United States
Environmental Protection
Agency
Office of Pesticides and
Toxic Substances
Washington DC 20460
EPA-560/12-80-005-A
Toxic Substances
Support Document
Economic Impact Analysis
of Proposed Section 5
Notice Requirements
Appendix: Volume I
Proposed Rule Section 5
Toxic Substances Control Act
-------
EPA-560/l2-80-005r4
September 1980
ECONOMIC IMPACT ANALYSIS OF PROPOSED
SECTION 5 NOTICE REQUIREMENTS
APPENDIX: VOLUME I
Contract No. 68-01-5878
Project Officer:
Sammy K. Ng
ECONOMICS AND TECHNOLOGY DIVISION
OFFICE OF TOXIC SUBSTANCES
WASHINGTON, D.C. 20460
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF PESTICIDES AND TOXIC SUBSTANCES
WASHINGTON, D.C. 20460
-------
PREFACE
The attached document is a contractor's study done with the supervision
and review of the Office of Pesticides and Toxic Substances of the U.S.
Environmental Protection Agency. The purpose of the study is to analyze the
potential economic impact of proposed section 5 notice requirements. These
requirements were prepared by the EPA Office of Pesticides and Toxic Substan-
ces to implement section 5 of the Toxic Substances Act.
This report was submitted in fulfillment of Task Order Number 3 of
Contract Number 68-01-5878 by ICF Incorporated. Work was completed as of
August 1980.
This report (consisting of a report and two volumes of appendices) is
being released and circulated to coincide with publication in the Federal
Register of a notice announcing the availability of the report and of a draft
regulatory analysis.
The study is not an official EPA publication. All comments received by
EPA will be considered in establishing the final analysis to be published
along with the final regulations. Prior to final promulgation of the section
5 requirements, the accompanying document shall have standing in any court
proceeding only to the extent that it represents the views of the contractor
who prepared it. The document cannot be cited, referenced, or represented in
any respect in any such proceedings as a statement of EPA's view regarding the
subject, the industry or the economic impact of the regulation.
-------
TABLE OF CONTENTS APPENDIX; VOLUME 1
Page
APPENDIX A: Cost of Forms A-l
Cost Estimations of Alternative Processor Notification
Requirements A-2
Estimated Cost of Compliance with the Minimum
Guidance Option A-ll
Use of ADL Methodology to Estimate the Cost of
the October 1979 Notice Requirement for Importers A-26
APPENDIX B: Assessment of the Feasibility of Developing a
Methodological Framework for Formal Impact Analysis.... B-l
Summary of Findings B-2
Introduction B~3
Literature Review B-6
Formal Analytical Approaches B-14
Bibliography B-23
APPENDIX C: Discounted Cash Flow Analysis C-l
Purpose and Scope of Arthur D. Little Work C-l
Alternative Analyses C-13
Summary of Results C-14
Economic Analysis Assumptions and Methodology C-14
Results of the Economic Analysis C-21
Supplement to Appendix C: Sales Data Used in Economic
Analysis C-37
APPENDIX D: Discussion of the Economic Burden of Section 5 Notice
Requirements on EPA and Society D-l
Introduction D-l
The Processor Reporting Rule D-2
Importers D-7
Importer Contact of Foreign Manufacturers/Suppliers D-ll
Exporter s D-14
Customer Contact D-15
Insufficient Submissions D-20
Supplemental Reporting D-22
The Confidentiality Options D-28
Summary D-4 3
-------
APPENDIX A
COST OF FORMS
-------
COST OF FORMS
In order to determine the most effective reporting requirements, EPA
sought the costs of various reporting requirements. Originally, Arthur D.
Little had estimated the cost of the January 10, 1979 proposed form and the
October 16, 1979 reproposed form. ICF was required to estimate the cost of a
minimum guidance reporting requirement, processor reporting requirements and
importer reporting requirements. The methodologies used for arriving at these
cost estimations had been established by Arthur D. Little, incorporated. Such
estimates, along with the methodologies and assumptions behind them, are
presented in the following three papers:
• "Cost Estimations of Alternative Processor
Notification Requirements,"
• "Estimated Cost of Compliance with the Minimum
Guidance Option," and
• "Use of ADL Methodology to Estimate the Cost of the
October 1979 Notice Requirement for importers".
The methodologies and assumptions used in these papers are the foundation
for the analysis presented in Chapter V of "Economic impact Analysis of
Proposed Section 5 Notice Requiements; Part 1: Analysis of the impacts on the
Chemical Industry of Proposed Section 5 Notice Requirements."
-------
A-2
I.
COSff ESTIMATIONS OF ALTERNATIVE PROCESSOR NOTIFICATION REQUIREMENTS
This section of the appendix estimates the cost to a company of complying
with any of three alternative processor reporting requirements. EPA
considered these three alternatiaves as options to the proposed processor
requirements outlined in the Federal Register (40 FR 54641). The proposal and
these three alternatives would require reporting by persons who process
chemical substances exempt from TSCA for a nonexempt commercial purpose.
These alternatives differ by the quantity of information processors are
required to submit and the cost estimates vary according to this factor. The
cost estimates are derived from Arthur D. Little incorporated's (ADL)
projection of the cost of completing the notice form proposed in October
1979. The alternative reporting requirements are:
Alternative 1:
Alternative 2:
This alternative would require that
processors fill out a form which
includes company and chemical
identity, amount processed,
categories of use, name of supplier,
worker exposure, consumer exposure,
and health and safety data.
This alternative would require that
some proportion of the reporting
firms provide the same information
required of manufacturers in the
October 1979 proposed form while
others provide the information listed
in alternative 1.
This alternative would require no
reporting and thus, would have no
direct cost. (The processor would,
of course, be subject to other
provisions of TSCA which are beyond
the scope of this analysis.)
In summary, the costs per submission of each alternative are:
Alternative 1: $724-$4,450
Alternative 2:*/ $724-$4,450
Alternative 3:
or
$l,155-$8,925
a/Under alternative 2, EPA would select the appropriate notice form for each
chemical depending on its potential risk to human health and the environment
(Premanufacture Review Program; Proposed processor Requirements, 45 FR 54642,
August 15, 1980).
-------
A-3
Alternative 3: $0-$0
In the subsequent pages the derivations of the estimates are explained.
ALTERNATIVE 1
This alternative requires that data on company and chemical identity,
amounts processed, categories of use, name of suppliers, worker exposure,
consumer exposure, and health and safety be provided.I/
in order to evaluate the cost of completing a processor notice under
alternative 1, estimates must first be made of the number of hours it takes to
provide this data. Clerical, technical and managerial personnel are expected
to complete this form with the necessary data. Therefore, separate time
estimates were obtained for each of these labor categories, and were
multiplied by appropriate wage rates to obtain the total cost of filing this
notice. The following section explains how the time estimates for managerial
and technical personnel were obtained.
Estimation of Technical and Managerial Hours
The number of technical and managerial hours needed to complete a form
providing this data was estimated using the following procedure. Each step
corresponds to a column in Exhibit AI-1 as indicated.
• Column (a); identify information requested under alternative 1.
• Column (b); identify parts of the proposed notice form that are
similar to the information requested.
• Column (c): Obtain estimated number of hours needed to complete
different parts of the proposed form..?/
• Column (d); Estimate fraction of these hours required to
complete similar parts of the alternative 1 form..?/
jl/Prentanufacture Notice; Domestic Manufacturers, 44 FR 59788.
J/Estimates taken from Arthur D. Little, Inc., Estimated Costs for
Ireparation and Submission of Reproposed Premanufacture Notice Form, prepared
for the EPA Office of Toxic Substances, September 1979.
.3/For example, if the hours required to complete a certain part of the
alternative 1 notice were estimated to be one-half of the hours required to
complete the corresponding part of the proposed form, then the estimated
fraction would be 0.5. This step is necessary because the information
requested under this alternative and under the proposed notice form are not
identical.
-------
Exhibit AI-1
Estimated Technical and Managerial Hours for Completing Notice Under Alternative 1
(3)
Information Required
Under Alternative 1
Certification Statement
(b)
Comparable Parts in Section 5 Notice
General Certification and
Confidentialitv Certification
(c) (d)
ADL Time Estimate Fraction of ADL Estimate
for Completing Attributable to providing
Section 5 Notice Information Reauirefl
(Hours) Under Alternative I
Technical Managerial Technical fanaaerial f
0 0 — V __V
(e)
Time Estimate
for Completing
Notice Under
Alternative 1
pohn, rBl MSnaoPrta-i
0
Technical Contact
Name/Address/Phone
A. Manufacturer Identification
{Contains Intended Manufacture Date and
Prenotice Communication Number in Addi-
tional to Form E Required Information)
1-8
~y
B. Health and Environmental
Effects Data
o-i
Chemical Name,
Structural Information, and
Supplemental Information
Amount Processed
Categories of Use
Name of Supplier
Worker Exposure
Consumer Exposure
Health and Safety
Data
Part I. General Information
B. Chemical Identity
1. Class I Chemical Substance 1- 5
2. Class II Chemical Substance
3. Polymers
Part II. Human Exposure and Environmental Release
A. Industrial Sites "Controlled 1- 4
by the Submitter
1. Process Information (Includes
Identity of Site, Type of Site,
Hours of Operation, Amount Processed)
Part I. General Information 1- 8
D. Production and Marketing Data
2. Category^ of Use
Part II. Human Exposure and Environmental Release 0- 2
B. Industrial Sites Controlled by Others
1. Process Information — Identity
of Site
Part II. Human Exposure and Environmental Release 4-32
A. Industrial Sites Controlled
by Submitter
3. Occupational Exposure
Part II. Human Exposure and Environmental Release 0-28
C. Consumer and Commercial User Exposure
Part III. List of Attachments 12-56
A. Physical and Chemical Properties Data
0 1 —I/ 1-5 0
2-6 0.5 0 0.5- 2 0
for entire
Section A
1-2 1 0.25 1- 8 0.25-0.5
for entire
Section D
0-21 0 0-20
for entire
Section B
2-6 1 0.5 4-32 1-3
for entire
Section A
0-2 1 1 0-28 0-2
3-12 1 1 12-56 3-12
Total Hours Under
Alternative 1
18.5-133 4.25-18.5
_ These fractions were not estimated because, regardless of what they might have been, they would have had no effect on the final cost estimation. This is
consistent with the fact that any number multiplied by zero is equal to zero.
—'Because of a peculiarity in the set of questions found in the comparable parts of the section 5 notice, this fraction was not estimated. The managerial
hours for completing this part of the Alternative 1 notice was estimated through direct examination of the Alternative 1 and section 5 questions. This
estimation procedure is explained in the text.
-------
A-5
• column (e); Multiply ADL time estimates by ICF
estimate of fraction of time required to complete
alternative 1 data requirements.
Total:
• Calculate sum of these products to obtain total
technical and total managerial hours needed to provide
information requested under alternative 1.
Estimation of Fractions
For each alternative 1 notice requirement, there was a similar part of the
proposed form for which ADL estimated technical and managerial hours. Each
part of the proposed form was chosen so that it incorporated all of the
information requested in the corresponding part of the alternative 1
requirements. Therefore, the technical and managerial hours needed to
complete the proposed form part would always be at least as great as the
number of hours needed to complete the corresponding part of alternative 1
notice requirements. In some cases, the hours needed to complete each
proposed form part could be much greater than the hours needed to complete the
corresponding alternative 1 part.
Each part of the Alternative 1 notice was assigned a fraction of the ADL
estimates for technical and managerial hours. The fraction represents that
portion of the ADL estimated hours needed to complete the counterpart sections
of the alternative 1 notice. For example, ADL estimated it would take from
one to four technical hours to complete a portion of the proposed form notice
related to process information. This particular portion of the form requests
information on site identity, site type, hours of operation, and amount
processed. However, alternative 1 would require information only on amount
processed. Thus, the number of technical hours required to complete that
portion of the alternative 1 notice is only a fraction of the ADL estimate for
the proposed form, in this case, the fraction assigned was 0.5, or one-half
of the estimated proposed form hours.
The following discussion explains the methodology used to estimate the
fractions. Each part of the Alternative 1 notice is considered separately.
1. certification Statement. The certification statement for Alterna-
tive 1 is quite similar in purpose and function to the general certification
and confidentiality certification statements of the proposed form. Both
require an authorized official to sign and date the form.
ADL did not include the certification statement in its cost estimate and,
therefore, did not assign to it any time estimates, zero or otherwise. In so
doing, ADL effectively attributed zero technical and zero managerial hours to
«-he certification statement. Thus, in being consistent with ADL's costing
methodology, zero technical and zero managerial hours were assigned here as
well.
-------
A-6
Because zero hours were assigned, the fraction of time needed to complete
the certification statement was not estimated. Regardless of what the
fraction is, the number of hours needed to complete this part of the
alternative 1 notice would be zero.
2. Company Name/Address and Technical Contact Name/Address/Phone. This
information, as required in alternative 1, was comparable to the information
requested in Section I (A) of the form proposed in October. The submitter must
provide its name and address and identify a technical contact. The comparable
proposed form data also requires identification of the parent company and
intended date of manufacture. TO this proposed form subsection, ADL assigned
zero technical hours and 1-8 managerial hours. The eight-hour maximum was
assumed to be mostly attributable to supplying intended date of manufacture,
while the rest of the section required only about an hour to complete. The
company name and technical contact information requested in alternative 1
required between 0 and 1 hours of managerial time, and no hours of technical
time.
3. Chemical Name, Structural information, and Supplemental information
on Chemical Identity. The alternative 1 requirements and the requirements of
the proposed form are identical. Therefore, the corresponding fractions were
estimated to equal 1.
4- Amount Processed. The part of the proposed form that requests this
information requests identity of the site, type of site, hours of operation,
and amount manufactured, processed, or used. Only two questions would require
a significant amount of time to complete: hours of operation, and amount
manufactured, processed, or used. These questions are dependent on each
other. It was estimated that each would take about the same amount of time,
on the average, to answer. Therefore, the fraction of technical time
attributed to the question on amount processed was 0.5. The amount of
managerial time required to complete this section was assumed to be zero
because 30 minutes to two hours of technical time (see column e of Table 1)
was considered sufficient to provide information on the amount processed.
5. Categories of Use. The alternative 1 requirements with respect to
this part are identical to the requirements in subsection I(D)(2) of the
proposed form.^/ Thus, one hour of technical time is estimated for this
section. ADL did not provide a separate estimate for the amount of managerial
time required to complete subsection I(D)(2), Category of Use. instead, ADL
estimated it would take from two to six managerial hours to complete the
entire section I(D). A comparison of the subsections of l(D) provides an
estimate of the fraction of time it would take to answer I(D)(2) independently
from the other subsections. Specifically, the length of time it would take to
answer I(D)(2) was estimated to be roughly the same as the amount of time it
would take to answer each of the following subsections: I(D)(1), I(D)(4), and
I(D) (5). The amount of time needed to answer I(D) (3) is assumed to be
negligible relative to the other four subsections.
i/st-tement by Ken Gray of the U.S. Environmental protection Agency, March
1980.
-------
A-7
Thus, there are four major parts - I(D)(1), (2), (4), and (5) - that take
about the same amount of time to answer, in other words, it would take
approximately one-fourth of the total time required to complete the entire
I(D) to answer subsection I(D)(2). Therefore, a fraction of 0.25 was
estimated for managerial hours.
6. Name of Supplier. Subsection II(B)(1) of the proposed form required
information comparable to that required in the alternative 1 notice. Thus,
the fraction of technical hours attributed to completing the section on
supplier's name was one. Because providing this information appeared to
require no managerial time, the fraction was estimated to be zero.
7. Worker Exposure. The Alternative 1 requirements with regard to worker
exposure are exactly those requested in subsection II(A)(3), Occupational
Exposure, of the proposed form. Thus, the fraction of technical hours
attributable to the alternative 1 requirements was estimated to be one.
The managerial hours for the subsection of the proposed form notice,
however, were not estimated by ADL. We estimated the subsection after
examining the entire section (II(A)) for which ADL had made an estimate. This
section contains four subsections: Process Information, Block Diagram,
Occupational Exposure, and Environmental Release and Disposal. Subsection 1
(process information) was considered negligible with respect to required
managerial hours, as was subsection 2 (the Block Diagram), which required
purely technical information. Subsection 4 (environmental release and
disposal) was considered to require roughly the same amount of managerial
hours as the subsection 3 (occupational exposure). Thus the total managerial
hours of the section could be divided equally between subsection 3 and
subsection 4. As a result, the fraction of managerial hours in Section II(A)
attributable to alternative 1 requirements on occupational exposure was
estimated to be 0.5.
8. Consumer Exposure. The information requirements of alternative 1
with respect to this category were defined to be those information
requirements in section II(C), Consumer and Commercial User Exposure, of the
form proposed in October. Therefore, the fractions were set at one for each
labor category.
9. Health and Safety Data. The information requirements of Alternative
1 with respect to this category were defined to be those information
requirements in section III(A), Physical and Chemical Properties Data, and in
(B), Health and Environmental Effects Data, of the October proposed form.
Therefore, the fractions were set at one for each labor category.
Range of Technical and Managerial Hours
AS shown in Exhibit Al-1 column(e) on page 4 the total technical hours
range froi-i 18.5-133. The total managerial hours range from 4.25-18.5. These
values represent the sum of the costs of the component parts.
-------
A-8
Estimation of Clerical Hours
Using AOL's time estimates for the section 5 notice,JL/ we computed the
ratios of clerical to technical and clerical to managerial hours. The ADL
time estimates for the proposed form, by labor category, ranged as follows:
Clerical 8-40 hours
Technical 27 - 267 hours
Managerial 8-37 hours
The ratios of minimum hours that appear in Exhibit AI-2 are simply the ratios
of the ADL lower bound estimates. Similarly, the ratios of maximum hours are
the ratios of the ADL upper bound estimates.
In order to estimate clerical hours, the ratios were multiplied by the
estimated total technical and managerial hours needed to provide the
information required by alternative 1. For example, the 5.48 minimum hours
estimate in Table 3 was obtained by multiplying .296 times 18.5 (the estimated
minimum technical hours to complete the alternative 1 notice). Once estimates
of the clerical hours were obtained in this fashion, an average was computed
to obtain an estimate of the number of clerical hours required to complete the
notice. That is, 5.48 plus 4.25, divided by two equals 4.87 (see Exhibit
AI-3. The estimated average ranged from 4.87 to 19.97 hours of clerical time.
Exhibit AI-2
Ratios of Clerical to Technical and Managerial Hours
Ratio of Ratio of
Ratio Minimum Maximum
Hours Hours
Clerical Hours to Technical Hours .296 .150
Clerical Hours to Managerial Hours 1.00 1.08
I/Table 4, Chapter VI, of Arthur D. Little, inc.. Estimated Costs for
Preparation and Submission of Reproposed Premanufacture Notice Form, September
1979.
-------
A-9
Exhibit AI-3
Estimated Clerical Hours
Based on clerical to Technical Hours Ratio
Based on Clerical to Managerial Hours Ratio
Approximate Average
Minimum Maximum
Hours Hours
5.48 19.95
4.25 19.98
4.87 19.97
Estimation of the Total Cost of Completing an Alternative 1 Notice
To estimate the total cost of completing a notice under alternative 1,
the average hourly wage rates by labor category were multiplied by the
respective number of hours needed to complete a notice. The average hourly
wages were the same as those established by ADL: $10/hour clerical, $25/hour
technical and $50/hour managerial. The results are presented in Exhibit
AI-4. The total cost of preparing and submitting a notice under alternative 1
ranged from $724 to $4,450.
Exhibit AI-4
Cost of Completing Alternative 1 Notice
Estimated Hours
Estimated Cost
Labor
Category
Clerical
Technical
Managerial
Total
to complete
Alternative 1 Notice
Minimum
4.87
18.50
4.25
Maximum
19.97
133.00
18.50
Average
Wage Rate
$10
$25
$50
of completing
Alternative 1 Notice
Minimum
$48.70
$462.50
$212.50
$723.70
Maximum
$199.70
$3,325.00
925.00
$4,449.70
or
$724
$4,450
-------
A-10
ALTERNATIVE 2
Alternative 2 would require processors of exempt substances for nonexempt
commercial purposes to submit one of two different types of notification. The
first type of notification would consist of those data requirements specified
in the Federal Register notice "Proposed Processor Requirements." EPA
proposed data requirements for processors that, with minor additions, are the
same as those proposed in October for domestic manufacturers. ADL estimated
that the. cost of submitting the October proposed form ranged from $1,155 to
$8,925.i/ The other type of notification a processor could be required to
file under this alternative consists of that information as outlined in
alternative 1. ICF's estimates for the cost of alternative 1 notification
range from $724 to $4,450.
The following equation may be used to estimate the range of costs when a
fraction of processors file the information required by the processor proposal
and the remainder file information required by alternative 1. The variable X
represents that fraction of all exempt substances processed for nonexempt
purposes that would be subject to the information requirements specified in
the July processor proposal.
Estimated Range of Notification Costs for Alternative 2 =
. [ADL's Estimated Cost Range! /T_X\(Estimated Cost Range for |
( ' [for Section 5 Notification) ( ~ ' [Alternative I Notification!
ERNC = (X) ($1,155 to $8,925) + (1-X)($724 to $4,450)
ERNC = (X)($1,155) + (1-X)($724) toX($8,925) + (1-X)($4,450)
Example; Suppose only 10 percent of all exempt substances used for
nonexempt purposes were subject to the requirements proposed for processors in
the Federal Register. Then:
X = .10
The estimated range of notification costs for alternative 2 would then be
$767 - $4,898.
Unfortunately, we have no information about the actual percentage. Thus,
it is impossible to estimate the average cost per submission.
ALTERNATIVE 3
This alternative does not require any reporting, therefore, the cost of
this alternative is zero. Unfortunately, we have no information about the
actual percentage. Thus it is impossible to estimate the average cost per
submission.
I/Arthur D. Little, Inc., Estimated Costs for Preparation and Submission of
Reproposed Premanufacture Notice Form, September 1979.
-------
A-ll
II. ESTIMATED COST OF COMPLIANCE WITH
THE MINIMUM GUIDANCE OPTION
DETERMINATION OF REQUIRED INFORMATION
In this paper, we assign costs to submitting a section 5 notice when EPA
does not provide a form, but does provide minimum reporting guidance. The
information required under a minimum guidance scheme is organized into the
following categories: Manufacturer's Identity, Chemical Identity, Use
Categories and Amount Produced per Use Category, Byproducts, Occupational
Exposure, Method of Disposal, Health and Environmental Effects Data, and
Federal Register Notice. For each of these categories we describe the
information required. This required information, by category, is displayed in
Exhibit AII-1.
Exhibit AII-1
Submission Requirements under Minimum Guidance Reporting
1. Chemical Identity; Included in the data would be specific chemical
name, CAS#, or molecular formula with structural diagram; and specific
identification of monomers at least to the two percent level.
2. Use Categories and Amount Produced per Use Category; Included in
these data would be amount produced per use category, types and levels of
exposures, decriptions of use, and some consumer exposure information.
3. Byproducts; These data would include a list of significant
byproducts at each stage in the substance's life cycle, including
manufacturing, processing, use, and disposal.
4. Occupational Exposure; This information would include the number of
workers, type and degree of exposure, and duration of exposure (duration of
daily exposure and duration of cumulative exposure by month or year)
expected. Information both on the submitter's work sites and work sites
controlled by others would be expected.
5. Method of Disposal; These data should include the identity of the
site where the submitter, and others known by the submitter, intend to dispose
of the new substance. They should also include the duration of release, mode
of release, media into which release occurs (air, water, land), amount
released, and pollution control equipment used in the disposal operations.
6. Federal Register Notice; This notice would include a name for the
chemical (actual or generic), manufacturer's identity (actual or masked),
categories of use (actual or masked), and test data.
-------
A-12
7. Manufacturer's Identity; These data would include the
organization's name, the name and address of the person filing the notice, the
technical contact's name and address, the parent organization's name and
address, and the intended date of manufacture for commercial purposes.
8. Health and Environmental Data; Included here would be data on risk
assessment, detection methods, environmental release and disposal, consumer
and commercial user exposure, physical and chemical properties, and health and
environmental effects.
-------
A-13
COSTING METHODOLOGY
A. OVERVIEW
ADL's costing methodology for the October reproposed form!/ is used as
a basis for the costing of the minimum guidance reporting requirements. The
reasons we chose this as the starting point are:
I. In its costing of the reproposal, ADL outlined the costs
per requirement category in detail. This allowed us to
assign costs accurately to similar requirement categories
under the minimum guidance option.
2. The reproposed form was the most recent source of detailed
cost data/ and therefore contained the most accurate
information about costs.
3. Without specific guidance from the agency, submitters will
tend to err on the side of a narrow interpretation of
statutory requirements.
In order to estimate the cost of meeting the requirements of the section
5 notice minimum guidance option, estimates are first made of the number of
hours necessary to provide the information. It is expected that providing the
required information will involve clerical, technical, and managerial
personnel. Therefore separate time estimates are given for each of these
labor categories. The time estimates are multiplied by appropriate wage rates
to obtain an estimate of the total cost. The following section explains how
the time estimates for managerial and technical personnel are obtained.
B. ESTIMATION OF TECHNICAL AND MANAGERIAL HOURS
The number of technical and managerial hours needed to comply with the
minimum guidance option are shown in Exhibit AII-2 and were estimated by the
procedure described below.
• Step 1; List each information category (or
submission requirement).
• Step 2; Identify parts of the reproposed form which
contain all of the information required by the
corresponding submission requirements under the
minimum guidance option.
i/The "reproposed form" or "reproposal" refers to the TSCA section 5
notification form entitled "EPA Premanufacture Notice: Domestic
Manufacturers," proposed on October 16, 1979, in the Federal Register, 44 FR
59788.
-------
Exhibit AII-2
Estimation of Required Technical and Managerial Hours
Minimum Guidance Option
Submission Requirements
Parts of Reproposed Form Comparable
to Submission Requirements
ADL's Estimated Hours
For Completion
of Section 5 Parts
Technical Managerial
Fraction of Those Hours
Needed to Complete
Minimum Guidance Option
Submission Requirements
Technical
Managerial
Hours Needed to Complete
Minimum Guidance Option
Submission Requirements
Technical Managerial
Chemical Identity
Use Categories and
Amounts/Use Category
I. General Information
B. Chemical Identity 1-5
1. Class 1 Chemical Substance
2. Class 2 Chemical Substance
3. Polymers
I. General Information
D. Production and Marketing Data 2-12
1. Production Volume
2. Category of Use
II. Human Exposure and Environmental
Release
C. Consumer and Commercial User
Exposure
1. Table - Route, Frequency and 0-16
Number Exposed
1-2 for
entire
section D
0-2 for
entire
section C
.5
1-5
2-12
.5-1
.25
0-16
0-.5
Byproducts
II. Human Exposure and Environmental
Release
A. Industrial Sites Controlled by
the submitter
2. Block Diagram
3. Occupational Exposure
3.5 Other Substances
1-24
2-6 for
entire
section A
2-16 for
subsections
3..3-3.5
.6
.6-14.4
.17
.3-1
1-8
C. Consumer and Commercial User
Exposure
4. Byproducts of Use
0-4
0-2 for
entire
section C
.25
0-4
0-.5
-------
Federal Register
Notice
Exhibit AII-2 (continued)
Estimation of Required Technical and Managerial Hours
a b
c
d
Fraction of Those Hours
ADL's Estimated Hours Needed to Complete
For Completion Minimum Guidance Option
Minimum Guidance Option Parts of Reproposed Form Comparable
Submission Requirements to Submission Requirements
Occupational Exposure II. Human Exposure and Environmental
Release
A. Industrial Sites Controlled
by the Submitter
3. Occupational Exposure
3.1 Identity of Site
3.2 Occupational Exposure at
Site
3.3 Direct Exposure
3.4 Physical State
B. Industrial Sites Controlled
by Others
3. Occupational Exposure
Method of Disposal II. Human Exposure and Environmental
Release
A. Industrial Sites Controlled
by Submitter
4. Environmental Release and
Disposal
B. Industrial Sites Controlled
by Others
4. Environmental Release and
Disposal
of Section 5
Technical
2-16 for
subsec-
tions
3.1-3.2
2-16 for
subsec-
tion
3.3-3.5
0-20
1-12
0-8
Parts Submission Requirements
Managerial Technical Managerial
2-6 for 1 .33
entire
section A
.5
0-2 for 1 .5
entire
section B
2-6 for 1 .5
entire
section A
0-2 for 1 .5
entire
section A
e_
Hours Needed to Complete
Minimum Guidance Option
Submission Requirements
Technical Managerial
2-16 .7-2
1-8
0-20 0-1
1-12 1-3
0-8 0-1
IV. Federal Register Notice
1-8
1-2
0-8
0-1
un
-------
EXHIBIT AII-2 (continued)
Estimation of Required Technical and Managerial Hours
Minimum Guidance Option
Parts of Reproposed Form Comparable
AOL's Estimated Hours
For Completion
of Section 5 Parts
Fraction of Those Hours
Needed to Complete
Minimum Guidance Option
Submission Requirements
Hours Needed to Complete
Minimum Guidance Option
Submission Requirements
Submission Requirements
Manufacturer
Identification
Health and Environmental
Data (Not Previously
Listed)
to Submission Requirements
I. General Information
A. Manufacturer Identification
I. General Information
F. Risk Assessment
G. Detection Methods
II. Human Exposure and Environmental
Release
C. Consumer and Commercial user
Exposure
2. Exposure Levels
Technical Managerial Technical
0 1-8
0-16 0-2 1
1-4 0 1
0-4 0-2 for 1
entire
section C
Managerial Technical
1 0
1 0-16
1-4
.5 0-4
Managerial
1-8
0-2
0
0-1
3 . Product Aspect Affecting
Consumer Exposure
III. List of Attachments
A. Physical and Chemical Properties
0-4
0-4
B.
Data
Health and
Ef f ects
Environmental
Data
4-16
8-40
1-4
2-8
1
1
1
1
4-16
8-40
1-4
2-8
TOTAL HOURS NEEDED TO COMPLETE MINIMUM GUIDANCE OPTION SUBMISSION REQUIREMENTS
22.6-215.4
7.5-35.0
-------
A-17
Exclude all reproposal parts that have previously
been listed in the column (so as to avoid the
double-counting of required information).
• Step 3; Give the estimated number of hours needed to
complete the listed parts of the reproposed form..2.7
• Step 4; Estimate the fractions of these hours
required to provide the information under the minimum
guidance option.2/
• Step 5; Multiply the hours estimates of column c
(for reproposal parts) by the fraction estimates of
column d to yield time estimates for the minimum
guidance option. Then, calculate the sum of these
products to obtain the total technical and managerial
hours needed to supply information required under the
minimum guidance option.
Two of the five steps required decision making on the part of ICF: the
listing of the parts of the reproposed form comparable to the minimum guidance
option (column b), and the estimation of the ratio of hours required to
provide categories of information under the minimum guidance option to hours
required to complete the comparable part of the reproposed form. The
rationale behind these decisions is presented below for each type of
information, and they should be read in conjunction with Exhibit AII-2.
I/Estimates taken from Arthur D. Little, Inc., Estimated Costs for
Preparation and Submission of Reproposed Premanufacture Notice Form, prepared
for the EPA Office of Toxic Substances, (Cambridge, Mass.: September 1979).
.I/For example, if the hours required to provide a certain category of
information were estimated to be one-half the hours required to complete the
corresponding part of the reproposal, then the estimated fraction would be
0.5. This step is necessary because the information requested under the
minimum guidance option and the information requested under the reproposal may
not be identical.
It would not take more time to complete a given category of information
under the minimum guidance option than to complete the corresponding
reproposal part, because each reproposal part was designed to incorporate all
of the information in the minimum guidance option.
-------
A-18
1. Chemical Identity. The minimum guidance option requires specific
chemical name; CASf or molecular formula with structural diagram; and the
specific identification of monomers, at least to the two percent level. The
subsections of the reproposed form containing this information are I.B.I.,
I.E.2., and I.E.3. Because the reproposed sections do not require
substantially more technical information than what is required under the
minimum guidance option, it was estimated that the technical hours for either
notice would be about the same. Thus, the minimum guidance technical hours
were assumed equal to the reproposal hours. Since the managerial hours were
zero, there was no need to estimate a fraction for managerial hours.
