect more than 8,600 small companies, and as many as 464 of them
would experience first- year compliance costs of at least 1 percent of their
annual revenues. Nevertheless, OPPTS again concluded that the rule would not
have a SEISNSE. Using data that we obtained from the Bureau of the Census,
we estimated that as many as 1,098 additional small manufacturing companies
could experience compliance costs of at least 1 percent of their annual
revenues, and as many as 78 small companies could experience a 3 percent
economic impact. Therefore, if OPPTS had used this analytic approach and the
discretion permitted in EPA's RFA guidance, it could have chosen not to
certify the rule. However, the Office's initial and draft revised analyses
and the conclusions that it based on those analyses were within the
discretion permitted under the RFA and the EPA guidance.

The RFA of 1980 requires federal agencies to examine the impact of their
proposed and final rules on small entities and to solicit the ideas and
comments of such entities for this purpose. Specifically, whenever agencies
are required to publish a notice of proposed rulemaking (NPRM),

3 Under Executive Order 12866, issued in September 1993, federal agencies
must submit their significant draft proposed and final rules to OMB before
publishing them in the Federal Register. Background

B- 284265 Page 4 GAO/ GGD- 00- 193 Regulatory Flexibility Act

sections 603 and 604 of the RFA require agencies to prepare an initial and a
final regulatory flexibility analysis (known as an IRFA and a FRFA,
respectively) when publishing a proposed and final rule. However, subsection
605( b) of the RFA says that sections 603 and 604 do not apply to any
proposed or final rule if the head of the agency certifies that the rule
will not have a SEISNSE. The RFA took effect on January 1, 1981.

On March 29, 1996, Congress passed SBREFA to strengthen the RFA's
protections for small entities. Among other things, SBREFA requires that
before publishing an NPRM that may have a SEISNSE, EPA must convene a small
business advocacy review panel for the draft rule. 4 Composed of
representatives from EPA and OMB's Office of Information and Regulatory
Affairs (OIRA) and the Small Business Administration's (SBA) Chief Counsel
for Advocacy, the panel must collect the advice and recommendations of
representatives of affected small entities about the potential impact of the
draft rule. The panel must report on the comments that it receives from
small entities and on the panel's recommendations no later than 60 days
after the panel is convened. EPA then decides whether to change the rule
pursuant to the panel's recommendations and to publish the NPRM. SBREFA also
amended the RFA to allow small entities that are adversely affected by a
rule to seek judicial review of the agency's compliance with certain
provisions of the RFA, including the agency's certification determinations
under subsection 605( b) of the act.

The RFA requires agencies to either certify that a proposed rule will not
have a SEISNSE or conduct an IRFA. It does not specifically require agencies
to determine whether or not the rule will have a SEISNSE. Therefore, EPA
officials said that the agency never determines that a proposed rule will
have a SEISNSE. The agency either certifies the rule will not have a SEISNSE
or prepares an IRFA. EPA officials emphasized that preparing an IRFA does
not mean that the agency has concluded that the rule will have a SEISNSE.

We have issued several reports in recent years on the implementation of the
RFA and SBREFA.

ï¿½ In our 1991 report on the RFA and small governments, we concluded that
each of the four federal agencies we reviewed had a different

4 These requirements are codified at 5 U. S. C. 609. The Occupational Safety
and Health Administration within the Department of Labor is also required to
convene advocacy review panels in these circumstances. Previous Reports on
the

RFA and SBREFA

B- 284265 Page 5 GAO/ GGD- 00- 193 Regulatory Flexibility Act

interpretation of key RFA provisions. 5 We said that the act allowed
agencies to interpret when they believed their proposed regulations affected
small governments. We recommended that Congress consider amending the RFA to
require SBA to develop criteria regarding whether and how to conduct the
required analyses.

ï¿½ In our 1994 report on the RFA, we noted that the act required the SBA
Chief Counsel for Advocacy to monitor agencies' compliance. However, we also
said that one reason for agencies' lack of compliance with the RFA's
requirements was that the act did not expressly authorize SBA to interpret
key provisions in the statute and did not require SBA to develop criteria
for agencies to follow in reviewing their rules. 6 We said that if Congress
wanted to strengthen the implementation of the RFA, it should consider
amending the act to (1) provide SBA with clearer authority and
responsibility to interpret the RFA's provisions; and (2) require SBA, in
consultation with OMB, to develop criteria as to whether and how federal
agencies should conduct RFA analyses.

ï¿½ In our 1998 report on the implementation of the small business advocacy
review requirements in SBREFA, we said that the lack of clarity regarding
whether EPA should have convened panels for two of its proposed rules was
traceable to the lack of agreed- upon governmentwide criteria as to whether
a rule has a SEISNSE. 7 We said that if Congress wished to clarify and
strengthen the implementation of the RFA and SBREFA, it should consider (1)
providing SBA or another entity with clearer authority and responsibility to
interpret the RFA's provisions and (2) requiring SBA or some other entity to
develop criteria defining a SEISNSE.

ï¿½ In 1999, we noted a similar lack of clarity regarding the RFA's
requirements for the review of existing rules under section 610 of the act,
and we recommended that Congress and OIRA take action to clarify those
issues. 8

To date, Congress has not acted on any of our recommendations to clarify key
terms in the RFA.

5 Regulatory Flexibility Act: Inherent Weaknesses May Limit Its Usefulness
for Small Governments (GAO/ HRD- 91- 61, Jan. 11, 1991). 6 Regulatory
Flexibility Act: Status of Agencies' Compliance (GAO/ GGD- 94- 105, Apr. 27,
1994). In this report, we noted that the SBA Chief Counsel for Advocacy had
repeatedly considered EPA to be in compliance with the RFA.

7 Regulatory Reform: Implementation of the Small Business Advocacy Review
Panel Requirements (GAO/ GGD- 98- 36, Mar. 18, 1998). 8 Regulatory
Flexibility Act: Agencies' Interpretations of Review Requirements Vary (GAO/
GGD- 99- 55, Apr. 2, 1999).

B- 284265 Page 6 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Although EPA has numerous headquarters and regional offices, its four major
program offices issue most of the agency's regulations. Those offices are as
follows:

ï¿½ OPPTS, which develops national strategies for toxic substance control,
promotes pollution prevention, and evaluates the safety of pesticides and
chemicals;

ï¿½ the Office of Air and Radiation, which oversees air and radiation
protection activities;

ï¿½ the Office of Solid Waste and Emergency Response, which provides policy,
guidance, and direction for the land disposal of hazardous wastes,
underground storage tanks, solid waste management, encouragement of
innovative technologies, source reduction of wastes and the Superfund
program; and

ï¿½ the Office of Water, which is responsible for water quality activities
including development of national programs, technical policies, and
regulations relating to drinking water, water quality, ground water,
pollution source standards, and the protection of wetlands, marine, and
estuarine areas.

Until October 1999, OPPTS was also primarily responsible for promoting the
public's right to know about chemical risks. 9 One key element of this
responsibility is the TRI program, which is a database created to inform the
public about chemical hazards in their communities. TRI reporting is
required by Section 313 of the Emergency Planning and Community Rightto-
Know Act of 1986 (EPCRA), 10 which was enacted in response to serious
chemical releases by plants in Bhopal, India, and West Virginia. Section 313
of the act generally requires facilities to report the amounts of various
toxic chemicals that they release to the environment and requires EPA to
make this information available to the public. EPCRA originally required
that reports be filed by owners and operators of facilities that (1) had 10
or more full- time employees; (2) were in Standard Industrial Classification
(SIC) codes 20 through 39 (manufacturing); 11 and (3) manufactured,

9 In October 1999, this responsibility was given to EPA's new Office of
Environmental Information. However, OPPTS officials and staff remained
involved in the proposed lead rule. To simplify the presentation of this
report, some of the actions taken after October 1999 by this new Office in
relation to the proposed lead rule are described as OPPTS actions.

10 42 U. S. C. 11001- 11050, 11023. Reporting is also required under the
Pollution Prevention Act of 1990 (42 U. S. C. 13101- 13109, 13106), which
added reporting requirements to EPCRA's reporting requirements beginning in
1991.

11 The SIC code classification system was used by the federal and state
governments, trade associations, and research organizations. On April 9,
1997, OMB decided to replace the SIC code system with the North American
Industry Classification System. However, the TRI program still uses the SIC
code EPA Program Offices

The TRI Program

B- 284265 Page 7 GAO/ GGD- 00- 193 Regulatory Flexibility Act

processed, or otherwise used any chemical listed in the act in quantities
greater than the established thresholds in the course of a calendar year.
EPCRA gave EPA the authority to add other industries that have to report, to
add chemicals for which reports have to be filed, and to change reporting
thresholds. Starting with reporting year 1998, EPA added seven
nonmanufacturing industries to the program's coverage. 12 Previously, EPA
had also added nearly 300 toxic substances to the more than 300 that were
originally listed in EPCRA. 13

The specific industries covered by the TRI program are identified by SIC
code. In the SIC code system, major groups of industries are coded with 2
digits, and additional digits are used hierarchically to classify different
industries within the major group. For example, SIC code 36 is for
manufacturers of “electronic and other electrical equipment and
components,” which includes all companies in SIC codes starting with
36, such as SIC code 367 “electronic components and
accessories.” Correspondingly, SIC code 367 contains all companies in
the more narrowly defined SIC code 3672 for “printed circuit
boards.”