2. Use Categories and Amount Produced per Use Category. The minimum
guidance option requires: amount produced per use category; types and levels
of exposure; descriptions of use; and use and some consumer exposure
information. The subsections of the reproposed form containing this
information are I.D.I., I.D.2., and II.C.I. Unlike the minimum guidance
option, the reproposed policy did not specifically request the quantity
produced per use category. It did request, however, both the total quantity
produced, in subsection I.D.I., and the production percent per use category,
in subsection I.D.2. Because such information can easily be used to derive
the quantity produced per use category, subsections I.D.I, and I.D.2. were
included in the costing of the present submittal requirement..!/ Types and
levels of consumer exposure, provided in subsection II.C.I., were also
included. Occupational exposure, however, will be discussed and included in
the costing of the occupational exposure requirement.
Technical hours; Except for question 2(c), all of the information
requested in subsections I.D.I., I.D.2., and II.C.I. of the reproposed form
appears to be required under the "use categories and amounts produced per use
category" of the minimum guidance reporting option. The time needed to
complete question 2(c) appears to be negligible relative to the time needed to
complete the rest of the subsection. For these reasons, it was estimated that
the number of hours needed to complete the three subsections would be about
the same as the number of hours needed to provide the information under the
the minimum guidance option. Therefore, technical hours for the minimum
guidance option were set equal to the hours estimated for these subsections of
the reproposal.
Managerial hours: The number of managerial hours for the subsections
were not provided in ADL's costing analysis. However, information was
available pertaining to the managerial hours estimated for the entire section
I.D., and the managerial hours estimated for the entire section II.C. To
estimate the managerial hours needed under the minimum guidance option to
provide the- information contained in the subsections, the proportions of each
section's managerial hours that could be attributed to the required
subsections had to be estimated. (This was consistent with the estimate that
the information provided under the minimum guidance option was essentially the
same as that in the subsections.)
I/Quantity produced times the percent production per use category equals
the quantity produced per use category.
-------
A-19
Section I.D. has three other subsections: I.D.3.—Previous Manufacture;
I.D.4.—Hazard Warnings; and I.D,5.—Number of Customers. Subsection I.D.3./
which required a yes, no, or don't know response to the question, "Has the
chemical substance been manufactured before?", would require an insignificant
amount of managerial time to complete. Subsections I.D.4. and I.D.5. seemed
to be roughly comparable, in the managerial hours that they might require, to
subsections I.D.I, and I.D.2. Therefore, the estimate was made that
subsections I.D.I, and I.D.2. together constituted about one-half of section
I.D. in terms of required managerial hours. As a result of this estimate, the
fraction 0.5 was determined to be the fraction of section I.D. managerial
hours necessary to provide information under the minimum guidance option.
This same methodology was used to estimate the fraction of managerial
hours for section II.C. that would be required to provide information under
the minimum guidance option. Section II.C. was composed of the subsections
II.C.I.—Table—Route, Frequency and Number Exposed; II.C.2.—Exposure Levels;
II.C.3.—Product Aspect Affecting Consumer Exposure; and II.C.4.—Byproducts
of Use. Since all of these sections were estimated to be comparable in terms
of managerial time required, a fraction of 1/4 (or 0.25) was determined to
reflect the managerial hours for subsection II.C.1..V
3. Byproducts. The minimum guidance option requires a list of
significant byproducts at each stage in the substance's life cycle including
manufacturing, processing, use and disposal. It is assumed that more
information (e.g., amounts generated) would be provided for those products
perceived to be more harmful. The subsections of the reproposal containing
similar information were those relating to the amount produced, the origin and
the release of each major byproduct (subsection II.A.2); those concerning
occupational exposure to byproducts (subsection II.A.3.5); and those
concerning consumer exposure (II.C.4).
Technical hours related to major byproducts; It was estimated that the
technical hours required to complete the information on byproducts in
subsection II.A.2. would be about 60 percent of the technical hours required
to complete the entire subsection. This rough estimate was based on the
following rationale:
• It was estimated that the technical hours required to
(1) identify the major unit operations and chemical
conversions, and (2) to provide the approximate mass
of all feed materials, byproduct materials, and
products entering and leaving each major unit
^/Because the minimum guidance option incorporated all of section II.C.,
section II.C. would have always contributed the range of 1-2 managerial hours
to the total hours required for the completion of the minimum guidance
option. However, fractions were determined for section II.C. in order to
maintain a consistent methodology.
-------
A-20
operation and chemical conversion, were about the
same, and that they were about twice the hours
required to identify those points in the block
diagram from which there would be releases of the new
chemical substance or byproducts into the environment.
• Thus, the ratio of the technical hours required to
complete parts of subsection II.A.2. corresponding to
the items mentioned above were 2:2:1, respectively.
• It was further estimated that, on the average, the
technical hours needed under the minimum guidance
option to provide information on byproducts would be
about half the time needed to (1) identify the major
unit operations and chemical conversions, and (2) to
provide the approximate mass of all feed materials,
byproduct materials, and products entering and
leaving each major unit operation and chemical
conversion, and would be almost all of the time
needed to identify those points in the block diagram
from which there would be releases of the new
chemical substances or byproducts into the
environment.
• The above estimate translates into the statement that
providing information on byproducts under the minimum
guidance option would take about one-half of
two-fifths, plus one-half of another two-fifths, plus
almost an entire fifth of the technical hours
attributed to the subsection.
• Therefore, the estimated fraction of technical hours
was (1/2) (2/5) + (1/2) (2/5) + (1) (1/5) = 0.6.
Technical hours related to occupational exposure to byproducts; The
technical hours required to provide information on occupational exposure to
byproducts were estimated to be about one-half the hours for all of II.A.3.3
to II.A.3.5. of subsection II.A.3. Because only occupational exposure data
were sought, the fraction of technical hours listed for subsection II.A.3.5
was 0.5 (see Table 2).
Managerial hours related to the above two topics; it was assumed that
relatively little managerial time would be spent on subsection II.A.2. (Block
Diagram) since the information contained in this subsection was primarily
technical. For this reason, it was assumed that the vast majority of
managerial hours for section II.A. would be spent on subsection 3,
Occupational Exposure, and on subsection 4, Environmental Release and
Disposal. Subsection 3 was estimated to require about the same number of
managerial hours as subsection 4. Thus, about half of all the managerial
hours for section II.A. would be spent on subsection 3.
-------
A-21
Three of the five questions in subsection 3 (3.2, 3.3, and 3.5) seemed to
require roughly the same number of managerial hours, while questions 3.1 and
3.4 appeared to require relatively few managerial hours. For these reasons,
question 3.5 was estimated to require about one-third of the managerial hours
needed for subsection 3. Since subsection 3 required about one-half of the
managerial hours needed for section II.A., it followed that question 3.5
required about one-third of one-half, or one-sixth, of the managerial hours
for section II.A. Because subsection II.A.2. was considered primarily
technical, it was ignored in the estimation of managerial hours. In
conclusion, the managerial hours for providing the information in section
II.A. under the minimum guidance option were estimated to be one-sixth, or
0.17 of the hours for the entire section.
Technical and managerial hours related to consumer exposure to
byproducts; Subsection II.C.4. was determined to require information that was
clearly called for under the minimum guidance option. Therefore, the fraction
of the technical hours assignable to minimum guidance was clearly 1. As
previously stated, it was estimated that the four subsections of section II.C.
were roughly comparable in terms of managerial hours. It follows that the
fraction of section II.C. managerial hours needed for subsection II.C.4. would
be one-fourth, or 0.25. Since ADL provided managerial hours only for the
entire section II.C., the managerial hours for II.C.4. were taken as 0.25 of
those hours.
4. Occupational Exposure. The minimum guidance option calls for
information on work sites other than the submitter's in addition to
information on the submitter's work sites; number of workers and duration of
exposure (duration of daily exposure and duration of cumulative exposure by
month or year); and the type and degree of exposure (Exhibit AII-2). The
subsections of the reproposal that contained this information and that were
not counted previously were related to occupational exposure at both the
submitter's work site and at work sites controlled by others. Work sites
where substances are made under contract to submitters would be excluded.
From examination of the subsections of the reproposal that require
similar information (I.A.3.1., II.A.3.2., II.A.3.3., II.A.3.4., II.B.3.), it
is clear that they call for the same information that is specified in the
minimum guidance option. Because all of the data in these subsections are
required under the minimum guidance option, the fraction of the technical
hours needed to complete submission requirements was 1. As already
discussed, j>/ subsections II.A. 3.3 and II.A. 3. 5 were each estimated to
require approximately one-half of the technical hours needed to complete
subsection II.A.3. Therefore, the technical hours for subsections 3.3 to 3.4
was estimated to be .5 of those hours attributed to subsections 3.3 to 3.5.
JL/See discussion of subsection II. A. 3.5 in "3. Byproducts'
-------
A-22
Also discussed previously was the estimate that subsection II.A.3.
constituted about one-rhalf of the managerial hours for section II.A., and that
subsection II.A.3. (3.5) constituted about one-third the managerial hours for
subsection II.A.3. Since II.A.3. consisted entirely of II.A.3. (3.1 to 3.5),
it followed that subsections II.A.3. (3.1 to 3.4) constituted two-thirds of
the managerial hours for subsection II.A.3. Thus, the managerial hours for
subsections II.A.3. (3.1 to 3.4) were one-half of two-thirds of the hours
attributed to all of section II.A., and so the estimated fraction was .33.
Managerial hours related to occupational exposure at work sites
controlled by others; Section II.B. requires kinds of data similar to those
required in section II.A., except that the data is for work sites not
controlled by the submitter instead of sites that the submitter controls.
Thus, the estimates of the proportion of managerial hours required for each
part were derived as discussed previously for section II.A. Subsections
II.B.I. and II.B.2. require virtually no managerial time. Subsections II.B.3.
and II.B.4., however, each appear to require a substantial amount of
managerial time, and the hours for each seem to be about the same. It was
estimated, then, that the fraction of section II.B. managerial hours
attributable to subsection II.B.3. was 0.5.
5. Method of Disposal. The minimum guidance reporting option calls for
the identity of the site where the submitter and others known to the submitter
intend to dispose of the new substance. It should also include the duration
of release; mode of release; media into which release occurs (air, water,
land); amount released; and pollution control equipment used in the disposal
operations. This is virtually identical to subsections II.A.4. and II.B.4. of
the reproposal form.
As previously stated, it was estimated that in the reproposal subsections
II.A.3. and II.A.4. each required about one-half of the managerial hours
needed for all of section II.A., and that subsections II.B.3. and II.B.4.
required the same with respect to section II.B. Therefore, the fractions of
the managerial hours for sections II.A. and II.B. attributable to subsections
II.A.4. and II.B.4., respectively, were estimated to be 0.5. Since ADL
provided estimates of technical hours for these subsections, and since these
subsections constituted the minimum guidance reporting, the fraction of listed
technical hours needed to complete these requirements was 1 for each
subsection.
6. Federal Register Notice. The minimum guidance reporting option
requires the same Federal Register notice that is required in part IV of the
reproposal (Exhibit AII-2). This would include a name for the chemical
(actual or generic); manufacturer's identity (actual or masked); categories of
use (actual or masked); and test data. The technical and managerial hours for
this part were estimated by ADL, and so the fractions corresponding to each of
these estimates were both listed as 1.
7. Manufacturer Identification. The minimum guidance option has the
same information requirement as section I.A. of the reproposed form (Table 1).
The submitter must supply the organization's name; the name and address of:
(1) the technical contact, (2)parent organization, and (3) the person filing
the notice, and the intended date of manufacture for commercial purposes.
-------
A-23
ADL estimated the technical hours for this requirement to be zero and the
managerial hours to be 1-8. Since the managerial hours were provided for this
requirement, the fraction for managerial hours was set at 1. No fraction
needed to be estimated for technical hours, since any fraction of zero is zero,
8. Health and Environmental Data. The minimum guidance option required
the same information on health and environmental data that was requested in
the reproposal (from Exhibit AII-2):
• Risk Assessment
• Detection Methods
• Occupational Exposure (previously counted under the
byproducts and occupational exposure requirements)
• Environmental Release and Disposal (previously
counted under the method of disposal requirement)
• Consumer and Commercial User Exposure (subsections
II.C.I and II.C.4 previously counted under the use
categories and by-products requirements respectively)
• Physical and Chemical Properties Data
• Health and Environmental Effects Data
• Federal Register Notice (previously counted under the
Federal Register Notice requirement).
In estimating technical and managerial hours, only sections and
subsections not previously counted in different requirements were considered
as part of the health and environmental data requirement. (See Exhibit
AII-2.) ADL provided estimates of the technical hours for each section and
subsection included in the requirement. Therefore, each fraction of the
technical hours was set at 1. The same was true for managerial hours except
for subsections related to consumer and commercial user exposure. Therefore,
the fractions of managerial hours were all set at 1 except for these
subsections (II.C.2., II.C.3.). As previously stated, each subsection of
section II.C. was determined to require approximately the same number of
managerial hours. Because the health and environmental data requirement
included two out of four such subsections, the managerial hours for the
combination of these subsections was estimated to be 0.5 of the managerial
hours for the entire section II.C.
Summary; Through the process described above, the total technical and
managerial hours needed for providing information under the minimum guidance
option were estimated to be 22.6-215.4 and 7.5-35.0, respectively. (See
Exhibit AII-2.) A more simplified process was used to estimate clerical
hours. This is presented below.
-------
A-24
C. ESTIMATION OF CLERICAL HOURS
Ratios of clerical to technical hours and of clerical to managerial hours
were estimated from ADL's time estimates for the reproposed form. These time
estimates, by labor category, were as followsrZ/
Clerical 8-40 hours
Technical 27-267 hours
Managerial 8-37 hours
Using the lower hours estimates, we created the first column of Exhibit
AII-3. Using the higher hours estimates, we created the second column of the
exhibit.
Exhibit AII-3
Ratios of Clerical to Technical Hours and of
Clerical to Managerial Hours
Ratio of Ratio of
Ratio Minimum Maximum
Hours Hours
Clerical Hours to Technical Hours .296 .150
Clerical Hours to Managerial Hours 1.000 1.080
Clerical hour estimates for the minimum guidance options were obtained by
multiplying these ratios by the estimated total technical and managerial hours
needed to provide information under the option. Once estimates of the cleri-
cal hours were obtained in this fashion, averages were taken to obtain a sin-
gle range for the number of clerical hours required (refer to Exhibit AII-4).
The estimated range was 7.1-35.1 hours.
Z/Estimated Costs for Preparation and Submission of Reproposed
Premanufacture Notice Form, p. 38.
-------
A-25
Exhibit AII-4
Estimated Clerical Hours
Ratio
Minimum
Hours
Based on Clerical to Technical Hours Ratio 6.69
Based on Clerical to Managerial Hours Ratio 7.50
Approximate Average 7.10
Maximum
Hours
32.30
37.80
35.05
D. ESTIMATION OF THE TOTAL COST OF COMPLETING THE MINIMUM GUIDANCE OPTION
To estimate the total cost of completing a notice under the minimum
guidance option, average hourly wage rates by labor category were multiplied
by the respective number of hours needed to complete the notice. The average
hourly wages were the same as those established by ADL: $10/hour clerical,
$25/hour technical, and $50/hour managerial. The results are presented in
Exhibit AII-5.
Exhibit AII-5
Cost of Completing Minimum Guidance Option
Labor
Category
Clerical
Technical
Managerial
Estimated Hours
to Complete
Alternative 1 Notice
Minimum Maximum
7.1 35.1
22.6 215.4
7.5 35.0
Average
Wage Rate
$10
$25
$50
Estimated Cost
of Completing
Alternative 1 Notice
Minimum Maximum
$ 71 $ 351
$565 $5385
$375 $1750
Total Cost
$1011 to $7486
Thus, the total cost of submitting a notice under the minimum guidance option
ranged from approximately $1,000 - $7., 500.
-------
A-26
III. USE OF ADL METHODOLOGY TO ESTIMATE THE POST OF THE
OCTOBER 1979 NOTICE REQUIREMENT FOR IMPORTERS
ADL estimated the cost of completing each section of the October 1979
notice form for domestic manufacturers. Using these section-by-section
estimates, the cost of completing the October 1979 notice for Importers can be
easily estimated. This is done through the following procedure:
• List each section of the Importer notice
• Beside each of these sections, list the section of the
notice form for domestic manufacturers that requests
roughly the same information. (This section of the
domestic manufacturers' notice form is termed a
"comparable section.")
• Record ADL's estimated ranges of clerical, technical
and managerial labor hours for each comparable section.
• Sum the minimum clerical hours across all comparable
sections and then the maximum clerical hours. Do the
same for technical and managerial hours.
• Multiply the minimum clerical hours by ADL's
prescribed wage rate for clerical labor, do the same
for the minimum technical and managerial hours, and
sum these products to obtain a minimum total cost.
• Repeat the above step for maximum hours to obtain a
maximum cost. The minimum and maximum costs comprise
the estimated cost range of completing the October
1979 notice for importers.
The above procedure cannot be done if a section of the importer notice
does not have a comparable section of the notice form for domestic
manufacturers. Fortunately, this is not the case. In fact, importer sections
are quite similar to domestic manufacturer sections in terms of required
information. Exhibit AIII-1 displays an example of the similarity between
comparable sections of both forms.
The comparable sections that appear to be the most dissimilar are the
sections of both forms dealing with transport. This is understandable, as it
is transport that distinguishes domestic manufacturers from importers.
However, the only difference between the sections is that the importer section
has an extra multiple-choice question on the mode(s) of transport into the
United States. It is estimated that this extra question requires a negligible
amount of labor relative to the section's other labor requirements.
Therefore, the manufacturer's transport section appears to serve as a good
-------
EXHIBIT AIII-1
EXAMPLE OF SIMILARITY BETWEEN COMPARABLE SECTIONS
Importer Notice Section
F»d«»l Regtrtn / Vol. 44, No. 201 / Tngcfay. October 18.1979 / Proposed Ruin
Comparable Section of Notice Form for Domestic Manufacturers
5(792 FeJml R«gtrt»r / VoL 44. No. 201 / Ttmday. October 18.1979 / Proposed Rule,
Secti.n D - U.S. IMPORT AND MARKETING DATA
If you date Mwrt Volne confidential. rear* (X) the box al the rial*
Tne artsweis to item 1 will be Included in this claiB.
1. Estimate Be irrinimum and raaxlnm annual tapert voluM for the first hree yean ol Import. Include in
your estimates import by others with whom you have contracted to import the new chemical substance.
c.TNrdye*
in*
If you claim Use Oat» confidential, nark (X) the box at the right
The answers to Item 2 will be included In the claim.
>D
'• Ust the categories) of use OB which you have based your Import estimates. (Example: solvent used tn automotive paint)
Ust partial information If complete information is not known. (Example: solvent.) Mark (X) the categories ef use as industrial,
cial, or consumer. Estimate the percent of total import volume for the first 3 years devoted to each category of use.
Category of use
(1)
[_J Mcffc thtm Dor It you attach * continuation shMf.
Mark (X) appropriate cotvmnfsj
Industrial
(3)
Commercial
(4)
Consumer
(5)
Conr«fen-
tia! code
». Urt ajy other categories) of ust that you haw actively explored
PI
M*t* A/a box If you attach a conNiuaffon sfteef.
«. 0« yw Intend or expect the new chemrcaf substance to be used* t
ing water supplies or to be used in product
f (hat will come in contact wnti drinking1 w
TOYe* * D "° J D Don'T llrtw
MTE - IT you claim the answers to items 3 or 5 confidential, place the letter(s) A-F in the box which indicates
OH basis of your claim and answer the linkage questions in appendix A, section II for categories A-E.
If you claim any Item sutwittetf tn in *ttacft«rU confidential, see SPECIAL INSTRUCTIONS, appendix A,
section II. part 8.
I. Hn the ditmical substance fae«n manufactured before?
4. ttaard warnings Attach to ttis notice a copy or reasonable facsimile of any hazard warning statement, label, labeling.
(WftmHiT instruct/ens, tecmical data sheet. material safely data sheel, and any other Infemation
which will be provided to any person regarding the safe handling, transport, use. disposal, treatment
upon accidental exposure, or the formulation, construction, or labeling of products containing the
chemical substance.
Q«enr irt/« tor tffwi ft txn « hmx w* *amtng.
S. Enter the number of customefs wfte hw* •ithn caatracJM t»
purchase, subnulted a purchase order, of wade any other firm
commitment to purchase the new chemical substance from you
for a category of ust unknown to you. Estimate the pafCMlaaj*
0* your import volume that wilt e« purchased by such
customers during the first 3 years of import.
Percentage Import
Section D - PRODUCTION AND MARKETING DATA
If you claim Production Volume confidential, mark (X) the box at the right. > I I
The answers to Mem 1 will tie included in this claim.
1. Estimate the minimum and maximum annual production volume for the first three years of production. Include in your estimates
production by others with whom you have contracted to manufacture Ihe new chenlcal substance.
Production year
ID
b. Second y
Production (Kg/yr|
Minimum
(2)
Maximum
(31
Confiden-
tial e<
2. Category of use __
If you claim Use Data confidential, mark (X) the box at the fight * LI
The answers to item 2 will be included in the claim.
a. List the category!IBS) of use on which you hav* based your production estimates. (Example: solvent used in automotive paint.) List
partial information if complete information is not known. (Example: solvent.) Mark (X) the categories of use as site limited, industrial,
commercial, or consumer. Estimate the percent of total production for the first 3 years devoted to each category of use.
Category of use
-(1)
Production
percent
Site limited
(3)
Uark (X) apfnoprlato c
Industrial
(4)
Commercial
(5)
Consumer
(6)
Wank this box It you attach a continuation sheet.
fe. List any other category,;ies| of use that you have actively explored
J Afar* this box II you attach a continuation sheet.
c. Do you Intend or expect the new chemical substance to be used to
treat drinking water supplies or to be used in products (e.g., paints
or coatings) that will come m contact with drinking water?
iQVes
3Q Don't krww
NOTE - !( you ctoim the answers to items 3 or S confidential, place the letter(s) A-F In the box which indicates
the basis of your claim and answer the linkage questions in appendix A, section II for categories A-E.
If you claim any item submitted in an attachment confidential, see SPECIAL INSTRUCTIONS, appendix A,
section (i, part 8.
3* Has the chemical substance been manufactured before?
2 Q No 3 Q rjon't know
4. Hazard warnings
'Attach to this notice a copy or reasonable facsimile of any hazard warning statement, label, labeling,
marking or instructions, technical data sheet, material safety data sheet, and any other information
which will be crowded to any person regarding the safe handling, transport, use. disposal, treatment
upon accidental exposure, or the formulation, construction, or labeling of products containing the new
cnermcal substance.
j Uerft thit to* llyov MfacA • nunrtf warning.
5. Enter the number of customers who have either contracted to
purchase, submitted a purchase order, or made any other firm
commitment to purchase the new chemical substance horn you
for a category of use unknown to you. Estimate the percentage
of your production volume that will be purchased by such
customer! during the lirst 3 years of production.
Cond den-
ial code
-------
A-28
basis for the labor estimates of the importer section. Both sections are
presented in Exhibit AIII-2.
Exhibit AIII-3 outlines the costing methodology and displays the results.
Note that in some cases, an importer will not have to fill out Section II-A
(e.g. for imports that are distributed in the U.S., and not processed by the
importer). The lowest cost of filling out Section II-A, estimated at $300, is
therefore subtracted from the minimum cost estimate. The cost estimate for
the importer's form is therefore $855-$8925.
-------
Importer Section
EXHIBIT AIII-2
IMPORTER AND DOMESTIC MANUFACTURER NOTICE REQUIREMENTS ON TRANSPORT
Comparable Section of Notice Form for Domestic Manufacturers
/ Vol. 44. No. 201 / Tutaday. Oclobn IB. 1978 / PropoMd Rule.
k SKttM E - TRANSPORT
If you clain the answer* to Items 1 or 2 ch.flchH.tixf. place the fetterf,s1 (A-F) In the box which indicates Urn
baits of your claw MM! animr ttw linkage questions in appendix A, lection H lor categori** A~E.
1. CnMt tta proper DOT sMppinf ft.wM and hazard elass of the MM etetfrriejl *ubstanc« (K appflcabtal.
a. Sh.pp.tig »*"»
b. Kazan) elau
2. HaA (X> the modets) of transpoft which you believe will be used for the new chemical substancs IB enter (he
US. and within the U.S.
a. T« enter the United States -
I Q Truck i Q Baf|«, «ss«l • Q Plawi
k. Within »m IMM States -
iQJ Truck sQBarft, vt»s*l tQPlaM
1 z Q Raifear 4(3 PHwlh* «D Ottar - «wc/.V
Confiden-
tial end*
t .Wctioii F . RISK ASSESSMENT
» Section E - TRANSPORT
Complete this section it you intend to ship (he new chemical substance horn iti site of manufacture.
if yflu claim the answer* to items 1 or 2 conlidential , place the .t?ner (__ ; Bar|B, vessel
«[-]Pio.»lim
Confiden-
tial code
NJ
<£>
-------
A-30
Exhibit AIII-3
Estimated Cost of Completing October Notice for Importers, Based on
ADL Estimates for October Notice Requirements for Domestic Manufacturers
a/ b/
Sections of PMN for Importers5' Comparable Sections of PUN for Manufacturers-
I. General Information, A. Importer
Identification, B. Chemical Identity,
C. Generic Names, D. U.S. Import and
Marketing Data, E. Transport, F. Risk
Assessment, G. Detection Methods.
II. Human Exposure and Environmental
Release, A. U.S. Industrial Sites
Controlled by the Submitter, (may not
be applicable to all importers) B. U.S.
Industrial Sites Controlled by Others
C. U.S Consumer and Commercial User
Exposure.
III. List of Attachments
IV. Federal Register Notice
Estimated Hours of Importer
Clerical Technical Managerial
I. General Information, A. Manufacturer
Identification, B. Chemical Identity,
C. Generic Names, D. Production and
Marketing Data, E. Transport, F. Risk
Assessment, G. Detection Methods.
II. Human Exposure and Environmental Release
A. Industrial Sites Controlled by the Sub-
mitter, B. Industrial Sites Controlled by
Others, C. Consumer and Commercial User
Exposure.
III. List of Attachments
IV. Federal Register Notice
Total Hours
X Cost/Hour
2-10
4-20
7-59
7-144
2-13
2-10
1-8
1-2
12-56
1-8
3-12
1-2
8-40
27-267 8-37
X 10
X 25
X 50
Minimum Total Maximum Total
Cost Cost
$1,155 $8,925
Note: Because importers may not
always submit Section II-A
(estimated to cost
$300-$2100), the minimum total
cost is reduced by $300.
Cost Range; $855-$8925
5/44 Federal Register 59852-59866, October 16, 1979.
•^44 Federal Register 59788-59802, October 16, 1979.
°/Estimates taken from Arthur D. Little, Inc., Estimated Costs for Preparation and Submission of Reproposed Premanufacture
Notice Form, prepared for the EPA Office of Toxic Substances, September 1979.
-------
APPENDIX B
ASSESSMENT OF THE FEASIBILITY OF
DEVELOPING A METHODOLOGICAL
FRAMEWORK FOR FORMAL ECONOMIC
IMPACT ANALYSIS
-------
ASSESSMENT OF THE FEASIBILITY OF DEVELOPING A METHODOLOGICAL FRAMEWORK FOR
FORMAL ECONOMIC IMPACT ANALYSIS
This report presents the results of a brief (two-month) small-scale review
undertaken by ICF Incorporated to assess the technical feasibility of
developing a formal economic approach to analyze the impact of TSCA's Section
5 notice requirements on the U.S. chemical industry. In conducting this
review, ICF began with, and drew heavily upon, a preliminary working paper by
Arthur D. Little, Incorporated. The ICF review was primarily directed toward
two interrelated goals:
(1) Assessing the technical feasibility of a formal analytical
approach to understanding the impacts of notice
requirements on the chemical industry—an approach that
could be used in the notice requirement economic analysis
to be completed within six months or so.
(2) Learning from the existing literature on innovation the
significant probable factors affecting chemical industry
innovation so that later tasks of the project could be
focused upon these factors.
A secondary goal of the task was to comment upon the analytical approaches
proposed in the working paper, approaches which have a longer-term and broader
objective than the notice requirement analysis: the construction of an
analytical tool or model of the U.S. chemical industry which could be used to
examine ex ante effets of a wide variety of TSCA reguations.
This report is organized as follows:
SUMMARY OF FINDINGS—summarizes the conclusions of the
report in response to five specific questions asked by EPA
in its work statement.
INTRODUCTION—defines the objective of the study and
discusses briefly the reasons for the emphasis of the
following sections.
LITERATURE REVIEW—briefly summarizes the relevant
literature.
FORMAL ANALYTIC APPROACHES—outlines the alternative
approaches for formal analytical modeling and discusses
their data requirements, benefits, shortcomings, and
costs.
BIBLIOGRAPHY of relevant sources.
-------
B-2
SUMMARY OF FINDINGS
This section summarizes the findings of this report in response to the
specific questions asked by EPA in Task Order 3 of November 10, 1979.
(1) We have argued against the formal analytic approaches recommended by
ADL (econometric models) because, most importantly, there is no apparent way
to use the tools suggested to test the effects of TSCA regulation on chemical
industry product innovation, the key issue not only for notice requirement
effects but probably for TSCA as a whole.
(2) We have reviewed the generally thorough ADL literature search and
other literature and summarized its utility. It provides some indication of
probable factors affecting chemical industry innovation, but is not very
useful in providing either theory, methods, or data to support formal
analytical approaches.
(3) We have concluded that neither the ADL approaches to formal analysis
nor the ICF alternative can be completed in time to study notice requirement
impacts.
(4) We have produced a conceptual model of the chemical innovation
process and outlined the data and relationships necessary to understand it
better, as an input to the notice requirement economic impact analysis task.
(5) We have summarized for EPA the longer-run feasibility of formal
technical approaches and reached the following conclusions:
(a) The best hope for a model of product innovation is a simulation
approach which relies upon simulating the R&D resource allocation decision and
commercialization decision, based upon risk, cost, expected payoffs, and
timing of these factors. TSCA regulatory impacts would be assessed through
changing the costs, timing, and probabilities of payoffs.
(b) If the product innovation decision can be modeled (see below),
there are a variety of interesting ways to trace economic impacts through
industry simulation or optimization models. We recommend deferring these
interesting questions until the feasibility of (a) is tested.
(c) The best approach to test the feasibility of (a) is to run a
small-scale pilot on one industry segment with high product R&D. This will
require original data collection from industry, take about nine months, and
cost about $100,000. It is by no means certain it will result in a useful
model (although it would necessarily shed some light on the R&D practices in
that segment).
-------
B-3
(d) if (c) is tested and fails, EPA could at a later date consider
formal industry models of the kind proposed either by ADL or ICF. Such a
project, realistically, is a three to five year (over $1 million) effort, and,
although it would be a useful tool for many purposes, could not be effectively
used to assess ex ante the economic impacts of TSCA regulations on innovation.
(e) if (c) is tested and succeeds, EPA could begin to build the
larger chemical industry model block by block, with the first work being to
construct the model of the industry segment for which (c) was built and, then
integrate the models.
INTRODUCTION
Objectives
The objective of this task is to assess the technical feasibility of
developing a formal economic approach to analyze the impacts of notice
requirements (and by implication those of other TSCA regulations) on the U.S.
chemical industry. As will become clear below, it is extremely important to
understand the ramifications of this objective:
• The notice requirements to be modeled apply to new chemical
products, which inevitably turns the focus of the study to
new product development; the innovation or research and
development process in the chemical industry.
• As amplified below in the literature summary, there is very
little in the literature which can help with this task.
• Because of the sophisticated tools available for industry
econometric modeling (and through input-output analysis for
tying industry effects to overall economic effects), there
is a strong tendency to propose the construction of a large
chemical industry econometric model. Without doubt, some
industry model is necessary to track the impacts of
innovation effects and change once these impacts are
understood. But, unless the innovation process can be
understood well enough to model so that the impacts of
regulation upon it can be understood, it is of no use to
have a sophisticated means to track the effects of these
impacts. Note that although a partial or general
equilibrium econometric model would have some use for TSCA
for issues other than notice requirements, it is our view
that TSCA effects on innovation will be such a fundamental
point of contention about impacts that any large modeling
effort must address innovation.
-------
B-4
• There are considerable dangers in over-aggregation. It is
by no means clear that there is a "chemical industry" in
the sense of, say, coal or Pharmaceuticals, but instead
perhaps a set of "industries," with little in common, that
must be addressed separately for many analytical purposes.
For example, the "Perfumes, Cosmetics, and other Toilet
Preparations" industry segment (SIC 2844) may have less in
common with, say, "Explosives" (SIC 2892) than the steel
industry has with, say, aluminum.
Brief Summary of Literature
Because the organization and conclusions of this report depend directly
upon the findings from the literature search, we have highlighted these
findings in this section. There are four primary bodies of literature of
possible relevance:
(1) The formal microeconomic literature on industrial innovation.
Neoclassical growth theory has struggled for many years to deal with
technological change. In the 1960s economists developed a means of
incorporating technical change in the neoclassical growth model as the "theory
of induced innovation"—but in a very formal manner which is of little use in
the current task. This is because, first, the theory deals with what we would
call "process innovation"—changes in the prices and quantities of input
factors for given outputs..?/ New product development falls completely
outside these theories.
Next, the theory has been developed largely at a macroeconomic level.
Finally, there seems to be little theory in the economics literature on the
role of uncertainty in the innovation process—either at the firm level,
industry level, or macroeconomic level. It is possible to distinguish between
"technical uncertainty" (the uncertainty that the scientific and engineering
development process will produce something of potential use) and "market
uncertainty" (the uncertainty about the degree of demand for a product in the
future and about how actions of competitors may affect payoffs). It is also
theoretically possible to model "technical uncertainty" through a probability
cash flow model, "market uncertainty" through a game-theoretic approach, and
the two together through incorporation into a portfolio-choice capital
budgeting model for R&D expenditures. But there is almost no body of
microeconomic theory to rely upon, unlike other areas of industry and firm
behavior.
I/Even here, there is much confusion and apparently no clear way
empirically to separate movement along an isoquant (factor substitution) and
shifts in the isoquant (technical change). Because of the difficulty in
separating the two (in either case, the optimal factor ratio changes in favor
of more use of the less expensive factor), several economists have recommended
abandoning the distinction entirely.
-------
B-5
Although economists are now moving into this field to close the gaps
mentioned, it is worth repeating what is to us a surprising finding: there is
at present very little economic literature or theory on product innovation.
On the one hand, this makes the problem doubly interesting from an academic
viewpoint (and because of this EPA should maintain a clear focus on new
advances in model-building); on the other hand, economic modeling of product
innovation (quite apart from the even more difficult issue of TSCA impact on
that innovation) puts us at a frontier of economic theory without tested or
accepted approaches upon which to rely.
(2) Empirical studies of innovation. There are extensive studies,
generally at the industry level, which attempt to analyze and understand the
industrial research and development process. These range from studies which
are quite analytical and use formal models and actual historical data (usually
time-series or pooled time-series/cross sectional multiple regression analysis)
to qualitative case studies. Industries covered are primarily drugs, chemi-
cals, glass, steel, or petroleum. Attempts are made to relate the amount of
R&D expenditures to a wide variety of firm and industry parameters and to
relate "outputs" such as sales of new products to amounts of R&D expenditures.
Despite the recent work of Mansfield, Nelson, Iverstine, and others, there is
not much empirical detail on the resource allocations to the various stages of
the R&D process, on how and why firms make R&D allocation decisions, on how
firms evaluate R&D expenditures, or on how firms (in a game-theoretic sense)
decide what to patent or bring to market. As discussed below (and in the ADL
survey), this literature provides some ideas about the determinants and
decision-making in the R&D process, but the bulk of it is not very useful for
several reasons:
• It is based upon old data (often, say, 1954-1959) and
thus upon structural and societal relationships and
forces which arguably no longer pertain.
• It is conflicting in many of its findings.
• It is often based upon questionable data and
surrogates for innovation (due to lack of useful data)
and at rather high levels of aggregation (i.e., even
at the large company level, there are many very
different "businesses" being aggregated).
• There are few studies of the chemical industry and few
significant conclusions from these studies.
-------
B-6
(3) Specific studies on the effects of government regulation on
innovation. There has been reasonably extensive study of the impacts of the
1962 Amendments to the Food, Drug and Cosmetic Act on innovation in the
pharmaceutical industry and to a lesser degree on the effects of the Federal'
Insecticide, Fungicide, and Rodenticide Act (FIFRA) on the pesticides
innovation process. As discussed below, although these studies highlight some
areas of interest to TSCA, they are of little use here. Either:
• the studies are generally post hoc analyses with
single equation approaches or international
comparisons, neither of which is particularly helpful
to TSCA's objective of developing an approach to
assess impacts ex ante;
• the studies deal with regulations very different from
notice requirements (and different from TSCA more
generally); or
• the studies are data constrained, in some cases use
questionable surrogates for innovation, fail to
separate the regulations themselves from concurrent
societal forces, and provide little data on how
regulations have affected R&D decision-making by
firms.
(4) Studies from the "business literature". There is a growing
literature on R&D, innovation, and technical change arising from managerial
sources, which provides some useful empirical insights. The literature ranges
from studies of R&D strategy to the extensive work done on capital budgeting,
decision-making under uncertainty (particularly Bayesian statistics), and
portfolio/risk theory. Although these are useful grounds upon which to model
certain business decision-making, much of this literature is too abstract and
complex to be applied to the variety of common business decisions: there
remains a significant gap between the case studies of individual firms and a
formal analysis of a business decision under uncertainty. Yet, in view of the
necessity to understand the effects of notice requirements on the chemical
industry and of the limited usefulness of the literature cited above in (1),
(2), and (3), it is apparent that how and why chemical industry decision-
makers allocate R&D resources, assess probabilities, and make decisions about
development and commercialization is the critical first factor which must be
understood or modeled if we wish to know how notice requirements may affect
this process.
LITERATURE REVIEW
We noted in the introduction that the existing literature has little to
offer in developing a formal approach to analyze the impacts of notice
requirements on the chemical industry. In this section a brief summary of the
literature is provided. Numbered citations in the text refer to the bibli-
ography at the end of this appendix.
-------
B-7
There are four areas of the literature:
• microeconomic theory of industrial innovation;
• empirical studies of innovation;
• specific studies of the effects of government regulation
on innovation; and
• "business" literature.
Microeconomic Theory of Industrial Innovation
In this category, the review by ADL is more than adequate; therefore, the
conclusions documented in the ADL report are summarized and restated:
• The theory of induced innovation that is incorporated
into neoclassical growth models addresses only
process innovation.
• Research and development costs are seldom even
included in models; when they are, there are serious
difficulties. Economists do not agree whether the
firm has a choice in the direction of technical
innovation; some believe that to admit that firms can
control the rate or direction of technical change
requires abandoning competitive models.
• Most past efforts have been in a macroeconomic
framework.
• Existing models do not deal either with technical
uncertainty or market uncertainty, or with how
allocation decisions would (or should) change with
changes in these factors. TSCA regulation is
expected to affect both technical uncertainty (by
increasing the number of technical criteria a product
must respond to) and market uncertainty (by adding
costs, increasing delays, weakening competitive
position, reducing payoffs, etc.); there is no
systematic means to deal with regulatory effects in
these models.
Empirical Studies of Innovation
In this category, the ADL review was also thorough. Conclusions in these
studies are often contradictory; for example, on how diversification affects
innovative activity: Grabowski finds a positive effect, Scherer finds no
systematic effect, and Weiss finds a negative effect. Even when there are
findings—e.g., that R&D intensity (as measured by R&D employment regressed
against sales) tends to rise with firm size in the chemical industry—there
are no straightforward implications for the task at hand.
-------
B-8
In broad summary, here are some findings from this literature:
• At the firm, industry, and economy level, there is
high correlation between R&D expense and economic
growth/ productivity. The ADL report cites R&D
intensity as a "source" of growth (although noting
that causality may run in both directions); more
conservatively stated, high R&D is associated with
high growth and positive changes in productivity.
• Chemical industry R&D intensity appears to rise with
firm size, yet R&D per sales dollar is higher with
smaller firms.
• There is no consensus on the relationship between
innovative activity and industry concentration or
degree of product differentiation.
• Several studies suggest after-tax average returns on
R&D capital of 15 to 20 percent.
• One expert (Kamien) estimated that more than
three-fourths of U.S. industrial R&D was directed
toward new products rather than new processes.
Specific Studies of the Effects of Government Regulation on Innovation
ICF looked closely at the previous work in the pharmaceutical and
pesticides industries because, despite its ex post nature and regulatory
differences from TSCA, the general approach used at least addressed directly
the relationship between regulation and innovation. Thus, the following
section provides a little more detail than the previous two summary sections;
without doubt, TSCA will be subject to analyses similar to these in the future.
This summary of the literature is a brief guide to what is known about
the effects of regulation on innovation in the pesticide and pharmaceutical
industries and how it may relate to the chemical industry. References are
made to major empirical studies in the bibliography.
The Pharmaceutical industry. A number of studies have been done on the
effects of the 1962 Amendments to the Food, Drug and cosmetic Act on
innovation and R&D in the pharmaceutical industry. The most significant
findings are summarized here in terms of the following four hypothesized
impacts of regulation on innovation. Regulation:
(1) lowers return on R&D due to increased costs, risk, and
development time associated with regulatory compliance,
leading to a lowered rate of investment in R&D and
potential decline in innovation;
-------
B-9
(2) changes the pattern of R&D and innovation (e.g., away from
basic research to emulative products or from product to
process innovation);
(3) changes market structure which may affect the rate and
pattern of innovation; and
(4) promotes international transfer of resources and
technology from countries with strict regulatory
requirements to those with less strict requirements.
!• Level of Innovation. Based on Baily's study (7) showing an increase
in new drug discoveries resulting from increased R&D expenditures, several
authors have investigated the expected rate of return on pharmaceutical R&D
investment (19, 82). Schwartzman, for example, finds a drop in the rate of
return from 11.4 percent in 1960 to 3.3 percent in 1972 due to increased costs
posed by government regulation.
Hansen (38) estimates a development cost of $54 million (1976 dollars)
for a new chemical entity (NCE) and about nine years for the NCE to reach the
market. Many others have documented the decline in introduction of NCE's
since 1962. (See 43, 55, 74, 90, 91.) The exact portion of that decline
directly attributable to regulation is, however, still open to question.
Both Baily (7) and Peltzman (74) find a statistically significant negative
effect of regulation on the number of new chemical entities introduced in the
U.S. after 1962, but serious criticisms of these empirical models exist. For
example, Baily's supply-side model ignores the effects of changes in demand,
and both Baily and Peltzman neglect to provide an adequate measure of "research
depletion" or to explore the possibility that any exhaustion of research
opportunities actually existed.
Using a comparative analysis of the British and American cases to get at
the same issue, Grabowski, Vernon, and Thomas (32) find a six-fold decline in
R&D productivity in the U.S. between 1960-1961 and 1966-70, compared to a
three-fold decline in the U.K. in the same time period. The difference between
the two countries may be part of the well-known "drug lag", suggesting that it
was the particular legal and institutional form of regulation in the united
States that has been most responsible for negative effects on innovation.
As for the three-fold decline in the U.K. rate, it is again not possible
to distinguish the effects of increased regulation in that country from
research depletion or from changes in demand due to a different social and
medical context.
The Arthur D. Little paper (4) emphasizes the importance of a simultaneous
equation approach to modeling innovation in the drug industry, rather than
single-equation models which model either the supply side or the demand side,
but not both together. The 1962 legislation in the U.S. (and regulation in
other countries) was one aspect of a change in attitudes directly related to
the thalidomide incident which arguably affected demand for new drugs. On the
supply side, the possibility of exhaustion of research opportunities, perhaps
due to gaps in scientific knowledge, is an important and controversial issue
that has not been effectively addressed in econometric studies of this subject.
-------
B-10
2. Type of innovation. The position taken by the FDA is that the
decline in new drugs since 1962 has been a result of the focus on new chemical
modifications of existing drugs, rather than the inhibition of new therapeutic
advances. This position is a controversial one both factually and in its
tacit assumption that society has not lost much from the decline in slightly
modified drug products. The nature of new drug innovation is dealt with in
several sources (see 39, 74, and 81).
3. Market Structure. A great deal of literature has dealt with
questions of market structure and innovation in the pharmaceutical and several
other industries. Most writers have explored the relationship of firm size to
innovation in an attempt to empirically assess Schumpeter's hypothesis that
large firms are associated with greater inventive output (80) . (See, for
example, 5, 12, 24, 35, 37, 59, 78, 81, and 88.) It appears that there is
still little consensus on the validity of this hypothesis.
Kamien and Schwartz (46) provide a useful survey of research done on this
question, noting the two aspects of the "scale" issue: the effect of firm size
on efficiency for a given size R&D facility, and the efficiency of different
sized R&D facilities within a firm. More recent studies have investigated the
effects of rivalry on the innovation process (31, 47, 66).
As in the pesticide industry, active ingredient R&D on Pharmaceuticals is
conducted by a limited number of firms (approximately 100) and is generally
confined to the larger companies because of high development and testing costs.
4. International Transfer. A fair amount of evidence has been amassed
indicating that there has been a geographical shift from the U.S. to overseas
in the location of pharmaceutical innovation since 1962 as a result of differ-
ential regulatory requirements. The pharmaceutical industry has a strong
multinational component, so that a shift in R&D and production can be effected
with some ease. It is difficult, however, to separate out the effects of regu-
lation from other cost-related factors on this trend. (See 32 and 33.)
The Pesticide industry. Very little literature exists on the effects of
regulation on innovation in the pesticide industry. The primary sources of
data are the industry surveys carried out by the National Agricultural
chemicals Association (69). These surveys, however, vary from year to year in
the number of firms participating and the categories used, so that a detailed
analysis of trends in innovation is not always possible. Some idea of recent
patterns in the industry can, however, be gleaned from the NACA data and other
sources.
Expenditures on pesticide research and development have continued to rise
in the last decade, keeping pace with the increase in pesticide sales. The
cost of bringing a new chemical to the market, though, has risen from estimates
-------
B-ll
of $5.5 to $6 million in the early 1970s to a current estimate of $15 to $20
million (including the cost of unsuccessful compounds). Total elapsed time
from discovery of a pesticide to marketing has also been increasing, from an
average of 60 months in 1967 to 110 months in 1977. Although the share of R&D
funds spent on discovery and commercialization of new products has been roughly
constant in the last five years (near 65 percent), a much higher proportion of
R&D expenditures is now devoted to regulation-related activities. The result
is that fewer new commercial products have been forthcoming, and it is assumed
that the number of annual new registrations of active ingredients with EPA
will not exceed 10 to 15 in the near future.
Unfortunately, no precise linkage of the changes in pesticide innovation
with the effects of FIFRA regulation is possible. The pesticide industry is
relatively mature (70) and it is generally believed that most major markets
have already been filled with effective products, though product obsolescence
is common in the industry. On the other hand, registration requirements have
probably exacerbated the trend toward increasing modification of existing
pesticides for new uses rather than developing new chemical entities.
There is, in addition, a wide variety of anecdotal evidence that direct
and indirect effects of regulation have hindered both the development of
products for minor crops and uses, and innovative work in the development of
biological alternatives to chemical pesticide control, it is believed that
the few small firms that are involved in pesticide R&D have been involved in
some of this innovative work (23)• The effects of regulation on market
structure, although limited, may therefore be indirectly injurious to some
sources of innovation.
The Chemical industry. The relevance of the experience of the pesticide and
pharmaceutical industries under premarket regulation to the effects of TSCA on
the chemical industry is severely limited. Apart from methodological problems
in the formal analytical approaches contained in previous studies which limit
their usefulness for studying the chemical industry, there are important
substantive differences among the industries.
Pesticides and Pharmaceuticals are not typical segments of the larger
chemical industry. The levels of R&D conducted in the industries differ
significantly, (about 10 percent of sales in Pharmaceuticals, eight percent in
pesticides, and two to four percent for chemicals). Although the lack of data
available for the chemical industry precludes definitive statements, it is
suspected that other important differences exist as well. Examples are the
expected value of sales of new products, the barriers to entry into R&D, and
in general, the effect of firm size and scale of R&D facility on innovative
potential (see 26 and 30). Whether or not the role of small firms in chemical
R&D and whether or not the research opportunities facing the chemical industry
are substantially different than the pesticides and Pharmaceuticals cases
could also modify the effects of regulation on the industry.
Variations in the regulatory process faced by different industries may
likewise be important. Premanufacture notification does not require specific
testing or explicit approval, and has a maximum review period of 90 days (may
be extended to 180 days) after which a chemical may be introduced into
commerce. The FIFRA pesticide registration program and FDA's investigative
New Drug (IND) approval system require explicit approval after review of test
results before a product can be manufactured, and the review period can extend
-------
B-12
indefinitely. Similarly, the treatment of confidential information in the
regulatory process may affect not only the rate of innovation but the type of
innovation (i.e., variations on existing chemicals or entirely new types of
chemicals) and the location of activity.
Measurement of innovation appears in the literature on Pharmaceuticals as
a serious obstacle to empirical study. The problem is aggravated in the case
of the chemical industry, where even the annual number of new chemical entities
introduced is unavailable from a single authoritative source. Some of these
issues are dealt with in (52), (66), (77) , and (78), and in the final section
of this appendix.
For these reasons, it is not likely that the major studies done on the
pharmaceutical industry or the experience of the pesticide industry can be
easily extrapolated to predict TSCA's effect on the chemical industry. No
doubt some of the determinants of innovation in the industries will be common.
However, it is not our expectation that the results of regulation in the
chemical industry as a whole will conform to the results in two of its rather
atypical sectors. For example, the nature of R&D in both Pharmaceuticals and
pesticides is more product-oriented than process-oriented, which is generally
believed untrue for at least the high-volume segments of the chemical industry;
most new products (active ingredients) in the former cases are expected to
penetrate large markets and have development costs of $20-30 million over the
course of 5 to 10 years.
What this points up is the heterogeneity of the chemical industry, and
the need to disaggregate the various sectors it encompasses. it appears that
pesticides and Pharmaceuticals are at one end of the spectrum of chemical
industry subgroups in terms of innovation. Other patterns of innovation and
of regulatory effects on innovation are more characteristic of the chemical
industry in general. A fuller discussion of the disaggregation of the
industry in order to properly model the effects of TSCA on innovation is
developed in the Formal Analytic Approaches section below.
The "Business" Literature
Apart from the extensive theoretical literature on capital budgeting,
decision analysis under uncertainty, and portfolio risk analysis, the business
literature also contains examinations of individual firms or innovations and
behavioral studies relating organization and management approaches in R&D to
successful outputs. Both these approaches look carefully at actual firm
performance. Some major conclusions:
• The bulk of R&D projects in large chemical firms are
relatively safe from a technical point of view (59).
• Expected rates of return (if successful) were about
30 percent—arguably about the same, after
probability adjustments for some failures, as other
capital investment (59).
• Firms devoted about nine percent of R&D funds to
"basic research," 45 percent to "applied research,"
and 46 percent to "development" (60).
-------
B-13
Even for research-intensive industries like
chemicals, innovations are often based upon
technology or research derived from organizations
other than the innovating firms (68).
There is considerable anecdotal evidence that in
significant parts of the chemical
industry—especially for small firms and in the
specialty chemicals area—a great deal of product
innovation is closely tied to customers and the
market (i.e., there may be little or no "market
uncertainty").
There has been considerable development of the
product life-cycle theory which may have application
in longer-term modeling of new chemical products.
In case studies of successful new product innovations
with major impacts, factors cited (49, 74) as sig-
nificant (ceteris paribus) in promoting innovations
are:
Effect on
innovation
firm size +
relative R&D spending: firm to industry +
industry growth rate +
stability of industry in terms of -
structural products
degree of involvement of R&D groups +
with potential users
degree of status, authority, and involvement +
of top R&D manager with the project
in-house initial support for the innovation +
on commercial grounds
relative sales effort for the new product, +
including publicity/advertising
Summary
As noted in the introduction, the existing literature serves mainly to
describe what we do not know about chemical industry innovation and how to
model it.
The conclusions we draw are:
• A major focus of any formal analytic approach must be to derive
a model of the R&D innovation process itself to determine how
notice requirements or other TSCA regulations would affect the
factors which determine how R&D input allocations are made, how
-------
B-14
the proess itself actually operates, and how commercialization
decisions are made—rather than treating R&D as one or more
production/cost function equations in a large, econometric
industry model.
• Whether building a formal model of the innovation process is
worth doing, under existing data and resource constraints, is
problematical.
• The chemical industry subcomponents are so different along the
major dimensions which might affect the R&D process that no
"general" aggregate model of innovation is likely to be useful.
• Developing a formal analytic approach, if it can be done
usefully, should begin by modeling one subcomponet of the
industry (say, a four-digit SIC code) and by focusing upon
innovation outputs.
• Because completing the step above, if it can be done, is about
a nine-month job, a formal analytical model cannot be in place
in time for use in the notice requirement economic impact
analysis.
• But, if the R&D/innovation area can be usefully modeled—so
that changes to the process caused by TSCA can be separately
identified and incorporated in such a way as to lead to output
changes—EPA would probably be well-advised to begin the long
complex task of modeling the industry (as discussed below).
FORMAL ANALYTIC APPROACHES
Before discussion of formal analytic alternatives, it is important to
look carefully at the chemical industry R&D process at the firm level—and
most particularly, at which parts of that process notice requirements (and
more broadly TSCA) might affect. Thinking through the process at the most
dis- aggregated level (one firm) and with the stages of innovation addressed
indi- vidually allows us to decide which combinations and aggregations may
later make analytical sense in testing TSCA impacts.
It appears evident, from the diverse literature addressing this process
as a whole as well as individual elements, that each block of activity may be
distinct and influenced by different factors. Further, the nature of the
activity will vary widely depending upon which segment of the chemical indus-
try is being considered and even by the individual firm within that segment.
We believe it is very important to look at each block of activity and to
hypothesize where notice requirements/TSCA effects might fall before any
consideration of modeling.
Although we will discuss each step in the process in more detail in a
following subsection on industry and microeconomic impact issues, let us
summarize the activities at each step of the process:
-------
B-15
The Innovation Process
Consider the following model of the chemical innovation process:
INNOVATION PROCESS
BLOCK 1
Resource Allocation
(Capital, Personnel Plans)
BLOCK 2
R&D Allocation (Basic,
Applied, Development:
Product/Process)
BLOCK 3
Creative Research Process
(Ideas, Plans, New, Non-Tested
Products/Processes)
BLOCK 4
Commercial Development
(Products/Processes)
BLOCK 5
Market Introduction
and Response
Outputs
Sales
Profits
Substitution
-------
B-16
1. Allocation of R&D Inputs. The firm, through some decision-
making process, decides to allocate dollars, people, office space, contract
monies, and other resources to research and development. Some of these
allocations may be influenced by existing R&D projects or expenditures; may be
required by specific market forces; may be a result of a formal decision
process or a rule-of-thumb; may be done at various levels of explicitness; and
may be on an annual or project budget. But the firm has a cost of capital and
alternative allocation possibilities (capital investment, marketing,
dividends, hiring Washington lawyers, raising executive salaries, etc.). In a
formal or informal manner, a decision maker in the firm is assessing the
potential return from R&D in relation to other uses of capital.
2. Allocation of R&D Resources to Projects. Possibly in
combination with Block 1, on an incremental or total basis, somebody in the
firm sets objectives, defines projects, and allocates resources to projects.
These projects may be long-term or short-term; "basic", "applied", or
"development" research, aimed at new products or processes, or at cost
reduction on existing processes; focused at a specific market opportunity
("demand-pull") or just an interesting product quality ("science-push").
3. The Research Process. According to some organization and
management rules or procedures, the project research proceeds. The activity
could range from a scientist alone at a bench for months, to close work with a
potential user, to work under contract at a university; it could be cost
engineering rather than "science"; it could be managed in a wide variety of
ways; it could be under intense time pressure or very loose; it could take
days or years.
4. Commercial Development. After a process, process change, or
product is created, a decision is made as to whether and how it should be
brought to market. Assessments will typically be made of the potential
market, the likely cost of production, the possible market price, what the
competition will do, the probability of success, timing, the firm's capital
availability in relation to other opportunities, whether it can or should be
patented (if it has not been in Block 3), and how it should be introduced or
promoted.
5. Market Introduction and Response. Initially, and over what may
be a very long product (or process) life cycle, the product is produced and
sold. Sales may go up, cost may decrease as volumes increase, a similar but
inferior product may be driven from the market, prices may rise or fall
depending upon demand and competition, and a host of impacts may be traced
throughout the industry and economy.
TSCA Impacts
We should note several points about TSCA's impacts on these blocks of
activity. First, TSCA could have important impacts on Blocks 1 and 2 (by
increasing costs and decreasing expected returns for product R&D). TSCA would
impact Block 3 by potentially requiring more tests which would delay produc-
tion of the new chemical and increase its cost; and by potentially causing a
company to rule out certain toxic families of substances as potential prod-
ucts. TSCA possibly would have a major impact both on the decision process in
-------
B-17
Block 4 (deciding wnat to bring to market); and on Block 5 (say, by competi-
tors getting to market more quickly as a result of obtaining previously
confidential information as a result of the notice requirements process).
Alternatively, it could have almost no effect, except perhaps the cost of
filing in the notice requirements form, importantly, though, the actual
decisions in Blocks 1, 2, and 4 will be made on a "marginal" basis, for that
project; publicly available data is average data on expenditures and returns,
and conclusions drawn from this average data about past projects are not
necessarily reliable in judging future response. This point—that expecta-
tions of returns are critical in R&D decisions—is a major reason that we are
pessimistic about the utility of econometric modeling. We are extremely
skeptical that all R&D could be captured for an entire industry segment by one
equation in a simultaneous equation model, and that even if a model could,
through use of a translog cost function, derive an elasticity of demand for
R&D based upon publicly available historical data, there would simply be no
way to test TSCA impacts with such a model: many of the more important
effects, for example, could occur on the margin and within the same R&D costs.
Thus we see the effects on these innovation process activities as being
consistent with the effects measured in our economic impact analyses. New
products will potentially not be introduced because TSCA may (1) increase the
cost of development; and (2) introduce delays in development. For those
products that do reach the marketplace, their prices will be higher (reflect-
ing their higher cost of production) and, therefore, demand for them will be
less than it otherwise would be. The extent to which these two changes (fewer
new chemicals introduced, higher prices for introduced chemicals) occur will
be the output of the R&D model.
Because of the need to trace industry and macroeconomic impacts of this
output, it is necessary to understand how the "outputs" of an innovation model
could be used to develop industry impacts before designing the innovation
model. We address this issue in the next section.
Industry and Macroeconomic impact issues
The major problem in producing a useful policy model for TSCA is the
construction of the "innovation" model described in outline form above and the
integration of this model with a model of the industry and economy which can
be used to determine economic impacts of changes in new product patterns. We
do not plan to discuss extensively the creation of the latter model(s), be-
cause without an effective innovation model the impact model cannot be of much
use to TSCA. However, several points need to be made on the issue of broader
industry models, because first, we must have some idea of how the outputs of
an "innovation model" would be used to assess overall industry economic pacts
and, second, we have been specifically asked by EPA to comment on the ADL
report and on approaches to longer-term broader modeling.
One thrust of this entire paper has been that econometric modeling
(whether by simultaneous equations or otherwise) does not hold out hopes for
understanding TSCA impacts on innovation, and we do not recommend any such
modeling. It seems particuarly clear that a simple version of a simultaneous
equation microeconometric model, built with publicly available data, is not
-------
B-18
worth doing to test notice requirement impacts. It is technically feasible,
though complex, to link any of the complex interrelationships in the chemical
industry in a broad partial or general equilibrium micro-econometric model.
And using the "production possibility frontier" as a generalization of the
production function, and corresponding cost functions is probably the best way
to approach the modeling. These functions are specified mathematically in
order of increasing flexibility by Cobb-Douglas, constant elasticity-of-
substitution, and translog functions. The translog specification is most
flexible and, although it requires a number of assumptions (e.g., competitive
product markets, marginal productivity pricing), it would probably be the best
framework for a micro-econometric approach.
Further, making the model a general equilibrium model through an input-
output linkage is technically feasible, although there is some question
whether any TSCA regulatory impacts will have great enough impacts outside the
chemical industry to make such an effort useful. But it is technically possi-
ble and intellectually fascinating. To do all of the above would consume ex-
tensive resources (possibly $1 million) and take about three years. The
ultimate tool could be a useful one for many EPA and other government analyses
affecting the chemical industry, and it would have the benefits of thorough
grounding in economic theory, flexibility, and avoiding the extensive diffi-
culties of modeling complex process/product relationships—but it will not be
able to assess the effects of TSCA reguation on chemical industry product
innovation endogenously, through econometric equations, in our view.
We have outlined in the section on the innovation process, and address in
somewhat more detail in the section below, a simulation model of the R&D
"product innovation" process in a particular industry segment. Apart from the
econometric approach discussed and dismissed above, we can see two alternative
modeling frameworks for a chemical industry model to test economic impacts of
changes in new product introduction profiles.
An Optimization (Linear Programming) Model. The innovation model outputs
could be new product types with associated cost curves in a variety of use
categories. Then, just as the ICF Coal Model allows for 40 types of coal and
new sources in each category, we could have an optimization model to minimize
costs as its. objective fuction, solving for prices, quantities, and other
factors in each use category. Clearly, because processes can be used to
produce different products, and because products can be produced by more than
one process, the "engineering economics" necessary to get the cost curves,
cross-elasticities and other relationships is complex. And there are serious
problems in modeling demand, which itself is not independent of new product
availability. But, as in the coal model, a wide variety of relationships can
be programmed in as constraints, and such an approach is at least technically
feasible. This approach has the advantage of extensive prior use in opera-
tions research models, so that many of the key software issues have already
been addressed, it also is flexible and could allow for simultaneous demand/
supply effects.
A Simulation Model. Here we would compile historical data on new products
in use categores, and attempt to simulate the relationships between new prod-
uct introduction and other economic and industry parameters (sales, prices,
etc.). if we could construct a baseline model, which could be based upon
-------
B-19
historical relationships and changes over time in these relationships, we
could then take the "outputs" of the innovation model (new products and
associated cost in each use category) and simulate their sales and economic
impacts over time. A particular problem is understanding to what degree new
products are substituting for existing products in a use type, as opposed to
enlarging markets overall. Although flexible, a drawback of this approach is
its complete dependence upon historical relationships and its lack of
theoretical underpinning.
Either of these model types would have to be contructed separately for
each industry segment or product-type category. Either could be tied to an
input-output model to test macroeconomic impacts (in either direction), but
tying chemical industry segments to each other and to an input-output
structure is a difficult and complex but technically possible task.
To construct an industry model of the chemical industry in either form
described above, is a time-consuming and expensive task, especially as the
number of product types, use types, or industry segments increase. Although
difficult to estimate, either would take 12 to 18 months and cost around
$500,000. Then, if the industry model were to be tied through an input-output
model to a macroeconomic model, another 12 to 18 months and approximately
$500,000 would be needed.
With these kinds of time and resource requirements, EPA must have
reasonable expectations that chemical industry innovation can usefully be
modeled at all before undertaking large industry modeling. Also, EPA should
first be considering a small-scale demonstration effort of the various models
to test their feasibility and then step-wise additions to the work with tests
at each stage.
The following section further explores the chemical industry innovation
process and how it might be modeled.
Modeling the Innovation Process
Before presenting our tentative ideas on how to attempt to model the
innovation process in a framework which allows for determination of
consideration of TSCA impacts, we should be clear on several points:
• No useful formal analytic tool can be constructed
using publicly available data alone—or in time for
the notice requirements economic impact analysis.
• It would be a bad idea to begin the "downstream"
modeling of the industry and economy without some
confidence that the innovation component can be
modeled.
• The appropriate approach would be a small-scale
(small-risk) attempt to develop an innovation model
for one high R&D industry segment and to develop
specificatons for its interface with an economic
simulation model of that industry segment. This would
-------
B-20
require original data collection from industry (as
detailed below); would take about nine months; and
should cost on the order of $100,000. And it may wind
up proving that because of lack of data and inability
to translate qualitative judgments and TSCA effects
into the model, a useful model cannot be constructed.
• Within the industry segment chosen, a great deal of
specific work must occur in three categories:
How the R&D/commercialization decisions (Blocks
1-4) have been made in the past and how they are
being made today (necessary for the innovation
model) .
How TSCA regulatons could or would affect these
decisions (necessary for the innovation model).