The TRI report, known as a Form R, must be at least five pages in length,
and consists of both facility identification information and
chemicalspecific information. The chemical- specific information that must
be included in the report includes (1) the quantity of the chemical entering
each environmental medium on- site (e. g., fugitive or nonpoint air
emissions, discharges to each receiving stream or water body, underground
injection on- site into wells, and disposing to land on- site); (2)
transfers of the chemical in wastes to off- site locations (e. g., publicly
owned treatment works); (3) on- site treatment methods and efficiency (e.
g., waste treatment methods sequence and range of influent concentration);
and (4) source reduction and recycling activities (e. g., the quantity
released, used for energy recovery, recycled on- site, and treated off-
site). Facilities must keep a copy of each report filed for at least 3 years
from the date of submission and must also maintain those documents,
calculations, worksheets, and other forms upon which they

categories because EPCRA specifically references SIC codes 20 through 39.
All of the data in this report use the 1987 SIC code definitions.

12 62 Fed. Reg. 23834 (1997). The added industries were metal mining (SIC
10), coal mining (SIC 12), electric services (SIC 4911), electric and other
services combined (SIC 4931), combination utilities (not elsewhere
classified) (SIC 4939), refuse systems (SIC 4953), chemicals and allied
products (not elsewhere classified) (SIC 5169), petroleum bulk plants and
terminals (SIC 5171), and solvent recovery services (SIC 7389). The TRI
program covers only certain types of facilities in some of these SIC codes.

13 59 Fed. Reg. 61432 (1994).

B- 284265 Page 8 GAO/ GGD- 00- 193 Regulatory Flexibility Act

relied to gather information for prior reports. EPA may conduct data quality
reviews of Form R submissions, and the agency said that a “partial
list” of records, organized by year, that a facility should maintain
includes

ï¿½ previous years' reports,

ï¿½ engineering calculations and worksheets,

ï¿½ purchase records from suppliers,

ï¿½ inventory data,

ï¿½ monitoring records and flowmeter data,

ï¿½ pretreatment reports filed with local governments,

ï¿½ invoices from waste management companies,

ï¿½ manufacturers' estimates of treatment efficiencies,

ï¿½ process diagrams that indicate emissions and other releases,

ï¿½ EPA permits and monitoring reports,

ï¿½ hazardous waste generator's reports and manifests under the Resource
Conservation and Recovery Act, and

ï¿½ records for those EPCRA Section 313 chemicals for which they did not file
TRI reports.

However, section 313( b) of EPCRA states that facilities “may use
readily available data (including monitoring data) collected pursuant to
other provisions of law, or, where such data are not readily available,
reasonable estimates of the amounts involved.” In 1994, EPA
established a less burdensome substitute for reporting Form R information
based on an alternate threshold for facilities with low amounts of EPCRA
Section 313 chemicals in waste. However, EPA said that the Form A was not
applicable to the proposed lead rule.

The first objective of our review was to describe the guidance that OPPTS
and EPA's other major program offices have used during the past 10 years to
determine whether their proposed rules could be certified as not having a
SEISNSE. To address this objective, we obtained and reviewed the three sets
of guidance documents that EPA had in effect at different times during those
10 years. We also obtained explanations from EPA officials of how the
guidance documents were used, why the guidance was changed, and whether the
current guidance would be changed.

Our second objective was to compare the rate at which OPPTS certified that
its substantive proposed rules published during calendar years 1994 through
1999 would not have a SEISNSE with the rates in EPA's other major program
offices. To address this objective, we first developed a list of proposed
rules published by each of the four major program offices from January 1,
1994, through December 31, 1999. We identified the Objectives, Scope, and

Methodology

B- 284265 Page 9 GAO/ GGD- 00- 193 Regulatory Flexibility Act

proposed rules for this list by searching the LEXIS- NEXIS database of the
Federal Register for any notices involving proposed rules issued by these
program offices during this period. EPA officials also identified some
proposed rules that had not been identified by LEXIS- NEXIS. We excluded
from our list of proposed rules a number of actions that were not
substantive proposed rules. For example, we excluded advance notices of
proposed rulemakings and notices that primarily announced meetings or the
availability of data. We included proposals to add, streamline, or amend
existing regulations and earlier proposed rules. We then reviewed the
content of the Federal Register notices for each of the substantive proposed
rules that we identified for the 1994 through 1999 period to determine
whether the relevant EPA program office certified the rule as not having a
SEISNSE. We considered a rule to be certified when the language in the
notice either specifically stated that the program office was certifying the
rule or more generally indicated that the rule would not have a SEISNSE.

Our third objective had multiple parts: (1) to describe key aspects of the
methodology that OPPTS used in the economic analysis for the proposed lead
rule that may have contributed to the Office's conclusion that the rule
would not have a SEISNSE; (2) to determine whether and, if so, how OPPTS had
changed its economic analysis since publication of the rule; and (3) to
determine whether additional data or analysis could have yielded a different
conclusion about the rule's impact on small entities. To address the first
part of this objective, we reviewed the economic analysis that OPPTS
prepared for the proposed lead rule and discussed the analysis with OPPTS
officials and staff. To address the second part of this objective, we
interviewed these officials and staff to determine how the economic analysis
had changed since the publication of the proposed rule and obtained draft
copies of documents summarizing those changes. However, neither OPPTS nor
OIRA officials would provide us with the draft final rule or the revised
economic analysis that OPPTS submitted to OIRA on July 13, 2000. An OIRA
official cited section 6( b)( 4)( D) of Executive Order 12866, which says
that OIRA must make available to the public all documents exchanged between
OIRA and the agency “( a) fter the regulatory action has been
published in the Federal Register or otherwise issued to the public, or
after the agency has announced its decision not to publish or issue the
regulatory action.” To address the third part of this objective, we
obtained and analyzed unpublished data from the 1997 Census of Manufactures
and reviewed OPPTS' initial economic analysis of the proposed lead rule and
a summary of the draft revised economic analysis. The specific methodology
and assumptions that we used in this part of our review are described later
in this report.

B- 284265 Page 10 GAO/ GGD- 00- 193 Regulatory Flexibility Act

We conducted this review from November 1999 until August 2000 in accordance
with generally accepted government auditing standards. We provided a draft
of this report to the OMB Director and the EPA Administrator for their
review and comments. OMB said they had no comments on the report. EPA's
comments are discussed at the end of this letter and reproduced in appendix
I.

Under the RFA, federal agencies have broad discretion regarding how key
terms in the act should be defined and how certification decisions should be
made. OPPTS and the other EPA program offices have adopted three very
different sets of guidance documents during the past 10 years to determine
if their proposed rules could be certified as not having a SEISNSE. The
current guidance was implemented in 1996 because of the enactment of SBREFA.
The guidance focuses on the magnitude of a rule's impact (e. g., estimated
annual compliance costs as a percentage of estimated annual revenues) and
the number and percent of affected small entities that are expected to
experience that impact. However, this guidance (and its predecessors) gives
agency personnel broad discretion in how it should be applied.

The RFA does not define what Congress meant by the terms “significant
economic impact” or “substantial number of small entities”
and does not give SBA, OMB, or any other entity the authority or
responsibility to define these terms governmentwide. Neither does the act
prescribe how federal agencies should determine whether a proposed or final
rule could be certified as not having a SEISNSE. Subsection 605( b) of the
act simply says that any certification must be published in the Federal
Register with the rule “along with a statement providing the factual
basis for such certification.” Therefore, agencies have broad
discretion to determine how these terms should be defined and how they
should reach certification decisions. EPA has used that discretion to
develop three different sets of RFA guidance documents.

EPA issued its first guidance document for assessing whether a proposed rule
could be certified as not having a SEISNSE in 1982- the year after the RFA
took effect. According to the guidance, a proposed rule was to be considered
to have a “significant economic impact” on small entities
whenever any of the following guidelines were satisfied:

ï¿½ The annual cost of complying with the proposed rule would increase the
total costs of production by more than 5 percent. EPA's Guidance for

RFA Certification Have Changed Over Time

EPA Has Used Three Different Sets of RFA Guidance Documents

EPA's 1982 RFA Guidance Document

B- 284265 Page 11 GAO/ GGD- 00- 193 Regulatory Flexibility Act

ï¿½ The annual cost of complying with the proposed rule as a percentage of
annual sales for small entities is at least 10 percent higher than
compliance costs as a percentage of sales for large entities.

ï¿½ Capital costs of compliance with the proposed rule represent a significant
portion of capital available to small entities, considering internal cash
flow plus external financing capabilities.

ï¿½ The requirements of the proposed rule are likely to result in closures of
small entities.

The 1982 guidance document generally defined “a substantial number of
small entities” as more than 20 percent of all small entities affected
by the proposed rule. Therefore, using the first of the “significant
economic impact” guidelines above, if the relevant program office
expected that a proposed rule would raise the cost of production by more
than 5 percent for more than 20 percent of the small entities affected by
the rule, the program office should not certify the rule and should conduct
an IRFA.

However, the 1982 guidance document gave the program offices substantial
discretion in how these guidelines should be applied. For example, the
guidance document said that the program office could consider other, similar
definitions for the term “significant economic impact” if the
above guidelines were not appropriate. It also said that the program office
should “use its best judgment on a case- by- case basis” in
deciding whether to certify proposed rules that do not strictly meet the
conditions shown in the guidance document.

This flexibility notwithstanding, EPA said that applying this guidance was
sometimes difficult and time consuming. Therefore, in April 1992, EPA issued
a new guidance document instructing agency program offices that they should
prepare an IRFA for any draft rule that the agency expected to have any
impact on any small entities. According to EPA, the change to what was later
referred to as the agency's “any/ any” policy was intended to
shift agency resources from determining whether a regulatory flexibility
analysis was required to “the more productive consideration of
regulatory options for small entities that are subject to the rule.”

The 1992 guidance document stated that program offices “have wide
latitude in determining the level of analysis that is appropriate for
individual rulemakings.” They also said that the offices could use
factors, such as “the quality and quantity of available data”
and “the severity of a rule's anticipated impacts on small
entities,” in deciding how much effort they needed to expend on the
analysis. In addition, the 1992 guidance gave EPA's 1992 “Any/
Any” RFA

Guidance Document

B- 284265 Page 12 GAO/ GGD- 00- 193 Regulatory Flexibility Act

program offices the option of continuing to follow the 1982 guidance for
rulemakings that had been initiated before the 1992 guidance was issued.

When the advocacy review panel requirements in SBREFA became effective in
June 1996, EPA instituted a new RFA policy that used the magnitude of the
proposed rule's economic impact (e. g., estimated annual cost of compliance
as a percentage of estimated annual revenues) and the number and percent of
small entities expected to experience that impact to help the program
offices determine whether rules could be certified. The new procedures were
published as written agency guidance in 1997.

According to EPA officials, the agency changed its RFA guidance because the
SBREFA requirement to convene an advocacy review panel for any proposed rule
that was not certified made the continuation of the any/ any policy too
costly and impractical. One EPA official also said the SBREFA provision
allowing judicial review of regulatory flexibility analyses meant that EPA
would have to make sure that all of its IRFAs could withstand judicial
scrutiny. Therefore, he said, EPA no longer had the option of doing a
limited analysis on rules that it believed would have a minimal impact on
small entities. The 1997 guidance document stated that the new approach
would allow EPA to “manage its scarce resources such that the Agency
can continue considering the potential small entity impacts of all its rules
while preparing full regulatory flexibility analyses for those rules
warranting such analyses under the RFA.”

The 1997 guidance document was updated in 1999, but the numerical guidelines
relating to the certification of rules as not having a SEISNSE were
essentially unchanged. However, the 1999 guidance document emphasized that
it “is not a binding agency procedural rule. In determining and
mitigating impacts on small entities, we anticipate that there will be many
situations in which agency staff and management must exercise considerable
judgment.” It also explicitly continued EPA's policy “that
program offices should assess the direct impact of every rule on small
entities and minimize any adverse impact to the extent feasible, regardless
of the magnitude of the impact or the number of small entities
affected.” Nevertheless, the guidance document said that it was
intended to provide agency staff and management with an “analytic and
sequential structure that should be sufficient for most covered
rulemakings.”

As figure 1 illustrates, EPA's current guidance identifies different mixes
of economic impacts and the number and percent of affected small entities.
All of these factors are intended to help program offices determine when to
certify a rule as not having a SEISNSE, when not to certify a rule, and
EPA's 1997 and 1999 RFA

Guidance Documents

B- 284265 Page 13 GAO/ GGD- 00- 193 Regulatory Flexibility Act

when to seek further guidance from EPA's small business advocacy chair (who
is responsible for implementing the agency's SBREFA small business advocacy
panel requirements).

Compliance costs as a percent of revenues of small entities

Number of small entities affected

Percent of small entities afftected

Program office's presumption about proposed rule

+ += Less than 1 percent

Any percent Any number

3 percent or more 1 percent or more

Less than 100 100 to 999 Less than 100

Any percent Less than 20 percent

Any percent Can certify rule

as having no SEISNSE (presumed eligible for certification)

1 percent or more 3 percent or more

1,000 or more 100 to 999

100 to 999 Must consult

EPA's small business advocacy chair in order to certify rule (no
presumption)

20 percent or more

Any percent Must prepare

IRFA analysis and convene a panel (presumed ineligible for certification) 20
percent or

more Less than 20 percent Any percent

3 percent or more 1,000 or more 100 to 999

Source: GAO analysis of EPA's 1999 Revised Interim Guidance on the RFA as
amended by SBREFA.

To illustrate how these numerical guidelines are applied, if a proposed rule
is expected to affect 1,000 small entities, and for 200 of those small
entities the rule's compliance costs represent 3 percent of their annual
sales, the program office should presume that the rule cannot be certified.
On the other hand, if the rule affects 10,000 small entities but their
compliance costs never rise above 1 percent of their annual revenues, the
guidance

Figure 1: EPA Numerical Guidelines to Help Determine Whether a Proposed Rule
Has a SEISNSE

B- 284265 Page 14 GAO/ GGD- 00- 193 Regulatory Flexibility Act

indicates that the program office should presume that the rule can be
certified as not having a SEISNSE.

Although EPA's current guidance document provides specific numerical
guidelines that program offices should use in determining whether a proposed
rule should be certified under the RFA, the guidance also gives agency
program offices substantial discretion regarding their application. For
example, the guidance indicates that program offices should consider not
certifying a rule if “the extent of the impact (measured in
quantitative or qualitative terms) is particularly severe, even though the
number of affected small entities totals fewer than 100 or 20% of all
affected small entities.” The guidance also says that “as the
number of small entities that will be affected by a rule by more than 1% of
sales or revenues approaches 1000 in number, the substantial number
guidelines of 20% of affected small entities may become less relevant in
determining whether a regulatory flexibility analysis or a certification
should be prepared.”

An EPA official told us that the agency is currently reevaluating its RFA
guidance document because it is concerned about how the numerical guidelines
could be viewed outside of the agency. For example, he said, the guidance
makes it appear that EPA believes a rule that would cost 99 small entities
100 percent of their annual sales would not have a SEISNSE. He said that EPA
is examining the feasibility of using compliance costs as a percentage of
profit margins (instead of annual sales) to measure economic impacts on
small businesses. However, he noted that EPA has not used profit margins to
measure the economic impacts of its rules in the past because valid data on
companies' profit margins are often not readily available.

OPPTS certified 86 percent of the substantive proposed rules that it
published during calendar years 1994 through 1999- about the same rate of
rule certification as EPA's three other major program offices. The RFA
certification rate in OPPTS and in each of the three other major program
offices increased after the implementation of SBREFA in 1996. By the end of
1999, OPPTS and one other program office had certified all of their
postSBREFA proposed rules as not having a SEISNSE. Two other EPA program
offices used the discretion afforded them in the guidance and did not
certify at least four proposed rules published during this period that the
numerical guidelines in EPA's guidance document would have allowed them to
certify. Guidance Document Provides

Discretion in How Numerical Guidelines Should be Applied

EPA's RFA Guidance May Be Revised Again

RFA Certification Rates Increased After SBREFA Implementation

B- 284265 Page 15 GAO/ GGD- 00- 193 Regulatory Flexibility Act

EPA's four major program offices published at least 654 substantive proposed
rules in the Federal Register during calendar years 1994 through 1999. Of
these, 249 (38 percent) were tolerance actions or significant new use rules
(SNURs) issued by OPPTS. Tolerance actions are rules that establish maximum
levels for the amount of a pesticide residue that is allowed on food or feed
crops, or establish an exemption from the statutory requirement for a
tolerance. SNURs require anyone who intends to manufacture, import, or
process a chemical substance in a manner designated in the SNUR as a
“significant new use” to notify EPA before beginning those
actions. OPPTS said that these types of rules are unlikely to have adverse
economic impacts on entities (and may have positive economic impacts)
because tolerance actions allow, instead of prohibit, the use of pesticides,
and SNURs apply only when and if someone decides to engage in a significant
new use. Both of these types of actions are issued on a chemical- by-
chemical basis or for groups of chemicals. Because the impacts of each type
of action are basically the same from chemical to chemical, EPA conducted a
general analysis for each type and established, in consultation with SBA's
Office of Advocacy, a generic certification for both types of actions. 14

We excluded these 249 tolerance actions and SNURs from our analysis because
we wanted to include only proposed rules that possibly could have had a
SEISNSE. We also excluded nine rules from the Office of Solid Waste that
proposed changes to EPA's National Priorities List for Uncontrolled
Hazardous Waste Sites. EPA said that any economic impact that results from
these rules would occur only indirectly through enforcement and cost-
recovery actions, and case law indicates that the RFA applies only to direct
regulatory impacts. 15 Therefore, our analysis focused on the remaining 396
substantive proposed rules that the 4 major program offices published during
calendar years 1994 though 1999.

Table 1 shows the number and percent of these substantive rules that EPA's
four major program offices proposed and certified between January 1, 1994,
and the implementation of SBREFA (June 28, 1996); between implementation and
December 31, 1999; and for the full 1994 through 1999 period.