Case studies of prior new product introductions
and their effects; creation of life-cycle
profiles (necessary to construct the linkage
between the innovation model and the industry
model) .
Framework for Modeling. We have argued that it is necessary to model the
"innovation" model separately from any general model of an industry segment,
and that to test TSCA impacts, the outputs of an innovation model must drive a
more general economic impact model, in our view, this effectively precludes
using a micro-econometric model structure, however large. instead, we envi-
sion an economic simulation model which simulates industry performance based
upon current relationships and trends in major economic variables and finan-
cial and economic production/use categories. We recognize the compexity of
this task as applied to the chemical industry, where production processes,
products, and uses form complex patterns.
INNOVATION
MODEL
ECONOMIC
SIMULATION
MODEL OF
INDUSTRY SEGMENT
TSCA Regulatory impacts
Output of
R&D
PROCESS
(NEW CHEMICALS)
The remainder of this section addresses how the innovation process could be
modeled.
-------
B-21
The innovation Model
Blocks I and II; R&p Resource Allocation. The first major set of decisions
we wish to model within a particular SIC code or other industry segment is the
capital budgeting decision by which resources are allocated, first to R&D vs.
other uses and then to product development R&D within all R&D resources. An
attempt would be made to couple historical data on R&D expenses with new
product payoffs and to solicit information from decision-makers and industry
experts as to:
• How R&D funds are allocated;
• How probabilities of project success are assessed and
what "hurdle rates" are employed;
• How decision makers mix projects with different
risk/payoff profiles into an R&D "portfolio";
• How timing is factored into the decision;
• Whether products would be patented, labels registered,
etc.; and
• What assumptions are made about competition if a
successful product is brought to market.
This would have to be done largely through personal interviews. We have been
unable to locate any previous work or set of data in these areas.
This initial allocation would focus upon three parameters which could be
affected by TSCA:
(1) the ex ante view of costs, at each stage over the
development process;
(2) the expected timing of events for new chemicals, and
how delays affect returns on investment; and
(3) the expected profits from successful projects, the
timing of these profits, and how profits might be
affected by competitors' knowledge at the various
steps of the development process.
This component of the model would produce an annual dollar allocation of
resources to new product development in the industry segment as a function of
the factors above. Historical data would be used to calibrate the model where
possible.
-------
B-22
Block III; Research. The assumption would be made that the research process
itself was unaffected by TSCA; historical probabilities of technical success
would be ascertained and used, either as fixed probabilities or dependent upon
external (non-TSCA) factors (such as market growth, degree of product dif-
ferentiation, etc.) A linear decline rate in probabilities, for example,
could be used if it reflected historical realities.
Block IV; commercialization. Then, the commercialization decision would be
modeled, based again upon the expected costs, timing, and expected payoffs.
It would be necessary here to examine patent protection, trade secrets,
customer loyalty, time-to-market, and other competitive factors in a game-
theoretic approach. This component of the model would attempt to replicate
the actual decision making on commercialization in the industry segment—as a
function of cost, delays, and payoffs.
Block V; Market Introduction. The final block would include how advertising,
promotion, or other marketing expenses affect success in product introduc-
tions; the expected growth curves over time in the various use categories; and
other post-commercialization effects.
The final result would be a simulation of the actual innovation process in
that industry segment. Then, analyses of TSCA effects would be tested by
inputting the estimated costs (at each time stage), delays (if any), and
changes in competitive payoffs (but changing the probability distribution on
payoffs). Ultimately, the objective would be to simulate the number of new
products and their costs and payoffs over time without TSCA and to compare
that to the number, costs, and payoffs with TSCA regulation. These results
would then be linked to the industry-segment micro-econometric simulation
model to test the economic impacts of greater or lesser numbers of new
products with particular cost pay-off profiles.
Can such a model be usefully built? it is very possible that the data neces-
sary to construct the model alone cannot be obtained from industry, and that
the many approximations made in quantifying costs, risks, and payoffs would
render the model useless as a tool, if so, in our view, the chemical new-
product innovation process cannot be formally modeled.
We believe that EPA should either attempt to build such a model on a very
small scale (one industry segment) or conclude that formal analytical appro-
aches are not appropriate in this case. It does not make sense, in our view,
to build a complex partial or general equilibrium model of the chemical indus-
try, based upon econometric relationships if the initial factor of interest,
product innovation—and the effects of TSCA on product innovation—cannot be
incorporated into the model.
The possible limited use of a model of the latter type would be to test
impact sensitivities to a range of gross assumptions about TSCA. For example,
if R&D were contained in one set of equations, we might, say, assume that
one-fourth, or one-half, or all R&D ceased as a result of TSCA and determine
what impact this would have over time on the industry and economy. Because
the impacts, at least in the shorter term, might be small, EPA might be able
to establish that TSCA would be having no greater negative economic effects
than those projected impacts.
-------
B-23
BIBLIOGRAPHY
Summaries of books and articles have been taken from Cooper (19) , Hill
(40), and Kelly, et al. (50).
1. Adams, John G., "Research and Development of a New Drug," Food Drug
Cosmetic Law journal, volume 26, January 1971.
FINDINGS/CONTENTS;
The golden age of drug discovery has reached a plateau and the
present trend of a decline in new product introductions will
continue. The former approach to drug research was to find a drug
and then search for an illness that it could combat. This approach
has now been reversed and this change from product-oriented to
disease-oriented research will have a serious impact on drug research
for both industry and government.
2. Allen, J.A., Studies in Innovation in the Steel and Chemical industries.
Manchester. England: Manchester University Press, 1967.
F PIPINGS/CONTENTS;
Three studies of innovations are presented in the context of the
prevailing state of the art in the areas in question, and the
contemporary social and political environment.
polyethylene and terylene, two new product innovations, and oxygen
steelmaking, a new process innovation, were chosen to show the
contrast between the two types of innovations and the subsequent
difference in their development.
3. American Chemical Society, Chemistry in the Economy. Washington, D.C.:
1973.
4. Arthur D. Little Working Paper on formal approaches to measuring TSCA
impact.
5. Ashford, N.A.; Heaton, G.R.; Priest, W.C.; and Lutz, H., The implications
of Health Safety and Environmental Regulations for Technological
Change, Center for policy Alternatives, Massachusetts Institute of
Technology. January 15, 1979.
-------
B-24
6. Backman, Jules. The Economics of the Chemical industry. Washington,
D.C.: Manufacturing Chemists Association, 1970.
7. Daily, Martin Neil, "Research and Development, Costs and Returns:
The U.S. pharmaceutical Industry," Journal of Political Economy,
Vol. 80, No. 1, January-February 1972.
8. Bender, R.J., "Air Pollution Control: Its Dual Effect on the Chemical
industry," Power, Vol. 116, July 1972.
FINDINGS/CONTENTS;
Environmental legislation has prompted companies to innovate in their
organizational structure; that is, to establish new departments devoted
to controlling pollution and protecting the environment. Examples are
given for DuPont, Water Resources Co., and Standard Oil.
9. Berenson, Conrad, The Chemical industry; Viewpoints and perspectives.
New York: John Wiley & Sons, 1963.
Case studies of existing CPI firms.
10. Bloom, Barry M., "The Rate of Contemporary Drug Discovery," Lex et
Scientia, Vol. 8, No. 1, January-March 1971.
Compares drug discovery rates for the U.S., U.K., Germany, France,
Italy and Japan, and assesses the approval problems.
11. Burhenne, W.E. and Schoenbaum, T.J., "The European Community and the
Management of the Environment," Natural Resource Journal, Vol. 13, 1973.
12. Burn, Duncan and Epstein, Barbara, Realities of Free Trade; Two industry
Studies. London: George Allen and Unwin Ltd., 1972.
Electrical engineering and the chemical industry.
13. Charpie, Robert, Technological innovation; Its Environment and
Management. Washington, D.C.: U.S. Department of Commerce, 1967.
14. "Big Drug Firms are More Innovative," Chemical and Engineering News,
Vol. 51, October 29, 1973.
-------
B-25
FINDINGS/CONTENTS;
Dr. David Schwartzman and Dr. Thomas R. Stauffer believe size will be
the key if a company is to remain innovative in an industry that
depends heavily upon innovation for survival. Also, research activity
increases more than proportionately with firm size and large firms
produce disproportionately more new drug innovations than smaller
firms. Conclusions are based on a statistical analysis of a) the size
of the research effort, b) the economies of scale in research, and
c) the number of discoveries in the drug industry.
15. Choa, William Wing Fai, "The Relevance of Market Structure to
Technological Progress: A Case Study of the Chemical industry." Ph.D.
Dissertation, Stanford University, 1977.
16. Clymer, Harold A., "The Changing Costs and Risks of innovation in Drug
Development," Research Management, Vol. 13, Sept. 1970.
FINDINGS;
The attrition rate of new drugs is increasing due to higher costs and
longer development periods, the search for therapeutic breakthroughs
(as a minor modification in an existing drug is subject to relatively
the same cost and development periods) and the development of superior
evaluation methodology.
17. Coleman, James S.; Katz, Elihu; and Menzel, Herbert, Medical innovation!
A Diffusion Study. Indianapolis: The Bobbs-Merrill Co., Inc., 1966.
18. Comanor, William S., "Research and Technical Change in the Pharmaceutical
Industry." Review of Economics and Statistics, vol. 47, May 1965.
A statistical examination of the pharmaceutical industry in an attempt
to answer several questions concerning the relationship between firm
size and propensity to conduct R&D, the relation between research and
technical change, and the possible existence of economies of scale in
R&D.
19. Cooper, Joseph D. (ed.). Regulations, Economics, and Pharmaceutical
innovation. Washington, D.C.: The American University, 1976.
20. Costello, Peter M., "The Tetracycline Conspiracy: Structure, Conduct,
and performance in the Drug industry," Antitrust Law and Economic
Review, Vol. 4, Summer 1968.
-------
B-26
FINDINGS/CONTENTS;
The broad spectrum antibiotic tetracycline is taken from its inception
to patent application through all its marketing phases as an example of
how patent and licensing agreements can and do lead to a decline in
competition and innovation.
21. DeHaen, P., New Product Survey, 1976; Newly Synthesized Drugs introduced
in the United States. New York: Paul deHaen, Inc., 1977.
22. Enos, John L., "invention and innovation in the petroleum Refining
Industry," in The Rate and Direction of inventive Activity, ed. by the
National Bureau of Economics Research, Princeton University Press, 1962.
This paper examines the processes of invention and innovation as they
have occurred in the petroleum refining industry, focusing on the
relation between the two processes, the intervals between them, the
returns to inventors and innovators, and the resulting changes in
factor proportions in the industry.
23. Evaluation of the Possible impact of Pesticide Legislation on Research and
Development Activities of pesticide Manufacturers, prepared by Arthur
D. Little, Inc. for Office of Pesticide Programs, U.S. Environmental
Protection Agency, EPA-540/9-75-018, February 1975.
24. Fisher, F.M. and Temin, P., "Returns to Scale in Research and Development:
What Does the Schumpeterian Hypothesis imply?," journal of political
Economy, vol. 81, No. 1, January-February 1973.
25. Freeman, Christopher, "Chemical Process Plant: innovation and the World
Market," National Institute Economic Review, No. 45, August 1968.
26. Freeman, Christopher, The Economics of Industrial Innovation. Middlesex,
England: Penguin Books, 1974.
27. Freeman, Christopher, "The Plastics industry: A Comparative Study of
Research and innovation," National institute Economic Review, No. 26,
November 1963.
28. Gibbons, M., "Factors Affecting Technical Innovation in British
industry," industrial Marketing Management, Vol. 2, February 1973.
-------
B-27
29. Grabowski, Henry G., Drug Regulation and innovation; Empirical Evidence
and policy Options. Washington, D.C.: American Enterprise institute,
1976.
30. Grabowski, Henry G., "The Determinants of Industrial R&D: A Study of
the Chemical, Drug, and petroleum industries," journal of political
Economy, Vol. 76, No. 2.
This paper reports the results of an empirical investigation into the
determinations of R&D expenditures in three industries—drugs,
chemicals, and petroleum refining. Determinants analyzed include past
research productivity, output diversification, and availability of
internal funding.
It is hypothesized that high past research productivity should have a
positive effect on propensity to undertake research, that high degree
of product diversification should also have a positive effect, and that
R&D expenditures should be highly dependent upon changes in cash flows
in the long run, but not in the short run. The hypothesis suggested is
largely confirmed.
31. Grabowski, Henry G., and Baxter, N.D., "Rivalry in industrial Research
and Development," journal of Industrial Economics, Vol. 21, No. 2,
July 1973.
32. Grabowski, Henry G.; Vernon, John M.; and Thomas, Lacy Glenn, "The Effects
of Regulatory Policy on the incentive to innovate: An International
Comparative Analysis", in Mitchell, Samuel A. and Link, Emery A.
(eds.). The impact of public policy on Drug innovation and Pricing.
Washington, D.C.: The American University, 1976.
33. Grabowskir Henry G. and John M. Vernon, "Structural Effects of Regulation
on Innovation in the Ethical Drug industry." in Masson, R.T. and
Quails, P. (eds.), Essays on Industrial Organization in Honor of Joe S.
Bain. Cambridge, Mass: Ballinger publishing Co.. 1976.
34. Greenberg, Edward, et al., Regulation, Market Prices, and Process
innovation; The Case of the Ammonia industry. Boulder: Westview
Press, 1979.
35. Grether, David, "Market Structure and R&D," in Noll, Roger G., et al.,
Government policies and Technological innovation, volume II-A,
State-of-the-Art Surveys, California institute of Technology.
36. Gruber, William and Marquis, Donald (eds.), Factors in the Transfer of
Technology. Cambridge: MIT Press, 1979.
-------
B-28
37. Hamberg, D., "Size of Firm, Oligopoly, and Research: The Evidence,"
Canadian journal of Economics and Political Science. Vol. 30,
February 1964.
38. Hansen, R.W., "The Pharmaceutical Development Process: Estimates of
Development costs and Times and Effects of Proposed Regulatory
Changes," in Chien, R.I. (ed.), jssues in Pharmaceutical Economics.
Lexington, Mass: D.C. Heath, 1979.
39. Helms, Robert B. (ed.), Drug Development and Marketing, Washington, D.C.
American Enterprise institute, 1978.
40. Hill, Christopher T., et. al., A State of the Art Review of the Effects
of Regulation on Technical Innovation in the Chemical and Allied
Products industries ("CAPI Project"), Vol. II, The State-of-the-Art,
prepared by Washington university for the National Science Foundation
under Grant No. RDA 74-200086) AOl (formerly DA-44092). St. Louis,
Missouri: Washington University, Center for Development Technology,
February 1975.
41. Hill, Christopher T. (ed.). Federal Regulation and Chemical innovation.
Washington, D.C.: American Chemical Society, 1979.
42. Hill, C.T. and J.M. utterback (eds.), Technical Innovation for a Dynamic
Economy, Elmsford, N.Y.: The Pergamon Press, 1979.
43. Jadlow, Joseph M., "Competition and 'Quality in the Drug industry: The
1962 Kefauver-Harris Drug Amendments as Barriers to Entry." Antitrust
Law & Economics Review, Winter 1971-1972.
Assesses the impact of 1962 Amendments to the Federal Food, Drug and
Cosmetic Act. Finds considerable evidence that the Amendments have
increased the cost of developing a marketable drug.
44. Jadlow, Joseph M., "Price Competition and the Efficacy of Prescription
Drugs," Nebraska Journal of Economics and Business, Autumn 1972.
Study of effects of 1962 Amendments to the Federal Food, Drug and
Cosmetic Act on price competition. The author finds: "The legislation
has been successful in reducing the flow of ineffective drug products
to the market, it has, however, prompted an increase in drug research
costs which has resulted in a decline in the private industrial
development of important new drugs." He finds also that research costs
of small drug firms have risen in greater proportions than those of
large firms.
-------
B-29
45. Jewkes, J.; Sawers, D.; and Stillerman, R., The Sources of inventiori.
New York: Norton, Second Edition, 1979.
46. Kamien, Morton I. and Schwartz, Nancy L., "Market Structure and
Innovation: A Survey," a report prepared for the National Science
Foundation, Evanston, ill.; Managerial Economics and Decision Sciences,
Graduate School of Management, Northwestern University, June 1974,
p. 94. (Nine-page "Summary" and 108-page "Supplement Precis: Annotated
Bibliography" are also available.)
A survey of economists' theories and empirical findings regarding
market structure and innovation, concludes the relationship is still
unresolved—better measures and data are needed. Finds that "many of
the issues regarding the determinants of technical advances are
intimately related to other facets of economic inquiry such as public
goods, imperfect competition, uncertainty and search, decentralization
and growth, which themselves are incompletely understood."
47. Kamien, Morton I. and Schwartz, Nancy L., "Risky R&D with Rivalry."
Annals of Economic and Social Measurement, Vol. 3, No. 1, January 1974.
48. Kane, E.R., "Concerning the Economic impact of Environmental Regulation
on the Chemical industry," Testimony before Joint Economic Committee,
U.S. Congress, November 21, 1974.
49. Kelly, p. and Kranzberg, M. (eds.), Technological innovation; A Critical
Review of Current Knowledge. San Francisco Press, 1978.
50. Kelly, p. and Kranzberg, M., et al., Technological innovation; A
Critical Review of Current Knowledge, Vol. IV; Selected Literature
Abstracts. Georgia institute of Technology, prepared for National
Science Foundation, National Technical information Service,
February 1975.
51. Kennedy. Charles, "induced Bias in Innovation and the Theory of
Distribution," The Economic journal, Vol. 74, September 1964.
52. Kuznets, Simon, "Inventive Activity: Problems of Definition and
Measurement," The Rate and Direction of inventive Activity, edited by
the National Bureau of Economic Research. Princeton University Press,
1962.
-------
B-30
53. Langrish, J.; Gibbons, M.« Evans, W.G.; and Jevons, F.R., Wealth from
Knowledge; Studies of innovation in Industry. London: McMillan
Press, 1972.
Case histories of British industrial.innovations.
54. Lasagna, Louis, "Research, Regulation, and Development of New
Pharmaceuticals: past, Present and Future," The American Journal of
Medical Science, vol. 263, January 1972.
55. Lasagna, L. and Wardell, W., "The Rate of New Drug Discovery," in Helms,
Robert, B. (ed.). Drug Development and Marketing. Washington, D.C.:
American Enterprise institute, 1975.
56. Layton, Christopher with Harlow, C., and DeHoughton, C., Ten innovations;
An international Study on Technological Development and The Use of
Qualified Scientists and Engineers in Ten industries. London: George
Allen and Unwin, 1972.
Compares British with U.S. and European innovation management.
57. Lebergott, S., Statement in U.S. Congress, Senate Subcommittee on
Monopoly of the Select Committee on Small Business, Hearings on
Competitive Problems in the Drug Industry, 93rd Congress. Washington,
D.C.: U.S. Government Printing Office, February 1973.
58. Machlup, F. The Production and Distribution of Knowledge in the U.S.
Princeton University Press, 1962.
59. Mansfield, Edwin, industrial Research and Technological innovation.
New York: W.W. Norton for the Cowles Foundation for Research in
Economics at Yale University, 1968.
60. Mansfield, Edwin; Rapoport, John; Schnee, Jerome; Wagner, Samuel; and
Hamburger, Michael, Research and Innovation in the Modern Corporation.
New York: W.W. Norton, 1971.
Study based on detailed firm data, one of a series of books by
Mansfield on the "economics of technological change." Much of the data
is drawn from the ethical drug industry.
61. Marquis, D.G., Innovation, Vol. 1, July 1969.
-------
B-31
62. Marschak, Thomas A., "Strategy and Organization in a System Development
Project," The Rate and Direction of inventive Activity, ed. by the
National Bureau of Economic Research. Princeton University Press, 1962.
63. Marshall, A. W., and Meckling, W.H., "Predictability of the Costs, Time
and Success of Development," The Rate and Direction of inventive
Activity, ed. by the National Bureau of Economic Research. Princeton
University Press, 1962.
64. Minasian, Jora R., "The Economics of Research and Development," The
Rate and Direction of Inventive Activity, ed. by the National Bureau of
Economic Research. Princeton University Press, 1962.
65. Montgomery, W. David and Quirk, James P., "The Market for innovations,"
Noll, Roger G., et al., Government Policies and Technical innovation,
Volume II-A, State-of-the-Art Surveys, California institute of
Technology, prepared for the National Science Foundation under contract
grant No. DA-39495-RDA-7307241, October 1974.
66. Morris, Simon Peter, "Models of Process-Diffusion and Entry in the U.S.
Chemical industry." Ph.D. Dissertation, University of Pennsylvania,
1975.
67. Mueller D.C., "Patents, Research and Development, and the Measurement
of inventive Activity." journal of industrial Economics, vol. 15, 1966.
68. Mueller, Willard F., "The Origins of the Basic inventions Underlying
DuPont's Major Product and Process Innovations, 1920-1950," The Rate
and Direction of Inventive Activity, ed. by the National Bureau of
Economic Research. Princeton University Press, 1962.
An examination of the sources of the basic inventions underlying 25 of
the most important of DuPont's product and process innovations, shows
that DuPont has been more successful in originating improvement
inventions than in discovering new products.
69. National Agricultural Chemicals Association, Industry Surveys, conducted
by Ernst and Ernst. Washington, D.C.: 1971, 1973, 1975, 1977 and 1978.
70. Nobel, Patricia (ed.), The Kline Guide to the Chemical industry,
Fairfield, New Jersey: Second Edition, 1974.
-------
B-32
71. Organization for Economic cooperation and Development, The Chemical
industry 1971/1972, Paris: OECD, 1973.
72. Organization for Economic cooperation and Development, The Measurement of
Scientific and Technical Activities, OECD DAS/SPR 70.40, 1970.
73. Pavitt, K. and Wald, S., "The Conditions for Success in Technological
Innovation," Paris: OECD, 1971.
74. Peltzman, Sam, Regulation of Pharmaceutical innovation; The 1962
Amendments, Washington, D.C.: American Enterprise Institute for Public
Policy Research, Second printing, April, 1975.
75. Robertson, A.R.; Achilladelis, B.; and Jervis, P., Success and Failure in
Industrial innovation; Report on Project Sappho, London: Centre for
the Study of industrial innovation, 1972.
76. Reuben, B. G., and Burstall, M.L., The Chemical Economy; A Guide to the
Technology and Economics of the Chemical industry, London: Longman
Group Limited, 1973.
FINDINGS/CONTENTS;
A lengthy account of the chemical industry in the U.K. Coverage in-
cludes: the development of the chemical industry; the economics of the
firms; the major sectors of the industry; problem with large scale
chemistry; and the future of the chemical industry.
77. Sanders, Barkev S., "Some Difficulties in Measuring inventive Activity,"
The Rate and Direction o£ Inventive Activity, ed. by the National
Bureau of Economic Research. Princeton University Press, 1962.
Examines the problems associated with using each of three variables—
inputs, rate of technological progress, and patent statistics—as
measures of inventive activity.
78. Scherer, P.M., "Firm Size, Market Structure, Opportunity, and the Output
of patented Inventions," American Economic Review, Vol. 55, 1965.
Statistical study of the relationships between inventive activity and
technological opportunity, firm size, product-time diversification, and
monopoly power.
79. Schmookler, Jacob, Invention and Economic Growth, Cambridge, Mass:
Harvard University Press, 1966.
-------
B-33
80. Schumpeter, J.A., Capitalism, Socialism and Democracy, New York: Harper
and Row, Third Edition, 1950.
81. Schwartzman, David, innovation in the Pharmaceutical industry, Baltimore:
The Johns Hopkins University Press, 1976.
82. Schwartzman, David, The Expected Return from pharmaceutical Research;
Sources of New Drugs and the Profitability of R&D investment,
Washington, D.C.: American Enterprise institute for Public policy
Research, 1975.
83. Foster D. Snell, Inc., Study of the Potential Economic impacts of the
Proposed Toxic Substances Control Act as illustrated by Senate Bill
S. 776, 26-31, February 20, 1975.
84. United Nations Educational, Scientific, and Cultural Organization,
Measurement of Output of Research and Experimental Development,
1970.
85. Utterback, James M., "innovation in Industry and The Diffusion of
Technology," Science, vol. 183, February 15, 1974.
State-of-the-art survey of factors which influence the process of
innovation, including organizational environment, acceptance in the
market, and diffusion of technology.
86. Vernon, John M. and Gusen, Peter, "Technical Change and Firm Size: The
Pharmaceutical industry." The Review of Economics and Statistics,
Vol. 56, No. 3, August 1975.
87. Waddams, A. Lawrence, Chemicals From petroleum; An introductory Survey,
New York: John Wiley and Sons, Inc., Third Edition, 1973.
An introduction to the petrochemical industry. Chemicals manufactured
are classified according to source; an explanation of their applica-
tions, and the relationship between the petroleum chemical industry and
related industries is given.
88. Wardell, William M., "British Usage and American Awareness of Some New
Therapeutic Drugs," clinical Pharmacology & Therapeutics, vol. 14,
No. 6, November-December 1973.
89. Wardell, William W., "Fluroxene and the Penicillin Lesson,"
Anesthesiology, vol. 38, No. 4., April 1973.
-------
B-34
90. Wardell, William M.; Hassar, M.; Anarekar, S.N.; and Lasagna, Louis, "The
Rate of Development of New Drugs in the United States, 1963 through
1975," Clinical Pharamacology and Therapeutics, vol. 24, No. 2,
August 1978.
91. Wardell, William M. and Lasagna, Louis, Regulation and Drug Development,
Washington, D.C.: American Enterprise institute for Public policy
Research, 1975.
92. Williams, G.J., Letter to Sen. J.V. Tunney, p. 319-325 of Hearings on
Toxic Substances Control Act of 1973 before Environment Subcommittee,
U.S. Senate Commerce Committee, 1973.
-------
APPENDIX C
DISCOUNTED CASH FLOW ANALYSIS
-------
DISCOUNTED CASH FLOW ANALYSIS
The analysis presented here is based, in part, on Arthur D. Little's (ADL)
preliminary economic analysis of the effects of premanufacturing notification
(notice).!/ This appendix first reviews the purpose and scope of ADL's work
and discusses the underlying methodologies as well as the results of their
analysis. Then it provides a refined analysis using the same methodology.
PURPOSE AND SCOPE OF ARTHUR D. LITTLE WORK
The specific purpose of ADL's work was to estimate the cost to the chemi-
cal industry of completing the various sections of the notice form, and then
to analyze the impact of the form. The analysis was hampered both by the lack
of a finalized format for the notice forms and requirements, and the lack of
appropriate economic and financial data on new chemical introductions. This
affected both the statistical validity and the predictive accuracy of the
analysis.
The ADL study was divided into four parts. The first was an industry
overview providing general characteristics of chemical products and markets,
types and sizes of firms, growth rates and the interaction ana dependence of
other industries on the chemical industry, it discussed company market share
concentration in various segments, the value of raw materials and goods sold,
and contained a description of the general flow of goods from primary raw
materials to finished retail products.
The second part segmented the industry into those having similar economic,
R&D and production characteristics. This was facilitated by using four-digit
SIC codes which were arranged according to differences in the lines of busi-
ness, relative size of the four-digit segment, and the degree of expected new
chemical innovation. These segments were then characterized by market struc-
ture, financial condition, new chemical development (including patent activ-
ity) , R&D expenditures, and new substance introduction, and by discussions
with industry representatives and analysts at ADL. This section included the
data on new chemical production and sales used in the impact analysis.
The third section described the regulatory requirements and estimated a
range of costs for completing each area of the proposed notice form. This
section used several major assumptions which affect completion costs, including
I/Arthur D. Little, Inc., impact of TSCA Proposed Premanufacturing
Notification Requirements (Cambridge, Mass.: December 1978).
-------
C-2
the "likely response of a prudent firm," differences in the chemical use of
the proposed product, and the ultimate distribution of the chemical..?/
The final section addressed the impacts of the notice requirements and
included an economic analysis based on sales data previously presented and on
notice completion costs. It discussed qualitatively the probable impacts on
selected segments, and in a general sense, probable changes to the industry
caused by notice requirements.
Results of the ADL Analysis
This ^section reviews the results of the ADL analysis and discusses the
major assumptions and procedures for notice cost estimation and impact
evaluation.
Exhibit c-1 summarizes the changes in the number of new chemicals intro-
duced for various levels of notice costs calculated by ADL. The ADL report
also stated that the impacts would be "relatively high" for small chemical
manufacturing companies, and suggested several segments that would be impacted
most severely because of a higher degree of innovation and smaller firm
size.;3/ These segments are listed in Exhibit C-2.
EXHIBIT C-1
IMPACT OF NOTICE COSTS ON NEW CHEMICAL INTRODUCTIONS
Percentage Reduction in
Type of cost cost Range New Chemical introductions
Minimum Mandatory
Form Completion $2,540 to $14,060 20 to 60
Maximum Mandatory
Form Completion (total) 3,740 to 22,160 25 to 70
Optional information
portion (total) 9,290 to 41,860 45 to 90
Source: Arthur D. Little, Inc., impact of TSCA Proposed Premanufacturing
Notification Requirements (Cambridge, Mass.: December 1978).
j/Ibid., p. v-2.
I/Ibid., p. VI-5.
-------
C-3
EXHIBIT C-2
INDUSTRY SEGMENTS MOST LIKELY TO FEEL NOTICE IMPACTS
Segment SIC Code
Soaps and Detergents 2841
Surface Acting Agents 2843
industrial Organic Chemicals, NEC 2869
industrial Inorganic Chemicals, NEC 2819
Plastic Material and Resins 2821
Synthetic Rubber 2822
Toilet Preparations, perfumes^/ 2844
Cyclic Crudes and intermediates 2865
l/ICF realizes this segment is not covered by TSCA.
Source: Arthur D. Little, impact of TSCA Proposed Premanu-
facturing Notification Requirements (Cambridge,
Mass.: December 1978).
Notice Cost Estimation
Because of the uncertainties associated with the format of the notice
form, ADL made several major assumptions when estimating the cost of
completing the proposed notice form. These assumptions include:
• the likely response of a "prudent" firm; ADL assumed the
firm will include as much information as necessary to make
the form acceptable in terms of the amount and nature of
the data supplied;
• the company size; ADL assumed a smaller firm will not be
expected to provide the level of detail on the proposed
notice form that a larger, more sophisticated firm could
and would supply;
• the use of the chemicals; ADL assumed chemicals used as
intermediates or with limited uses will not require as
much detail on the notice form;
• the ultimate user of the chemical product; ADL assumed no
regulating of ultimate users, which reduces the magnitude
and complexity of the data gathering effort from secondary
users of the chemical;
• the cost of preparing test data; ADL assumed that this
was added to the cost of completing the form, even though
the actual testing costs were not; and
-------
C-4
• the hourly rate of employees engaged in completing the
form; ADL estimated these at: managerial, $50/hour;
professional/technical, $25/hour; clerical, $lO/hour.
The actual cost estimation procedure grouped related activities together
and estimated ranges for the amount of time necessary to complete the form for
each pay category. Factors that affected the estimation of costs were grouped
into four activity categories including data development, data collection,
administration, and managerial review and supervision. These factors were
then applied to each major segment of the proposed three-part form to provide
a range of costs for Part I (General information), Part II (Risk Assessment
Data), and part in (Risk Analysis and Optional Data). Alternative methods,
such as grouping the efforts by type of information required, regardless of
which part was affected, were not addressed.
ADL stated in their December 1978 report 4/ that their estimation meth-
ods resulted in uncertainties of up to "50% or more" in the stated results.
In their October 1979 report, the estimated cost range for the revised notice
was $1,000 to $9,000 (over 50 percent less than the January form). These cost
estimates, based on the January 10 estimates of the hours and pay rates and
the October 16, 1979 form, provide a reasonable estimate of the probable com-
pletion costs for the reproposed form.
There may be additional non-direct costs associated with the completion
of the form. These costs could include the loss of future revenues caused by
a lack of confidentiality, additional testing, and other delays in the intro-
duction of a new product. However, the variability and magnitude of these
additional non-direct costs may be substantial and could radically affect new
product innovation. It is beyond the scope of this portion of the discussion
to address the probable magnitude of these costs.
Economic Analysis Assumptions and Methodology
The impact evaluation was preceded by economic estimates of the number of
new chemicals developed annually, their rates of unit production growth, and
their sales volume. ADL developed the estimates using a number of underlying
assumptions which influenced the results and may have affected the impact
evaluation.
ADL's estimate of the number of new chemicals introduced annually was
based on several sources including the results of a Foster D. Snell survey,
(Foster D. Snell, inc.. Study of the Potential impact of the Proposed Toxic
Substances Control Act as illustrated by Senate Bill S776, February 20, 1975).
the number of new chemical patents issued, the EPA inventory of Chemicals, the
U.S. Department of Commerce-Current Industrial Reports, interviews with the
publishers of "several leading commercial (chemical) directories" whose titles
were not specified, and interviews with industry observers and representa-
tives. 5/ pour independent statistics were mentioned in the ADL report as a
i/Ibid., p. v-9.
I/Ibid., p. III-7.
-------
C-5
framework for the estimation of new chemical introductions, but the interrela-
tionship of these data to each other and to the estimation methodology was not
stated nor could it be inferred, in particular, these statistics, estimated
from a sample of data from the inventory of Chemicals and Current industrial
Reports, were:
• fewer than an estimated 600 chemicals in EPA's Inventory
of Chemicals (approximately 55,000 chemicals) had annual
sales in excess of one million dollars per year;
• about an estimated one-half (or 30,000) of the chemicals
in the Inventory of Chemicals had sales of less than
10,000 pounds per year;
• three percent of chemicals listed in "major chemical
directories" (10,000 chemicals listed) are new each year,
resulting in 300 new commercial chemicals whose annual
sales exceed 1,000 dollars or 1,000 pounds;
• from the Current industrial Reports, the four hundred six-
tieth ranked organic chemical had sales of $100,000 and
the fifty-ninth ranked inorganic chemical had sales of
1,000,000 dollars.
From these data and the aforementioned sources, ADL estimates of new chemical
introductions are summarized in Exhibit C-3. ADL states that "very little
information is available on this specific subject (of new chemical introduc-
tions) , and that these estimates are presented solely for purposes of this
preliminary analysis of TSCA premanufacturing notification impacts."I/
EXHIBIT C-3
ADL ESTIMATES OF DISTRIBUTION OF NEW CHEMICALS INTRODUCED ANNUALLY^/
Use
R&D
Commercial
Total
Under 1,000 pounds
2,000
700
2,700
Over 1,000 pounds
300
300
Total
2,000
1,000
3,000
^/Estimates subject to ± 30% uncertainty.
Source: Arthur D. Little, Inc., impact of TSCA Proposed Premanufacturing
Notification Requirements (Cambridge, Mass.: December 1978), p.
III-8.
6/ibid., p. III-7.
-------
C-6
The volume of a chemical's sales growth was assumed by ADL to follow a
typical product life cycle path. Exhibit C-4 describes this path, indicating
the sales per time period throughout the chemical's product life cycle.
EXHIBIT C-4
TYPICAL CHEMICAL SALES VOLUME LIFE CYCLE
MATURE
LATE QIOWTH
Sales
Volume
Per
Year
TIME
The underlying theory and empirical evidence supporting the product life
cycle concept is well documented elsewhere and will not be discussed here, it
is important to note, however, that the shape and duration of the curve is
strongly affected by the product under discussion. For example, the basic raw
material chemicals such as oxygen, chlorine, and inorganic acids may never
reach a declining growth phase, because there are no substitute products. At
the other extreme, for new products in an industry sector where innovation is
high and product substitution substantial, the growth curve may be nearly ver-
tical for several years and suddenly fall to zero without a gradual decline.
Thus, the use of this curve and associated growth functions as a model must be
tempered by a reasonable understanding of the characteristics of the products
and markets under analysis.
ADL selected 25 chemicals appearing in the Chemical Week Buyers' Guide
which were added between 1973 and 1978. ADL acknowledged the bias of this
sample: the Buyers' Guide is oriented toward the industrial segments, which
implies higher sales for those chemicals sampled than the average of all 1,000
new commercial chemicals introduced each year. ADL did not adjust for this
assumed bias. Exhibit C-5, taken directly from the ADL report, shows the data
ADL used for analysis. ADL acquired the data on sales and first year of sale
from telephone conversations with the manufacturers. The purpose of selecting
a sample of chemicals was to arrive at a sales distribution which would indi-
cate number of chemicals sold versus sales volumes in dollars.
-------
C-7
EXHIBIT C-5
RESULTS OF TELEPHONE SURVEY OF "NEW" CHEMICAL PRODUCTION/SALES VOLUME^/
Column
B
Chemical Manufacturer
Numberjb/ NumberjV
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1977 Sales Estimate
Dollars
Pounds (thousands)
1,000,000
N/A
85,000
46,000,000
N/A
N/A
225
150
N/A
N/A
100,000
5,800
N/A
N/A
65,000
400
1,000
650
150
404
11,040
N/A
N/A
24
116
N/A
10
155
119
N/A
10
520
24
45
D
Estimated
Value Per
Pound
(dollars)
.65
N/A
4.75
.24
N/A
N/A
107
775
N/A
N/A
1.55
20.50
N/A
N/A
8
60
45
First Year
of Sale
1965, 1978
1969
1950s
1950s
1968
1968
1967
1965
1958
N/A
1968
1972
1968
N/A
1967
1975
N/A
17-25 No longer manufactured, or sold only for R&D purposes.
new chemical for purposes of this survey was defined as one newly listed
in the Chemical Week Buyers' Guide between 1973 and 1978. R&D sales of the
chemicals were excluded.
^/Numbers presented here are nominal. Different manufacturers, however, may
each be denoted by "1".
Source: Arthur D. Little, inc., impact of TSCA Proposed Premanufacturing Noti-
fication Requirements (Cambridge, Mass.: December 1978), p. III-ll.
-------
C-8
To normalize the sales to a single year for analysis, ADL adjusted the
sales amounts. They state:
In order to place these chemicals on a consistent basis for
purposes of the impact analysis, the sales values have been
normalized to a 10th year sales figure. This has been done by
assuming linear growth during this 10-year period. . . . The
estimating procedure implies that the sales after five years
would be half of those shown in [(Exhibit C-5, column 1)].27
It is apparent, however, that ADL did not use a linear normalization
technique. A linear normalization technique achieves results different from
ADL estimates.JL/ A modified compound annual growth rate technique yields
approximately the same normalized tenth year sales estimates ADL presented.9/
In the absence of yearly sales data, it is unclear which technique yields more
accurate estimates. Often, the early stage of product sales growth is quite
rapid, and an exponential growth function is more appropriate. In later years
a linear function may be more appropriate due to slower growth. The use of an
exponential (or compound annual) growth function implies that the sample chem-
icals are enjoying an upward sloping growth curve and that they are still in
the early stages of their life cycle—a reasonable assumption.
Exhibit C-6 summarizes these findings, showing the ADL 1977 sales esti-
mates (as shown previously in Exhibit C-5), the ADL "linear growth" normalized
tenth year sales estimate as presented in their report, an ICF linear growth
tenth year sales calculation, and an ICF compound annual growth tenth year
sales calculation with appropriate annual compound growth rates.
.I/Ibid., p. 111-10.
.§/A linear extrapolation/interpolation technique assumes:
(1) 0 sales at introduction, year 0;
(2) 1977 sales estimates at the end of 1977, (Column C, Exhibit C-5);
(3) equal yearly growth increments calculated by dividing the 1977
sales estimates by total years of sales: (Column C, Exhibit C-5)
divided by (1977 - Column E, Exhibit C-5).
(4) multiplying this yearly growth increment by 10 to calculate
tenth year sales.
I/A modified compound annual growth rate technique assumes:
(5) sales at the end of year 1 are equal to one linear yearly growth
increment as calculated above in (3);
(6) 1977 sales estimates are at the end of 1977;
(7) compound growth in sales occur at the appropriate constant
compound rate between the end of year 1 and 1977; and
(8) the tenth year sales estimate is the sum of the sales at the end
of Year 1 as calculated in (5) and nine years of growth at the
compound annual rate which yields the 1977 sales estimate found
in column C, Exhibit C-5 as calculated in (7).
-------
EXHIBIT C-6
COMPARISON OF ADL AND ICF NORMALIZED GROWTH RATES
Column
B
Total Years Soldi/
10th Year
Sales Estimate
ICF 10th
Linear Growth
Estimate
E
ICF 10th
Compound Annual
Growth Estimate
1
1
2
3
6
7
10
11
14
15
16
650
150
404
11,040
24
116
155
119
520
24
45
13
9
23
23
11
13
10
6
11
3
—
($000)
367
267
67
—
17
83
200
833
183
N/A
50
($000)
500
166
176
4,800
22
89
155
198
473
80
—
Sales Growth Rate %
($000)
359
259
69
1,875
19
75
200
715
183
3,115
—
21.81
31.67
14.61
14.60
24.36
23.83
29.15
43.15
27.10
44.22
__
o
vo
I/Source: Arthur D. Little, Inc., Impact of TSCA Proposed Premanufacture Notifications Requirements,
(Cambridge, Mass.: December 1978).
jVsource: ICF estimates.
-------
C-10
ADL constructed a cumulative distribution curve of annual dollar sales
versus the cumulative percentage of chemicals sold. (ADL report Figure III-3,
p. iu-13) . A major assumption made was that each of the 10 chemicals sampled
fell into the respective decile of the probability distribution. ADL's result-
ing curve shows an S-shaped distribution.!^/ Although the theories and sta-
tistical evidence for lognormal cumulative probability distributions and com-
pound growth rates of products are fully accepted, we must recognize that these
results are based on only 10 data points. Since ADL's figures are based on
limited data, the uncertainty attached to their estimates is relatively large.
Impact Analysis Methodology
In order to analyze the impact of the notice, ADL employed a discounted
cash flow analysis of the 10-year earnings streams of each of the sample chem-
icals. Major financial assumptions included a six percent return on sales, a
15 percent discount rate and a constant compound annual growth in sales of each
chemical at the rate previously used in their normalization calculations, and
depreciation charges whose net present value equaled the investment. Because
they had no useful data about elasticity of demand, they assumed relatively
elastic demand. Thus, the increased cost of the notice requirement did not
result in a price increase; and a decrease in the quantity demanded did not
occur.
Within the impact analysis are imbedded several major methodological
assumptions:
1. The present value of the depreciation cash flow stream
from capital investment in production facilities for the
chemical, which is not captured in a discounted earnings
stream, must have no impact on the decision to introduce a
new chemical; Standard NPV analysis considers all after
tax cash inflows and normally includes operating income,
depreciation, depletion and amortization allowances,
deferred taxes (if any), investment tax credits and the
like. ADL did not include any cash inflows other than
after tax operating income in their analysis, which may
lead to underestimating actual returns and overstating
notice impacts.
2. Notice costs must be on an after-tax basis to be reason-
ably compared to after-tax earnings streams: ADL used the
before-tax notice costs, which will overstate the impacts
by a substantial amount.
3- The sales of a particular chemical are not linked to the
sales of other chemicals. Any two chemicals are independ-
ent in terms of sales growth, manufacturing, and prices;
therefore, the removal or non-introduction of any product
— This curve has the general form: log :_ where p is the
cumulative probability of occurrence and varies between zero and one.
-------
C-ll
assumes no change in the overall distribution of products
measured in terms of price/pound, it is possible that the
imposition of small costs may not perturb the original
distribution of chemicals by price/pound; but large costs,
which could create large percentage changes in the number
and type of new chemicals introduced in the market, may
have unforeseen and unpredictable effects on the product/
price distribution used in the analyses. It may cause a
shift of the distribution to higher price/pound products
or lower price/pound products. Given ADL's relatively
severe impact, the methodology they employed, which
assumed no shifts, may have been inappropriate.
4. New products face perfectly elastic demand. ADL assumed
that regulatory costs would result in no change in the
price of the new product and concomitantly no reduction in
the quantity demanded. Evidence about new product pricing
is contradictory. Comments by representatives of the
specialty manufacturers about the uniqueness of, and
immediate need by purchaser for, some new chemicals sug-
gests that this assumption is wrong. Other public comment
about the marginal technical improvement represented by
some new chemicals and therefore the necessity to price
them carefully, suggests this assumption is right. ICF
believes that some new products face elastic demand and
others do not.
By assuming elastic demand for all new products, ADL
neither understated nor overstated the impact of section
5. The reason for this is that in a competitive
marketplace the increased price associated with the added
costs will reduce the quantity demanded. Depending on the
elasticity of demand it is possible that revenues could
rise, fall, or stay the same at the new equilibrium.
The present values of the earnings stream of each chemical were arrayed
in a cumulative distribution plot, which assumed each of the 10 chemicals
represented 10 percent of new chemical introductions. This plot is shown as
Figure vi-1, p. VI-11 of the report.
The methodology then compared the present values to the estimated notice
completion costs. Where the notice costs exceeded the present value of the
earnings stream, the chemical was assumed not to be introduced. From this
comparison, and using the estimate of 1,000 new commercially introduced chem-
icals each year, ADL estimated the impact in terms of number of new chemicals
that would be introduced. Their results are summarized in Exhibit C-7.
impact Analysis Results
ADL's impact analysis was qualitative in nature. Except for the analysis
on reduction in new chemicals introduced presented earlier in this section,
there was no apparent analytical methodology associated with the results. In
-------
C-12
EXHIBIT C-7
NEW CHEMICAL INTRODUCTION AS A FUNCTION OF NOTICE COMPLETION COSTS
TSCA (Notice) Costs Number of Commercial introductions^/
$ 0 1,000
5,000 700
10,000 500
20,000 300
40,000 100
a/Annual rate of introduction in the absence of notice
estimated to be 1,000 ± 30%.
Source: Arthur D. Little, Inc., impact of TSCA Proposed
Premanufacture Notification Requirements (Cambridge,
Mass.: December 1978), Table VI-3, p. VI-12.
summary, they suggested that the following impacts might occur as a result of
notice cost imposition:
• As additional costs of new chemical introduction rise, the
requirements for financial success will be larger, thus
causing fewer new chemicals to be commercialized.
• The reduced number of new chemicals will, in turn,
increase the required investment per new chemical because
a smaller proportion of successful chemicals will support
a larger proportion of unsuccessful chemicals.
• There may be changes in the R&D process.
• There will be increased spending for compliance (testing
and administrative activitie's) .
• There will be economic impacts in the chemical producing
industries, the chemical consuming industries, and the
national economy including the balance of payments. ADL
did not specify the type, direction, magnitude, or nature
of these impacts.
Summary
This section provided a summary of the ADL approach used in January 1979
and noted weaknesses in their methodology. The general directions of the
impacts identified by ADL appear to be reasonable. Nevertheless, the apparent
weaknesses in the analysis have led ICF to conclude that the magnitude of the
impacts cannot be accurately quantified without first obtaining a comprehensive
set of data about historical new chemical introduction. In the next section,
we provide an alternative analysis of the available data having the benefit of
more time and hindsight. Although we consider both analyses to be practically
-------
C-13
useless for identifying the impacts of notice requirements, the alternative
analysis is presented in order to illustrate what can be done given that the
necessary data are available.
ALTERNATIVE ANALYSIS
This section provides quantitative economic analysis similar to ADL's.
The analysis can be replicated by those wishing to change the underlying
assumptions, when and if additional data warrant such changes. Second, within
the constraints of available data, this chapter presents a sensitivity anal-
ysis so the reader can appreciate the range of possible results.
A model which would accurately predict the impact on new product develop-
ment would ideally have as its basis the capability to determine the impacts on
product innovation. Such a model might consider the changes notice requirement
options could produce in the cost and probability of success of commercializa-
tion of new chemicals and the consequent effects on R&D resource allocation.
While changes in product innovation from notice requirement options could be
conceptualized for heuristic purposes, the technical feasibility of this type
of analytic approach presently is doubtful.
Today the majority of corporate investment decisions are made by formal
or informal net present value (NPV) analysis.ll/ This allows for maximizing
returns on investments while considering the timing of investments and returns,
the uncertainty of returns, and competing alternative investment proposals.
In the case of new chemical product investment, the NPV methodology abstracts
from a myriad of complex issues on resource allocation, innovation, and uncer-
tainty, and yields a rational economic analysis based on readily available
financial and market data. For these reasons, ICF has chosen a discounted
cash flow (DCF) analysis which includes cash flow streams from both earnings
and depreciation. Because the actual amounts of front-end investments in the
sample chemicals used for this analysis are not known, the traditional NPV
analysis will be modified. This model allows for varying the costs of notice
completion, discount rates, sales growth functions, and returns on sales to
test the sensitivity of new chemical introduction to changes in these factors.
Ideally, the scope of our analysis would include a statistically signifi-
cant sample of chemical sales data which reflects the range of current and
proposed product development and is believed to parallel the expected commer-
cial success/failure ratio for new products in every segment of the industry.
It would encompass a representative sample of sales data that allows for an
accurate probability distribution of sales volumes for all sectors of the
chemical industry. In reality, we realize the limitations of the ADL data and
the few additional pieces of information available from industry responses.
Yet, the methodology we are using easily allows for "updating" the analysis as
additional appropriate data are made available.
more complete discussion of NPV analysis may be found in the sub-
section: NPV Methodology.
-------
C-14
SUMMARY OF RESULTS
Exhibit C-8 summarizes the results of the economic analysis. It shows
both the "base case", considering the average impact that would be felt by the
entire chemical and allied products (SIC 28) industry, and the results of sen-
sitivity analyses on the major influential factors (discount rate, notice cost,
and return on sales). Because of data limitations the range of these impacts
could be larger. Moreover, as discussed earlier, it is unclear whether the
sample data is biased towards high or low volume chemicals, and thus inaccu-
rately reflects notice impacts.
The results in Exhibit C-8 are highly sensitive to changes in the assump-
tions used. For example, if a different data sample was used where the median
chemical revenue stream was lower, the results could indicate a larger reduc-
tion in annual new chemical introductions. Although the varied parameters in
the sensitivity analysis are independent of each other, industrial and market
charactersties of particular sections of the chemical industry can be most
accurately evaluated by changing several parameters at once. The effect of
jointly varying several parameters is synergistic, and can change the expected
reduction of new chemical introductions from two percent to 40 percent or more.
The remainder of this section discusses the assumptions, methodology, and
results of the economic analysis.
EXHIBIT C-8
SUMMARY OF IMPACTS OF NOTICE COSTS TO NEW CHEMICAL
INTRODUCTIONS—BASE CASE AND VARYING MAJOR ASSUMPTIONS
After-tax
Notice Cost
$2,500£/
$500 - 4,500
$2,500
$2,500
$4,500
$20,000
Discount Rate
15%
15%
10 - 45%
15%
45%
15%
Return on Sales
6%
6%
6%
4 - 10%
4%
6%
Percent Reduction
In New Chemical
Introductions
0.5 - 5.0
1.0 - 7.0
0.5 - 5.0
24
30
a/Base Case, notice cost is expressed in after-tax cost (see subsequent
pages).
ECONOMIC ANALYSIS ASSUMPTIONS AND METHODOLOGY
The utility of an economic analysis based on limited data hinges on how
accurately the sample's distribution of data represents the actual distribution
of the population under study. For example, assuming a population of 1,DOT) new
-------
C-15
chemicals (or larger) that have a typical lognormal distribution of sales
growth, if we have a random sample of 30 or more chemicals, the distribution of
the sampling statistics may approach the actual population distribution quite
closely. Ultimately, the only way to test a particular sample's accuracy is
by empirical tests—gather more random data and see if the statistics change
drastically. We must assume from the outset that the chemical samples we are
using for this analysis are randomly chosen and accurately represent the popu-
lation of new chemical introductions. Without this basic underlying assump-
tion, no further analysis can be performed. For this reason, we feel that the
sample we are using is limited and thus may not accurately predict economic
behavior. Nevertheless, the understandable reluctance of the industry to pro-
vide additional data has necessitated that we use the data at hand. The use of
sensitivity analysis may compensate for any bias in the sample data by creating
a range of possible outcomes which lie at the far ends of possible outcomes.
Understandably, a subjective element is introduced when deciding what is a rea-
sonable and relevant range for the variables in the analysis. The following
subsections discuss the assumptions and coefficients of the important variables
used in the analysis.
NPV Methodology
The net present value (NPV) method of investment analysis discounts all
cash inflows (the returns) at the marginal cost of capital, and subtracts the
initial or discounted outlay as shown in Equation (1).ii/ For this analysis,
the value of the investment (I) is not known, and as a result the actual NPV
or changes in NPV as a result of additional investment costs from notice com-
pletion cannot be calculated. Assuming that I is zero, then the present value
of cash inflows calculated would be the net present value. Whenever this NPV
is exceeded by a notice cost, the chemical would not be introduced, invest-
ments are greater than zero, however, so for any given discount rate the NPV
must be less than the PV of cash inflows, although this value cannot be calcu-
lated with the available data. Using the assumption that 1=0 will, of course,
understate the impacts of notice costs somewhat for any given discount rate.
Rl R2 Rn
(1) ^- + =-r + ... + I = NPV
(1+r)1 (1+r)2 d+r)n
Where R = Return (revenues-expenses)
I = investment
r = marginal cost of capital (or desired
return rate set by company)
n = year of final return.
.12/The marginal cost of capital is used by firms who follow a value-
maximizing strategy. A higher discount rate, often called a "hurdle" rate,
leads to a less-than-maximum investment strategy, and is often used where the
company faces capital rationing.
-------
C-16
For example, suppose the PV of cash inflows from chemical A were $10,000.
If an investment of zero is assumed, a notice cost of $10,000 could be imposed
before the chemical would not be introduced. Of course, investment is greater
than zero so the NPV of chemical A is less than $10,000. If the accounting
method of depreciation of assets were known, the investment from depreciation
cash flows could be calculated. Several methods of accelerated and straight
line depreciation are currently in use, however, and without knowing which
method is used the original investment cannot be accurately calculated.
Because the depreciation cash flow here is estimated as a percent of sales, a
further estimation of the original investment cannot be made with a high degree
of assurance.
There is a modified methodology which will provide another way to look at
the impact of notice costs, however. This is the internal rate of return (IRR)
method, which applies a discount rate that equates the present value of cash
outflows (the investment) with net cash inflows (the returns), as shown in
Equation 2.13/
R-, R2 Rn
(2) ±-r- + s-r+... + 2— - I - 0
(1+r)1 (1+r)2 (l+r)n
where R = Return (revenues-expenses)
I = investment
r = internal rate of return (IRR)
n = final year of returns
M/Both the IRR and NPV methods are based on the principle of discounting,
which assumes the opportunity, but not the necessity to reinvest all return
cash flows at the discount rate. The NPV method implies the opportunity (but
not the necessity) exists to reinvest at the company's chosen discount or
hurdle rate. The IRR method implies the opportunity (but not the necessity)
exists to reinvest at the calculated rate of return for the proposed invest-
ment. Both criteria give the same accept-reject decision for specific proj-
ects, but under certain conditions of different amounts and timing of return
cash flows may rank projects differently. Weston and Brigham (F. Weston and
E. Brigham, Managerial Finance, 5th ed., p. 274, Dryden Press, Hinsdale,
Illinois, 1976) argue that if management is seeking to maximize the value of
the firm, the project with the higher NPV will contribute more to the firm.
Because there are also questions whether reinvestment opportunities will actu-
ally be available at the calculated IRR, the NPV method is used more often by
industry.
-------
C-17
We can use the IRR method to show changes in the assumed internal rate of
return in the following fashion:
• Assume the IRR equals some value;
• Calculate the PV of net cash inflows from net income and
depreciation. The resulting number, by definition, is
equal to the original investment;
• Add notice costs to the original investment amount;
• Calculate the new IRR that will make the PV of net cash
inflows equal the new investment amount; and
• The difference in IRRs is the change in return that would
occur if notice costs were imposed.
This allows us to see the relative impact of notice costs on proposed new
chemical introductions as a change in IRR.
For example, assume an initial IRR of 15 percent and a PV of cash inflows
of $10,000.
From equation 2, we know:
$10,000 -1=0
(3)
or I = $10,000
If we impose a notice Cost of $1,000, then:
(4) I1 = $11,000
From this, we can find the new IRR (IRR1) that gives a PV of cash inflows of
$11,000. in this example, IRR1 might equal 13.5 percent. The change
(IRR - IRR1), or -1.5 percent, is the relative economic impact of the notice
cost. For smaller PVs, the relative impact would be larger, for larger PVs,
the impact would be smaller. Additionally, the relative impact will depend on
which initial IRR is chosen.
Discount Rate
The discount rate for ongoing projects is a function of the firm's average
cost of capital. If we assume that the technical and commercial risks of new
projects do not differ from those of ongoing projects, we could use this cost
of capital for new projects when discounting. However, it is readily apparent
that new product introductions vary widely in terms of commercial and technical
risk, from relatively minor chemical modifications of an established product to
radically new compounds with no sales guarantees. Firms typically establish a
higher discount rate for higher perceived risk introductions. This explicitly
takes into account the higher probability of commercial failures. Only pros-
pects with expected high rates of return will be commercially introduced where
the procability of poor sales is large, in order to cover the cost of a larger
number of failures. For a firm, this requires factoring into an individual
-------
C-18
commercialization analysis some failure rate or probability distribution of
expected failure, which results in the adjustment to the required rate of
return. Discus'sions with industry-representatives and comments received in
response to Reproposal of Premanufacture Notice Form and Provisions of Rules
(44 Federal Register 59764, October 16 1979) indicate a range of 10 to 45 per-
cent discount rates are used by firms for new chemical introduction analysis.
We assume these to be reasonable and have chosen 10 percent, 15 percent, 30
percent and 45 percent as representative discount rates for analysis purposes.
ADL used 15 percent across all segments.
Sales
As discussed earlier, the use of a particular growth function for sales
must be tempered by a reasonable understanding of the markets and products
under analysis, if the sample is assumed to be random the sales behavior of
the cheiticals will mirror the industry aggregate behavior. In lieu of actual
data on yearly sales for sample chemicals, the use of a general product life
cycle growth model is then reasonable, if it can be ascertained where in the
life cycle the chemical fits. For the sample data this is not possible. On
the other hand, linear or exponential growth models may be more appropriate
for portions of the product's life, especially in the early phases of growth.
Caution must be exercised when using exponential or linear estimators, because
these models have no practical limit to growth, which can radically affect the
present value analysis where the growth rate differs widely from the discount
rate, or the data are extrapolated to several times the known period of sales.
industry experts such as Henry Grabowski, John Vernon and Christopher
Hill (see Appendix B Bibliography, especially 30, 32, 40) differ on what the
average life cycle of a new chemical looks like; therefore, both linear and
exponential growth models, normalized to 10 (linear) or 11 (exponential)
years, have been chosen for use here, to provide some sensitivity to changes
in growth rates. In order to calculate a 10- or 11-year stream of returns,
estimates of sales for each of the first 10 or 11 years of sales were made as
follows:
• Linear interpolation/extrapolation; The 1977 sales fig-
ures are divided by the number of years the chemical has
been sold. The resultant number is considered to be the
sales at the end of year 1, and is increased by that same
amount each succeeding year. For example, if a chemical
had been sold for 13 years, sales in Year 1 are equal to
one-thirteenth (1/13) of 1977 sales, sales in Year 2 equal
two-thirteenths (2/13) of 1977 sales, and so forth.
Exhibit C-9 illustrates this example.
• Exponential interpolation/extrapolation; A modified
compound growth function is created by using a one year
linear growth rate for one year. Knowing the sales at the
end of Year 1 and 1977, we solve for the particular com-
pound rate which gives both end points. For example, if a
chemical had sales of 100 in 1977 and had been sold for 13
years, sales at the end of year I would be 7.69, and the
compound annual rate of growth would be 23.83 percent for
the 12 years between the end of Year 1 and Year 13. By
-------
C-19
EXHIBIT C-9
LINEAR GROWTH MODEL
ANNUAL
SALES
YEARS
10
EXHIBIT C-10
EXPONENTIAL <3*OWTH MODEL
ANNUAL
SALES
YEARS
10
-------
C-20
interpolation, sales in Year 10 (nine after the end of
Year 1) are 52.66, using 7.69 as starting sales, at the
end of year 1 compounded at 23.83 percent per year.
Exhibit C-10 illustrates this example.
From the sale figures, net income is calculated by multiplying sales by a
return on sale (ROS) factor. This factor varies from firm to firm and from
product to product. ADL reported a range of after-tax ROS from 2.3 percent to
12 percent with the industry average 7.1 percent in 1977.M/ We will use
.04, .06, and .10 ROS factors to test the sensitivity of new chemical intro-
duction to changes in returns.
Depreciation
Cash in-flows also occur from depreciation tax deductions on capital
equipment. Exhibit C-ll illustrates the depreciation after-tax cash flows
from selected basic/commodity chemical manufacturers (23 firms) and selected
specialty chemical manufacturers (20 firms) for the period 1974-1978. The
depreciation cash flow is calculated (adjusted by the appropriate tax rate) as
a function of sales, making it directly comparable to the after-tax return on
sales figures.
The current year sales of the majority of sample chemicals indicate they
are probably high volume, low-cost commodities in the current year. The nor-
malized tenth year estimates by either linear or compound growth method result
in nine out of 12 chemicals with sales in excess of $100,000. Therefore, we
suspect that these are capital intensive chemicals and the more capital inten-
sive depreciation after-tax cash flow factor of 2.62 percent of sales has been
used in this analysis. The depreciation cash flow is added to the ROS flow,
and the sum is discounted at the discount rate.
We could have ignored the depreciation stream by assuming it equaled the
investment. However, this would not be analytically sound.
The addition of this cash flow without subtracting the investment tends
to understate impacts. The underestimate is relatively small in comparison
with the impact of other factors. Without data on the level of the ivestment
it is more accurate to include this cash flow than to exclude it.
Notice Costs
No attempt has been made to perform any cost analysis of the notice com-
pletion costs other than that mentioned in Chapters 5, 6, 7, and Appendix A.
The most up-to-date estimates provide a notice cost range of $1,000 to $9,000,
which does not include testing of any type. Because this was a before-tax
cost and we are performing after-tax analysis, we assumed a 50 percent tax
rate and obtained the after-tax cost range of $500 to $4,500.
M/Arthur D. Little, inc., impact of TSCA Proposed Premanufacturing
Notification Requirements (Cambridge, Mass.: December 1978).
-------
e-2i
EXHIBIT C-ll
REPRESENTATIVE DEPRECIATION CASH FLOWS OF CHEMICAL MANUFACTURERS
AS A PERCENTAGE OF NET SALES: 1974-1978
23 Basic Chemical Companies
Range of Depreciation
Average
Range of Tax Rates
Average Tax Rate
After-Tax Depreciation Cash Flow
5.54 to 6.60% of sales
6.30%
41.3 to 41.8%
41.6%
.416 x 6.3% = 2.62% of sales
20 Specialty Chemical companies
Range of Depreciation
Average Depreciation
Range of Tax Rates
Average Tax Rate
After-Tax Depreciation Cash Flow
2.56 to 2.73% of sales
2.59%
44.3 to 46.1%
44.9%
.449 x 2.59% = 1.16% of sales
Source: Value Line investment Survey, A. Bernhard & Co., New York, N.Y.,
1980.
RESULTS OF THE ECONOMIC ANALYSIS
Overview/Base Case Results
The base case results provide the most probable results of notice cost
imposition. These results can be used as a standard to assess the relative
changes in the results due to sensitivity analysis. As shown in Exhibit C-14,
if the after-tax cost of completing the notice is $10,000, fifteen percent of
all new chemicals that would have been introduced will not be. Because the
expected after-tax cost of the manufacturer's notice is $2,500, the expected
reduction in new chemicals is two percent, in the subsequent paragraphs the
derivation of the curve shown in Exhibit C-14 is provided.
The base case variables are assumed to be:
• 15 percent discount rate;
• 15 percent IRR;
• six percent ROS;
-------
C-22
• $2,500 (average) notice completion after-tax cost;
• 2.62 percent of sales depreciation cash flow; and
• exponential sales growth model compounded for 11 years.
Exhibit C-12 summarizes the data taken from the ADL report and the Reilly
Tar and Chemical company, giving sales or sales estimates in the first and
tenth years, and 1977.IS/ it shows the ADL compound growth estimate normal-
ized to the tenth year, the ICF compound growth estimate for the tenth year
and the number of years the chemical has been sold, the number of years it was
compounded to achieve the ICF estimate, and the resulting compound interest
rate expressed in percent growth per annum. Exhibit C-13 presents the pv
analysis for this data, it summarizes the first and tenth year sales esti-
mates, present value of cumulative sales for 11 years, the resulting ROS for
each chemical, and the adjusted return after a $2500 cost imposition.