14 For the blanket certifications for tolerance actions and SNURs, see 46
Fed. Reg. 24950 (1981) and 62 Fed. Reg. 29684 (1997), respectively. 15 For a
discussion of these cases, see Clean Water Act: Proposed Revisions to EPA
Regulations to Clean Up Polluted Waters (GAO/ RCED- 00- 206R, June 21,
2000). All EPA Program Offices

Certified Most of Their Proposed Rules

B- 284265 Page 16 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Pre- SBREFA (Jan. 1, 1994 - June 27, 1996)

Post- SBREFA (June 28, 1996 - Dec. 31, 1999)

Total (Jan. 1, 1994 – Dec. 31, 1999) Number of rules Number of rules
Number of rules

EPA program office Proposed Certified Percentage

of rules certified Proposed Certified

Percentage of rules certified Proposed Certified

Percentage of rules certified

OPPTS 30 22 73 % 26 26 100 % 56 48 86 % Office of Air and Radiation

82 70 85 151 147 97 233 217 93 Office of Solid Waste 27 20 74 21 21 100 48
41 85 Office of Water 19 12 63 40 35 88 59 47 80

Total 158 124 78 238 229 96 396 353 89

Source: GAO analysis of EPA's proposed rules for calendar years 1994 through
1999.

Overall, the four program offices certified the vast majority (89 percent)
of the proposed rules that they published during this period. OPPTS' rate of
RFA certification throughout this period (86 percent) was within the range
of certification rates in the three other EPA program offices (80 percent to
93 percent). Notably, the rate of RFA certification in all four program
offices was substantially higher during the post- SBREFA period (96 percent)
than before the act took effect (78 percent). 16 OPPTS and the Office of
Solid Waste certified all of their proposed rules after the implementation
of SBREFA. EPA officials told us that the change in the agency's
certification guidelines (from the “any/ any” policy to the
policy adopted in 1996) led to the increase in the frequency of RFA
certification from what they described as the “artificially low”
levels during the preSBREFA period.

Two of EPA's program offices did not certify (i. e., they conducted an IRFA
and held an advocacy review panel) 9 of the 191 proposed rules that they
published during the post- SBREFA period of our review. 17 The Office of
Water issued five of these nine rules, and the Office of Air and Radiation
issued four of the rules. For at least 4 of these 9 rules, the relevant
program

16 Our review of the NPRMs indicated that the program offices made RFA
determinations for at least 30 of the proposed rules published before SBREFA
using EPA's 1982 guidelines rather than the 1992 guidelines. As noted
previously, EPA's 1992 guidelines expressly allowed the use of this earlier
guidance in rules that were under development before 1992.

17 The Office of Air and Radiation and the Office of Water convened other
advocacy review panels during this period but had either not published a
proposed rule by the end of 1999 or certified that the rules would not have
a SEISNSE.

Table 1: Number and percentage of proposed rules certified as having no
SEISNSE by four EPA program offices between 1994 and 1999

EPA Program Offices Could Have Certified More PostSBREFA Rules

B- 284265 Page 17 GAO/ GGD- 00- 193 Regulatory Flexibility Act

offices estimated that fewer than 100 small entities would experience
compliance costs of 1 or 3 percent of their annual revenues. 18

ï¿½ The Office of Air and Radiation published two proposed rules on October
21, 1998, dealing with the transport of ozone. 19 The Office estimated that
one of these rules would impose at least a 1 percent economic impact on 42
of the 153 affected small entities, and 22 small entities would experience
economic impacts of greater than 3 percent. The Office of Air and Radiation
estimated that the other ozone transport rule would impose economic impacts
of at least 1 percent on 39 of the 145 affected small entities, and 22 would
experience greater than 3 percent economic impacts.

ï¿½ The Office of Water published a proposed rule on June 25, 1998, dealing
with transportation equipment cleaning. 20 The Office estimated that the
rule would impose at least a 1 percent economic impact on 75 of the 87
affected small entities, and 64 of these small entities would experience at
least a 3 percent economic impact. The Office of Water published another
rule on January 13, 1999, dealing with centralized waste treatment. 21 The
Office estimated that the rule would impose at least a 1 percent economic
impact on 45 of the 63 affected small entities, and 23 of them would
experience at least a 3- percent impact.

Applying EPA's numerical guidelines (1- percent or 3- percent impacts on
fewer than 100 small entities) to each of these rules, the Office of Air and
Radiation and the Office of Water could have presumed that the rules would
not have a SEISNSE. Instead, the program offices used the discretion
afforded them in the guidance documents, decided not to certify the rules,
prepared IRFAs for the rules, and convened advocacy review panels. The
preambles to the proposed rules did not indicate why these offices decided
not to certify the rules, and the RFA does not require the offices to
provide these explanations. However, EPA officials told us during this
review that in making their certification decisions, the offices considered
factors other than just the number of companies expected to

18 One of the other five proposed rules met the numerical guidelines for
being presumed ineligible for certification, and two other rules had impacts
that fell into the guidelines' middle category- no presumption about whether
they should be certified as not having a SEISNSE. The Federal Register
notices for two rules did not contain enough information about the impacts
of the proposed rules for us to determine whether the rules met the
numerical guidelines for being presumed eligible for certification.

19 63 Fed. Reg. 56292 (1998) and 63 Fed. Reg. 56394 (1998). 20 63 Fed. Reg.
34686 (1998). 21 64 Fed. Reg. 2280 (1999).

B- 284265 Page 18 GAO/ GGD- 00- 193 Regulatory Flexibility Act

experience 1 percent and 3 percent economic impacts. For example, in the
centralized waste treatment rule, EPA said that the range of impacts ran so
high above 3 percent for a number of firms that the office considered the
size of the impact more important than just the number of small firms
affected.

OPPTS made a number of assumptions in the original economic analysis for the
proposed lead rule. As a result of those assumptions, OPPTS estimated that
small manufacturers had annual revenues of $4 million. These and other
assumptions and methods led OPPTS to conclude that although the rule would
affect more than 5,600 small businesses, it would not have a SEISNSE. OPPTS
recently revised some of the assumptions in its economic analysis, estimated
that the rule would affect more than 8,600 small companies, and said that as
many as 464 of them would experience first- year compliance costs of between
1 and 3 percent of their annual revenues. Nevertheless, OPPTS again
concluded that the rule would not have a SEISNSE. Using data that we
obtained from the Bureau of the Census, we estimated that as many as 1,098
additional small manufacturing companies could experience compliance costs
of at least 1 percent of their annual revenues, and as many as 78 small
companies could experience a 3 percent economic impact. If OPPTS had used
this analytic approach and applied the flexibility afforded in EPA's RFA
guidance, it could have chosen not to certify the rule and conducted an
advocacy review panel. However, the Office's initial and draft revised
analyses and the conclusions that it based on those studies were within the
discretion permitted agencies under the RFA and the EPA guidance.

In April 1998, the Vice President announced three initiatives designed to
expand the amount of information available to the public about the health
effects of chemicals. In one of those initiatives, he said that some
chemicals accumulate in our bodies and that some substances linked to
significant public health concerns were not subject to TRI reporting or are
reported only at levels far exceeding those linked to health effects.
Therefore, he said that EPA would review these “persistent
bioaccumulative toxics” (PBTs) and determine if they should be subject
to TRI reporting or lower reporting thresholds. The Vice President said any
regulatory changes would be finalized by December 1999 and would be
“fashioned in a way that minimizes cost and other burdens on
business.”

In January 1999, OPPTS published an NPRM adding a number of PBT chemicals to
the TRI program and lowering the reporting thresholds for Different
Assumptions,

Data, and Analysis Yield Different Estimates of Lead Rule's Effects

Development of the Proposed Lead Rule

B- 284265 Page 19 GAO/ GGD- 00- 193 Regulatory Flexibility Act

other PBT chemicals already covered by the TRI program. 22 According to
OPPTS, the rule had an expedited development schedule in order to meet the
Vice President's committed final issuance date. OPPTS said that it received
more than 35,000 public comments on its January 1999 PBT proposal, many of
which requested that the Office include lead and lead compounds as PBT
chemicals to be covered by the rule.

Rather than delay the issuance of the final PBT rule, OPPTS decided to
publish a separate proposed rule covering lead and lead compounds. At the
time of the proposal, a facility had to report under the TRI program if it
manufactured or processed 25,000 pounds of lead, or if it “otherwise
used” 10,000 pounds of lead. The NPRM, published in August 1999,
proposed lowering the TRI reporting threshold for lead and lead compounds to
10 pounds. During the development of the rule OPPTS also considered
reporting thresholds of 1,000 pounds, 100 pounds, and 1 pound.

A number of the public comments on the proposed lead rule questioned OPPTS'
conclusion that the rule would not have a SEISNSE. On November 15, 1999,
OPPTS announced three public meetings to obtain comments on issues relating
to the proposed lead rule and requested comments on specific issues related
to its RFA certification. OPPTS held those public meetings in November and
December 1999, at which several of the participants questioned the Office's
conclusions regarding the effect of the rule on small entities. On July 13,
2000, OPPTS sent the draft final lead rule and a revised economic analysis
to OIRA for final review under Executive Order 12866. OPPTS officials told
us that they hoped to publish the final rule in the Federal Register by the
end of October 2000.