Exhibit C-14 shows graphically a cumulative distribution of the present
values for the sample chemicals (from Exhibit C-12, Column D) with the assump-
tion that each sample chemical represents approximately eight percent of total
chemicals sold, with each sample falling into its respective eight percent
bracket from lowest to highest sales.
Exhibit C-15 shows how the imposition of a $2,500 additional cost would
change an assumed 15 percent IRR for the range of PVs found in the base case
(from Exhibit C-12, Column D). For example, if the present value of returns
was $10,000 before imposition of notice costs, the imposition of those costs
would reduce the IRR by approximately five percent. Thus the new IRR would be
10 percent. As expected, the impact of the $2,500 cost is most significant for
chemicals with smaller sales volume histories and lower present values.
M/Data are from the Reilly Tar and Chemical Company response to the
reproposal notice (44 Federal Register 59746, October 16, 1979).
basic sales growth data from which all subsequent calculations
were made are contained in the Supplement to this appendix which follows, it
is not necessary to refer to these data for the discussion of the results of
the analysis, however.
-------
C-23
EXHIBIT C-12
SUMMARY OF DATA USED IN BASE CASE ANALYSIS
Column
B
D
1st Year
Estimated
Chemical Sales
1 $2,180
2 8,920
3 17,560
4 16,630
5 15,500
6 8,000
7 16,660
8^/ 72,900
9 50,000
10 19,800
ll^/ 155,040
12 480,000
1977 Sales
$24,000
116,000
404,000
183,000
155,000
24,000
150,000
650,000
119,000
—
11,040,000
Years
Sold
11
13
23
11
10
3
9
10
13
6
10
23
ADL
10th Year
Estimated
Sales
$17,000
83,000
67,000
183,000
200,000
N/A
267,000
N/A
367,000
833,000
N/A
N/A
ICF
10th Year Annual
Estimated Years Growth
Sales Compounded Rate
$19,300
75,600
69,000
183,000
200,000
311,000
260,000
30,400
359,000
715,000
189,000
1,875,000
11
12
23
10
9
3
8
13
5
—
23
24.36%
23.83%
14.60%
27.10%
29.15%
44.22%
31.67%
21. 81%
43.15%
—
14.61%
N/A = Not available.
from Reilly Tar and Chemical Company, no model used. All other
chemicals from ADL report identified below.
Source: Arthur D. Little, inc., Impact of TSCA Proposed Premanufacture Noti-
fication Requirements (Cambridge, Mass.: December 1978); and ICF
estimates.
-------
C-24
EXHIBIT C-13
SUMMARY OF BASE CASE ECONOMIC ANALYSIS RESULTS
Column
Chemical
1
2
3
4
5
6
7
as/
9
10
ll£/
12
A
First Year
Sales
2,180
8,920
17,560
16,630
15,500
8,000
16,660
72,900
50,000
19,800
155,040
480,000
B
10th Year
Sales
19,300
75,600
69,000
183,000
200,000
311,000
260,000
30,400
359,000
715,000
189,000
1,875,000
C
Present Value
of Sales
36,560
145,950
189,830
316,990
325,500
348,450
394,540
549,488^
745,510
818,430
1,325,838^/
5,191,370
D
Net Cash
inf lows**/
3,151
12,581
16,363
27,325
28,058
30,036
34,009
47,366
64,263
70,549
114,287
447,496
E
Return After
Added $2,500
Cost
646
10,081
13,863
24,825
25,558
27,536
31,509
44,866
61,763
68,049
111,787
444,796
I/Include six percent ROS + Depreciation Cash Flows (2.62 percent of Sales).
b/Compounded for 10 years.
£/Reilly Tar and Chemical Company data.
-------
EXHIBIT C-14
BASE CASE RESULTS
CUMULATIVE PERCENTAGE OF CHEMICALS NOT INTRODUCED AS PMN COST RISES ASSUMING 15% RETURN REQUIREJD
CONDITIONS
• 15% DISCOUNT RATE
• 6% ROS + 2.62% OF SALES =
NET CASH INFLOW
• COMPOUND GROWTH MODEL
N>
en
AFTER-TAX
1000
(thousands of dollars)
-------
EXHIBIT C-15
CHANGE IN IKR CAUSED BY $2500 PMN COST IMPOSITION
$2,500 PMN COST
6% ROS
COMPOUND GROWTH MODEL
o
NJ
cn
PRESENT VALUE OF RETURNS (BEPQRE ADDED PMN COST)
Thousands of Dollars
-------
C-27
Sensitivity to changes in Discount Rate
In this portion of the analysis, we use discount rates of 10 percent, 15
percent, 30 percent, and 45 percent. This range is extremely wide, and it is
unlikely that any firm's required rate of return will exceed these limits.
Exhibit C-16 presents the present values for each chemical at the four differ-
ent discount rates, before the $2,500 notice cost addition. Exhibit C-17
illustrates the change in shape and value of the cumulative sales curve for
the four rates. This graphically illustrates the percentage of new chemicals
that might not be introduced as a result of additional cost increases as the
required rate of return increases. For example, varying the discount rate
from 15 percent to 45 percent changes the reduction in new chemical introduc-
tions from two percent to seven percent if the notice cost is $4,500 after
tax. Exhibit C-18 supports this illustration by showing the variation in the
sensitivity of changes in IRR from a $5000 notice cost for different starting
IRRs. This is above the maximum expected notice cost, but was chosen to
clearly illustrate the adverse impact.
EXHIBIT C-16
SENSITIVITY OF CASH INFLOW PV TO DIFFERENT DISCOUNT RATES!/
PRESENT VALUE
Chemical
1
2
3
4
5
6
7
8
9
10
11
12
First Year
Sales ($K)
2.18
8.92
17.56
16.63
15.50
8.00
16.66
72.90
50.00
19.80
155.04
480.00
Tenth Year
Sales ($K)
19.3
75.6
69.0
183.0
200.0
311.0
260.0
34.4^
359.0
715.0
189. O^/
1,875.0
10%
4,112
15,709
20,602
35,974
37,180
41,405
45,398
57,312
78,801
97,018
141,296
563,561
(Discount
15%
3,151
12,581
16,363
27,325
28,058
30,036
34,009
47,366
64,262
70,549
114,287
447,496
Rate)
30%
1,672
6,430
9,585
14,122
14,226
13,444
16,852
29,444
34,968
31,822
67,874
262,105
45%
1,076
4,157
6,678
8,887
8,802
7,385
10,213
20,401
22,987
17,628
45,700
182,572
5/Cash inflow includes depreciation cash return at 2.62 percent of sales.
^/Actual Sales data provided by Reilly Tar and Chemical Corporation.
-------
EXHIBIT C-17
SENSITIVITY OF RESULTS TO CHANGES IN REQUIRED RATES OF RETURN (DISCOUNT RATE)
DISCOUNT RATES OF
10%,15%,30% AND 45%
6% ROS + 2.62%
DEPRECIATION CASH
FLOWS
COMPOUND GROWTH
MODEL
O
10
100
AFTER-TAX NOTIFICATION COST (Thousands of Dollars)
-------
EXHIBIT C-18
SENSITIVITY OF CHANGES IN IRR FROM $5,000 PMN COST IMPOSITION
Chemical IRR
1 S/
2 4.5
3 6.6
4 6.8
5 7.0
6 8.5
7 7.6
8 7.3
9 8.5
10 9.2
11 9.4
12 9.8
10%
AIRR
-6.5
-3.4
-3.2
-3.0
-1.5
-2.4
-2.7
-1.5
-.8
-.6
-.2
PV
$15,700
20,600
36,000
37,200
41,400
45,400
57,300
78,800
97,000
141,300
563,600
IRR
a/
6.50
9.50
11.25
12.25
12.75
12.50
12.00
13.20
12.75
14.10
14.70
15%
AIRR
-8.50
-5.50
-3.75
-2.75
2.25
-2.50
-3.00
-1.80
-2.25
-.90
-.30
PV
$12,100
16,400
27,300
28,000
30,000
34,000
47,400
64,300
70,500
141,300
447,500
IRR
a/
15.75
18.50
23.75
22.80
25.50
23.80
25.50
26.25
27.50
27.75
29.50
30%
AIRR
-14.25
-11.50
-6.25
-7.20
-4.50
-6.20
-4.50
-3.75
-2.50
-2.25
-.50
PV
$6,400
9,600
14,100
14,200
13,400
16,900
29,400
35,000
31,800
67,900
262,100
IRR
a/
£/
24.5
30.2
30.8
31.5
32.5
34.8
37.2
38.6
39.4
43.8
45%
AIRR
—
-20.5
-14.8
-14.2
-13.5
-12.5
-10.2
- 7.8
- 6.4
- 5.6
- 1.2
PV
—
$6,700
8,900
8,800
7,400
10,200
20,400
23,000 9
NJ
17,600 ^
45,700
182,600
— IRR is negative.
-------
C-30
Sensitivity to ROS
Because return on sales (ROS) can vary widely between firms and among
individual products, we have chosen to examine the sensitivity of the present
value of casn inflows to a relatively wide range of returns. Returns for
individual firms were reported by ADL and ranged from two to over 12 percent
of sales. It is reasonable to expect for all chemical firms the range will be
greater for all products in a single year, as some products will lose money
and some will be extremely profitable. Over a longer period of time, however,
the average profits from a wide portfolio of chemicals must approach the
industry average if the chemicals are a representative sample of chemicals
sold. Profit margins have not changed drastically over the past two decades,
as Exhibit C-19 indicates. Because profit margins have not changed greatly,
the values of four percent, six percent, and 10 percent have been selected as
a reasonable and representative range for ROS.
EXHIBIT C-19
PROFITABILITY OF THE U.S. CHEMICAL INDUSTRY: 1964-1977
Year
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Net Sales
j$ million)
36,300
40,100
44,500
47,500
52,000
55,500
58,100
62,020
69,338
83,581
84,928
88,168
101,809
113,389
income
Before Tax
($ million)
N/A
N/A
N/A
N/A
N/A
N/A
N/A
6,700
7,788
9,978
10,285
9,668
11,268
11,806
Income
After Tax
($ million)
2,900
3,200
3,500
3,260
3,530
3,590
3,430
3,778
4,422
5,686
7,152
6,703
7,610
8,047
income
After Tax as a
Percent of Sales
7.9
7.9
7.8
6.9
6.8
6.5
5.9
6.1
6.4
6.8
8.4
7.6
7.5
7.1
Source: Arthur D. Little, inc., impact of TSCA Proposed Premanufacturing
Notification Requirements, (Cambridge, Mass.: 1978).
-------
C-31
Exhibit C-20 indicates the sensitivity of cash inflow present values to
changes in ROS. As might be expected, a change from four to 10 percent (a
two-and-one-half times increase in profits) has a significant effect on PV and
subsequently on the cumulative distribution plot as shown in Exhibit C-21.
For example, changing the ROS from six percent to 10 percent changes the
reduction in new chemical introductions from five percent to two percent in
the base case, in Exhibit C-20 and Exhibit C-21, the base case discount rate
(15 percent) has been used, and only ROS varied. The contribution from depre-
ciation has not changed.
Sensitivity to Growth Rates
Ideally, the actual sales histories for each sample chemical should be
used in the analysis, in lieu of this, the use of the generalized product
life cycle sales concept is reasonable, as discussed earlier and shown again
here in Exhibit C-22. This model can accurately predict the average growth
and decline in sales during a product's life time, but it must follow the
actual life cycle of the product, as the model's growth rate changes drasti-
cally over the life cycle, it is thus very sensitive to time and may, if
applied without a reasonable understanding of the product and its markets,
produce a misleading forecast.
EXHIBIT C-20
SENSITIVITY OF THE PRESENT VALUE OF CASH INFLOWS
TO CHANGES IN RETURN ON SALES (ROS)
(15% DISCOUNT RATE)
Present Value ($)
Chemical 4% ROS 6% ROS 10% ROS
1 2,100 3,151 5,251
2 8,057 12,085 20,141
3 10,909 16,363 27,272
4 18,217 27,325 45,542
5 18,705 28,058 46,763
6 20,024 30,036 50,060
7 22,671 34,006 56,677
8 31,577 47,366 78,943
9 42,841 64,262 107,103
10 47,037 70,555 117,592
11 76,191 114,287 190,478
12 298,329 447,493 745,822
-------
EXHIBIT C-21
SENSITIVITY OF RESULTS TO CHANGES IN ROS
100
100
1000
AFTER-TAX NOTIFICATION COSTS (Thousands of Dollars)
-------
C-33
EXHIBIT C-22.
PRODUCT LIFE CYCLE/EXPONENTIAL/LINEAR SALES GROWTH MODEL COMPARISON
Annual
Sales
Linear Sales
Growth Model
Exponential Sales
Growth Model
Product Life
Cycle Model
Years
-------
C-34
As shown in Exhibit C-22, the embryonic and early growth phases have
positive exponential growth, and thus can be closely modeled by a exponential
(or compound) growth function, if the assumption is made that the maturity or
late growth phase has not occurred then the use of an exponential model is
appropriate. If the late growth phase is assumed to have occured, then a
linear model might be less misleading. For these reasons, the sensitivity of
the chemicals' PVs of cash inflows are being tested with both the compound and
linear growth models, although there is no assurance which, if either, model
should give a higher or lower PV or which should be more reasonable.
Exhibit C-23 shows the results of the linear model, giving the first and
tenth year sales estimates, and PVs of cash inflows for the four discount
rates, assuming a six percent ROS. A comparison with estimates using the com-
pound growth model (see Exhibit C-14) shows that the linear model gives higher
present values for returns for seven of 10 chemicals. Exhibit C-24 shows the
cumulative sales distribution curves for the four discount rates using the
linear growth model. A comparison with the results of the compound growth
model (see Exhibit C-17) shows that a linear growth model assumption would
predict lesser impacts at any notification cost.
Sensitivity to Changes in Notice costs
At a minimum, a new chemical will not be introduced whenever the notice
cost exceeds the present value of cash inflows. TO the extent that our
approximation of depreciation charges discounted to the present understates
actual investment costs, the present values given in Exhibits C-16, C-20, or
C-23, and the impact on new chemicals introduced will be underestimated by
some unknown, but relatively small, amount.
The impact of any notice cost may be estimated from Exhibits C-17, C-21,
and C-24 as follows: calculate the notice cost, find the amount on the hori-
zontal axis that corresponds to the notice cost, move up to the sales curve,
and read left to the vertical axis from the curve's intercept point. In this
case, the number at that intercept point is the percent of new chemcals that
will not be introduced.
-------
C-35
EXHIBIT C-23
SENSITIVITY OF CASH INFLOW PV TO USE OF THE LINEAR GROWTH SALES MODELS/
Present value
(discount rate)
Chemical
1
2
3
4
5
6
7
8
9
10
11
12
First Year
Sales ($K)
2.2
72.9^/
8.0
8.9
15.5
16.6
17.6
155. O^/
19.8
47.3
50.0
480.0
Tenth Year
Sales ($K)
22.0
30. 4^/
90.0
89.0
155.0
166.0
176.0
IB. 9^
198.0
473.0
500.0
4,800.0
10%
5,000
57,312
20,027
22,283
38,790
41,548
44,048
141,296
49,551
118,381
125,134
1,201,384
15%
3,793
47,366
15,171
16,881
29,394
31,477
33,373
114,287
37,540
89,691
94,820
910,200
30%
1,892
29,444
7,571
8,423
14,668
15,703
16,651
67,874
12,492
44,767
55,929
454,260
45%
1,111
20,401
4,445
4,945
8,611
9,223
8,341
45,700
11,001
26,277
27,785
266,688
.2/Cash inflow includes depreciation cash return at 2.62 percent of sales.
^/Actual sales data provided by Reilly Tar and Chemical Corporation, no model
used.
-------
EXHIBIT C-24
PRESENT VALUES OF CASH INFLOWS USING LINEAR GROWTH MODEL
100
1000
AFTER-TAX NOTIFICATION COST (Thousands of Dollars)
-------
C-37
SUPPLEMENT TO APPENDIX C
SALES DATA USED IN ECONOMIC ANALYSIS
This supplement contains the sales data generated for each chemical sample
used in the analysis. Data on two of the chemicals were provided by the Reilly
Tar and Chemical Company, as part of their public response to the reproposal
(44 Federal Register 59764)- For these two samples, sales volumes (in pounds)
for the years 1967 through 1976 were converted into dollars by multiplying by
the average price received during that year. The appropriate price and resul-
tant sales value in dollars are shown in Exhibits C-33 and c-36. For these
two chemicals, no sales growth model was used. For the remaining 10 samples,
both linear and exponential (compound interest) model results are shown.
These tables do not include cash flows from depreciation. This cash flow,
calculated as a percent of sales as discussed in the subsection Depreciation,
can be found for any particular chemical and year by multiplying the value
found in any appendix table by .0262 (2.62 percent). All calculations per-
formed in the section Analysis of PMN Costs impacts on New Chemical
Introductions include depreciation cash flows.
Linear Model Data
As previously discussed, the linear model calculated sales in each year
by dividing 1977 sales by the total years of sales. The resulting number was
sales in Year 1, and each subsequent year simply added this number for total
sales. Thus, if sales in year one were "X", sales in year two were 2X; year
three, 3X; and so on. The values for the chemicals in the 10th year of sales
are shown in Exhibit C-25.
Because'the numbers of direct interest are the present value of sales and
return on sales, only present values at the four discount rates (10 percent,
15 percent, 30 percent, and 45 percent) are shown for sales summed for 10
years. For the linear model, sales in any year can be derived from the non-
discounted year 10 sales by dividing by the appropriate factor; thus, only
10th year sales estimates are given. The present values of the sum of 10 years
of sales and net income after tax at four percent, six percent, and 10 percent
return on sales (ROS) are shown as well.
Exponential (Compound Interest) Model Data
Data for the 10 chemical samples from the Arthur D. Little report as well
as the Reilly Tar and Chemical Company data are shown in Exhibits C-26 through
C-37. Each exhibit contains the present value of sales at each discount rate
(10 percent, 15 percent, 30 percent, and 45 percent) for each year, the total
10 years present value of sales, and the returns for each ROS (four percent,
six percent, and 10 percent). The growth rate, calculated as discussed in the
subsection Sales 'Growth, is shown at the top of each table.
-------
EXHIBIT C-25
LINEAR GROWTH MODEL (PRESENT VALUE SUMMED FOR TEN YEARS OF SALES)
($1000)
ICF 10th Year
Linear Growth
Chemical Estimate
1 20
2 80
3 89
4 155
5 166
6 176
7 198
8 500
9 473
10 4,800
10% Discount
Return on Sales
.04
2.323
9.291
10.340
18.000
19.280
20.440
23.000
58.080
54.940
557.490
.06
3.484
13.940
15.510
27.000
28.920
30.660
34.490
87.100
82.400
836.230
.10
5.807
23.230
25.840
45.010
48.200
51.100
57.490
145.180
137.340
1,393.700
15% Discount
Return on Sales
.04
1.760
7.039
7.831
13.640
14.610
15.490
17.420
44.000
41.620
422.370
.06
2.64
10.56
11.75
20.46
21.91
23.23
26.13
66.00
62.43
633.55
.10
4.40
17.60
19.58
34.10
36.52
38.72
43.56
110.00
104.05
1,055.91
30% Discount
Return on Sales
.04
.8783
3.5130
3.9080
6.8070
7.2900
7.7290
8.6950
21.9580
20.7700
210.7900
.06
1.317
5.270
5.863
10.210
10.930
11.590
13.040
32.930
31.160
316.190
.10
2.196
8.783
9.771
17.020
18.220
19.320
21.740
54.900
51.930
526.980
45% Discount
Return on Sales
.04
.5156
2.0630
2.2950
,1,
3.9960
4.2800
4.5380
5.1050
12.8900
12.1900
123.7500
.06
.7734
3.0940
3.4420
5.9940
6.4200
6.806
7.657
19.340
18.290
185.630
.10
1.289
5.156
5.736
9.990
10.700
11.340
12.760
32.230
30.490
309.380
n
i
u
CO
Source: ICF Estimates.
-------
C-39
EXHIBIT C-26
PRESENT VALUE OF CHEMICAL #1 SALES OVER TEN YEARS
(growth rate = 24.36 percent per year)
End of Year
1
2
3
4
5
6
7
8
9
10
11
Total
4% ROS
6% ROS
10% ROS
10%
2.180
2.456
2.786
3.150
3.561
4.026
4.552
5.146
5.818
6.577
7.436
47.700
1.908
2.862
4.770
Discount Rate
15%
2.180
2.357
2.549
2.757
2.981
3.224
3.486
3.770
4.077
4.409
4.767
36.560
1.462
2.193
3.656
30%
2.180
2.085
1.995
1.908
1.826
1.746
1.671
1.598
1.529
1.462
1.399
19.400
.776
1.164
1.940
45%
2.18
1.870
1.604
1.375
1.180
1.012
.8676
.7441
.6382
.5474
.4694
12.4900
.4995
.7492
1.2490
-------
C-40
EXHIBIT C-27
PRESENT VALUE OF CHEMICAL #2 SALES OVER TEN YEARS
(growth rate = 23.83 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
8.920
10.040
11.300
12.730
14.330
16.130
18.150
20.440
23.010
25.900
29.150
15%
8.920
9.605
10.340
11.140
11.990
12.910
132.900
14.970
16.120
17.360
18.690
30%
8.920
8.497
8.093
7.709
7.343
6.995
6.663
6.347
6.045
5.758
5.485
45%
8.920
7.618
6.505
5.556
4.745
4.052
3.450
2.955
2.524
2.155
1.841
Total 190.090 145.950 77.860 50.330
4% ROS
6% ROS
10% ROS
7.604
11.410
19.010
5.838
8.757
14.500
3.114
4.671
7.786
2.013
3.020
5.033
-------
C-41
EXHIBIT C-28
PRESENT VALUE OF CHEMICAL #3 SALES OVER TEN YEARS
(growth rate = 14.60 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
17.560
18.290
19.060
19.860
20.690
21.550
22.450
23.390
24.370
25.390
26.450
15%
17.560
17.500
17.440
17.380
17.320
17.260
17.200
17.140
17.080
17.020
16.960
30%
17.560
15.480
13.650
12.030
10.600
9.348
8.241
7.265
6.404
5.645
4.977
45%
17.560
13.880
10.970
8.669
6.852
5.415
4.280
3.383
2.673
2.113
1.670
Total 239.060 189.830 111.200 77.460
4% ROS
6% ROS
10% ROS
9.563
14.340
23.900
7.593
11,390
18.980
4.448
6.672
11.120
3.098
4.648
7.746
-------
C-42
EXHIBIT C-29
PRESENT VALUE OF CHEMICAL #4 SALES OVER TEN YEARS
(growth rate = 27.10 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
16.630
19.220
22.202
25.650
29.640
34.250
39.570
45.730
52.830
61.050
70.540
15%
16.63
18.38
20.31
22.45
24.81
27.42
30.31
33.50
37.02
40.92
45.22
30%
16.630
16.260
15.900
15.540
15.200
14.860
14.520
14.200
13.880
13.570
13.270
45%
16.630
14.580
12.780
11.200
9.818
8.606
7.543
6.612
5.796
5.080
4.453
Total 417.310 316.99 163.830 103.090
4% ROS
6% ROS
10% ROS
16.690
25.040
41.730
12.68
19.02
31.70
6.553
9.830
16.380
4.124
6.186
10.310
-------
C-43
EXHIBIT C-30
PRESENT VALUE OF CHEMICAL #5 SALES OVER TEN YEARS
(growth rate = 29.15 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
15.50
18.20
21.37
25.09
29.45
34.58
40.60
47.67
55.97
65.71
77.15
15%
15.50
17.41
19.55
21.95
24.66
27.69
31.10
34.92
39.22
44.05
49.46
30%
15.500
15.400
15.300
15.200
15.100
15.000
14.900
14.800
14.710
14.610
14.520
45%
15.500
13.810
12.300
10.950
9.755
8.689
7.739
6.893
6.140
5.469
4.871
Total 431.29 325.50 165.030 102.110
4% ROS
6% ROS
10% ROS
17.25
25.88
43.13
13.02
19.53
32.55
6.601
9.902
165.030
4.084
6.127
10.210
-------
C-44
EXHIBIT C-31
PRESENT VALUE OF CHEMICAL #6 SALES OVER TEN YEARS
(growth rate = 44.22 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
8.00
10.49
13.75
18.03
23.64
30.99
40.63
53.27
69.85
91.58
120.07
15%
8.00
10.03
12.58
15.78
19.79
24.82
31.12
39.03
48.95
61.38
76.98
30%
8.000
8.875
9.846
10.920
12.120
13.440
14.910
16.540
18.350
20.360
22.590
45%
8.000
7.957
7.914
7.872
7.829
7.787
7.745
7.704
7.662
7.621
7.580
Total 480.30 348.45 155.970 85.67
4% ROS
6% ROS
10% ROS
19.21
28.82
48.03
13.94
20.91
34.85
6.239
9.358
15.600
3.427
5.140
8.567
-------
C-45
EXHIBIT C-32
PRESENT VALUE OF CHEMICAL #7 SALES OVER TEN YEARS
(growth rate = 31.67 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
16.66
19.94
23.87
28.57
34.20
40.94
49.00
58.66
70.21
84.05
100.60
15%
16.66
19.07
21.84
25.01
28.63
32.78
37.53
42.97
49.20
56.33
64.50
30%
16.660
16.870
17.090
17.310
17.530
17.760
17.990
18.220
18.450
18.690
18.930
45%
16.660
15.130
13.740
12.470
11.330
10.290
9.340
8.482
7.702
6.994
6.351
Total 526.72 394.54 195.500 118.490
4% ROS
6% ROS
10% ROS
21.07
31.60
52.67
15.78
23.67
39.45
7.820
11.730
19.550
4.739
7.109
11.850
-------
EXHIBIT C-33
PRESENT VALUE OF REILLY CHEMICAL #2 OVER TEN YEARS
(Chemical #8 in text of this section.)
Present Value
(discount rate)
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
Sales (Ibs.)
2,700
2,500
16,800
22,600
1,000
400
11,000
22,000
5
40
Price/lb.
$27.00
35.00
11.40
9.60
43.00
66.00
13.70
10.00
760.00
760.00
Total:
Value
72,900
87,500
191,520
216,960
43,000
26,400
150,700
220,000
3,800
30,400
10%
66,273
72,314
143,892
148,187
26,700
14,902
77,333
102,632
1,612
11,721
665,566
15%
63,391
66,163
125,928
124,048
21,379
11,413
56,654
71,918
1,080
7,514
549,488
30%
56,077
51,775
87,173
75,964
11,581
5,469
14,017
26,970
358
2,205
341,589
45%
50,276
41,617
62,822
49,080
6,709
2,841
11,182
11,258
134
740
236,659
Total x .04
Total x .06
Total x .10
ROS
ROS
ROS
26,622
39,934
66,557
21,980
32,969
54,949
13,664
20,495
34,159
9,466
14,200
23,666
n
it*
CTi
Source: Reilly Tar & Chemical Company and ICF Estimates.
-------
C-47
Total
EXHIBIT C-34
PRESENT VALUE OF CHEMICAL #9 SALES OVER TEN YEARS
(growth rate = 21.81 percent per year)
Discount Rate
End of year
1
2
3
4
5
6
7
8
9
10
11
10%
50.00
55.37
61.31
67.90
75.18
83.26
92.20
102.09
113.06
125.19
138.63
15%
50.00
52.96
56.10
59.42
62.94
66.66
70.61
74.79
79.22
83.91
88.88
30%
50.00
46.85
43.90
41.13
38.54
36.11
33.84
31.71
29.71
27.84
26.08
45%
50.000
42.000
35.290
29.640
24.900
20.920
17.570
14.760
12.400
10.420
8.752
914.19
745.51
405.71
266.660
4% ROS
6% ROS
10% ROS
36.57
54.85
91.42
29.82
44.73
74.55
16.23
24.34
40.57
10.670
16.000
26.670
-------
C-48
EXHIBIT C-35
PRESENT VALUE OF CHEMICAL #10 SALES OVER TEN YEARS
(growth rate = 43.15 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
19.80
25.77
33.53
43.64
56.79
73.90
96.17
125.16
162.88
211.96
275.84
15%
19.80
24.65
30.68
38.19
47.54
59.17
73.66
91.69
114.13
142.07
176.85
30%
19.80
21.80
24.01
26.44
29.11
32.06
35.30
38.87
42.80
47.13
51.90
45%
19.800
19.550
19.300
19.050
18.810
18.570
18.330
18.100
17.870
17.640
17.410
Total 1,125.43 818.43 369.21 204.420
4% ROS
6% ROS
10% ROS
45.02
67.53
112.54
32.74
49.11
81.84
14.77
22.15
36.92
8.177
12.270
20.440
-------
EXHIBIT C-36
PRESENT VALUE OF REILLY #7 CHEMICAL OVER TEN YEARS
(Chemical #11 in text of this section.)
Present Value
(discount rate)
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
Sales (Ibs.)
11,400
26,600
32,400
60,200
84,500
75,800
113,000
186,000
53,000
135,000
Price/lb.
$13.60
9.20
8.60
6.40
3.30
5.20
13.70
1.30
6.90
1.40
Total:
Value
155,040
244,720
278,640
385,280
278,850
394,160
150,700
241,800
365,700
189,000
10%
140,945
202,248
209,346
263,151
173,144
222,493
77,333
112,801
155,093
72,868
1,639,069
15%
134,817
185,043
183,210
220,285
138,638
170,406
56,654
79,045
103,955
46,718
1,325,838
30%
119,262
144,805
126,827
134,897
75,102
81,661
14,017
29,642
34,485
13,710
787,404
45%
106,924
116,395
91,399
87,157
43,504
42,410
11,182
12,374
12,836
4,600
530,176
Total x .04
Total x .06
Total x .10
ROS
ROS
ROS
65,562
98,344
163,907
53,034
79,550
132,584
31,496
47,244
78,740
21,207
31,810
53,017
o
Source: Reilly Tar & Chemical Company and ICF Estimates.
-------
C-50
EXHIBIT C-37
PRESENT VALUE OF CHEMICAL #12 SALES OVER TEN YEARS
(growth rate = 14.61 percent per year)
Discount Rate
End of Year
1
2
3
4
5
6
7
8
9
10
11
10%
480.00
500.12
521.08
542.91
565.67
589.37
614.07
639.81
666.62
694.56
723.67
15%
480.00
478.37
476.75
475.13
473.52
471.92
470.32
468.72
467.13
465.55
463.97
30%
480.00
423.18
373.08
328.91
289.97
255.64
225.38
198.70
175.18
154.44
136.15
45%
480.00
379.40
299.88
237.03
187.35
148.09
117.05
92.52
73.13
57.80
45.69
Total 6,537.88 5,191.37 3,040.63 2,117.93
4% ROS
6% ROS
10% ROS
261.52
392.27
653.79
207.65
311.48
519.14
121.63
182.44
304.06
84.72
127.08
211.79
-------
APPENDIX D
DISCUSSION OF THE ECONOMIC BURDEN OF SECTION
NOTICE REQUIREMENTS ON EPA AND SOCIETY
-------
APPENDIX D
INTRODUCTION
The section 5 notice (hereafter notice) policy may impose economic burdens
on three principal actors:
• the chemical industry;
• the EPA; and
• society (i.e., the intended beneficiaries of
TSCA—including public and special interest groups).
The purpose of this document is to identify the potential costs to EPA and
to society associated with these notice policy issues:i/
• Processor Reporting Rule;
• Importer Definition Options;
• Importer Contact of Foreign Manufacturers/Suppliers;
• Exporter Options;
• Customer Contact;
• Insufficient Submissions;
• Supplemental Reporting; and
• Confidentiality Options.
These issues are examined individually.I/ The criteria used to identify the
costs to EPA and society vary from issue to issue depending on the options.
This paper is designed to assist EPA in making better decisions in each
case by constructing a framework from which to evaluate the proposed rule and
its alternatives.
A/Costs imposed on the chemical industry have been assessed. See
"Economic Impact Analysis of Proposed Section 5 Notice Requirements," Part
II-Issue Papers (Washington, D.C.: ICF Incorporated, September 1980). This
Appendix should be read in parallel with Part II.
.2/The Possession or Control issue is not discussed here.
-------
D-2
Because many of the costs associated with these issues do not lend
themselves easily to quantification, the major outcome of this discussion is
an ordinal ranking of options by expected economic costs to EPA and society.