The methodology that OPPTS used to conclude that the proposed lead rule
would not have a SEISNSE was extremely complicated, involving numerous steps
and a variety of assumptions within each step. However, the overall process
essentially involved (1) estimating the number of companies that would have
to file new TRI reports at the new 10 pound reporting threshold, (2)
estimating the cost of complying with the rule in the first year and in
subsequent years, and (3) estimating the revenues of the companies that
would have to file the new reports.

Data are generally not available on the amount of lead and lead compounds
that individual facilities and companies manufacture, process, or otherwise
use each year. Therefore, OPPTS used six different methods

22 64 Fed. Reg. 688 (1999). OPPTS published the final PBT rule on October
29, 1999, at 64 Fed. Reg. 58666 (1999). OPPTS' Economic Analysis

for the Proposed Lead Rule Estimating the Number of Companies Filing New
Reports

B- 284265 Page 20 GAO/ GGD- 00- 193 Regulatory Flexibility Act

to develop SIC code- specific estimates of the number of facilities and
companies that would have to file additional TRI reports for lead and lead
compounds. Different methods were used in different industries. For example,
in the “air emission factor” method, OPPTS used lead and lead
compound air emissions as a proxy for minimum lead and lead compound use in
the pulp mill, asphalt paving, iron foundry, and primary and secondary
metals smelting industries.

Each of these estimation methods involved a number of assumptions. For
example, in many of the methods, OPPTS used the number of employees to
approximate a distribution of lead use within an industry. OPPTS also
assumed that lead use was proportional to the cost of materials or the value
of shipments and that certain facilities in each industry manufacture,
process, or otherwise use lead in their operations. In three of the six
methods, OPPTS used lead emissions as a proxy for minimum lead use- a
process that OPPTS said in the economic analysis underestimates the total
amount of lead used by a facility. However, OPPTS told us during this review
that in practice, this method generally does not result in underestimates of
the number of facilities that will use 10 pounds of lead. They said that
affected facilities generally appear to emit more than 10 pounds of lead, so
their lead use would be above that level.

As a result of these methods and assumptions, OPPTS estimated that 15,043
facilities from 27 SIC codes and SIC code groupings would have to file
additional TRI reports at the proposed 10 pound reporting threshold- a
nearly 800- percent increase from the 1,902 facilities that reported for
lead and lead compounds in 1998 at the 25,000 and 10,000- pound thresholds.
23 OPPTS also estimated that these 15,043 facilities were in 8,175 parent
companies, of which 6,874 (84 percent) were manufacturers (SIC codes 20
through 39). (See fig. 2.) The other 1,301 companies that OPPTS said would
be affected by the rule were in 7 other SIC codes.

23 Twenty- four individual SIC codes were at the 2, 3, and 4- digit levels,
and three SIC code groupings were for coal, oil, and wood- fired combustion
in SIC codes 20 through 39.

B- 284265 Page 21 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Note 1: OPPTS estimated that a total of 8, 175 parent companies would be
affected by the proposed lead rule at the 10 pound reporting threshold.

Note 2: The petroleum bulk stations and terminals sector is SIC code 5171,
and coal mining is SIC code 12. “Other” includes electric
services (SIC code 4911), electric and other services combined (SIC code
4931), combination utilities not elsewhere classified (SIC code 4939),
refuse systems (SIC code 4953), and business services not elsewhere
classified (SIC code 7389) .

Source: GAO analysis.

OPPTS also estimated that 5,620 (69 percent) of the 8,175 affected companies
were small companies. Several of the public comments that OPPTS received on
the proposed rule maintained that the Office's estimates of the number of
affected facilities and companies, and the number of small entities
affected, were too low.

In addition to the 15,043 new TRI reports that OPPTS estimated would come
from companies in these 27 SIC codes and code groupings, the Office
identified 42 other 4- digit SIC codes in the economic analysis that it said
might also be affected by the proposed rule. Forty of these 42 SIC codes
each had more than 5 facilities reporting lead or lead compounds at the
current thresholds, which OPPTS said “may indicate that additional
facilities in these SIC codes use lead or lead compounds at levels below
current thresholds but above the proposed thresholds.” 24 However,
OPPTS

24 The other two SIC codes were listed on the basis of lead consumption data
from the U. S. Geological Survey (SIC code 3443- fabricated plate work) or
because of the sector's potential for processing chemical products that
contain lead as a trace constituent (SIC code 5169- wholesale distribution
of chemicals).

Figure 2: Most of the Parent Companies Affected by the Proposed Lead Rule
Were Manufacturers

B- 284265 Page 22 GAO/ GGD- 00- 193 Regulatory Flexibility Act

did not estimate the number of additional lead TRI reports that it expected
to be filed from facilities in these 42 SIC codes because of “lack of
data on lead consumption or emissions at the facility and sector
level.”

OPPTS developed its estimate of the average annual cost of filing a TRI
report for lead and lead compounds by multiplying an estimate of the average
number of hours needed to complete the form (111 hours in the first year for
new TRI filers) times an estimated average hourly wage rate (about $69 per
hour across managerial, technical, and clerical occupational categories). 25
As a result, OPPTS estimated that first- year compliance costs for small
manufacturers who had never filed a TRI report would be about $7,500. 26
OPPTS estimated that first- year reporting costs for companies currently
filing TRI reports for other substances would be about $5,200, and that lead
reporting costs in subsequent years would be about $3,600.

OPPTS used a variety of methods and assumptions to estimate the annual
revenues of the companies that would have to file the new TRI reports for
lead and lead compounds. However, three of these methods and assumptions
appeared to be key to this portion of the analysis and were the subject of
several of the public comments. 27

ï¿½ OPPTS assumed that small manufacturing companies filing their first TRI
reports under the proposed lead rule would be similar in terms of revenues
and employment to small manufacturing companies that already filed TRI
reports under other rules. OPPTS officials said that it used this assumption
because (1) many small businesses that were expected to file under the
proposed lead rule already filed other TRI reports; and (2) small businesses
that did not already file TRI reports were likely to come from capital-
intensive industries, which were the types of industries that already filed
TRI reports. Therefore, OPPTS constructed revenue profiles for these new
filers using revenue data from parent companies of existing filers who were
manufacturers. Some of the comments on the proposed rule

25 OPPTS estimated that for first time TRI filers, the paperwork burden
associated with the first year of reporting for the PBT rule would be nearly
140 hours. 26 In the original economic analysis for the proposed lead rule,
OPPTS estimated small manufacturers' average first- year compliance costs at
about $7, 400, assuming that parent companies have some facilities that are
current filers and some that are first- time filers. In its draft revised
analysis, OPPTS estimated those costs at $7, 700, assuming that the parent
company has one facility that is a first- time filer. To simplify the
presentation of this report, we used a uniform $7, 500 reporting figure.

27 OPPTS used similar assumptions in its economic analysis of the PBT rule
and in earlier TRI rules. However, EPA officials said that the proposed lead
rule was the first time EPA received significant adverse comments on these
issues. Estimating the Cost of

Complying With the Proposed Rule

Estimating Company Revenues

B- 284265 Page 23 GAO/ GGD- 00- 193 Regulatory Flexibility Act

indicated that businesses that would have to report under the new 10pound
threshold were likely to be smaller than those filing for lead and other
substances at the 25,000 and 10,000 pound reporting thresholds. As a result,
they suggested that OPPTS had overestimated the annual revenues of new
filers under the proposed lead rule and that OPPTS' certification that the
rule would not have a SEISNSE may be incorrect.

ï¿½ As noted in figure 2, OPPTS estimated that 84 percent of all companies
affected by the proposed lead rule would be manufacturers (SIC codes 2039),
with the other 16 percent representing seven nonmanufacturing industries.
Although OPPTS estimated the impact of the rule on each of these seven
nonmanufacturing groups separately, it did not examine whether the rule
would have different impacts on the different types of manufacturers that
were expected to comprise the bulk of new filers. Some of the comments that
were submitted to OPPTS on the proposed rule said that the various
industries within the 20 major groups in the manufacturing sector are
unrelated in most aspects and, therefore, should not have been analyzed as a
group. Doing so, they said, may mask variation in small business impacts
within the manufacturing industries.

ï¿½ OPPTS modeled the revenues of affected small businesses on the basis of
revenue quartiles for the manufacturing industries. To do so, OPPTS first
arrayed from lowest to highest the revenues of all the small manufacturing
companies currently filing TRI reports. It then selected the revenues of
companies at the 25th, 50th, and 75th percentiles to represent the revenues
of companies in the lowest third, middle third, and highest third revenue
level, respectively. OPPTS officials said that prior economic analyses had
used only one revenue level- averages or medians- to represent the revenues
of expected TRI filers, and using three revenue levels was intended to
better reflect the variation of revenues in the group. Nevertheless, in the
November 15, 1999, announcement of the public meetings, OPPTS specifically
asked for comments on whether the 25 th revenue quartile was an appropriate
revenue level for considering the potential impacts on the smallest of the
small businesses affected by the proposed rule.

As a result of these and other methods and assumptions, OPPTS estimated that
small manufacturing companies filing new TRI reports for lead would have
annual revenues of $4 million. To determine whether the economic impact of
the proposed lead rule on small businesses was significant, OPPTS divided
the average annual cost of filing a TRI report (e. g., about $7,500 for
first- time TRI filers in the manufacturing sector) by the estimated annual
revenue level of filers (e. g., $4 million for the smallest manufacturers).
The results of these calculations (0.2 percent for small manufacturers)
caused OPPTS to conclude that no small entities would

B- 284265 Page 24 GAO/ GGD- 00- 193 Regulatory Flexibility Act

experience a 1 percent economic impact from the rule. Using this information
in concert with EPA's RFA guidance, OPPTS certified the rule as not having a
SEISNSE.

OPPTS revised its economic analysis for the proposed lead rule earlier this
year in response to the concerns raised in the public comments. In June
2000, we obtained a summary of the changes that OPPTS made in a draft of the
revised analysis and a summary of the results of the draft revised analysis.
On July 13, 2000, OPPTS sent the draft final lead rule and the revised
economic analysis to OIRA for final review. OPPTS officials said they could
not provide us with a copy of the economic analysis that they sent to OIRA
because of restrictions on the distribution of documents related to a rule
undergoing final review. OIRA officials also said that, under Executive
Order 12866, they could not provide us with a copy of the revised economic
analysis until the final rule was published in the Federal Register. OPPTS
officials said that the results of the analysis that they provided to OIRA
were only marginally different from the summary that they provided to us,
but they cautioned that the results may change due to OIRA's review.

According to the summary of the draft revised analysis, OPPTS changed its
analytic approach in a number of ways. For example, to determine the number
of affected facilities, OPPTS contacted more trade organizations and
potentially affected businesses and used additional information sources that
had been identified as a result of public comments received on the proposed
rule.

OPPTS also changed several of the key assumptions that it used in the
initial analysis to estimate the annual revenues of filers. Those changes
included the following:

ï¿½ OPPTS did not assume that new filers under the lead rule would be similar
to current TRI filers in terms of revenues and employment. Instead, OPPTS
used the revenues of companies that did not already file under the TRI
program to characterize the revenues of small manufacturing companies that
would be required to file a TRI report for the first time under the proposed
lead rule.

ï¿½ OPPTS examined whether the rule would have different impacts on different
types of manufacturers. Specifically, OPPTS developed revenue profiles for
each of the 20 2- digit SIC code groupings within the manufacturing category
(SIC codes 20 through 39) and used those SIC code- specific values to
estimate the rule's impact. OPPTS' Draft Revised

Economic Analysis for the Lead Rule

OPPTS Changed Key Assumptions to Estimate Companies' Annual Revenues

B- 284265 Page 25 GAO/ GGD- 00- 193 Regulatory Flexibility Act

ï¿½ OPPTS used more revenue levels from the population of representative small
entities to characterize the revenues of small manufacturing companies that
will have to file under the lead rule. For the draft revised analysis, OPPTS
used every 10th percentile level (10th through 90th) and the 95th percentile
to represent 10 different levels of revenues. For example, it used the
revenues at the 10th percentile to represent the revenues of the bottom 10
percent of the companies expected to file.

Changes in these and other methods and assumptions resulted in very
different data being used in the revised economic analysis. As previously
noted, in the original economic analysis, OPPTS estimated that the smallest
manufacturers (represented by companies at the 25th revenue percentile of
current small manufacturing TRI reporters) had annual revenues of $4
million. However, as table 2 illustrates, revenues of small manufacturing
companies not currently reporting were quite different from the revenues of
those companies currently reporting at the 25th revenue percentile, and even
more different at the 10th revenue percentile. There were also substantial
differences in annual revenues among the different SIC codes within the
manufacturing sector.

B- 284265 Page 26 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Annual revenues of small manufacturing companies SIC codes

Reporting to TRI at the 25 th revenue

percentile Not reporting to TRI

at the 25 th revenue percentile

Not reporting to TRI at the 10 th revenue

percentile

20 15,721,258 1,156,884 620,000 21 a 822,512 470,000 22 8,542,745 1,071,000
688,823 23 12,541,717 871,328 500,000 24 6,000,000 850,000 560,000 25
9,238,569 970,181 610,000 26 15,000,000 1,700,000 1,000,000 27 6,698,920
790,000 500,000 28 4,914,041 1,500,000 1,000,000 29 5,800,000 2,400,000
1,400,000 30 2,367,047 1,300,000 800,000 31 8,200,000 790,000 500,000 32
6,000,000 1,000,000 680,000 33 5,000,000 1,300,000 830,000 34 3,300,000
1,000,000 710,000 35 9,199,323 1,000,000 680,000 36 5,950,000 1,300,000
810,000 37 7,000,000 1,200,000 800,000 38 6,000,000 1,200,000 800,000 39
4,850,000 870,000 570,000 a Insufficient data.

Source: OPPTS.

In the draft revised economic analysis, OPPTS estimated that the lead rule
would affect 14,586 facilities in 10,133 companies. As table 3 illustrates,
OPPTS said that 8,637 small companies- up from 5, 620 in the initial
analysis- would be affected by the proposed rule. OPPTS also said that as
many as 464 of those small companies (about 5 percent of all affected small
companies) could experience compliance costs of at least 1 percent of their
annual revenues in the first year of implementation-- up from zero in the
initial analysis. 28 However, OPPTS said that no small company would
experience compliance costs of 3 percent. After evaluation of the revised
economic analysis, OPPTS officials said that they still believed the rule
would not have a SEISNSE, and therefore they planned to certify that the
final rule would not have a SEISNSE. 28 OPPTS said that 273 (59 percent) of
these 464 companies were in SIC Code 36 (electrical equipment).

Table 2: Differences in Data Used in OPPTS Economic Analyses of Revenues of
Small Manufacturing Companies

OPPTS Revised Its Estimates of Rule's Impact

B- 284265 Page 27 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Number of small companies affected

Number of small companies with economic impacts of at least 1 percent of
revenues

in the first year b Type of affected

small companies Initial analysis Draft revised

analysis Initial analysis Draft revised analysis

Manufacturing 4,673 7,952 0 458 Other a 947 685 0 6

Total 5,620 8,637 0 464

a Other small companies include metal mining, coal mining, electrical
utilities that combust coal and/ or oil, hazardous waste treatment and
disposal facilities, chemicals and allied products wholesale distributors,
petroleum bulk plants and terminals, and solvent recovery services. b In
both the initial and draft revised economic analyses, OPPTS estimated that
no companies would

experience economic impacts of 1 percent of annual revenues in subsequent
years of the rule's implementation.

Source: OPPTS.

OPPTS officials and staff said that their revised estimates of the number of
small entities that would experience at least a 1 percent economic impact
from the lead rule were probably higher than the number of entities that
would actually experience those impacts. In particular, they said that
firstyear compliance costs in a small company could reasonably be expected
to be less than the cost of reporting at a larger facility because (1)
reporting in the smallest facilities may take only half as long as the
average of all facilities (111 hours), and (2) wage rates in those
facilities may be less than the $69 per hour average used. Therefore, OPPTS
estimated that the firstyear cost of compliance in small firms could be
$3,360 (56 hours times $60 per hour)- well below the $7,500 estimate for all
companies that was used to estimate the impacts on affected small entities.
OPPTS officials and staff also noted that on the basis of a review of actual
reporting, their estimates of the number of additional reports and affected
facilities for previous TRI rules had been overestimates.

The 464 small companies that OPPTS indicated in its draft revised economic
analysis could experience compliance costs of at least 1 percent of their
annual revenues were primarily in the 27 SIC codes and code groupings for
which OPPTS developed estimates in the original economic analysis. In that
analysis, OPPTS said that companies in 42 other 4- digit SIC codes could
also be affected by the proposed rule, but lack of data prevented it from
estimating how many additional TRI reports for lead would be filed by
companies in those SIC codes. As previously noted, 40 of these 42 SIC codes
each had more than 5 facilities reporting lead or lead compounds at the
current thresholds. In the revised analysis, OPPTS developed estimates for 9
of these 42 SIC codes on the basis of information

Table 3: Comparison of the Results of OPPTS' Initial and Draft Revised
Economic Analyses of the Proposed Lead Rule

Hundreds of Other Small Manufacturers Could Experience a 1 Percent Economic
Impact From the Lead Rule

B- 284265 Page 28 GAO/ GGD- 00- 193 Regulatory Flexibility Act

received during the public comment period. 29 Therefore, OPPTS did not
estimate the economic impact of the rule on companies in 33 SIC codes that
it said could be affected.

All but 1 of the 33 SIC codes in which OPPTS did not estimate the lead
rule's economic effects were in SIC codes 20 through 39- manufacturing. We
attempted to provide a more complete picture of how the lead rule might
affect small manufacturing companies by estimating how many companies in
these 32 manufacturing SIC codes could experience a firstyear economic
impact of at least 1 percent of revenues. To do so, we first obtained
unpublished data from the Bureau of the Census' 1997 Economic Census of
Manufactures on the number of companies in SIC codes 20 through 39 that had
10 to 499 employees and annual revenues (shipments) of $750,000 or less. 30
We used the 10 to 499 employee size limitation because the TRI program
applies only to facilities with at least 10 full- time employees, and any
company with fewer than 500 employees is considered “small”
according to SBA's size categories. 31 We selected the $750,000 revenue
threshold because (1) OPPTS estimated that the proposed lead rule would
impose average first- year compliance costs for new TRI filers of about
$7,500, and (2) EPA's current RFA guidance indicates that a rule that
imposes compliance costs of less than 1 percent of annual revenues on any
number of companies should be presumed not to have a SEISNSE. In order for a
company's cost of compliance with the lead rule in the first year to
represent 1 percent of its revenues, the company would have to have annual
revenues of $750,000 or less.