THE PROCESSOR REPORTING RULE
Under the joint statutory authority of TSCA sections 5(a)(1) and 5(a)(2),
EPA has proposed to require the person^/ who first processes an exempt,
non-inventoried substance for a nonexempt purpose to submit a section 5
notice. Without such a rule, persons could process for a TSCA commercial
purpose an exempt substance which was not on the Inventory and which was never
reviewed for toxic effects.
Options
For the purposes of this discussion, EPA specified the following options
for the processor reporting rule:
Option 1; Proposed Processor Reporting Rule with section 5
notice similar to the October 16 form proposed
for manufacturers.
Option 2; January 10 Proposal—Manufacturer Reporting
requirement.
Option 3; No reporting for commercial processing of exempt
substances.
Option 4; Section 8 Processor Notification.
Option 5; Processor rule as proposed with exemption for
one-time processing.
Option 6; Combination of section 5 and section 8(a)
notices.
Option 7; Section 5 notification under minimum guidance
option.
Each of these options is evaluated below according to the costs it imposes on
EPA and on society.£/
I/For purposes of this report, "person" can be an individual or a corpo-
rate entity.
I/Thorough descriptions of these options are contained in: "Economic
Impact Analysis of Proposed Section 5 Notice Requirements," Part II - Issue
Papers, Chapter 8—Processor Reporting (Washington, D.C.: ICF Incorporated,
September 1980).
-------
D-3
EPA Costs
A portion of the economic costs of a processor rule will fall on EPA in
the form of notice reviews and expenditures related to requests for
supplemental information.
Under the proposed processor rule, costs to the Agency would be identical
to those incurred for manufacturers and importers under the premanufacturing
notification proposal. The major difference between the processor rule and
the January 10 and October 16 proposals is coverage; with a processor notice
rule, EPA would have additional notices to review. However, the processor
rule as proposed with an exemption for one-time processing would reduce the
number of notices submitted for review.
A section 8 rule would represent less of a burden on EPA than a section 5
processor rule to the extent that a section 8 notice requires less
information. The review process is made complicated when a section 5 rule is
proposed in combination with a section 8 rule, but the marginal cost to EPA
associated with such a combination is unclear. In addition to review costs, a
combined rule would impose other significant costs on EPA in the form of
identifying chemical categories that would be subject to section 5 notices.
Finally, the "no reporting" alternative would not impose any costs on
EPA. Exhibit D-l presents a ranking of the options by two types of costs to
EPA: administrative and insufficient information. A "1" corresponds to the
least costly option.
Administrative Costs. The first column of Exhibit D-l ranks the options
by the per-notice administrative cost of review. The amount of information
requested in the notice was the key factor in deciding which option would
require longer and more careful review by EPA. Thus, because separate
processors would not submit information under Options 2 and 3, these options
were ranked as the least costly.
Section 8 notices could request less information than section 5 notices,
and EPA would presumably require less time to review them. Thus, Option 4 was
ranked as the second least costly alternative. A notice submitted under the
minimum guidance option may contain more information than a section 8 notice,
and thus Option 7 ranked higher than Option 4 in review costs.
EPA would have the most information to review under any of the options
requiring a section 5 notification. Thus, costs of review are highest when a
section 5 notice is submitted.
Options 5 and 6 may rank similarly to Option 1 or Option 4, depending on
whether a section 5 or a section 8 notice applies.
Insufficient Information Costs. The second column of Exhibit D-l ranks
the options in terms of information submitted to EPA. If a notice submitted
under an option does not provide sufficient information for EPA to make an
-------
D-4
EXHIBIT D-l
RANKING OF PROCESSOR NOTIFICATION OPTIONS
BY COSTS TO EPA
Cost
Option
Review
Insufficient Information
to Support Regulation
1. Proposed Processor
Notice Rule
2. January 10 Proposal
3. No Reporting
4. Section 8 Notice
5. Processor Rule
With Exemption
1 or 3
1 or 2
6. Combined Section 5
and Section 8 Notice
1 or 3
1 or 2
7. Minimum Guidance
Option
A "1" indicates the least costly alternative.
-------
D-5
optimum decision, then that represents a cost to the extent that there may
exist more effective ways for the Agency to carry out its statutory
responsibilities.
On the other hand, information requested under another option may help the
Agency make a more qualified decision on how to regulate the substance in
question. In Exhibit D-l, a rank of "1" indicates that of the options con-
sidered, EPA will have the best available information on which to base its
decisions. A higher rank means that EPA will have less information on which
to base its decision. Because EPA would receive no information under Option
3, the cost of insufficient information is highest here relative to the other
options.
Society Costs
Promulgation of a processor notice rule may impose costs on society in the
form of foregone chemicals and time delays in being able to use processed
substances. These costs will equal the corresponding costs under the
premanufacturing notification rule as proposed on January 10 and reproposed on
October 16. Although these society costs may be significant, of greater
importance is the tradeoff encountered by society for different levels of
protection from toxic substances.
The tradeoff varies across alternative processor rules. The least
stringent alternative (no reporting for commercial processing of exempt
substances) affords the least amount of protection. In other words, without
any mechanism for reviewing exempt substances prior to nonexempt commercial
processing, the probability of a toxic substance (i.e., one that may present
an "unreasonable risk of injury to health or the environment") entering
commerce is maximized.
The probability of an "unreasonably risky" toxic substance entering
commerce is decreased when substances undergo a review of toxicity prior to
processing. In options that require notice submission (i.e., Options 1, 5, 7,
and to an extent 2 and 6), that probability presumably is further diminished.
Exhibit D-2 presents a ranking of processor notification options by costs
to society in terms of foregone chemicals, time delays, and unreasonable risks.
Foregone Chemicals and Time Delay Costs. The January 10 proposal and the
option of no reporting received the lowest rank with respect to foregone
chemicals and time delays for several reasons. First, the risk of an adverse
determination would be absent because no separate processor notice would be
submitted. The absence of such a risk would keep a person's decision to
process unchanged; thus, society would not have to forego the use of any
processed chemicals. Secondly, because no separate processor notice is
submitted under either alternative, delays in being able to use the substances
would be nonexistent. Both proposals rank high in terms of exposure risks for
these reasons. Because the January 10 proposal contains provisions for
reviewing a few processed substances, it ranked lower in terms of exposure
risks than the option of no reporting.
-------
D-6
EXHIBIT D-2
RANKING OF PROCESSOR NOTIFICATION OPTIONS
BY COST TO SOCIETY
~~^ — «^^_^ Cost Foregone Time
Option -^^^ Chemicals Delays
1. Proposed Processor 3 3
Notice Rule
2. January 10 Proposal 1 1
3. No Reporting 1 1
4. Section 8 Notice 2 2
5. Processor Rule 2 or 3 2 or 3
With Exemption
6. Combined Section 5 2 or 4 2 or 3
and Section 8 Notice
7. Minimum Guidance 3 3
Option
Unreasonable
Risks
1
6
6
4
2
3
1
A "I" indicates the least costly alternative.
-------
D-7
The degree and extent of coverage under each option were used to rank the
remaining options. For example, a section 8 notice covers fewer substances
and requires less information than a section 5 notice. Thus, it would re-
sult in fewer foregone chemicals and shorter delays in processing. Delays and
chemicals foregone under Option 5 (the processor rule with an exemption for
one-time processing) depend on whether a section 5 or a section 8 notice is
submitted. The same is true for Option 6.
Costs of Unreasonable Risks. Under each option, coverage and the type of
information submitted to support regulation was used to rank the options in
terms of unreasonable health and safety risks. Unreasonable risks are a
function of exposure and toxicity. When coverage is widest, EPA has a chance
to control a broader array of exposures. When valuable information to support
regulation is submitted, EPA has a chance to determine whether the chemical's
toxicity presents an unreasonable risk. Without information on toxicity, EPA
cannot make a knowledgeable decision on whether or not to exercise its
regulatory authority, and society may face unnecessary and unreasonable risks.
The proposed rule and the minimum guidance option would present fewer
unreasonable risks among the options because they cover all nonexempt
commercial processing and because available health and safety studies may be
submitted. EPA would thus be able to make a more qualified decision on
whether or not to take regulatory actions, and the amount of unreasonable risk
to society is minimized.
Under the option of no reporting, unreasonable risks may be higher because
EPA would have little information on which to base its regulatory decisions.
IMPORTERS
Under section 720.10(a)(2) of the January 10, 1979 proposed rules, "any
person who intends to import" a new chemical substance is required to submit
notification to EPA. Because of the variety of participants in importing and
the technical nature of some of the information required in the notice, EPA is
considering the question of which parties should be responsible or required to
submit notification for imported new chemical substances.
Options
We have evaluated five alternatives for designating persons responsible
for submission of section 5 notification on imported new chemical
substances:V
Option 1; Define importer as the consignee.
j>/These options are described more completely in: "Economic Impact
Analysis of Proposed Section 5 Notice Requirements", Part II - Issue Papers,
Chapter 3 - Definition of Importers (Washington, D.C.: ICF Incorporated,
September 1980).
-------
D-8
Option 2; Define importer as consignee, and impose
upon him an obligation to obtain the
information required by the rules and forms
from other persons.
Option 3; Use January 10 proposal importer definition.
Option 4; Define importer as "the person who imports,
or knowingly causes to be imported a
chemical substance."
Option 5; Clarify January 10th proposal by changing
definitions so that principal importer must
submit section 5 notice.
Each of these options is evaluated on the basis of costs to EPA and society.
It is assumed that the notification form submitted for all five options is the
October 16th reproposed form.
EPA Costs
Costs incurred by EPA include (1) administrative costs of reviewing
notices, and (2) expenditures related to requests for supplemental information.
Administrative Costs. The likelihood of submission of notices to EPA
under each option is used as the measure of administrative costs. Since the
same form will be submitted under each option, the administrative costs to EPA
depend on how many forms actually are submitted; the more notices submitted,
the more reviewing by EPA will be required.
The definition of importers under Option 5 is intended to reach a single
individual. However, the parties involved will face some uncertainty as to
who will submit notification in different types of import transactions unless
these circumstances are defined in great detail.
Under Options 3 and 4, a single person with responsibility for submission
is not identified. Since no one party has primary responsibility, it is more
likely that a new chemical will be imported without any notification being
submitted. In addition, these alternatives burden the parties involved with
negotiating an agreement on who should submit notification. This increases
the uncertainty faced by the parties, and thus decreases the likelihood of
submission of a notice.
Exhibit D-3 shows the relative administrative costs to EPA under each
option. If all importers comply with the law the administrative costs would
be identical across all options.
Supplemental Information Costs. To measure the amount of resources EPA
must expend requesting supplemental information, we assume that the more
technical knowledge the submitter has, the better will be the quality of
information submitted. Presumably, EPA will devote more resources requesting
-------
D-9
EXHIBIT D-3
THE RELATIVE COSTS OF IMPORTER DEFINITION OPTIONS
Costs
Options
EPA Cost 1 EPA Cost 2 Society Cost
(Administrative) (Supplemental
Information)
1. Importer Defined
as Consignee
2. Importer as Consignee
With Obligation To
Obtain Information
From Others
3. January 10 Proposal
4. Importer as "Person
Who Imports Or Knowingly
Causes To Be Imported A
Chemical Substance"
5. Principal Importer Must
Submit Section 5 Notice
A "1" indicates the least costly alternative.
-------
D-10
supplemental information when the quality or quantity of information submitted
in the notice is poor.
Some of the options evaluated here do not specify how many people will be
involved in the notification process, nor whom these persons will be. It is
difficult, therefore, to assess the level of technical knowledge a submitter
has, and then to estimate the quality of information submitted. Under any of
the options, the behavior of the parties involved cannot be predicted with
certainty. For example, under Option 1, the submitter is the consignee of the
imported chemical, and may be a retailer or a broker. Retailers who intend to
process, use or distribute the imported chemical themselves, presumably would
have the technical knowledge to submit a section 5 notice. On the other hand,
brokers rarely would be aware or able to determine that chemical shipments in
a consignment are "new". As a result of this uncertainty as to the level of
technical knowledge of the consignee, we will evaluate each option according
to the opportunity for the most knowledgeable person to submit notification.
As discussed above, under Option 1, the consignee may not be the most
knowledgeable person (if the consignee is a broker) to submit a section 5
notice. A broker does not have the technical knowledge to adequately submit
the notification form. By relying on brokers to submit notification under
Option 1, EPA would be limiting itself to the minimum of information
available; EPA would be most likely to require supplemental information under
Option 1.
Option 2 notices would contain better information than Option 1 notices.
The submitter of Option 2 notices is required to obtain information from other
persons, including identifiable future users and processors of the chemical.
These other persons would be expected to possess some technical knowledge, and
improve the quality of information available to EPA.
Option 3 requires each of the parties included in the definition of
importer to be responsible for submission. If all the parties with technical
knowledge are responsible for submitting a notice to EPA, the quality of
information would be high. However, there is no assurance provided by this
regulation that the most knowledgeable party would agree, or could be
persuaded, to submit a Section 5 notification.
The Option 4 definition of importer requires the most knowledgeable person
to submit notification. In some cases, it might prove difficult for both the
Agency and the parties involved to identify the most knowledgeable person.
Finally, Option 5 is designed to get the most knowledgeable person to
submit notification. Thus, Option 5 notices may contain better information
than the other options; the necessity of EPA requesting supplemental
information will be minimized under Option 5. Exhibit D-3 shows the relative
supplemental information costs of the five importer options.
Society Costs
Costs to society are incurred when the public is not protected from toxic
substances. The probability of an "unreasonably risky" toxic substance
-------
D-ll
entering commerce is decreased when substances undergo a review for toxicity
prior to processing. In the case of imported chemicals, the factors deter-
mining their review include: (1) whether or not a section 5 notice is
submitted, and (2) the quality of information submitted. Presumably, society
is well protected (and hence costs to society are low) when a section 5 notice
is submitted that has been filled out by the most knowledgeable person
involved with importing the chemical.
According to these criteria, Option 5 affords the greatest protection to
society (and the least cost in terms of unreasonable risk). Under Option 5,
the information included in the notice will be high in quality, and only one
person is responsible for submitting the notice.
It is difficult to predict the relative consequences of choosing among
Options 2, 3, and 4. They are ranked equivalently in terms of society costs.
Option 1 notices will provide the lowest quality information of all the
options.
IMPORTER CONTACT OF FOREIGN MANUFACTURERS/SUPPLIERS
This issue concerns obtaining information from foreign manufacturers and
suppliers about new chemical substances being imported into the United States.
Options
The following four alternatives are evaluated according to the costs they
impose on EPA and society:^/
Option 1; mandatory contact and the Foreign Manufacturers/
Suppliers Form outlined in the January 10
proposal;
Option 2; mandatory contact and a Foreign Manufacturers/
Suppliers Form revised to be consistent with the
forms in the October 16 reproposal;
Option 3; mandatory contact and the request for only health
and environmental effects data and risk
assessments from foreign manufacturers/suppliers;
and
JL/A separate document contains thorough descriptions of these options:
"Economic Impact Analysis of Proposed Section 5 Notice Requirements," Part II
- Issue Papers, Chapter 4-importer Contact of Foreign Manufacturers/Suppliers
(Washington, D.C.: ICF Incorporated, September 1980).
-------
D-12
Option 4; no mandatory contact by importer.
An assessment of the costs to EPA and society imposed by any of the four
options is made difficult by the fact that cooperation on the part of foreign
manufacturers/suppliers is voluntary. An assessment of the relative costs of
each option must depend on some assumptions about the behavior of industry
parties. It is difficult to predict, for example, whether a foreign
manufacturer will submit more information if asked for it, or whether the more
onerous request will discourage submission of some or possibly any
information. Currently, many foreign manufacturers must submit information to
their own governments which is equivalent or similar to that which would be
required under the proposed regulations.
EPA Costs
We examine EPA's costs in terms of the ease of administration of various
options. Ease of administration will be high when the foreign manufacturer/
supplier form is consistent with the domestic manufacturer form. EPA then
will have only one notice format to review. Assuming the domestic
manufacturer form will be the October 16 proposal, Option 2 will be the least
costly to EPA.
Option 3 is ranked less costly to EPA than Options 1 or 2, since the
amount of information EPA must review is less under Option 3; the foreign
manufacturers/suppliers form outlined in the January 10 proposal (Option 1)
requests more information than that requested in Options 2 or 3.
Finally, under Option 4, a lack of adequate information hinders EPA's
administration of the program. Mandatory contact by the importer is of
practical benefit to EPA as a first step in obtaining necessary information.
Because EPA cannot forego contact with the foreign manufacturer, Option 4 is
most costly to the Agency. Exhibit D-4 displays these relative costs to EPA.
Society Costs
Costs to society are assessed on the basis of how much information EPA has
available. A more informed decision by EPA, presumably, will protect society
to the greatest degree.
-------
D-13
EXHIBIT D-4
RELATIVE COSTS OF IMPORTER CONTACT OF FOREIGN MANUFACTURERS/SUPPLIERS
Costs
Options
EPA Cost
(Ease of
Administration)
Society Cost
(Risks)
1. January 10 Proposal
2. October 16 Proposal
Only Health and
Environmental Effects
Data and Risk Assess-
ment
No Mandatory
Contract
A "1" indicates the least costly alternative.
-------
D-14
The amount of information submitted to EPA is highest under Option 1. The
January 10 proposal includes several questions which are omitted in the
October 16 reproposal. The other three Options, 2, 3, and 4, provide EPA with
successively less information, and are successively more costly to society.
Exhibit D-4 shows the progression from the least to the most costly option.
EXPORTERS
This issue pertains to the manufacture in the U.S. of new chemicals solely
for the purpose of export abroad. EPA is concerned about whether unreasonable
risk is posed by these new chemicals during manufacture and transport in the
U.S.
Options
We have evaluated three alternatives concerning new chemicals manufactured
solely for exporti2/
Option 1; Section 5 Notice.
Option 2; Section 8(a) Notice.
Option 3: No Notice.
Each of these options is evaluated on the basis of costs to EPA and
society.
EPA Costs
The criterion used to assess EPA costs is the amount of information
submitted to EPA under each option.
Both Options 1 and 2 would allow EPA to require supplemental reporting of
information on a new chemical substance. The costs associated with this
possibility would be the same under both options. Option 3 has no
supplementary reporting costs.
The small business exemption under Option 2 proposes to exempt exporters
with total annual sales of less than $1 million from reporting requirements.
It is difficult, however, to estimate precisely how many firms will be
exempted.
Z/These options are described more completely in: "Economic Impact
Analysis of Proposed Section 5 Notice Requirements," Part II - Issue Papers,
Chapter 5-Exporters (Washington, D.C.: ICF Incorporated, September 1980).
-------
D-15
Option 3 is ranked as the least costly to EPA. Under Option 3, exporters
are exempt from any reporting requirements for new chemicals manufactured
solely for export. Thus, EPA will not incur any costs of reviewing notices
for exporters. These rankings are summarized in Exhibit D-5.
Society Costs
The criterion used to assess society costs is the degree to which society
is exposed to unreasonable risks under the various options. The more
(quantity and quality) information available to EPA, the more knowledgeable a
decision the Agency can make. In this way, society has the most protection
from unreasonable risks.
Under the option of no reporting—Option 3—unreasonable risks posed to
society may be higher because EPA will have no reported information about new
chemicals manufactured solely for export.
Option 1 entails the widest coverage of exporters, differing from Option 2
in that Option 2 includes a small business exemption. To the extent that
action under section 5 is more likely than action under other sections of
TSCA, Option 1 offers EPA more authority to take action against new chemicals
once notice has been submitted. Thus, Option 1 is ranked as the least costly
alternative to society. For a summary of these rankings, see Exhibit D-5.
CUSTOMER CONTACT
To fully evaluate the environmental risks associated with the introduction
of any new chemical, EPA may find it necessary to obtain information from
those who purchase the chemical. The types of information sought would
include worker exposure during processing, environmental release, disposal,
and use by the general population.
Options
In this paper, we will assess the relative costs to EPA and to society of
these options for customer contact:
Option 1; January 10 Proposal—Submitter must contact all customers
in writing.
Option 2; October 16 Proposal—Submitter must contact all customers
in writing who have made a firm commitment to purchase the
substance.
Option 3; Provision of Estimated Number of Customers.
Option 4; Provision of Potential Customer List.
Option 5; Elimination of All Customer Contact Provisions.
-------
D-16
EXHIBIT D-5
THE RELATIVE COSTS OF EXPORTER OPTIONS
-^^^ Cost EPA Cost
Option ~^^^ (Administrative)
1. Section 5 Notice 1
2. Section 8 (a) Notice 1
3. No Notice 0
Supplemental Society Cost
Costs (Risks)
1 1
1 2
0 3
Note: "0" indicates no cost, "1" indicates lowest cost, etc.
-------
D-17
A discussion of customer contact in isolation from the supplemental
reporting and confidentiality issues is difficult. Therefore, in this
discussion of customer contact, it is assumed that EPA can obtain supplemental
information if it desires, and that confidentiality provisions sufficient to
protect confidential data from disclosure have been adopted. That changes in
these assumptions could alter the conclusions, however, must be recognized.
EPA Costs
Obtaining a Customer List, in some options, EPA must obtain a customer
list from the producer before contacting customers. From the experience
gained in processing the notices submitted thus far, and from industry
comments, producers appear to be reluctant to identify customers. As a result
of this reluctance, EPA may have to spend time negotiating with producers.
Because EPA anticipates that the majority of its attempts to contact customers
will be associated with those substances which it chose to review in detail
(about 20 percent of the total), and because the amount of time devoted to
interaction with the producer should not be great, the cost of obtaining
customer lists should be minimal.
There are no added costs to EPA under Options 1 and 4, because the
customer list is already provided. Under Options 2 and 3, the costs of
persuading chemical producers to provide customer lists should be small, since
customer lists will be requested only about 20 to 25 percent of the time.
Contacting the Customers. Under some options, EPA will incur the direct
cost of contacting customers. The resources that EPA expends in contacting
customers depends to a certain extent on the degree to which customers
voluntarily provide information. Customer contact is anticipated to occur in
only about 20 percent of all cases and EPA suggests that phone conversations
with the customers should last about 30 to 45 minutes, so this expense should
not be a major one.
The cost of contacting customers under Option 1 will be less than the
costs under the other options at any given level of cooperation on the part of
customers. Under Option 1, EPA will not have to expend resources to collect
information on those customers willing to respond on the initial customer
contact form.
Processing the information. EPA will devote resources to evaluating the
information obtained from the customers.
Under Options 2, 3, and 4, EPA will obtain information from customers in a
maximum of 20 to 25 percent of the cases. Discussions with Agency personnel
suggest that the cost of processing this information will be relatively small.
-------
D-18
Under Option 1, EPA will receive information from customers in all cases
(depending, of course, on the degree to which customers provide information
voluntarily). This suggests that processing costs under Option 1 would be
four to five times greater than under the other options. EPA could disregard
the information gathered for those chemicals which do not undergo detailed
review, thereby reducing its costs under Option 1 to the same level as for the
other options.
The Option-Cost matrix in Exhibit D-6 displays the relative costs to EPA
of the Customer Contact Options.
Society Costs
In accordance with each option, coverage and the type of customer
information submitted are used to rank the five options in terms of unreasona-
ble health and safety risks. In theory, the more information that is avail-
able, the better the decision EPA and the public will make; and the less
likely it is that dangerous chemicals will be mistakenly distributed in
commerce, and that safe chemicals will be mistakenly restricted. Therefore,
the inclusion of as broad a range of information from customers as possible
will increase the probability that a good decision is made.
Assuming that under Options 2, 3, and 4, the Agency would contact custom-
ers in all cases in which it felt their information to be necessary, these
three options are ranked equivalently. Within the parameters of Option 1,
customer information would be obtained in all cases, resulting in more infor-
mation available to EPA. Presumably, this would enhance the possibility of a
good decision by EPA. It should be noted that if EPA contacts customers in
accordance with Options 2, 3, and 4 whenever necessary and is able to obtain
the same information submitted by use of Option 1, then the useful information
available in accordance with Options 1, 2, 3, and 4 would be the same. Option
5, by which use EPA is prevented from obtaining any information through cus-
tomer contact provisions of section 5 of TSCA, would incur the highest society
cost in terms of the quality of a decision.
It is possible to devote so many resources to evaluating the safety of new
chemicals, delaying their introduction up to 180 days, and consuming society's
scarce evaluative resources, that more harm than good is done to society by
the evaluation process. Therefore, the cost to society in terms of foregone
chemicals will be assessed. In terms of foregone chemicals, the use of Option
1 results in the highest costs. Option 1 involves processing customer
information in all cases, and possibly delaying chemicals needlessly as a
result. Options 2, 3, and 4 may involve some delay in processing chemicals
through the customer contact procedure; however, the vast majority of cases in
which EPA would request information from customers by use of Options 2, 3, and
4 would involve chemicals it chose to review in detail. Option 5 will result
in no foregone chemicals.
The only difference between Options 2 and 3 is that in accordance with
Option 3, the submitter is required to indicate on the notice form the
estimated number of customers for all categories of use, including unknown
-------
D-19
EXHIBIT D-6
THE RELATIVE COSTS OF CUSTOMER CONTACT OPTIONS
^X. Cost
Option ^s.
^X^
1. January 10 Proposal
(Contact All Customers)
EPA Cost 1
(Obtaining
List)
0
EPA Cost 2
(Contacting
Customers)
1
EPA Cost 3
(Processing
Information)
2
Society Cost
(Quality of
Decision)
1
Society Cost
(Foregone
Chemicals)
3
2. October 16 Proposal
(Contact Customers With
Firm Orders)
3.
4.
5.
Provision of Estimated 1 2
Number of Customers
Provision of Potential 0 2
Customer List
Elimination of All 2 3
Customer Contact
Provisions
122
122
031
Note: "O" indicates no cost, "1" indicates lowest cost, etc.
-------
D-20
categories; b.y use of Option 2, the number of customers for unknown uses would
be estimated only. Thus, the difference in cost between Options 2 and 3 is
trivial. As a result of this small difference, these options are ranked
equivalently according to EPA and society costs. Exhibit D-6 summarizes the
relative costs of the Customer Contact Options.
INSUFFICIENT SUBMISSIONS
Section 5 of TSCA requires that a notice contain information and meet
standards as prescribed in the Act. If submissions fail to meet these
standards, EPA informs the submitter of the deficiencies and establishes the
framework within which the submitter can correct the deficiencies.
Options
We have analyzed these four policy options associated with insufficient
submissions:^/
Option 1; January 10 Proposal—major deficiencies result
in notice period not starting.
Option 2; modified January 10 Proposal—major deficiencies
result in notice period not starting but sub-
mitter can appeal agency finding.
Option 3; deletion of general insufficient notice
provisions from rules but agency can on
case-by-case base determine individual
submission insufficient.
Option 4; no insufficient notice provisions.
These options will be analyzed according to the relative costs they impose
on EPA and on society.
The costs of Options 3 and 4 are difficult to assess, because they are
heavily dependent on the behavior of the industry. In addition, the lack of
rules under Option 3 and Option 4 means the industry will face considerable
uncertainty until EPA establishes a pattern of behavior. Thus, the costs of
Options 3 and 4 depend on EPA's behavior, as well as on industry's behavior.
The uncertainty associated with these options makes the estimation of their
relative costs difficult.
^./Complete descriptions of these four options can be found in:
"Economic Impact Analysis of Proposed Section 5 Notice Requirements," Part II
- Issue Papers, Chapter 7-Insufficient Submissions (Washington, D.C.: ICF
Incorporated, September 1980).
-------
D-21
EPA Costs
Costs of Persuading Submitters to Voluntarily Correct Deficiencies. In
accordance with Options 1, 2, and 3, EPA can refuse to accept a section 5
notice (the criteria vary among the options). By use of Options 2 and 3, EPA
can only persuade the submitter to correct deficiencies voluntarily. This
persuasion entails an expenditure of resources on the part of EPA. According
to Option 1, the correction can be mandatory or the Agency may refuse to
review the submission immediately.
Because the level of authority varies according to Options, EPA would
incur costs trying to persuade firms to correct their submissions by use of
Options 2 and 3, whereas by using Option 1, there would be no persuasion
costs. If the relatively high level of voluntary compliance experienced thus
far indicates the compliance level that will be achieved after final
regulations are promulgated, persuasion costs using Options 2 and 3 are likely
to be low.
Under Option 4, EPA may also use persuasion to convince the submitter to
correct deficiencies. If Option 4 is used, the resources devoted to
persuasion are likely to be greater than if Options 2 and 3 are used. By
using Option 4, the agency must persuade the submitter to correct all
deficiencies, whereas under Options 2 and 3 EPA must persuade the submitter to
correct only minor ones.
Cost of Hearing Appeals. Under Option 2, submitters may appeal the
Agency's determination that a submission is insufficient. EPA must expend
resources to review the appeals. If the high degree of voluntary cooperation
on notices submitted thus far indicates the degree that will be achieved after
final regulations are promulgated, it is not likely that the frequency of
appeals will be great. The possibility of an adjudicatory hearing in response
to Option 2 use is an additional cost to consider.
Cost of Section 5(e) Action and Supplemental Reporting. Using Option 4,
the remedies available to EPA would be to request information according to the
proposed supplemental reporting provisions or by taking action in accordance
with section 5(e) of the Act. The costs of supplemental reporting are
discussed in the next section of this Appendix. In addition, the next section
includes a description of the significant costs to EPA imposed by taking
action under section 5(e). The provision in Option 1 for suspension of the
notice period until the deficiency is corrected would be less costly to EPA
than taking action in accordance with 5(e). Exhibit D-7 displays the relative
costs to EPA of the four options.
-------
D-22
Society Costs
The intent of the section 5 notice requirements is that EPA, with the
assistance of the public should be able to judge whether or not a chemical is
potentially harmful. The more time that is available to society to evaluate
the information provided by the notice process, the better that judgment
should be. TSCA specifies a 90-day period plus a possible 90-day extension as
the time allowed for evaluation. The options will be evaluated in terms of
how the notice period is altered under each of the options.
According to Options 1, 2, and 3, the initiation of the review period can
be delayed if the submission is found to have major deficiencies (the
initiation of the review period can be delayed for minor deficiencies as well
using Option 1).
One reason for delaying the initiation of the review period is that the
submission contains deficiencies so severe that the chemical substance cannot
properly be evaluated. If this were so, and if the notice period clock were
not set back to the beginning pending receipt of corrections, those evaluating
the notice (EPA and society) might not have the time necessary to do a proper
evaluation.
In accordance with Option 1, the clock is suspended for minor deficien-
cies. Therefore, some delays in the initiation of the review period using
Option 1 may not be necessary to provide EPA and society with the full TSCA
mandated review period. Under Option 2, the clock is not stopped for any
reason. Thus, there will be no delays in the initiation of the review period
for minor deficiencies.
By using Option 3, delays in the initiation of the review period would
result if EPA determines that a notice is incomplete. No well-designed
guidelines would exist for doing so. Therefore, no a priori assessment can be
made as to whether delays occurring under Option 3 would provide society with
the full TSCA-mandated review period.
Similarly, Option 4 presents uncertainties to society, and costs for this
option are difficult to assess. The relative costs of insufficient
submissions options are reviewed in Exhibit D-7.
SUPPLEMENTAL REPORTING
In evaluating the danger a new chemical substance poses for public health,
EPA may require information other than that contained in the initial notice
submission.
-------
D-23
EXHIBIT D-7
THE RELATIVE COSTS OF INSUFFICIENT SUBMISSIONS OPTIONS
Cost
EPA Cost 1
EPA Cost 2
Option
Persuasion Hearing Appeals
EPA Cost 3
Section 5(e)
Society Cost
Review period
1. Major Deficiencies 0
Result in Notice
period Not Start-
ing
2. Same as 1
with Appeal
3. Case-by-Case
4. No Provisions
Note: "0" indicates no cost, "1" indicates lowest cost, etc.
-------
D-24
Options
We have analyzed three policy options for supplemental reporting. They
are:V
• Option 1; January 10 proposal— supplemental data can be
requested;
• Option 2; October 16 proposal—specific supplemental data can
be requested; and
• Option 3; No supplemental reporting rule.
Each of these options will be evaluated according to the costs imposed on
EPA and on society.
EPA Costs
The costs of the supplemental reporting provisions to EPA are dependent on
the behavior of the chemical industry in response to the regulation. The
greater the voluntary cooperation on the part of industry, the less EPA will
have to use formal procedures which consume more of the agency's resources.