The Bureau of the Census data indicated that nearly 18,000 manufacturing
companies had 10 to 499 employees and annual revenues of $750,000 or less.
32 Of these, 1,108 were in the 32 manufacturing SIC codes that OPPTS said
might be affected by the rule but for which it did not provide an estimate
of how many companies could be affected. OPPTS told us that 10 companies in
the 32 SIC codes with 10 to 499 employees and annual

29 OPPTS concluded that 85 small companies in these 9 SIC codes would
experience a first- year economic impact of at least 1 percent of revenues.
30 The Economic Census of Manufactures is conducted every 5 years on the
universe of businesses in the manufacturing industry. It is to collect facts
about the structure and functioning of the economy and features unique to
this industry, such as materials consumed, inventories held, the number of
establishments, payroll, and geographic location.

31 The SBA size standard for “small” businesses is 500 for about
75 percent of manufacturing industries; and either 750, 1, 000, or 1, 500
for the other industries. Therefore, the use of the 500- employee limit is a
conservative measure of the number of small manufacturing companies.

32 About 99 percent of the nearly 18, 000 companies had fewer than 50
employees, and about 84 percent had fewer than 20 employees.

B- 284265 Page 29 GAO/ GGD- 00- 193 Regulatory Flexibility Act

revenues of $750,000 or less filed TRI reports in 1998. As previously noted,
OPPTS estimated that first- year compliance costs for companies that already
filed reports for some other TRI- covered substance would be about $5,200.
Therefore, these companies would have to have annual revenues of $520,000 or
less in order for the proposed lead rule to have a 1 percent economic impact
on them in the first year. Although some of the 1,108 companies may have had
annual revenues that low, we eliminated 10 companies from the total to
ensure that all of the resultant 1,098 companies were first- time TRI
reporters. 33

Two other adjustments to the data may also be necessary before the rule's
effects on small entities are estimated.

ï¿½ As implemented by OPPTS, the TRI program covers companies with 10 or more
“full- time equivalent” (FTE) employees, not just 10 employees.
34 Although the Bureau of the Census data reflects companies with at least
10 employees, some of those employees could be part- time workers. As a
result, the number of employees in some of the smaller firms could fall
below the 10 FTE threshold for the TRI program. For example, if a company
had 12 employees but 6 of them worked only 20 hours per week, the company
would have only 9 FTE employees. In order to account for this possibility,
we also developed estimates of the number of companies that could experience
a 1- percent impact assuming that three- quarters, one- half, and one-
quarter of the companies in the 32 SIC codes met the 10 FTE threshold.

ï¿½ Although OPPTS indicated that the companies in these 32 SIC codes may use
lead in their manufacturing processes and could be affected by the proposed
rule, it is not clear how many of these companies use at least 10 pounds of
lead each year- the proposed reporting threshold. We were unable to locate
any data identifying the amount of lead used by each company or within each
SIC code. In order to account for the possibility that some of the companies
in these 32 SIC codes use less than 10 pounds of lead each year, we
developed estimates of the number of companies that could experience a 1
percent economic impact assuming that annual lead use in three- quarters,
one- half, and one- quarter of the companies within these SIC codes met the
10- pound threshold.

33 Therefore, we assumed that all 33 TRI filers in these 32 SIC codes had
annual revenues of between $520, 000 and $750, 000. 34 EPCRA (the statute
creating the TRI program) states that the TRI program covers facilities with
10 or more full- time employees. OPPTS has defined the program as including
facilities with 10 FTEs, even though some facilities with 10 FTEs may not
have 10 full- time employees.

B- 284265 Page 30 GAO/ GGD- 00- 193 Regulatory Flexibility Act

The results of these possible adjustments to the Census data are presented
in table 4. If no adjustments are made to the Census data (i. e., assuming
that all of the companies in the 32 manufacturing SIC codes have 10 FTE
employees and all use 10 pounds of lead annually), then all 1,098 companies
in these SIC codes with 10 to 499 employees and annual revenues of $750,000
or less could experience a 1- percent impact from the lead rule. 35 On the
other hand, if we assume that half of the companies in these 32 SIC codes
have 10 FTE employees and that half of those companies use 10 pounds of lead
annually, we estimated that 275 additional companies could experience a 1-
percent impact in the first year of implementation.

All companies have 10 FTE

employees Three- quarters

of companies have 10 FTE

employees One- half of

companies have 10 FTE

employees One- quarter of

companies have 10 FTE

employees

All companies use 10 pounds of lead

1,098 824 549 275 Three- quarters of companies use 10 pounds of lead

824 618 412 206 One- half of companies use 10 pounds of lead

549 412 275 137 One- quarter of companies use 10 pounds of lead

275 206 137 69 Note: The estimates in this table were developed assuming no
interrelationships between companies' lead usage and their FTEs. The
estimates could be higher or lower depending on the strength or nature of
any relationships.

Source: GAO analysis of Bureau of the Census data.

We also used Bureau of the Census data to estimate the number of companies
in these 32 SIC codes that could experience first- year compliance costs
equal to 3 percent of their annual revenues. To do so, we focused on
companies with 10 to 499 employees and annual revenues of $250,000 or less.
The data indicated that 78 companies in the 32 SIC codes met that employment
and revenue profile. Because we could not presume that all 78 companies had
10 FTE employees or used 10 pounds of lead in the course of the year, we
again adjusted the data assuming that threequarters, one- half, and one-
quarter of the companies met those

35 Even this number could be an underestimate. As previously noted, EPA
estimated that 85 companies in 9 other SIC codes could experience 1 percent
economic impacts; however, the Bureau of the Census data indicated that up
to 544 companies in those 9 SIC codes could experience 1 percent economic
impacts.

Table 4: Estimated Number of Small Companies Experiencing a 1 Percent
Economic Impact From the Lead Rule in 32 Manufacturing SIC Codes During the
First Year of Implementation

B- 284265 Page 31 GAO/ GGD- 00- 193 Regulatory Flexibility Act

parameters. The results, presented in table 5, indicate that between 5 and
78 companies in these 32 SIC codes could experience a 3 percent economic
impact from the lead rule. 36

All companies have 10 FTE

employees Three- quarters

of companies have 10 FTE

employees One- half of

companies have 10 FTE

employees One- quarter of

companies have 10 FTE

employees

All companies use 10 pounds of lead 78 59 39 20 Three- quarters of companies
use 10 pounds of lead

59 44 29 15 One- half of companies use 10 pounds of lead

39 29 20 10 One- quarter of companies use 10 pounds of lead

20 15 10 5 Note: The estimates in this table were developed assuming no
interrelationships between companies' lead usage and their FTEs. The
estimates could be higher or lower depending on the strength or nature of
any relationships.

Source: GAO analysis of Bureau of the Census data.

Neither our estimates of the number of small companies that could be
affected by the rule nor the estimates that OPPTS developed take into
account behavioral changes that may occur as a result of the rule's
implementation. For example, a company might change its manufacturing
processes in order to reduce the amount of lead it manufactures, processes,
or otherwise uses below the 10 pound reporting threshold. Also, a company
might reduce its employment below the 10- FTE threshold and avoid the rule's
requirements. We also did not attempt to determine the effects of the rule
after the first year of implementation or to identify impacts above the 3
percent level.

The numerical guidelines in EPA's current RFA guidance document establish
what appears to be a high threshold for what constitutes a SEISNSE. For
example, an EPA rule can impose $10,000 in compliance costs on 10,000 small
businesses, but the guidelines indicate that the program office issuing the
rule can presume that the rule does not have a SEISNSE as long as those
costs do not represent at least 1 percent of the businesses' annual
revenues. The numerical guidelines also suggest that a

36 We again assumed that all 33 TRI filers in these 32 SIC codes had annual
revenues of between $520, 000 and $750, 000. Because table 5 only includes
companies with revenues of $250, 000 or less, no adjustments were made to
this table for current filers.

Table 5: Estimated Number of Small Companies Experiencing a 3 Percent
Economic Impact From the Lead Rule in 32 Manufacturing SIC Codes During the
First Year of Implementation

Conclusions

B- 284265 Page 32 GAO/ GGD- 00- 193 Regulatory Flexibility Act

rule can be presumed not to have a SEISNSE if as many as 99 small businesses
experience compliance costs amounting to 100 percent of their annual
revenues- essentially putting them out of business.

However, EPA's current RFA guidance document also gives the agency's program
offices substantial discretion regarding how those numerical guidelines
should be applied. The guidance suggests that those offices not certify
proposed rules in which the extent of a rule's expected impact is severe or
widespread, even if , according to the numerical guidelines, the rule can be
presumed not to have a SEISNSE. Some EPA program offices have used that
discretion and have conducted IRFAs and held advocacy review panels for
proposed rules that met the conditions in the numerical guidelines for being
presumed not to have a SEISNSE. However, by the end of 1999, OPPTS and the
Office of Solid Waste had certified all 47 of the proposed rules that they
published after the implementation of the SBREFA advocacy review panel
requirements. It is impossible to determine with certainty whether EPA
program offices are implementing RFA requirements in a consistent manner
without a detailed examination of the circumstances surrounding each of the
four offices' rules.

One of the 26 proposed rules that OPPTS certified after the implementation
of SBREFA was its August 1999 lead rule. In both the original and draft
revised economic analysis, the Office estimated that the lead rule would
result in about 15,000 new TRI reports for lead and lead compounds- an
increase of nearly 800 percent from reporting at the current 25,000 pound
and 10, 000 pound reporting thresholds. However, the actual increase in TRI
lead reports as a result of the rule could be more than 800 percent. The
Office's estimate of the number of new reports did not include any companies
in 32 manufacturing SIC codes that OPPTS said might be affected by the rule.
We did not estimate the total number of new TRI reports for lead that could
come from companies in these 32 SIC codes. However, it is reasonable to
presume that the number will be greater than zero- the number that OPPTS
implicitly assumed in both its original economic analysis and its draft
revised analysis.

OPPTS initially estimated that no small entities would experience 1 percent
economic impacts from its proposed lead rule. However, using new assumptions
and new data, OPPTS estimated in the draft revised analysis that as many as
464 small entities could experience those impacts. OPPTS believes that 464
is the maximum number of small companies that will experience 1 percent
economic impacts and that 464 is likely to be an overestimate. However, that
estimate- like the office's estimate of the number of additional reports-
does not include companies in 32

B- 284265 Page 33 GAO/ GGD- 00- 193 Regulatory Flexibility Act

manufacturing SIC codes that OPPTS said could be affected by the rule. Using
data from the Bureau of the Census, we estimated that as many as 1,098 small
manufacturing companies in these 32 SIC codes could face first- year
compliance costs of at least 1 percent of their annual revenues. Although it
is unclear exactly how many companies in these SIC codes will experience 1
percent economic impacts, it is reasonable to presume that the number will
be greater than zero- the number that OPPTS implicitly assumed in both its
original and its draft revised analysis. By combining the OPPTS estimate and
our estimate, we believe that the lead rule could have a 1 percent economic
impact on as many as 1,500 small companies. We also estimated that as many
as 78 companies in these 32 manufacturing SIC codes could experience 3
percent economic impacts from the rule.

According to EPA's current guidance, if a proposed rule has a 1 percent
economic impact on 1,000 or more small entities, the program office should
consult with EPA's small business advocacy chair if the office wants to
certify the rule. The guidance also indicates that a program office can
conclude that a rule has a SEISNSE if it has a 1 percent economic impact on
nearly 1,000 small entities- regardless of whether it has that impact on 20
percent of all affected small entities. As previously noted, some EPA
program offices did not certify proposed rules that had at least a 1 percent
economic impact on less than 100 companies. Therefore, if OPPTS had used our
analytic approach and applied the flexibility allowed in the guidance (and
that other EPA program offices have used), it could have concluded that the
rule should not be certified, prepared an IRFA, and convened an advocacy
review panel for the proposed lead rule.

Nevertheless, we believe that the analytic methods that OPPTS used in both
the original economic analysis and the draft revised economic analysis, as
well as the conclusions that the Office drew as a result of those analyses,
were within the discretion provided by both the RFA and EPA's guidance. As
we have said previously, concerns about the RFA compliance are often
traceable to differences in those interpretations and the fact that no
federal agency is responsible for developing uniform definitions. During the
past 10 years we have recommended that Congress either define what it
intended these key RFA terms to mean or consider giving some other entity
the authority and responsibility to define them. 37

37 Regulatory Flexibility Act: Inherent Weaknesses May Limit Its Usefulness
for Small Governments (GAO/ HRD- 91- 61, Jan. 11, 1991); Regulatory
Flexibility Act: Status of Agencies' Compliance (GAO/ GGD- 94- 105, Apr. 27,
1994); Regulatory Reform: Implementation of the Small Business Advocacy
Review Panel Requirements (GAO/ GGD- 98- 36, Mar. 18, 1998); and Regulatory
Flexibility Act: Agencies' Interpretations of Review Requirements Vary (GAO/
GGD- 99- 55, Apr. 2, 1999).

B- 284265 Page 34 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Congress has thus far not acted on those recommendations. We continue to
believe that clarifying what Congress intends the term “significant
economic impact on a substantial number of small entities” to mean
would make the implementation of the RFA more consistent and help to prevent
concerns about how agencies are implementing the act.

We provided a draft of this report to the Director of OMB and the
Administrator of EPA for their review and comments. OMB officials said they
did not have any comments on the draft report. We obtained technical
comments on the draft report from EPA officials and staff on August 30,
2000, and September 1, 2000, and we made changes to the report where
appropriate. We also eliminated a section of the draft report in which we
used information from Federal Register notices for proposed rules to
conclude that the RFA had been inconsistently applied among EPA's four major
program offices. EPA officials indicated that the program offices' RFA
determinations were sometimes based on qualitative factors that were not
always clearly reflected in the Federal Register notices. We therefore
concluded that it was impossible to determine with certainty whether EPA
program offices were implementing RFA requirements in a consistent manner
without a detailed examination of the circumstances surrounding each of the
four offices' rules. We will address the transparency of EPA's RFA
determinations in separate correspondence with the Administrator of EPA.

On September 1, 2000, the Acting Assistant Administrator of OPPTS provided
written comments on the draft report, which are reproduced in appendix I.
The Acting Assistant Administrator said that EPA takes its responsibilities
under the RFA very seriously, and that the agency appreciated the report's
conclusion that its determinations were within the discretion permitted
under the RFA. In addition, the Acting Assistant Administrator noted that
EPA had twice extended the public comment period and conducted a series of
public meetings on the proposed lead rule. She stated that this process
allowed small business representatives to present their concerns directly to
EPA and to submit information that EPA used to improve its assessment of
potential small business impacts.

We agree that EPA's actions after the proposed lead rule was published
helped the agency improve its estimate of the rule's impact on small
businesses. However, if EPA had determined that the proposed lead rule could
have a SEISNSE, it would have been required by SBREFA to convene a small
business advocacy review panel. During the panel process, small businesses
would have been afforded the opportunity to provide their views on the
proposed lead rule before it was published in Agency Comments and

Our Evaluation

B- 284265 Page 35 GAO/ GGD- 00- 193 Regulatory Flexibility Act

the Federal Register. This early involvement of small businesses was what
Congress envisioned when it enacted the advocacy review panel requirements
in 1996. Nevertheless, EPA's determination that the lead rule did not have a
SEISNSE (and thus was not required to convene an advocacy review panel) was
within the broad discretion that federal agencies have under the RFA. EPA's
revised economic analysis and the agency's exclusion of certain SIC codes
from both its original and revised analysis demonstrate that choices made in
exercising the discretion available to agencies can have a significant and
even determinative effect on SEISNSE determinations.

As arranged with your office, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days after the
date of this report. At that time, we will send copies of this report to
Senator John F. Kerry, Ranking Minority Member of the Senate Committee on
Small Business; Representatives James M. Talent and Nydia M. Velazquez,
Chairman and Ranking Minority Member, respectively, of the House Committee
on Small Business; the Honorable Jacob J. Lew, Director of OMB; and the
Honorable Carol M. Browner, Administrator of EPA. We will make copies
available to others on request.

If you have any questions regarding this report, please contact me or Curtis
Copeland on (202) 512- 8676. Key contributors to this assignment were Ellen
Grady and Kathy Peyman.

Sincerely yours, Michael Brostek Associate Director, Federal Management

and Workforce Issues

Appendix I Comments From the Environmental Protection Agency

Page 36 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Page 37 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Page 38 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Page 39 GAO/ GGD- 00- 193 Regulatory Flexibility Act

Ordering Copies of GAO Reports The first copy of each GAO report and
testimony is free. Additional copies are $2 each. Orders should be sent to
the following address, accompanied by a check or money order made out to the
Superintendent of Documents, when necessary. VISA and MasterCard credit
cards are accepted, also. Orders for 100 or more copies to be mailed to a
single address are discounted 25 percent.

Order by mail: U. S. General Accounting Office P. O. Box 37050 Washington,
DC 20013

or visit: Room 1100 700 4 th St. NW (corner of 4 th and G Sts. NW) U. S.
General Accounting Office Washington, DC

Orders may also be placed by calling (202) 512- 6000 or by using fax number
(202) 512- 6061, or TDD (202) 512- 2537.

Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone phone. A recorded menu
will provide information on how to obtain these lists.

Viewing GAO Reports on the Internet For information on how to access GAO
reports on the INTERNET, send e- mail message with “info” in the
body to:

info@ www. gao. gov or visit GAO's World Wide Web Home Page at: http:// www.
gao. gov Reporting Fraud, Waste, and Abuse in Federal Programs To contact
GAO FraudNET use: Web site: http:// www. gao. gov/ fraudnet/ fraudnet. htm
E- Mail: fraudnet@ gao. gov Telephone: 1- 800- 424- 5454 (automated
answering system)

United States General Accounting Office Washington, D. C. 20548- 0001

Official Business Penalty for Private Use $300

Address Correction Requested Bulk Rate

Postage & Fees Paid GAO Permit No. G100

(410525)
*** End of document. ***