Obtaining the Information; Because firms cannot be required to provide
information by use of Option 3, EPA may make some effort to persuade firms to
provide the desired information. Initial requests would be made for voluntary
submissions. If the information is not provided voluntarily, EPA can initiate
a section 5(e) action if the criteria for such an action are met.
Taking action under section 5(e), however, requires a large commitment of
resources by the Agency.iP/ In order to take action under section 5(e), the
Administrator must find that the information available is insufficient to
permit a reasoned evaluation of the health and environmental effects of a
chemical substance, and either:
(1) that the manufacture, processing, distribution in
commerce, use, or disposal of the chemical substance
may present an unreasonable risk of injury to health
or the environment, or
I/Thorough descriptions of these three options can be found in:
"Economic Impact Analysis of Proposed Section 5 Notice Requirements," Part II
- Issue Papers, Chapter 6-Supplemental Reporting {Washington, D.C.: ICF
Incorporated, September 1980).
iP_/Conversations with EPA staff.
-------
D-25
(2) that the chemical substance is or will be produced in
substantial quantities, and the substance enters, or
may be reasonably anticipated to enter the environment
in substantial quantities, or there is or may be
significant or substantial human exposure to the
substance.
The burden of establishing these findings imposes significant costs
on the Agency.
Other Potential Costs. Under Options 1 and 2, EPA will still make an
initial request for voluntary submission. If the needed information is not
provided voluntarily, EPA will be able to pursue paths other than initiating a
section 5(e) action, thereby incurring different costs. These different costs
include:
• Preparing Justifications for Requests for Supplemental
Information.
By using Options 1 and 2, EPA must make certain findings before
it can request supplemental information; additionally, EPA must
send a statement of the findings, along with a detailed
description of the information, to the person from whom the
information has been requested.
• The Appeal Process
By using Option 2, EPA must review appeals to requests for
supplemental information.
• Cost of Processing Information
EPA must expend resources in processing and evaluating the
information received. These costs presumably are proportional
to the amount of information received.
Analysis of EPA Costs; The costs to EPA for each information request
using Option 3 would be somewhat greater than those costs using Option 1.
When using Option 2, EPA would have to prepare a statement of findings that
would be sent along with each request for supplemental information, and would
have to hear appeals to its requests for information; neither of these are
necessary under Option 1. If the appeal process resulted in an adjudicatory
procedure, the costs to EPA of Option 2 in relation to Option 1 would increase
significantly.
It should be noted that the additional cost of preparing a statement of
findings using Option 2 should reduce the number of requests for information
made by EPA. [Information that has a marginal benefit just equal to or less
than the marginal cost of gathering and processing using Option 2 should not
be sought, thereby decreasing total processing costs.] Whether this will be
enough to offset the increased cost per request is unclear.
-------
D-26
Using Option 3, there are no costs of appeals or providing justification,
but EPA incurs the cost of persuading firms to supply information that is not
supplied at the first request. The amount of information provided EPA by use
of Option 3 will be less than when Options 1 and 2 are used, and the cost of
processing and evaluating it will be less. Using Option 3, EPA may also incur
costs of section 5(e) actions that the Agency would not incur using Options 1
and 2.
Exhibit D-8 summarizes the relative costs to EPA of the supplemental
reporting options.
Society Costs
Quality of Notice Decision. From the standpoint of society, the least
costly option is the one which minimizes:
1) the probability of a toxic substance entering
commerce, and
2) the number of non-dangerous foregone chemicals.
For this analysis, the costs to society are based on the quality of a notice
decision; the quality of a notice decision presumably is proportional to the
amount of information available to EPA. (The more information that is
available to EPA, the better the decision will be.)
Analysis of Society Costs. There will be slightly less information
available by using Option 2 than by using Option 1, both because EPA will
request less information and the industry may win some appeals. Option 2 will
result in EPA making decisions based on less information and lower quality
information than that available using Option 1. Thus, Option 2 is ranked
higher in costs to society than Option 1.
The differences between the use of Options 1 and 2 and the use of Option 3
are dependent on the degree to which industry voluntarily complies with
requests for supplemental information. At one extreme, if firms comply with
every request for information, there is no difference in costs among the
options. At the other extreme, if the firms refuse all voluntary requests,
the cost differences are maximized. Industry spokesmen assert that the
information will be voluntarily provided in the majority of cases.1JL/ If
the number of recalcitrant firms is small, the costs of Options 1 and 2 should
be low, and the difference between these two and Option 3 should be small.
jj/Comments of the Chemical Manufacturers Association on EPA's
Reproposal of Forms and Rules for the Submission of Premanufacture Notices
Under Section 5 of TSCA, November 30, 1979, pp. 36-37.
-------
D-27
EXHIBIT D-8
THE RELATIVE (POSTS OF SUPPLEMENTAL REPORTING OPTIONS
^^ Cost EPA Cost 1
^^v^^^ Preparing a
Option ^Vs-->. Justification
1. January 10 proposal. 2
Supplemental data
can be requested
EPA Cost 2
Obtaining
Information
1
EPA Cost 3
Hearing
Appeals
1
Society Cost
Quality of
Notice
Decision
1
2. October 16 proposal.
Specific supplemental
data can be requested
3. No rule
Note: The option with the highest number (i.e., 3) has the highest relative
cost. Where there are no significant differences in the costs imposed
by the options, they are equally marked.
Source: EPA guidance; industry and public interest group proposal comments.
-------
D-28
Under Option 3, uncertainty exists over whether or not vital information
will be made available by industry. As a result, it is more likely that EPA
will lack the necessary information to make a good decision under Option 3
than under the other two options.
As illustrated in Exhibit D-8, Option 3 represents the greatest society
costs, and Options 2 and 1 represent successively lower society costs.
THE CONFIDENTIALITY OPTIONS
The issue of which notice information is confidential and how it should be
handled is addressed by a set of options specified by EPA. There are
competing interests affected by the treatment of confidential business
information. On the one hand, industry's interest in maintaining the
confidentiality of business and trade secrets is based on the competitive
advantage it derives from its work. On the other hand, society has a clear
interest in obtaining information about chemicals so that potential harmful
exposures and misuses can be avoided.
Subissues
EPA has specified policy options for the confidentiality issue according
to subissues:i2/
• Methods of Assertion and Substantiation
• Timing for Substantiation;
• Generic Masking Options; and
• Options for Disclosing Chemical Identity (ID) as Part
of a Health and Safety Study.
These subissues will be analyzed separately in terms of the costs they impose
on EPA and on society. The costs attributed to the various options were
derived from industry comments on the proposal and reproposal, and ADL
^./Thorough descriptions of these options are found in "Economic Impact
Analysis of Proposed Section 5 Notice Requirements," Part II - Issue Papers,
Chapter 1 - Confidentiality (Washington, D.C.: ICF Incorporated, September
1980).
-------
D-29
estimates derived for the reproposal.13/
EPA Costs. There is one EPA cost considered in this analysis:
administrative costs. EPA must process notices, decide about the adequacy of
confidential substantiation, consult with submitters, and so on. All of these
activities consume the time of EPA staff, and the more activities there are,
the greater the administrative cost to EPA.
Society Costs. The TSCA confidentiality policy entails costs for society
as well as for EPA. TSCA's primary purpose is to strengthen the government's
ability to protect the public from the hazards of toxic chemicals.il/ This
purpose is achieved, in part, by permitting the public to participate in the
review of a new chemical's effect on humans and the environment. It is
necessary for the public to have information about the new chemical if
meaningful public participation is to occur. The confidentiality policy,
however, may block the public's access to information that would be helpful,
or even crucial, to meaningful review of a new chemical. Insofar as the
confidentiality policy has this effect, it imposes a cost or leads to costs
that the public otherwise would have avoided. The costs should not be
considered insignificant, because there is at stake the possibility of perma-
nent and disabling injury to humans and long-term degradation of the environ-
ment. A labor union may be unable to warn its members of precautions that
should be taken in dealing with a certain chemical, a citizens' group may be
handicapped in learning about and preventing a particular use of some chemical
which seriously damages the environment. Although many of the costs imposed
are types which traditionally have been difficult or impossible to evaluate in
monetary terms, the importance of these costs should not be
jil/Cominents of the Manufacturing Chemists Association on EPA Proposed
Regulations for Premanufacture Notification Under Section 5 of TSCA
(hereafter: MCA proposal comments), March 26, 1979; Comments of the Chemical
Manufacturers Association on EPA's Reproposal of Forms and Rules for the
Submission of Premanufacture Notices Under Section 5 of the Toxic Substances
Control Act (hereafter: CMA reproposal comments), November 30, 1979; EPA's
Interim Policy (44 Federal Register 28564-28572); EPA's Reproposal (44 Federal
Register 59764-597-83); and Arthur D. Little, Inc., Estimated Costs for
Preparation and Submission of Reproposed Premanufacture Notice Form, September
1979 (hereafter: A.D. Little Study).
ii/TSCA, Section 2(b)(2) indicates: "It is the policy of the United
States that adequate authority should exist to regulate chemical substances
and mixtures which present an unreasonable risk of injury to health or the
environment, and to take action with respect to chemical substances and
mixtures which are imminent hazards."
-------
D-30
underestimated. Finally, public costs can be imposed on society in
general or on.particular institutions representing public interests
(for example, labor unions or public interest groups). Three costs
to the public are considered in the analysis.
• Cost 1; Probability of being denied access to infor-
mation to which the public is entitled. A certain
amount of information is necessary before an adequate
public review can be made of the harm, .if any, posed
by a new chemical substance. This threshold may not
always be met because Congressional policy' generally
forbids public disclosure of Confidential Business
Information (CBI), and CBI may include information
important to review of a chemical's toxicity. This
does not mean that the public is left unprotected.
EPA receives and evaluates all relevant information
including CBI, and has the authority to take
appropriate action to safeguard human health and the
environment. Nevertheless, there also may be
information to which the public ordinarily would have
access, but is denied through some oversight or mis-
judgment. There seems to be at least one contributing
factor to this cost: the rigor of confidentiality re-
views may tend to decrease as EPA's limited resources
are occupied with increased administrative duties.
• Cost 2; Time delay. The timing of public disclosure
of information can often be as important as the sub-
stance of the disclosure. If information is-not re-
ceived in time to avoid exposures (human or environ-
mental) to harmful substances, then when disclosure
does occur it is of limited use. If information is-
revealed to the public during the 90 day review
period, the public can participate in the review
process. Thus, failure to obtain timely disclosure is-
a significant cost to" society. This cost also may
have a synergistic relation to EPA's administrative
burden.
• Cost 3: Out-of-pocket expenses. This cost would
normally be incurred by those institutions which
represent the public interest. For example, a public
interest group may institute legal action to compel
EPA to disclose information. Such action may require
substantial expenditures. Any option selected by EPA
which makes it more difficult for the public to obtain
information will increase the out-of-pocket costs for'
society, it should be recognized that if out-of-
pocket expenses become too great, it may boost the
extent of the other two public costs as well.
-------
D-31
Methods of Assertion and Substantiation
The options associated with this subissue include:
• Option 1; Assert item by item; substantiate by
responding to a series of questions.
• Option 2: Assert by categories and linkages; substan-
tiate by making a general certification (which would
cover manufacturer's ID),!§/ answering a series of
questions for specific chemical ID and "other"
information, and answering no more than two questions
each for use data, process, mixture, and production
information.iZ/
EPA Costs. Neither option involves clearly lower relative costs
for EPA's administrative efforts.j-g7 The same kind of substantive
material is required to be submitted by both options; however,
Option 2 involves less duplication in confidentiality claims, and
requires submitters to focus on the key elements in EPA's
determination of whether an item should be accorded the confidential
status claimed by the submitter. This facilitates EPA's review
process. On the other hand, the industry has intimated that the
linkage system will require the submitter to provide considerably
greater amounts of material to EPA than would be the case using
Option 1. If so, EPA would have to review this greater quantity of
material, thereby increasing its relative administrataive cost.
Neither option, therefore, can be said to present a clearly lower
cost alternative in terms of EPA administrative expenses.
2J>/For a complete description of the subissues and policy options, see
"Economic Impact Analysis of Proposed Section 5 Notice Requirements," Part II
- Issue Papers, Chapter 1-Confidentiality (Washington, D.C.: ICF
Incorporated, September 1980).
i5./This certification also would cover confidentiality claims for use
data, production volume, and process information if neither manufacturers'
identity nor chemical identity have been claimed confidential.
iZ/These questions must be answered only if manufacturer's identity or
chemical identity is claimed and held confidential.
^./Industry claims the reproposal assertion and substantiation require-
ments are more onerous than the proposal (CMA reproposal comments, p. 99),
while EPA claims these requirements avert duplication (Reproposal, 44 Federal
Register 59774).
-------
D-32
Society Costs.
• Cost 1; Option 2 lowers the cost to society by reduc-
ing the probability that information to which the
public is entitled will be withheld. Option 2
requires the submitter to make clear the association
between an item of information claimed confidential
and the category under which it is claimed
confidential. Option 1, on the other hand, involves
only a general substantiation of a category of
confidentiality, and does not require the submitter to
focus on the link between a piece of information and
the alleged harm its disclosure would bring. Option
2, therefore, minimizes the risk that the public will
be denied access to information that was unnecessarily
included in an assertion of confidentiality.
• Cost 2; Neither option presents a clearly lower
relative cost in terms of reducing delay in the
public's access to information. In general, these two
options affect how the submitter will assert and
substantiate a claim of confidentiality, not the
timing of disclosure.
• Cost 3; Option 2 may involve a lower relative cost
than Option 1 in terms of reducing the public's
out-of-pocket expenditures. Insofar as the second
option reduces the information improperly withheld from the
public (see above), it also reduces the public's cost of trying
to obtain the information which should not have been granted
confidential status.il/
The relative costs of the two options are summarized in Exhibit D-9.
Timing of Substantiation
Perhaps more than the question of how to substantiate, the issue of when
to substantiate presents clearly defined alternatives with certain costs. EPA
presented four options to be considered for analysis.
• Option 1: Requires substantiation only upon request
from EPA.
il/There is not necessarily a one-to-one relationship between this cost
and the cost of improperly withholding information, if information is
withheld, it is less likely that the public will be aware of the relevance of
the information to a confidentiality claim. In such a situation,
out-of-pocket costs may be minimized simply because the public is unaware of
which disclosure it should be seeking.
-------
D-33
EXHIBIT D-9
THE RELATIVE COSTS OF OPTIONS FOR
ASSERTING AND SUBSTANTIATING CONFIDENTIALITY
Cost EPA Cost Society Cost 1 Society cost 2 Society Cost 3
Administrative Withhold Delay Access Out-of-pocket
Option ^^ information to information Costs
1. Check-off
Series
2. Linkage/
Reproposal
Questions
Note: A "1" indicates the least costly alternative. Where there are no clear
differences in the costs imposed by the options, they are equally
ranked.
-------
D-34
• Option 2; Requires substantiation when the notice is submitted
for chemical ID, and health and safety study data;
substantiation of all other information upon request.
• Option 3; Requires substantiation of section 5{d)(2) notice
information when notice is submitted; substantiation of all
other information upon request.
• Option 4; Requires substantiation of all confidentiality claims
when notice is submitted.
EPA Costs. The administrative costs to EPA are minimized by Option 1 and
maximized by Option 4.127It is certain that Option 1 initially involves a
lower administrative burden than any of the other options, simply because EPA
would have less information to review. On the other hand, if there were many
Freedom of Information Act (FOIA) requests or other reasons for requiring
subsequent substantiation, EPA would experience efficiency losses in reviewing
the information in a piecemeal fashion. The efficiency difference between
options, however, is not known. Therefore, the point at which it becomes less
efficient to provide piecemeal substantiation, as opposed to substantiation
with the notice, is unclear. Whether or not that point is crossed depends
upon the amount of subsequent substantiation that is required.li/ Thus,
based on the limits of present knowledge, Option 1 represents the smallest
administrative burden for EPA, and Options 2, 3, and 4 represent successively
greater administrative burdens.
l£/The industry has suggested that EPA's attitude toward substantiation
at the time of submission of required material has not been the Agency's
historical stance. (See CMA reproposal comments, p. 118, footnote 3.) EPA's
estimate that administrative savings will result if substantiation is submit-
ted with the notice is based on an assumed high percentage of FOIA requests
for submitted notices' confidential material (44 Federal Register 59775).
21/For example, if there is one item of information each under manufac-
turer's identity, chemical identity, and use data that is claimed confiden-
tial, initial administrative costs obviously are lower if EPA does not have to
review substantiations for the three claims when the notice is submitted. In
the long term, however, substantiation may become necessary for some or all of
the claims. Because there is some minimum administrative effort required to
review the substantiation, whether submitted with the notice or later, it is
not necessarily more costly for EPA to review one substantiation not submitted
with the notice than to review three substantiations with a notice. A similar
conclusion is reached about two subsequent substantiations compared to three
substantiations submitted with the notice. The only definitive statement that
can be made is that, using Cption 1, EPA's administrative costs will be higher
if EPA must subsequently review all the substantiations that would have been
submitted with the notice using one of the other three options. The above
example is simplistic, but it illustrates the uncertainty of what long-term
administrative advantages actually will accrue to EPA if some or all substan-
tiation is required to be submitted with the notice.
-------
D-35
Society Costs.
Cost 1; Relative to the other three options. Option 1
raises the cost to the public by increasing the proba-
bility that information to which the public is enti-
tled will be withheld. If the submitter can delay in
providing substantiation (which is permitted in vary-
ing degrees by Options 1, 2, and 3), information
claimed confidential (but not necessarily to be accor-
ded confidential status when appropriate substantia-
tions are reviewed) will be withheld from the public
until a request (at the instigation of EPA or the
public) for substantiation is made. This failure to
provide substantiation when the notice is submitted
may result in public ignorance of information to which
the public could have gained access. On the other
hand, it may be that substantiation submitted with the
notice will no longer be accurate when a subsequent
request for the protected information is made. If EPA
relies on the outdated substantiation to deny disclo-
sure of the information, the public would unnecessar-
ily have been denied access to information.^/
Thus, if substantiation upon request (which is present
in varying degrees in Options 1, 2, and 3) had been
utilized instead of substantiation with the notice,
information would not have been withheld from the
public. This latter situation which postulates an
out-of-date substantiation seems less likely and more
easily remedied than the former situation postulating
public unawareness. Similarly, the level of cost
seems higher for the first situation, because it may
unnecessarily prevent effective public participation
in the review notice, while the second situation
suggests denial of information at some later date when
effective review participation no longer is possible.
Therefore, Option 1 represents the greatest cost to
the public in terms of probable withholding of in-
formation to which the public is entitled; in
general,2_3/ the other three options, 2, 3, and 4,
are successively less costly.
.22/CMA reproposal comments, pp. 123-124.
21/There may be particular situations where one specific item of infor-
mation is most important—for example, production volume. Because only Option
4 requires that substantiation of this information be submitted with the no-
tice, Options 1, 2, and 3 are equally costly (in this instance) in terms of
withholding public information. As a general rule, however, the more confi-
dential claims that are not substantiated with the notice, the greater is the
public cost in terms of withheld information.
-------
D-36
Cost 2; Option 1 involves the highest relative costs
in terms of delay in providing the public with infor-
mation, and Option 4 involves the lowest relative
" — -' • • - •' ii • • • JL .. I ' " " ™'
costs in these terms..£i/ Substantiation submitted
with the notice enables EPA to determine the validity
of confidentiality claims early in the review
process. This, in turn, enables EPA to disclose in-
formation at an earlier time than if the substantia-
tion could be obtained from the submitter only on
request. Similarly, substantiation submitted with the
notice enables EPA to give more timely notice to re-
questors of confidential information that the informa-
tion is unavailable. These requestors can then take
whatever action is necessary from their standpoint
without experiencing the delays inherent in the sub-
stantiation-by-request situation. Therefore, as more
confidentiality claims are substantiated with the
notice, this particular cost is minimized. Option 4
represents the lowest relative cost in terms of delay
in release of public information, and Options 3, 2,
and 1 represent successively higher relative costs.
Cost 3; Option 4 involves the lowest relative cost in
terms of public out-of-pocket expenses, and Option 1
represents the highest relative cost in these
terms.jjyOne of the ways that the public may
obtain information not disclosed by EPA is by some
group acting in the public interest to institute FOIA
proceedings. This involves out-of-pocket expenditures
on the part of the public interest group. When more
information is disclosed as a result of EPA's review
of the notice, there is less necessity for these FOIA
proceedings and the accompanying out-of-pocket costs
to the public.
^A/Industry suggested that delay would not be serious if FOIA request
frequency were low (CMA reproposal comments, p. 133). Nevertheless, all
information claimed confidential, but subsequently determined by EPA not to be
confidential, should have been available to the public when it was submitted.
Failure to substantiate with the notice submission delays EPA's determination
of confidentiality and delays the public's access to the information (see
Federal Register 59774-59775). Of course, this latter point assumes there
will be some information erroneously claimed confidential.
21/See 44 Federal Register 59775. Substantiation submitted with the
notice makes it more likely that information not entitled to confidential
treatment will "be made available to the public during the 90-day notice
period." This could save the public the expense of unnecessary FOIA requests.
-------
D-37
Option 4 enables EPA to make the broadest determination of
confidentiality, thereby facilitating more disclosure of
non-confidential material than the other three options and
reducing the level of public out-of-pocket costs. The other
three Options, 3, 2, and 1 represent successively lesser amounts
of non-confidential information disclosure and correspondingly
higher public out-of-pocket costs.
The relative costs of the form options are summarized in Exhibit D-10.
Generic Masking Options
In some instances when a confidentiality claim is asserted, the submitter
would be required to substitute a generic name or information to replace the
confidential information. Three generic options have been considered and are
described below..?JL/
• Option 1; January 10 Proposal—Requires generic name if
chemical identity or use data is claimed confidential.
• Option 2; October 16 Reproposal—Requires considerable generic
information if confidentiality is claimed for chemical identity,
use data, physical and chemical properties, or manufacturer's
identity.
• Option 3; No Generic Information.
EPA Costs. Option 3 imposes the lowest relative administrative costs for
EPA.JLZ/EPA1s resources must be expended in evaluating the submitter's
generic information—whether at the pre-notice consultation stage or when the
notice is submitted. The less information that EPA has to review, or the less
time EPA has to devote to negotiating generic information, the lower the cost
of an option to EPA in terms of administrative expenditures. Option 3 would
require the submission of no generic information and, consequently, represents
the lowest relative administrative costs for EPA. Using Option 2, EPA is
required to evaluate more specific information than under Option 1, and also
may have to compose the generic description for three of the four generic
items. Option 2, therefore, is relatively more costly to EPA in terms of EPA
administrative resources.
Society Costs.
• Cost 1; Option 2 represents the lowest relative
public costs in terms of the amount of information
which will be withheld. Because of the requirements
of confidentiality, the public potentially will be
denied access to information which is useful in
If/These options are described in detail in "Economic Impact Analysis of
Proposed Section 5 Notice Requirements," Part II - Issue Papers, Chapter 1 -
Confidentiality (Washington, D.C.: ICF Incorporated, September 1980).
27/See CMA reproposal comments, pp. 104, 151, 155-157.
-------
D-38
EXHIBIT D-10
THE RELATIVE COSTS OF OPTIONS FOR TIMING SUBSTANTIATION
CostEPA CostSociety Cost 1Society Cost 2Society cost 3
Administrative Withhold Delay Access Out-of-pocket
Option ^^-v^ information to information Costs
1. Nothing 14 44
with Notice
2. Chemical ID and
Health and
Safety Data
with the Notice,
the Rest on
Request
3.
4.
5(d)(2) 3 2
Information
with the
Notice, the
Rest on
Request
Everything 4 1
with
the Notice
2 2
1 1
Note: A "I" indicates the least costly alternative. Where there are no clear
differences in the costs imposed by the options, they are equally
ranked.
Source: CMA reproposal comments; EPA reproposal comments; and A.D. Little
Study.
-------
D-39
reviewing the effects of a new chemical on humans and on the
environment. Although Options 1 and 2 seek to redress this
inherent problem of the confidentiality issue, Option 2 goes
further in providing information for the public assessment of
the new chemical. Instead of being limited to generic
descriptions of a chemical's identity and its categories of use,
Option 2 provides additional generic information about the
identity of the manufacturer and the nature of the chemical's
physical and chemical properties. This additional information
enhances the public review of a chemical's toxicity. Option 3,
which provides no generic information, clearly represents the
most costly of the three options in terms of information
withheld from the public.
• Cost 2; Neither Option 1 nor 2 has a relative advantage in
avoiding the delay in releasing information to the public.jj*/
In accordance with both Options 1 and 2, pre-notice consultation
is encouraged so that generic information release can be
expedited. Option 3, which requires no generic submittals, has
no effect on the delay costs.
• Cost 3; None of the three options has a distinct effect in
lowering out-of-pocket costs for the public. The public cannot
compel submission of generic information. Therefore, regardless
of which option is selected, the public will not incur
option-dependent out-of-pocket costs.
The relative costs of the three options for generic masking are summarized
in Exhibit D-ll.
Disclosing Chemical ID as Part of Health and Safety Study
If a health and safety study is submitted as part of the notice, the
following subissues must be decided:
• timing options: when the chemical ID will be disclosed; and
• inventory options: how the chemical ID should be added to the
inventory.
Timing Options
• Option 1; when the notice is submitted;
• Option 2; when manufacture begins; and
• Option 3; when the new chemical substance is
distributed in commerce.
.M/See the EPA comment regarding chemical identity and generic
information, 44 Federal Register 59776.
-------
D-40
EXHIBIT D-ll
THE RELATIVE COST OF OPTIONS FOR GENERIC MASKING
Cost EPA Cost Society Cost 1 Society Cost 2 Society Cost 3
Administrative Withhold Delay Access Out-of-pocket
Option ^v^ Cost information to information Costs
1. Chemical ID 2 2
and Chemical
use
1 1
Chemical ID,
Chemical Use,
Manufacturer' s
ID, and
Physical/
Chemical
Properties
3. No Generic
Information
Note: Higher numbers indicate higher relative costs. Where there are no
significant differences in the costs imposed by the options, they are
equally ranked.
Source: CMA reproposal comments; MCA proposal comments; EPA reproposal
comments.
-------
D-41
EPA Costs. EPA's costs are not significantly affected by which one of the
three options is selected. There might be some minor administrative
efficiency achieved if dislosure was in conjunction with receipt of the notice
(i.e., Option 1); but disclosure itself is a relatively mechanical act and
would require no additional evaluation, and later disclosure should involve
little efficiency loss.
Society Costs.
• Option 1; The disclosure is earlier than the other
two options, and the public receives the information
for more timely input to its assessment of the
chemical's potential effect on humans and the
environment. Thus, the time delay cost is kept at a
minimum.i£/ Other public costs are not
significantly affected by this set of options.
• Option 2 is between the other two options. Informa-
tion is provided to the public more slowly than Option
1, but more quickly than Option 3. Thus, depending on
the option to which it is compared, Option 2 either
increases or reduces the cost to society in terms of
delayed receipt of information.
* Option 3 imposes the highest costs for the public.
Option 3 protects the chemical identity from disclo-
sure longer than the other two options. This could
result in disclosure more than six months after the
Option 1 disclosure date. Those six months can be a
period during which substantial advances could have
been made in assessing the chemical's impact on humans
and on the environment.
Exhibit D-12 displays the relative costs of timing options.
Inventory Options. The final set of options regards placement of chemical
ID on the inventory.
Option 1; by specific chemical name; and
Option 2; by generic chemical name.
EPA Costs. Neither option involves lower relative costs for EPA's
administrative efforts.
.?_2/The importance of avoiding this time delay is emphasized by EPA.
See, e.g., EPA comments in its Interim Policy announcement 44 Federal Register
28564 et. seq. at pp. 28567-28568, and EPA comments accompanying the repropo-
sal, 44 Federal Register 59775.
-------
D-42
EXHIBIT D-12
THE RELATIVE COSTS OF OPTIONS FOR TIMING THE DISCLOSURE OF
CHEMICAL ID AS PART OF A HEALTH AND SAFETY STUDY
Cost EPA Cost Society cost 1 Society Cost 2 Society Cost 3
Administrative Withhold Delay Access Out-of-pocket
Option ""^-^ Cost information to information Costs
1. Disclose at 1 1 1 1
Time Notice
Submitted
2. Disclose at
Time Manufac-
ture Begins
3. Disclose at
Time Distri-
buted in Com-
merce)
Note: Higher numbers indicate higher relative costs. Where no significant
differences exist in the costs imposed by the options, they are equally
ranked.
Source: MCA (CMA) proposal comments; CMA reproposal comments; EPA comments
accompanying the Interim Policy and the Reproposal.
-------
D-43
Society Costs. The generic option would impose some costs on society
calculated in terms of additional time delay in obtaining the information
(i.e., specific chemical ID would be more readily available to the public as
part of the inventory—an available, separate document) and in terms of the
additional out-of-pocket costs to obtain the information (i.e., it will be
more expensive to obtain the specific chemical ID from the public record than
to obtain a single document—the inventory—with the desired information).
Exhibit D-13 illustrates the increased cost to society of the generic name
option.
SUMMARY
The purpose of this report is to present to EPA a framework from which to
evaluate these issues. The costs to EPA and to society involved with each of
these issues are dependent on the regulatory option chosen, the way EPA
implements the regulations, and the way industry responds to the regulations.
The sensitivity of the costs to the uncertain behavior of both regulators and
the regulated reduces the precision with which costs can be estimated. By
estimating the cost differences among the options rather than absolute cost
levels, we have minimized this problem. Nevertheless, in several parts of
this report it was necessary to make certain assumptions about the behavior of
parties towards the regulation. It should be noted that changes in these
assumptions may affect the estimated cost rankings of the policy options.
-------
D-44
EXHIBIT D-13
THE RELATIVE POSTS OF OPTIONS FOR PLACING CHEMICAL ID
(CONTAINED IN A HEALTH AND SAFETY 'STUDY) ON THE INVENTORY
Cost EPA Cost Society Cost 1 Society cost 2 Society Cost 3
Administrative Withhold Delay Access Out-of-Pocket
Option ^*^ Cost information to information Costs
1. Place on
Inventory by
Specific Name
2. place on
Inventory by
Generic Name
Note: Higher numbers indicate higher relative costs. Options are ranked
equally when no significant cost differences exist.
21759 *D.S. OOVEHNMEHT PRINTING OFFICE: 1980-0-341-085/11605
-------
TECHNICAL REPORT DATA
(Please read fnanictions on the reverse before completing]
. REPORT NO.
EPA-560/12-80-OQ5
3. RECIPIENT'S ACCESSIOONO.
4. TITLE AND SUBTITLE
Economic Impact Analysis of Proposed
Section 5 Notice Requirements
5. REPORT DATE
September 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
Robert Dresser, James Edwards, Joseph Kirk,
Stuart Fribush
3. PERFORMING ORGANIZATION REPORT NO.
3. PERFORMING ORGANIZATION NAME AND ADDRESS
ICF Incorporated
1850 K Street, N.W., Suite 950
Washington, D.C. 20006
10. PROGRAM ELEMENT NO.
2LS811
11. CONTRACT/GRANT NO.
68-01-5878
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Pesticides and Toxic Substances
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Proposed Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
EPA Project Officer: Sammy K. Ng
16. ABSTRACT
This report presents the analysis of the economic impact of TSCA section 5 rules on
the chemical industry. The industry will be impacted when it introduces new chemicals.
Of the six distinguishable consequences for the chemical industry, the most important
are the nonquantifiable uncertainty consequences. The more unclear EPA's rationale in
making section 5 notice decisions, the greater are the uncertainties.
There will likely be a short-run drop in the number of new chemicals introduced into
commerce as chemical companies shift their innovation activities into "safe" chemicals.
Current data do not allow a quantitative estimate to be made of the rate of chemical
introductions, or the extent of the reduction caused by the section 5 notice require-
ments; and, even if the data were available, it is doubtful that accurate quantitative
predictions could be made.
Smaller companies will face greater uncertainties and the direct costs will more often
be a factor in company decisions. In the long run, this regulation may cause the chemi-
cal industry to be composed of a fewer number of larger competitors better able to absorb
the direct costs and regulatory uncertainty associated with the requirements.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDEDTERMS
COSATl Field/Croup
TSCA Section 5 Notice Requirements
Economic Impact Analysis
3. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |