ANALYSIS OF TSCA SECTION 8(a)
SMALL MANUFACTURER EXEMPTION
Submitted to
U.S. Environmental Protection Agency
April 1982
ICF INCORPORATED 1850 K Street, Northwest,
Suite 950, Washington, D C 20006
-------
ANALYSIS OF TSCA SECTION 8(a)
SMALL MANUFACTURER EXEMPTION
by
Michael J. Gibbs
Joseph L. Kirk
Kenneth B. Kolsky
Frank Lerman
Contract No. 68-01-6287
Task Order No. 3
Project Officers
Scott Peters
Michael Shapiro
Economics and Technology Division
Office of Toxic Substances
Washington, D.C. 20460
OFFICE OF PESTICIDES AND TOXIC SUBSTANCES
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
ICF INCORPORATED
-------
DISCLAIMER
This report was prepared under contract to an agency of
the United States Government. Neither the United
States Government nor any of its employees,
contractors, subcontractors, or their employees makes
any warranty, expressed or implied, or assumes any
legal liability or responsibility for any third party's
use or the results of such use of any information,
apparatus, product, or process disclosed in this
report, or represents that its use by such third party
would not infringe on privately owned rights.
Publication of the data in this document does not
signify that the contents necessarily reflect the joint
or separate views and policies of each sponsoring
agency. Mention of trade names of commercial products
does not constitute endorsement or recommendation for
use.
- ii -
ICF INCORPORATED
-------
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 provide
support of EPA's development of a small manufacturers definition required for
Section 8(a) of the Toxic Substances Control Act (TSCA).
This report is being released and circulated before publication in the
Federal Register of a proposed rule defining small manufacturers for
reporting purposes under Section 8(a) of TSCA. The study is not an official
EPA publication. It will be considered along with any comments received by
EPA before or during the proposed rulemaking proceedings in establishing the
final rule. Prior to final promulgation of the final small business
definition, the accompanying study shall have standing in any EPA proceeding
or court proceeding only to the extent that it represents the view of the
contractor who performed the study.
- iii -
ICF INCORPORATED
-------
EXECUTIVE SUMMARY
This report presents analyses in support of EPA's development of a small
manufacturers definition required for Section 8(a) of the Toxic Substances
Control Act. EPA provided production data for over 1,400 chemicals and sales
and employment data on the over 3,000 locations at which the chemicals are
manufactured. These data were used to estimate the amount of information that
EPA would receive for Section 8(a) reporting rules under various small
manufacturer definitions.
Four proxies are used to indicate the "amount of information" EPA would
receive:
Number of reports required to be prepared;
Number of locations required to report;
Percent of total production covered by the reporting
requirement; and
Percent of the chemicals for which sufficient
information is received.
How these proxies vary as a function of the parameters of the definition is
described.
A proxy for total cost to industry is also used. Total cost for various
small manufacturer exemptions is indicated by the sum of the number of
locations that have to report and the total number of reports that have to be
prepared. How this proxy for total cost varies with the parameters of the
definition is also described.
The analysis investigates two sets of parameters: (1) the annual
production at a location of the chemical upon which a report is required (in
kilograms) and the annual sales of the parent firm of the location; and (2)
the annual production at a location of the chemical upon which a report is
required (in kilograms) and the total employment at the parent firm of the
location. Additionally, two structures of the exemption are considered; one
for which locations must be below both parameters to be exempted, and the
second for which locations need only be below one parameter to be exempted.
The results of the analysis indicate that an exemption based on annual
sales and production quantity which requires locations to be below both
parameters to be exempt, is preferred to the other exemption structures
- v -
ICF INCORPORATED
-------
examined. Requiring locations to be below both parameters more accurately
targets reductions in reporting costs to small manufacturers. Total
employment was found not to be a good substitute for annual sales as an
indicator of firms' abilities to absorb reporting costs.
The choice of the level of the exemption requires judgments about the
relative value of obtaining different types of information from various
segments of the chemical industry. These valuations must be balanced off
against the cost to industry of providing the information. This cost (in
dollars), will depend on the information requested in the rule. Therefore, an
optimum exemption level has not been defined. Rather, this report identifies
the tradeoffs between information and cost which occur at different exemption
levels. EPA can use this information to identify appropriate exemption levels
for the types of information it expects to require in future Section 8(a)
rules.
- vi -
ICF INCORPORATED
-------
TABLE OF CONTENTS
PAGE
EXECUTIVE SUMMARY V
TABLE OF CONTENTS vii
1. INTRODUCTION 1-1
2. FRAMEWORK FOR ANALYSIS
2.1 Introduction 2-1
2.2 Estimating Benefits 2-1
2.3 Estimating Costs 2-3
3. FUTURE SECTION 8(a) RULES
3.1 Introduction 3"!
3.2 Group Chemical Rules 3-2
3.3 Individual Chemical Rules 3-4
4. DATA ANALYSIS
4.1 Introduction 4-1
4.2 Relationships Between the Benefit and Cost Proxies 4-3
4.3 Missing and Confidential Data 4-8
5. ANALYSIS OF OPTIONS
5.1 Introduction 5-1
5.2 Choice of Number and Type of Parameters 5-2
5.3 One-Level Exemption 5-8
5.4 Multi-Level Exemption 5-14
6. CONCLUSIONS 6-1
Appendix A: Description of Data Base A-l
Appendix B: Data Analyses B-l
Appendix C: Analysis of a Hypothetical Reporting Rule C-l
- vii -
ICF INCORPORATED
-------
1. INTRODUCTION
The Toxic Substances Control Act (TSCA) authorizes EPA to develop or
obtain data on new and existing chemical substances. Section 8 of the Act
empowers the Administrator to require manufacturers, processors, and certain
other persons to collect, retain, and/or report certain information. The Act
also requires that EPA exempt small manufacturers from Section 8(a) reporting
and recordkeeping rules. This report analyzes how small manufacturers should
be defined for the purpose of this Section 8(a) exemption.
Ideally, the small manufacturer exemption should balance the benefits and
costs of obtaining information. To attempt this balance, we developed a
framework for evaluating the "benefits" and "costs" of various levels of small
manufacturer exemptions.
Balancing of benefits and costs must take place in the context of
expected future Section 8(a) rules. The marginal value of obtaining
incremental amounts of information may vary with the type of information
reported and the purpose for which it is being collected. One of the proxies
for "benefits" may be more appropriately applied to evaluate some types of
future rules than other proxies. Therefore, we examined expected future
Section 8(a) rules to assess which of the proxies are most appropriate to use
to evaluate the benefit of the information expected to be received by EPA
under the future rules.
Having considered the expected Section 8(a) rules in the context of the
framework, we evaluate the options for the small manufacturer exemption. This
evaluation is aided by data analyses performed on 19,485 reports of production
of 1459 chemicals by 3044 manufacturing locations. The data analyses describe
how the proxies vary under different exemptions. The examination of exemption
options included:
The number and type of parameters for the exemption;
The structure of the exemption, one-level or multi-
level ; and
The level of the exemption itself, i.e., at what
values should the chosen parameters be set.
The remainder of this report is divided into five chapters:
2. Framework for Analysis;
3. Future Section 8(a) Rules;
4. Data Analysis;
5. Analysis of Options; and
6. Conclusions.
ICF INCORPORATED
-------
2. FRAMEWORK FOR ANALYSIS
2.1 INTRODUCTION
Ideally, the small manufacturer definition should balance the costs and
benefits of obtaining information. The "marginal benefit" of obtaining
information is an elusive concept, and in this case, its estimation is
complicated by the fact that the purposes for obtaining the information are as
yet not fully identified. For example, if information were collected to
estimate the magnitude of annual production of certain chemicals, information
from a large number of small firms that account for less than ten percent of
total production would not be very valuable. However, if information were
collected to identify the total number of sites producing certain chemicals,
then information on even small producers could be needed.
To address this problem, four proxies for "benefit" are used:
Number of reports required to be prepared;
Number of locations required to report;
Percent of total production covered by the
reporting requirement; and
Percent of the chemicals for which sufficient
information is received.
To balance the "benefits" and "costs" of different rules, a proxy for
cost was also developed. The total cost to industry of a reporting rule is a
function of the number of reports that must be prepared, the number of sites
at which the reports are prepared, and possibly the number of parent firms to
which the sites belong. The actual cost in dollars will depend on the
complexity of the rule and the type of information requested. Generally,
however, the greater the number of reports, sites, and possibly firms, the
greater the cost of the rule.
Each of the proxies is discussed below.
2.2 ESTIMATING BENEFITS
Four proxies will be used to evaluate the relative benefit of exempting
fewer, as opposed to a larger number of small manufacturers from having to
report under Section 8(a). The proxies themselves do not strictly measure
"benefits." Instead, they measure the relative amounts of different types of
information required to be reported under various small manufacturer
exemptions. The amounts of information received must be valued relative to
each other in order to have a final benefit measure. We have not assigned
values to the different types of information, although we have speculated
ICF INCORPORATED
-------
regarding potential future Section 8(a) rules to determine which types of
information, i.e., which proxies, may be most appropriate to use to evaluate
them. This assessment of future Section 8(a) rules is in Chapter 3 of this
report.
Each of the proxies for benefits is described below.
Reports. The number of reports that would have to be prepared can be
used as a proxy for the amount of information received under a given small
manufacturer exemption. The assumption is that the greater the number of
reports received, the more is learned about the chemicals' annual production,
uses, or whatever type of information is requested. The marginal value of
receiving reports from "small manufacturers" may not be high if small
manufacturers are not able to furnish new information about chemicals that
large manufacturers cannot also provide. Therefore, although the marginal
value of additional reports may vary, generally, the greater the number of
reports prepared, the greater the "benefits" of the rule.
The number of reports that are required to be prepared under a given
small manufacturer exemption has meaning only in relation to the maximum
possible number of reports that would be required if there were no exemption.
The percentage of total possible reports that must be prepared measures the
information received in relation to the maximum possible amount of
information. The percentage of total possible reports that is not prepared
under a given exemption measures the amount of information not received
relative to the maximum possible. This last measure, percentage of total
possible reports not prepared, is used as a proxy for the amount of
information received under various small manufacturer exemptions. The greater
the value of this proxy, the less information is received.
Locations. The number of locations that would have to prepare reports
is the second proxy for benefit. In a situation in which information is
requested on only one chemical, the reports and locations proxies are
identical; each location that produces the chemical can submit one report.
When information is obtained on a group of chemicals, which is the primary
focus here, the proxies diverge because a single location may submit any
number of reports. A small percentage of locations may account for a large
percentage of reports. Therefore, both proxies are examined.
The greater the number of locations that must report, the more
comprehensive the information received. The marginal value of obtaining
information about even "small manufacturers" may be high if the purpose of the
rule is to identify all the locations. The assumption is that for many toxic
chemicals, comprehensive reporting may be necessary.
As with the "reports" proxy, the number of locations required to report
has meaning only in relation to the total number of locations that would have
to report were there no exemption. Therefore, a percentage measure is used;
the percentage of total possible locations not required to report. The
greater the value of this proxy, the less information is received.
ICF INCORPORATED
-------
2-3
Production. Amount of production is the third proxy for benefit. The
greater the amount of production covered, the more comprehensive the reporting
requirement. For example, if ,a Section 8(a) rule were designed to identify
opportunities for environmental releases and human exposure by requiring the
reporting of information on the uses and disposal methods of chemical
substances, then the amount of total production covered would indicate the
extent to which the information obtained was complete. Again, a percentage
measure is used; the percentage of total possible production not covered by
the reporting rule. The greater the value of this proxy, the less information
is received.
Chemicals. The final proxy for benefit is the number of chemicals for
which insufficient information is reported. The production proxy measures how
much of the total production for a group of chemicals is covered. However, it
gives no indication of whether some chemicals are missed; i.e., needed
information is not received for them at all. This proxy measures the
percentage of chemicals about which insufficient information is obtained under
various small manufacturer exemptions.
"Sufficient information" can be defined for this proxy in various ways,
including:
cover a given percentage of total production of the
chemical; and
* cover a given percentage of the locations that
produce the chemical.
For simplicity, we defined sufficient information about a chemical as having
at least one location report on it. Therefore, this proxy measures the
percentage of the total number of chemicals that are missed entirely.
These four proxiesreports, locations, production, and chemicalsmake
up the framework for evaluating the benefits of requiring various amounts of
information to be reported. Their use and interpretation vary with the types
of reporting rules to which they are applied. The next section discusses
proxies for costs.
2.3 ESTIMATING COSTS
To balance the "benefits" and "costs" of different rules, a proxy for
cost is needed. The total cost to industry of a reporting rule is a function
of the number of reports that must be prepared, the number of locations at
which the reports are prepared, and possibly the number of parent firms to
which the sites belong. The actual cost in dollars will depend on the
complexity of the rule and the type of information requested. Generally,
however, the greater the number of reports, locations, and possibly firms, the
greater the cost of the rule.
ICF INCORPORATED
-------
2-4
The total cost of the rule can be divided into two parts: "fixed cost"
and "variable cost." The fixed cost refers to the resources of the location
and firm required to become familiar with the reporting requirements. We have
assumed that this cost is wholly borne by locations that actually have to
report. Therefore, for purposes of our analysis, the fixed cost of reporting
is indicated by the number of locations required to report.
It is not evident how much involvement the parent firm will have in the
reporting procedure. If, for instance, the reports are cleared through the
parent company, additional costs will be incurred there. If these resources
are constant for each additional location controlled by one company, this cost
will be captured in the number of locations. However, if a firm must incur a
fixed cost regardless of how many locations it controls, the number of parent
firms should be a separate parameter in the cost equation.
The variable cost refers to the cost per report prepared. This includes
collecting the data on each chemical and filling out the form. For each
additional report that a site prepares, it bears an additional cost.
Therefore, the total number of reports prepared is an indicator of the
variable costs of reporting.
Because the total cost to industry is a function of the cost at the
location and/or firm and the cost per report prepared, we set up the following
equation:
Total Cost = ( a) (Number of Firms) + ( b) (Number of Locations)
~ + ( c) (Number of Reports).
If the parent firm is not involved in the reporting, a = 0 and the equation
becomes:
Total Cost = ( b) (Number of Locations) + ( c) (Number of Reports).
To estimate the coefficients b and c, we reviewed estimates of fixed costs
and variable costs of reporting for two different EPA rules. The first
estimate, for the inventory reporting rule where the manufacturers were
required to report production volume, set fixed costs at $300 per location and
variable costs at $100 per report.1 The second estimate was for the
proposed Section 8(a) Preliminary Assessment Information Rule (discussed in
Chapter 3). Under this rule, the manufacturer provides more information about
each chemical and this is reflected in the higher estimate for variable costs
($420 per report). The fixed costs for this rule were estimated at $480 per
site.
1A.D. Little, Analysis of Options for Definition of Small Business and
Estimated Cost of the Initial Section 8(a) Reporting Requirements (Cambridge,
MA; November 1977).
ICF INCORPORATED
-------
2-5
Of the two rules, we assume the costs for future Section 8(a) rules will
be closer to those estimated for the Preliminary Assessment Information Rule.
Therefore, given the information available, the best values for the
coefficients are a = 0, b = 8, and c = 7. To facilitate the analysis, we
set a = 0, b = 1, and c = 1, and the proxy for cost becomes simply the
number of reports prepared plus the number of locations reporting. This is
neither a conservative nor a non-conservative assumption regarding costs
because the actual cost in dollars is not addressed in the analysis. Only
the shape of the cost curve (as a function of the exemption parameters) is
affected by the coefficients, and the sensitivity of the results to the choice
of coefficients is examined in Section U.2.
The next chapter assesses future Section 8(a) rules in the context of the
framework described in this chapter.
ICF INCORPORATED
-------
3. FUTURE 8(a) RULES
3.1 INTRODUCTION
The language of Section 8(a) of TSCA is very broad. The EPA
Administrator may require manufacturers to maintain records or submit reports
as necessary for the effective enforcement of the Act. The types of 8(a)
rules may vary from requiring general information about many chemicals to
requiring very detailed information about a small number of chemicals.
Small manufacturers and processors are not subject to Section 8(a)
reporting and recordkeeping unless this is specifically required by a rule
promulgated by the Administrator when:
a chemical is subject to a testing rule under
Section 4;
the Administrator rules that a chemical may present
an unreasonable risk of injury to health or the
environment under Section 5(b)(4);
additional information is required by the
Administrator during the premanufacturing
notification (PMN) review process in order to decide
whether to regulate a new chemical under Section 5(e);
a chemical is subject to a control action under
Section 6; or
a chemical is the subject of litigation under civil
action provided for in Section 5 or 7.
For other types of actions, small manufacturers and processors are
exempt. Section 8(a)(3)(B) requires the Administrator to prescribe standards
for determining which manufacturers and processors qualify for this exemption.
EPA is considering proposing a general small manufacturers exemption to
apply for all Section 8(a) rules unless the chemical substance or mixture is
the subject of rules outlined above.1 The possible variation in the scope
and level of detail of 8(a) rules complicates the selection of a general
exemption for small manufacturers. An exemption set at a level to achieve
very detailed information in one rule might be too inclusive for another rule,
resulting in more reporting than necessary for the purposes of the rule, and
x45 YR 66180, 10/6/80.
ICF INCORPORATED
-------
3-2
unnecessarily increasing the cost to industry. For the purpose of this study,
we have speculated about the kind of rules that might be used.
This chapter examines two types of potential future rules: group
chemical rules and individual chemical rules. Group chemical rules may be
used to obtain general information, not available from public sources,
regarding a large number of chemicals. These data may be useful for
evaluating potential hazards, for setting priorities for further
investigation, or for other purposes. An individual chemical rule may be used
to obtain detailed information about a particular chemical that is not
available from public sources. This information may be useful for evaluating
risks and potential risk reduction options.
These potential rules illustrate the kind of information manufacturers
may have to report in future 8(a) actions. We briefly examine these rules
with the proxies for information that were developed for this analysis and
discuss which of the proxies is the most appropriate to use to evaluate the
benefits of information received under each rule.
3.2 GROUP CHEMICAL RULES
Group chemical rules may require manufacturers to report general
information such as that regarding production processes, exposure potential,
and consumer uses. Chemicals could be selected for such a rule on the_basis
of potentially significant toxicity and the unavailability of the needed
information from public sources.
To determine which of the "value of information received" proxies is most
approprate for defining small manufacturers for group chemical rules, a more
precise understanding of the potential rule is required. For purposes of this
analysis we shall assume the following:
The primary purpose of a group chemical rule would
be to aid in the setting of priorities for further
investigation.
The information received on each chemical in a
group chemical rule should be representative (to the
extent possible) of all of its potential risks (i.e.,
representative of the various manufacturing
processes, uses, and disposal problems).
The Agency could tolerate not receiving sufficient
information1 on a limited number of chemicals in the
group chemical rule assuming that the Agency has the
means and the opportunity to supplement the
information received with data from other sources.
Alternatively, in some cases the lack of reported
data may indicate that the chemical is of little
concern (e.g., it is only produced in very small
quantities and therefore all of its manufacturers
were not required to report) so that additional
investigation is not required.
ICF INCORPORATED
-------
3-3
The "chemicals missed" proxy is of obvious importance in evaluating this
type of rule. Clearly, the fewer chemicals missed the better, and this proxy
allows EPA to set a small manufacturer exemption to minimize, for a given cost
to industry, the number of chemicals for which no reports would be received.
However, this approach could also lead to an unnecessary increase in the total
cost to industry of the reporting rule. A balance is required between the
additional cost to the Agency of filling in data on missed chemicals and the
additional reporting cost imposed on industry in order to reduce the number of
chemicals missed. Therefore, to make full use of the chemicals missed proxy,
information is required on what steps the Agency would take (if any) to obtain
information on chemicals missed by the rule, and what the cost of those steps
would be. For example, in a given range of small manufacturer definitions,
missing a chemical may indicate that the chemical is of too little importance
to be considered further. This could occur if the exemption were structured
such that a chemical could be missed only if it were produced in very small
quantities. In this case the Agency may not be very interested in any
chemicals that were missed, and it would not take steps to fill in data on the
missed chemicals. Therefore, in this example, the chemicals missed proxy
would not be appropriate for evaluating the value of information received
because there would be very little incremental value of missing fewer
chemicals. How few chemicals are missed is a judgment for EPA to make based
on the value of the information received and the costs to industry of
providing such data.
The percentage of production covered proxy could indicate the extent to
which all uses of the chemicals are covered. If a large portion of total
production is covered, it may also be likely that most uses would also be
covered. However, this proxy is biased toward large production processes.
Therefore, relying on this proxy may lead to not receiving sufficient
information on manufacturing processes of various sizes. This lack of
coverage of manufacturing processes could be an important deficiency if
processes of various sizes have different exposure potentials, or if
manufacturers of smaller quantities tend to supply different uses than
manufacturers of larger quantities.
The production process u$ed by a firm for a particular chemical may be a
function of the size of the firm (as measured by annual sales) and the
production volume of the chemical. The data do not indicate that small
quantities of chemicals are produced only by locations with lower sales and
fewer employees. Forty-one percent of the reports of chemical production
listed in the lowest production category (less than one-thousand pounds) are
attributable to manufacturers in the largest sales category (over $100
million). While it can be seen that large companies, in many cases, produce
small volume chemicals, it is not known how the production processes used by
those firms compare to the processes used for the same chemicals produced in
similar quantities by smaller firms. Authoritative information outlining the
extent to which production processes and equipment vary across segments of the
chemical industry is not readily available. It is believed, however, that the
potential is very great; in particular if systematic differences exist in the
age and type of equipment used, and disposal methods employed. This may be
due, in part, to variances in the availability of capital across segments of
the industry.
ICF INCORPORATED
-------
3-4
To the extent that these systematic variances do occur, the exposure
potentials from different segments of the industry may also vary, and
information from "large manufacturers" may not be applicable for extrapolation
to describe "small manufacturers." To ensure that sufficient information is
obtained on all segments of manufacturers, the information proxies could be
evaluated on the basis of their ability to provide reports across both
production level and firm size. However, if production processes and
equipment do not vary systematically throughout the industry, then
representativeness across all the production and sales categories may not be
an important criterion to consider. Due to its potential for importance,
representativeness is used in the analysis below to illustrate the effect it
may have on the choice of the small manufacturer exemption.
Both the reports and locations proxies may be useful for evaluating group
chemical rules. They are both neutral to the representativeness problem in
that neither of them is biased in its construction. However, in using them to
help choose a particular level for the exemption, the representativeness of
the information that would be received as a result of maximizing one or the
other proxies for a given cost should be examined. This is discussed more
fully in Chapter 5.
3.3 INDIVIDUAL CHEMICAL RULES
While future 8(a) reporting may include group rules, these rules are not
in place. Several rules have been prepared for individual chemicals.
PBB/Trisฆ This final 8(a) rule2 requires chemical manufacturers and
importers with sales above $500,000 to report any manufacture or import of
polybrominated biphenyls (PBBs) or tris (2,3-dibromopropyl) phosphate. All
companies which manufacture or import more than 10,000 pounds per year must
report.
While both of these chemicals are toxic, they are both apparently out of
production. Section 6 action is not currently contemplated. A reporting rule
gives EPA notice of the resumption of production and a chance to initiate
action if it determines that an unreasonable risk exists at that time.
Asbestos. This proposed 8(a) rule would require manufacturers,
importers and processors of asbestos to report information about activities
involving use of and exposure to asbestos.1 The rule was proposed because
the information available from other agencies is limited in various ways. For
instance, the Bureau of Mines estimated that information from its asbestos
consumption figures is accurate to only + 50 percent. Much of the
information from other agencies is not available because of confidentiality
restrictions.
245 I]? 70728, 10/24/80.
ฎ46 FR 8200, 1/26/81.
ICF INCORPORATED
-------
3-5
Although it has not proposed any such action, EPA might use Section 8(a)
authority for requiring reporting on new chemicals as a supplement to the
reporting required under the PMN process. This reporting would be used to
obtain more information on production, use, exposure and available test data
for a new chemical which did not warrant regulation based on its PMN
submission.k An alternative method of obtaining some of this information is
through Section 5(a)(1)(B) significant new use authority.
All of these proposed or anticipated single chemical rules are collecting
information for assessment purposes. The information will be used to decide
whether to take further action and if so, what type of action to take.
The chemicals missed proxy is not relevant to individual chemical rules.
The proxy could be interpreted as the percent chance that a randomly selected
chemical (from the chemicals included in the data used in this report) would
be missed. However, it is likely that preliminary investigations by the
Agency would be sufficient to identify whether adequate reports would be
received on a particular substance that was the subject of an individual
chemical rule.
The production proxy is biased toward large production processes, as was
discussed above in relation to group chemical rules. For an individual
chemical, this production proxy may be an adequate measure of the information
received under a given reporting rule. However, it is not possible to
generalize regarding its application for evaluating all future individual
chemical rules.
For individual chemical rules the reports received and locations
reporting proxies are equivalent. The appropriate level at which to set the
exemption will depend on the type of information requested Tinder the rule.
Nevertheless, in general there appears to be no reason to believe that the
exemption level for individual chemical rules should differ substantially from
the exemption level for group chemical rules.
''Office of Toxic Substances, U.S. Environmental Protection Agency,
Priorities for OTS Operation (January 1982), pp. 111-12 and 111-13.
ICF INCORPORATED
-------
4. DATA ANALYSIS
4.1 INTRODUCTION
To analyze the options for setting the small manufacturer exemption, we
analyzed data on over 3,000 manufacturing locations of over 1,400 different
chemicals. EPA provided sales, employment, and production data specific to
each location and production data specific to each chemical. This information
was obtained by EPA from the TSCA Chemical Substance Inventory and Dun and
Bradstreet, Incorporated. Because a significant portion of the sales and
employment data was incomplete, we supplemented the information provided by
EPA by manually searching through reference materials for missing sales and
employment data. The final data base included 3,044 manufacturing locations
of 1,459 chemicals. The data used in this report are described in Appendix A.
The data were used to estimate how the proxies for "benefits and costs
change for various small manufacturer exemptions. We calculated the percent
of reports not prepared, percent of locations not required to report, percent
of production not covered, and percent of chemicals missed entirely for
combinations of two sets of parameters for the small manufacturer exemption:
* Total annual sales for the corporation and
production quantity of individual chemicals at each
manufacturing location; and
Total employment for the corporation and production
quantity of individual chemicals at each
manufacturing location.
The percent of chemicals missed was estimated using a simulation that is
described in .Appendix B.
We also examined two structures of the exemption: one for which the
location must be below both parameters in the set in order to be exempted, and
the second for which the location need be below only one parameter in the set
-in order to be exempted. The choices of which parameters to use and how to
structure the exemption are discussed in Chapter 5 of this report. The data
analyses are described in Appendix B.
The results for the proxies for benefits and costs for small manufacturer
exemptions with parameters of production quantity and sales are presented
first. The values of the proxies are reported in matrix format (Exhibits 4-1
through 4-9), then they are graphed (Exhibits 4-10 through 4-18), and then the
slopes of the graphs are reported in matrix format (Exhibits 4-19 through
4-27).
ICF INCORPORATED
-------
4-2
The values for the locations, reports, and production proxies are
reported for both structures of the exemption: when the location must be
below both parameters to be exempted, and when the location need only be below
one parameter to be exempted. The percent of chemicals missed was only
calculated for all the combinations of parameters for the case when the
locations must be below both parameters in order to be exempted (see Appendix
B for a full discussion of this analysis).
As a rough approximation for the percent of chemicals that would be
missed if the locations were required to be below only one parameter to be
exempted, we calculated the percentage of the chemicals whose highest report
of production falls into each of the production quantity categories. The
cumulative sum of these percentages is the lower bound on the percent of
chemicals that would be missed as a function of the production quantity
parameter alone, when locations need only be below one parameter to be
exempted. These estimates, shown in Exhibit 4-28, reveal that a large
percentage of chemicals would be missed for even low production quantity
parameters when the locations need only be below one parameter in order to be
exempted.
The results for the proxies for benefits and costs for small manufacturer
exemptions with parameters of production quantity and employment are presented
next. The values of the proxies are reported in matrix format (Exhibits 4-29
through 4-37), then they are graphed (Exhibits 4-38 through 4-46), and then
the slopes of the graphs are reported in matrix format (Exhibits 4-47 through
4-55).
The values for the locations, reports, and production proxies are
reported for both structures of the exemption: when the location must be
below both parameters to be exempted, and when the location need only be below
one parameter to be exempted. As mentioned above, the percent of chemicals
missed was only calculated for the case when the location must be below both
parameters in order to be exempted. An approximation of the percent of
chemicals that would be missed if, the locations were required to be below only
one parameter to be exempted is shown in Exhibit 4-28.
Several observations and conclusions can be drawn regarding these proxies
for benefits and costs. For example, Exhibit 4-5 shows that for the case when
the locations must be below both parameters to be exempt, over 99 percent of
total production is covered when the production parameter is set at or below
45.4 million kilograms (100 million pounds), and the sales parameter is set at
or below 100 million dollars. Almost all production is covered for exemptions
even in this high range because, predictably, large quantity chemicals make up
nearly all production and locations with very large sales (or employment)
produce nearly all the large quantity chemicals. This conclusion is borne out
in Exhibit 4-6 which shows (in its top row) that locations with annual sales
under 100 million dollars account for only 8.3 percent of total production.
In the first column of Exhibit 4-6, we also see that only five percent of all
production was reported in quantities under 22.7 million kilograms (50 million
pounds). Therefore, if the small manufacturer exemption requires that
ICF INCORPORATED
-------
4-3
locations be below-both parameters in order to be exempted, then nearly all
the production (of the chemicals in our data) would be covered if the
parameters of the exemption were set within the range considered here. A
significant portion of total production can be missed if locations need only
be below one parameter to be exempt. The choice between these two structures
for the definition is discussed in Chapter 5.
The behavior of the other proxies as a function of the parameters of the
definition can also be analyzed from the results of the data analyses. In
particular, the behavior of the benefit proxies relative to the behavior of
the cost proxy is of interest. The remainder of this chapter discusses the
relationships between the benefit and cost proxies and the effects that
missing data and information held confidential may have on the results.
4.2 RELATIONSHIPS BETWEEN THE BENEFIT AND COST PROXIES
To examine the relationships between the benefit and cost proxies, the
behavior of the various proxies as a function of the exemption parameters must
first be determined. Isoquant curves, a convenient method for displaying
information about the relationship among three variables, are presented here
to facilitate this analysis. The employment parameter is not considered in
this section because its effects on the level of information received are very
similar to the effects of the sales parameter. Therefore, only isoquant
curves showing the relationship between annual sales, production volume, and
level of information received are included.
Exhibit 4-56 displays how sales and production volume trade off to
maintain a constant number of locations reporting a measure of information
obtained from the reporting rule. This exhibit was constructed for the case
when locations must be below both parameters of the definition to be
exempted. For a small manufacturer definition of less than $100 million in
sales and 454,000 kilograms (1 million pounds) of production, approximately 15
percent of the locations would be exempted. Fifteen percent of the locations
could also be exempted for the other combinations of sales and production that
lie along the isoquant labeled 15 percent, e.g., $10 million and 227 million
kilograms (500 million pounds). The fact that the number of locations stays
the same does not mean that the information would be received from exactly the
same manufacturers. For example, a location with annual sales of $40 million
and production of 22,000 kilograms would not have to report under the $100
million, 454,000 kilogram definition. It would, however, have to report under
the $10 million and 227 million kilogram definition.
The total cost to industry is represented in Exhibit 4-57. This exhibit
shows total cost as the number of locations reporting plus the number of
reports prepared, (the coefficients in the cost equation presented in Chapter
2 are set at a = 0, b = 1, and c = 1). In this exhibit, a location must be
below both criteria to qualify as a small manufacturer.
ICF INCORPORATED
-------
4-4
If the cost curves are put on the same graphs as the information curves,
one can identify which small manufacturer definitions provide a given amount
of information for the least cost, or alternatively, the most information for
a given cost. As shown in Exhibit 4-58, in order to get the most locations
for a given level of cost (or spend the least for a given number of
locations), the definition should be set in the "northwest" corner of the
graph. For a given cost constraint, say 15 percent reduction, we can
approximate how many locations would be exempt under different definitions.
The set of definitions, however, must fall on the cost isoquant labeled 15
percent. Following this cost isoquant, one sees that anywhere from
approximately 17 percent to 8 percent of the locations could be exempted
within this cost constraint. To maximize the number of locations reporting at
a given cost constraint, one would move to the western (left) and northern
(top) portion of that isoquant. Because the other isoquants are of the same
general shape, to maximize the number of locations reporting for any given
cost constraint,- one would be driven toward the "northwest" portion of the
cost isoquants, i.e., "high" sales parameter and "low" production quantity
parameter.
Exhibit 4-59 displays isoquants of the percent of reports exempted along
with isoquants of the cost proxy. These reports and cost isoquants are very
similarly shaped because the cost function is driven primarily by the number
of reports that must be prepared. However, upon close examination one sees
that to get the most reports for a given level of cost, the definition should
be set in the "southeast" portion of the curves, i.e., "low" sales parameter
and "high" production quantity parameter.
Exhibit 4-60 displays isoquants of the percent of chemicals missed along
with isoquants of the cost proxy. These curves clearly demonstrate that the
most cost-effective small manufacturer definitions in terms of this proxy for
information are in the "southeast" portion of the curves, i.e., "low" sales
parameter and "high" production quantity parameter.
Exhibits 4-58 through 4-60 also display the effect that maximizing a
proxy at a given level of cost has on the other proxies. For example, at a
ten percent reduction in cost, 3.8 percent of locations will be exempted when
the exemption is set at 4,540 kilograms (10,000 pounds) and $100 million
(i.e., the "northwest" of Exhibit 4-58). At this exemption level, 11.7
percent of reports would be exempted and 9 percent of chemicals would be
missed. However, more reports can be obtained at the cost of losing some
locations by moving to the "southeast" of Exhibit 4-59. At 45.4 million
kilograms (100 million pounds) and $5 million, the cost of reporting is still
reduced ten percent, but 11.2 percent of locations and 9.6 percent of reports
would be exempted, and 5.6 percent of chemicals would be missed.
Although the reporting costs in these two examples are the same, the
information obtained varies. When locations are maximized, locations
producing moderate quantities of chemicals (greater than 4,540 kilograms) must
report in all sales ranges and the coverage of locations is representative
across the sales ranges. When reports are maximized, however, three times as
many locations are exempted, some of which produce chemicals in large
ICF INCORPORATED
-------
4-5
quantities (up to 45.4 million kilograms). More reports are received from the
lower production ranges of locations with moderate sales (greater than $5
million), reducing the total number of chemicals missed. There is almost no
representation in the low sales ranges.
As mentioned above, nearly all production is covered for the full range
of exemptions considered here for the case when locations must be below both
parameters to be exempted. Therefore, isoquants are not presented here for
the production proxy. However, for a given cost constraint, the percent of
production not covered would be minimized in the "northwest" portion of the
curves, i.e., "high" sales parameter and "low" production quantity parameter.
The difference, however, in the relevant ranges considered here, are very
small.
In Exhibits 4-58, 59 and 60, a is set equal to 0. We also investigated
the case in which a = 1, b = 1, and c = 1. The shapes of the cost
isoquants are almost identical to the cases shown in Exhibits 4-58, 59, and
60, demonstrating that the rate of change of the total cost of reporting as a
function of sales and production volume is not very sensitive to assumptions
concerning the parent firm's role. This is expected because, as mentioned
above, the total cost is primarily driven by the number of reports that are
prepared.
The above analysis displays the cost-effective solutions for a particular
set of coefficients, and for a given hypothetical reporting rule. However,
not all reporting rules may be of the type discussed above. As discussed
above in Chapter 3, EPA may develop various types of Section 8(a) rules. As
shown by the difference between the cost estimates for the inventory reporting
rule and for the preliminary assessment rule, there is no guarantee that the
ratio between fixed and variable costs will be approximately 1:1 for different
reporting rules.1
The conclusions for the hypothetical reporting rule analyzed above are
unaltered by variations in the relative values of the coefficients. This can
be illustrated first for the case in which a manufacturer must be under both
criteria to be considered a small manufacturer. Exhibit 4-61 displays both
report and location isoquants. As can be seen, moving to the left, the
location isoquants curve upward more rapidly than the report isoquants.
Therefore, a location isoquant which intersects a report isoquant will be
above it at the left of the graph, and below it at the right. But the cost
function is simply a linear combination of the report and location functions,
and the cost isoquant will lie somewhere between the report and location
isoquants (assuming a = 0). The cost isoquant will therefore lie above the
report isoquant at the left and below it on the right, keeping the lowest cost
solutions in the "northwest" and "southeast" portions of the graphs for
locations and reports respectively.
*It does not matter if the absolute values of the fixed and variable
costs change; only the relation between the two can influence this analysis.
ICF INCORPORATED
-------
4-6
The addition of firm-specific fixed costs leaves the conclusion unaltered
if information is measured by reports prepared, but could affect the
conclusions if information is measured by locations covered. This can be
illustrated in Exhibit 4-62. One would move a short distance from point A,
where a location, report, and firm isoquant (the latter not shown here)
intersect, to point B, where the level of each isoquant is a small increment
lower than its level at point A. One could then move directly upward, and
eventually one report will move from exempt to non-exempt status. However,
the change of one report may not necessarily move a location from exempt to
non-exempt status: it depends on whether that is the only report made by that
location. Similarly, as the upward movement continues, a location will
eventually move from exempt to non-exempt status, but a firm will not
necessarily move as well. Therefore, as one moves upward along line BC, one
will first encounter the reports isoquant, then the location isoquant, and
then the firm isoquant. By applying the analysis of changes in the relative
values of b and c to the three coefficient case, one can easily show that
the addition of firm-specific fixed costs also does not change the conclusion
when information is represented by reports prepared.
However, when information is represented by the locations proxy, the
existence of firm-specific costs means that changes in the relative values of
the coefficients can change the conclusions. As c, the reports coefficient,
increases relative to the other two coefficients, the cost isoquants are
approximated by the reports isoquants, keeping the cost-effective solution for
locations in the "northwest" corner of the graph. As a, the firm coefficient,
increases relative to the other two coefficients, the cost isoquants are
approximated by the firm isoquantsr moving the cost-effective solution toward
the "southeast" corner of the graph. Because the number of reports vastly
outweighs the number of locations and the number of firms, the relative size
of a to b and a to c would have to be quite large to cause this shift for
the mix of chemicals and manufacturers examined here; at least in the range
above 5:1. Therefore, this shift is considered unlikely.
The cost-effective solution for missed chemicals is also not sensitive to
the relative values of the coefficients b and c in the cost equation. Even
if b is much larger than c , or if c is much larger than b , the fewest
chemicals are missed for a given cost constraint toward the "southeast"
portion of the curves in Exhibit 4-60.
When the locations need only be below one parameter, the shapes of the
curves change. For this type of exemption, the curves bend the other way and
the proxies for which the cost-effective solutions were "southeast", become
"northwest" (i.e., "low" production quantity and "high" sales), and
vice-versa. Again, the relative value of b to c does not change the results.
We have demonstrated that the conclusions regarding cost-effective
solutions are insensitive to alterations in the relative coefficient values
for the hypothetical reporting rule analyzed here. However, the conclusions
are not insensitive to changes in the type of reporting rule. For the
purposes of the analyses conducted here, the differences between reporting
rules are fully incorporated in one of two categories:
ICF INCORPORATED
-------
4-7
differences in relative coefficient values; and
differenced in the mix of locations and chemicals
covered by the reporting rule (the joint distribution
of sales and production figures over the locations
covered by the rule).
We have already shown that changes in relative coefficient values are unlikely
to affect the conclusions. However, changes in the mix of firms and chemicals
covered by the reporting rule can affect the conclusions. As shown in
Appendix C, one can construct a different hypothetical reporting rule with a
certain mix of locations and chemicals in which the cost-effective solution
for reports is in the center of the graph rather than in the "southeast"
corner for a rule in which a location must satisfy both criteria to qualify as
a small manufacturer.2
It is also important to realize that the reporting rules actually
promulgated may seldom, if ever, differ substantially from our primary
example. Our conclusions are likely to hold for any rule in which the mix of
locations and chemicals covered approximates the mix of locations and
chemicals in our data. Only when the mix is markedly different than our
example will the conclusions be altered.1
The above discussion is necessarily rather vague about the ways in which
the mix of firms subject to the reporting rule, might influence the shapes of
the isoquants, which in turn might influence the conclusions. It is difficult
to analyze the complex interactions represented by the joint distribution of
sales and production and determine under what conditions our conclusions will
hold. In order to fully investigate the degree to which our conclusions are
generalizable to the types of rules which may actually be promulgated, it is
necessary to perform these same data analyses on the chemicals and
manufacturers covered by the individual rules. The exact structure and
purposes of future Section 8(a) rules are as yet not fully identified;
although, it is likely that future Section 8(a) rules will relate to a subset
of those chemicals included in the data used for this analysis. Nevertheless,
at a later date it may be useful to analyze individual rules to generate
conclusions that may be more generalizable to all reporting miles.
2It should also be noted that for a different reporting rule
incorporating a different mix of chemicals and firms, the cost-effective
solution when percent of chemicals missed is used as a proxy may no longer be
resistant to changes in the relative values of b and c.
*tfhen a reporting rule covers only one chemical substance, the
location, report, and cost isoquants will coincide, and as long as a = 0 (or
if each location is owned by a different firm), there will be no unique
cost-effective solution for the reports, locations, or chemicals missed
proxies.
ICF INCORPORATED
-------
4-8
4.3 MISSING AND CONFIDENTIAL DATA
The data used to generate the results of the analyses presented in this
chapter exclude a significant portion of the information provided by EPA on
the manufacturers of the 1459 chemicals chosen for this study. Some of the
data provided by EPA were incomplete; thus preventing their inclusion. Other
information was held confidential by the manufacturers; again preventing the
inclusion of the data in this report. This section discusses whether the
exclusion of these data is likely to bias the results and conclusions drawn
here.
EPA identified 3778 manufacturing locations of the 1459 chemicals used in
this study. The locations were identified from reports of production
submitted by manufacturers to the Agency in 1977 when the TSCA Chemical
Substance Inventory was compiled. From these reports, EPA determined in which
of twelve possible production categories each chemical was produced at each
location. These data may be in error either due to reporting errors or
because production of these chemicals has changed since 1977.
Random reporting errors in so large a sample are not likely to cause a
bias in the results presented here. However, if the accuracy of the
information provided by certain segments of manufacturers, e.g., manufacturers
that are small in terms of annual sales, was consistently below the accuracy
of other segments of manufacturers, then some biases may develop. There has
been no indication of such differentials in accuracy or of any systematic
reporting bias in this sample. Therefore, the data are assumed to accurately
reflect the production of this given list of chemicals in 1977 for the
purposes of this report. (The initial reporting rule for the Inventory
included a small manufacturer exemption that may affect the results of this
analysis in the lower ranges of sales and production. This problem is
discussed in Appendix A.)
If production of these chemicals has changed significantly since 1977,
then the production data used here do not reflect the current status of the
industry. For example, the development of a new, capital intensive production
process may shift the majority of the- production of a particular chemical from
many small to a few large manufacturers. There is no way to tell from the
data themselves whether production patterns have changed, or whether the
general level of all production has increased or decreased. It is likely that
the production of various chemicals fluctuated up and down, while the
production of other chemicals increased or decreased steadily over the past
four years. Therefore, in the absence of specific information on the changes
in production of the 1459 chemicals used in this report, we assume that the
1977 production data provided by EPA are sufficiently accurate for the
purposes of this study.
EPA also provided sales and employment information on 3778
manufacturers. This information was incomplete for 2386 locations. We
completed the needed data for 1652 of these locations, leaving 734 locations
incomplete. (Appendix A describes the process used to obtain the missing
data.) Exhibit 4-63 displays the distribution of the reports of production
ICF INCORPORATED
-------
4-9
across the twelve production categories for the 3044 locations used in the
analysis and for the 3778 locations on which data were provided by EPA. The
production statistics for the 3044 locations appear to be representative of
the overall group of 3778 locations because:
The distributions of reports across production
categories are very similar; and
The percentage of reports in each production
category included in the analysis is not
statistically different from the percentage of total
locations included (80.6 percent). (This was
determined by performing a chi-squared test on column
(5) in Exhibit 4-63.)
It was not possible to determine whether the sales and employment
statistics of the 3044 locations were representative of the overall group of
3778 locations because the lack of sales and/or employment data was the factor
that excluded the 734 locations from the analysis. There is no obvious bias
in the 2386 locations on which EPA did not provide complete information. In
our search for data, we would not have found sales and employment data for any
firms that were very small, because our data sources did not cover firms with
very small assets. However, we were also unable to obtain the needed
information for many locations that appeared to be part of larger corporations
because they were identified by names such as "Plant #2." Therefore, the 3044
locations used in the analyses are very likely to be representative of the
overall group of 3778 manufacturing locations of the 1459 chemicals.
Exhibit 4-63 also shows that approximately one-third of the reports of
production are in catgories BLANK and N. Because the production quantities
are unknown for these reports, it is not possible to test them against small
manufacturer exemptions, and they were dropped from the analysis. A total of
658 locations submitted reports only in these two categories, reducing the
total number of locations in the analysis to 2386.
Category N includes those chemicals on the list reported as not produced
in 1977, but produced prior to 1977, and expected to be produced in the
future. These chemicals are often produced in large batch processes on an
infrequent basis. Information requested under future Section 8(a) rules is
likely to be desired for these types of chemicals and production processes.
However, a small manufacturer exemption with a production quantity parameter
greater than zero will exempt many of these reports of production in any given
year. In 1977, 6.5 percent of the chemicals analyzed in this report were
only reported in production category N. Therefore, EPA may wish to vary the
reporting period on a rule-by-rule basis in order to obtain this information.
An alternative may be to require that reporting on chemicals produced
infrequently in batch processes only be exempt on the basis of the sales (or
employment) parameter of the exemptions.
ICF INCORPORATED
-------
4-10
Exhibit 4-64 shows that reports in production category N are slightly
more concentrated in the smaller sales ranges than reports of production in
categories 0 through 9. However, the distributions are generally similar, and
the results of the analyses would vary insignificantly with the inclusion of
these reports because the shapes of the isoquant curves would not change very
much.
A larger percentage of total reports were held confidential by the
manufacturing locations in 1977; category BLANK. A total of 8.8 percent of
the chemicals examined in this report were only reported in category BLANK in
1977. EPA may exempt locations reporting confidentially held information on
the same basis as other manufacturers, production quantity and sales (or
employment). The results may not vary significantly because reports in
category BLANK are distributed across the sales categories similarly to the
reports of production in categories 0 through 9 (see Exhibit 4-64). The
effect of the production parameter cannot be determined, however.
Chapter 5 analyzes options for setting the small manufacturer exemption.
ICF INCORPORATED
-------
EXHIBIT 4-1
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT
FOR VARIOUS SALES AND PRODUCTION QUANTITY PARAMETERS
(location must be below both parameters to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (lr000)
0.00
0.04
0.38
0.67
0.88
1.0
1.2
1.4
1.6
1.6
1.7
4,540 (10,000)
0.00
0.04
0.54
1.1
1.8
2.6
2.9
3.2
3.5
3.5
3.8
45,400 (100,000)
0.00
0.17
1.1
2.5
4.2
5.9
7.1
7.8
8.3
8.6
9.1
454,000 (1 million)
0.00
0.17
1.5
3.6
6.2
9.2
11.2
12.5
13.4
14.1
IE.2
4.54 million (10 million)
0.00
0.17
1.6
4.0
6.9
10.6
13.4
15.2
16.5
17.6
19.4
22.7 million (50 million)
0.00
0.25
1.8
4.2
7.3
11.1
14.0
16.2
18.0
19.4
21.6
45.4 million (100 million)
0.00
0.25
1.8
4.2
7.3
11.2
14.3
16.6
18.5
19.9
22.5
227 million (500 million)
0.00
0.25
1.9
4.3
7.5
11.5
14.9
17.5
19.6
21.3
24.4
454 million (1 billion)
0.00
0.25
1.9
4.4
7.6
11.6
15.0
17.7
19.8
21.6
24.8
-------
EXHIBIT 4-2
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT FOR
VARIOUS SALES AND PRODUCTION QUANTITY PARAMETERS
(tocation need only be below one parameter to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.25
1.9
4.4
7.6
11.6
15.1
17.7
19.9
21.8
25.1
454 (1,000)
2.8
3.0
4.3
6.4
9.5
13.4
16.7
19.0
21.2
22.9
26.2
4,540 (10,000)
6.7
7.0
8.1
10.0
12.5
15.8
18.9
21.2
23.2
25.0
26.6
45,400 (100,000)
16.0
16.1
16.9
17.9
19.5
21.8
24.0
26.0
27.7
29.2
32.1
454,000 (1 million)
31.5
31.6
31.9
32.3
32.9
34.0
35.4
36.7
38.1
39.2
41.5
4.54 million (10 million)
47.9
48.0
48.2
48.3
48.7
48.9
49.7
50.5
51.4
52.1
53.4
22.7 million (50 million)
64.0
64.0
64.1
64.2
68.3
64.5
65.0
65.5
66.0
66.4
67.5
45.4 million (100 million)
71.3
71.3
71.4
71.4
71.6
71.8
72.1
72.5
72.8
73.1
74.0
227 million (500 million)
89.9
89.9
90.0
90.0
90.1
90.1
90.2
90.2
90.3
90.4
90.7
454 million (1 billion)
95.1
95.1
95.1
95.1
95.1
95.1
95.2
95.2
95.2
95.3
95.5
-------
EXHIBIT 4-3
PERCENT OP TOTAL REPORTS NOT REQUIRED TO BE PREPARED FOR VARIOUS
SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.05
0.39
1.9
2.2
3.0
4.6
5.1
6.4
6.7
7.5
4,540 (10,000)
0.00
0.06
0.70
2.4
3.4
5.0
7.4
8.2
9.6
10.4
11.7
45,400 (100,000)
0.00
0.10
0.94
3.2
4.9
7.4
10.7
12.0
13.9
15.1
17.1
454,000 (1 million)
0.00
0.10
1.1
3.6
5.6
8.9
12.7
14.5
16.8
18.3
20.8
4.54 million (10 million)
0.00
0.10
1.2
3.8
5.9
9.4
13.5
15.5
18.1
19.8
22.9
22.7 million (50 million)
0.00
0.12
1 k 2
3.9
6.0
9.6
13.8
16.0
18.8
20.6
24.0
45.4 million (100 million)
0.00
0.12
1.2
3.9
6.1
9.6
14.0
16.3
19.1
20.9
24.5
227 million (500 million)
0.00
0.12
1.3
4.0
6.2
9.8
14.2
16.6
19.5
21.5
25.5
454 million (1 billion) 0.00 0.12 1.4 4.0 6.2 9.9 14.3 16.7 19.6 21.6 25.7
-------
EXHIBIT 4-4
PERCENT OP TOTAL REPORTS NOT REQUIRED TO BE PREPARED FOR VARIOUS
SALES AND PRODUCTION QUANTITY PARAMETERS
(location need only be below one parameter to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0*0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.12
1.4
4.0
6.2
9.9
14.3
16.7
19.7
21.7
25.9
454 (1,000)
12.7
12.8
13.7
14.9
16.7
19.6
22.4
24.4
26.1
27.7
31.2
4,540 (10,000)
20.2
20.3
20.9
21.8
23.1
25.1
27.1
28.8
30.3
31.5
34.5
45,400 (100,000)
31.1
31.1
31.5
31.9
32.4
33.5
34.7
35.8
36.9
37.6
39.9
454,000 (1 million)
44.7
44.7
44.9
45.1
45.3
45.7
46.3
46.9
47.6
48.0
49.8
4.54 million (10 million)
60.0
60.0
60.2
60.2
60.4
60.5
60.8
61.3
61.6
61.9
63.0
22.7 million (50 million)
74.3
74.3
74.4
74.5
74.5
74.6
74.8
75.0
75.2
75.4
76.3
45.4 million (100 million)
81.5
81.5
81.6
81.6
81.7
81.8
81.8
82.0
82.1
82.2
82.9
227 million (500 million)
95.5
95.5
95.5
95.5
95.5
95.5
95.5
95.6
95.6
95.6
95.9
454 million (1 billion)
98.0
98.0
98.0
98.0
98.0
98.0
98.0
98.0
98.1
98.1
98.2
-------
EXHIBIT 4-5
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS SALES
AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
ANNUAL SALES
(millions oC dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4,540 (10,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
45,400 (100,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454,000 (1 million)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.02
0.02
4.54 million (10 million)
0.00
0.00
0.00
0.00
0.01
0.03
0.05
0.06
0.07
0.08
0.11
22.7 million (50 million)
0.00
0.00
0.03
0.04
0.07 _
0.09
0.15
0.22
0.29
0.31
0.43
45.4 million (100 million)
0.00
0.00
0.06
0.07
0.08
0.15
0.27
0.43
0.57
0.65
0.94
227 million (500 million)
0.00
0.00
0.25
0.26
0.43
0.64
1.0
1.5
1.8
2.2
3.8
454 million (1 billion) 0.00 0.00 0.50 0.68 0.95 1.2 1.6 2.3 2.7 3.2 5.7
-------
EXHIBIT 4-6
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS
SALES AND PRODUCTION QUANTITY PARAMETERS
(location need only be below one parameter to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.00
0.50
0.70
0.90
1.2
1.8
2.5
3.3
4.2
8.3
454 (1,000)
0.00
0.00
0.50
0.70
0.90
1.2
1.8
2.5
3.3
4.2
8.3
4,540 (10,000)
0.00
0.00
0.50
0.70
0.90
1.2
1.8
2.5
3.3
4.2
8.3
45,400 (100,000)
0.00
0.00
0.50
0.70
1.0
1.2
1.8
2.5
3.3
4.2
8.3
454,000 (1 million)
0.10
0.10
0.60
0.80
1.0
1.3
1.9
2.6
3.4
4.2
8.3
4.54 million (10 million)
0.80
0.80
1.3
1.4
1.7
2.0
2.6
3.2
4.0
4.8
8.9
22.7 million (50 million)
5.0
5.0
5.5
5.7
5.9
6.2
6.7
7.3
8.0
8.9
12.9
45.4 million (100 million)
11.4
11.4
11.8
12.0
12.2
12.5
12.9
13.5
14.1
14.9
18.7
227 million (500 million)
52.8
52.8
53.0
53.2
53.3
53.4
53.6
53.8
54.2
54.7
57.2
454 million (1 billion)
75.4
75.4
75.4
75.4
75.4
75.4
75.6
75.6
76.0
76.4
77.9
-------
EXHIBIT 4-7
PERCENT OF TOTAL CHEMICALS MISSED FOR VARIOUS SALES
AND PRODUCTION QUANTITY PARAMETERS
(location must be below both parameters to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.00
0.38
1.66
1.95
2.53
3.83
4.08
5.16
5.45
6.30
4,540 (10,000)
0.00
0.00
0.46
1.87
2.63
3.70
5.48
5.95
7.20
7.90
9.05
45,400 (100,000)
0.00
0.02
0.53
2.22
3.26
4.92
7.14
7.95
9.50
10.52
12.74
454,000 (1 million)
0.00
0.02
0.55
2.30
3.43
5.40
7.75
8.76
10.52
11.72
14.40
4.54 million (10 million)
0.00
0.02
0.55
2.34
3.49
5.57
8.06
9.13
10.98
12.26
15.05
22.7 million (50 million)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.21
11.07
12.36
15.16
45.4 million (100 million)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.23
11.09
12.37
15.23
227 million (500 million)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.23
11.09
12.37
15.23
454 million (1 billion)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.23
11.09
12.37
15.23
-------
EXHIBIT 4-8
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS! SALES AND PRODUCTION QUANTITY PARAMETERSฎ/
(location must be below both parameters to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.05
0.39
1.65
1.98
2.61
3.98
4.42
5.47
5.73
6.42
4,540 (10,000)
0.00
0.06
0.66
2.20
3.12
4.55
6.57
7.23
8.45
9.09
10.18
45r400 (100,000)
0.00
0.11
0.96
3.08
4.76
7.15
10.01
11.24
12.84
13.90
15.58
454,000 (1 million)
0.00
0.11
1.16
3.64
5.75
8.92
12.39
14.11
16.15
17.49
19.78
4.54 million (10 million)
0.00
0.11
1.25
3.84
6.07
9.62
13.45
15.42
17.79
19.38
22.24
22.7 million (50 million)
0.00
0.14
1.34
3.96
6.28
9.88
13.87
16.07
18.65
20.34
23.52
45.4 million (100 million)
0.00
0.14
1.37
3.98
6.31
9.95
14.04
16.33
19.00
20.76
24.15
227 million (500 million)
0.00
0.14
1.43
4.05
6.43
10.13
14.34
16.79
19.56
21.44
25.29
454 million (1 billion)
0.00
0.14
1.46
4.10
6.50
10.22
14.42
16.90
19.70
21.59
25.56
>rt qu tc pr ปd] Nu ol at rซ ed
-------
EXHIBIT 4-9
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS t SALES AND PRODUCTION QUANTITY PARAMETERS!/
(Location need only be below one parameter to be exempted)
ANNUAL SALES
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds) 0.0 0.1 0.5 1.0 2.5 5.0 10 20 30 50 100
0 (0)
0.00
0.14
1.46
4.10
6.50
10.2
14.4
16.9
19.8
21.7
25.8
454 (1,000)
10.8
11.0
11.9
13.3
15.4
18.5
21.3
23.4
25.1
26.8
30.2
4,540 (10,000)
17.7
17.9
18.5
19.6
21.1
23.4
25.6
27.4
29.0
30.3
33.3
45,400 (100,000)
28.2
28.3
28.7
29.3
30.0
31.3
32.7
33.9
35.2
36.0
38.4
454,000 (1 million)
42.2
42.2
42.5
42.6
42.9
43.5
44.2
45.0
45.8
46.4
48.2
4.54 million (10 million)
57.7
57.8
57.9
58.0
58.2
58.3
58.7
59.2
59.7
60.0
61.3
22.7 million (50 million)
72.4
72.4
72.5
72.5
72.6
72.7
73.0
73.2
73.5
73.7
74.6
45.4 million (100 million)
79.6
79.6
79.7
79.7
79.8
79.9
80.0
80.2
80.3
80. 5.
81.2
227 million (500 million)
94.4
94.4
94.4
94.5
94.5
94.5
94.5
94.6
94.6
94.6
94.9
454 million (1 billion) 97.5 97.5 97.5 97.5 97.5 97.7 97.5 97.5 97.5 97.6 97.7
5/cost = (Number of reports required to be prepared) + (Number of locations required to report)
-------
EXHIBIT 4-10
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT
FOR VARIOUS SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
40
o
ฆn
o
o
I
30
Percent of
Information
Missing
20
10
45v4 million kg..
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,360 kg.(100,0001b.)
454 kg.(1000 lb.)
10Q0
-ฆ .L
10,000 10QfQQQ 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale*
-------
EXHIBIT 4-11
100
80
60
Percent of
Information
Missing
40
20
o
n
3
0
a
s
1
o
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT
FOR VARIOUS SALES AND PRODUCTION QUANTITY PARAMETERS
(Location need only be below one parameter to be exempted)
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,360 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
1000 10,00Q 100,Q0Q 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale)
-------
EXHIBIT 4-12
PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED
FOR VARIOUS SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
40
30
Percent of
Information
Missing
o
-n
z
o
0
30
s
1
20
10
45.4 million kg.
(10O million lbl)
4.54 million kg.
(10 million lb.)
45,360 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
10Q0 10,Q0Q 10QfQQQ 1 million 10 million 100 million
Annual Sale? Parameter in Dollars llog scale!
-------
EXHIBIT 4-13
100
80
60
Percent of
Information
Missing
40
20
8
0
30
3
1
o
PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED
FOR VARIOUS SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below only one parameter to be exempted)
*45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,360 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
1000 10,000 100,Q0Q 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale)
-------
EXHIBIT 4-14
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS SALES
AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be- exempted)
upper bound
\
454 million kg.
(1 billion lb.)
1227 million kg.
(9 00 million lb.)
22.7 milliop kg.
(50 million lb.)
10Q0
10,QOQ
10(1,000
1 million 10 million 100 million
Annual Sale? Parameter in Dollars Clog scale)
-------
EXHIBIT 4-15
100
80
60
Percent of
Information
Missing
40
20
o
0
39
s
1
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS SALES
AND PRODUCTION QUANTITY PARAMETERS
(Location need only be below one parameter to beexempted.)
upper bound
454 million kg!
*(1 billion lb.)
--lower bound
upper bound
227 million kg.
(500 million lb.)
lower bound
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
1000 10,000 100,QQQ 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale)
-------
EXHIBIT 4-16
PERCENT OF TOTAL CHEMICALS MISSED FOR VARIOUS
SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
20
15
Percent of
Chemicals
Missed
10
o
n
I
o
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,400 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
1000 10,000 100,000 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale)
-------
EXHIBIT 4-17
50
40
30
Percent
Reduction in
Total Cost
20
10
o
z
o
0
3
s
1
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS: SALES AND PRODUCTION QUANTITY PARAMETERSฎ./
(Location must be below both parameters to be exempted)
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,400 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
=ฆi
1000 10,00Q 100,QQ0 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale)
-------
EXHIBIT 4-18
100
Percent
Reduction in
Total Cost
80
60
40
20
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS: SALES AND PRODUCTION QUANTITY PARAMETERSฎ/
(Location need only be below one parameter to bp exempted)
-ปป-
45.4 million kg.
- (100 million lb.)
-t ซ 1-
.ป-
4.54 million kg.
(10 million lb.)
45,400 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
1000 10,000 100,000 1 million 10 million 100 million
Annual Sales Parameter in Dollars (log scale)
a/Post
'NurK~- Of * Dr t ~ ~ "ปqu1 1 tO-hซ* ororsrod1 Of; InmMnna n'red re' ')
-------
EXHIBIT 4-19
PRODUCTION QUANTITY
kilograms (pounds)
0 (0)
454 (1,000)
4,540 (10,000)
45,400 (100,000)
454,000 (1 million)
4.54 million (10 million)
22.7 million (50 million)
45.4 million (100 million)
227 million (500 million)
454 million (1 billion)
CHANGE IN THE PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT
PER A ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location must be below both parameters to be exempted)
SHIFT OF ANNUAL SALES PARAMETER '
(millions of dollars)
0 0.1 0.5 1.0 2.5 5.0 10 20 30 50
.1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.40
0.85
0.58
0.14
0.05
0.04
0.02
0.02
0.00
0.00
0.4
1.2
1.1
0.47
0.32
0.06
0.03
0.03
0.00
0.01
1.7
2.3
2.8
1.1
0.68
0.24
0.07
0.05
0.02
0.01
1.7
3.3
4.2
1.7
1.2
0.40
0.13
0.09
0.04
0.02
1.7
3.6
4.8
1.9
1.5
0.56
0.18
0.13
0.06
0.04
2.5
3.9
4.8
2.1
1.5
0.58
0.22
0.18
0.07
0.04
2ป5 3.9 4.8 2.1 1.6 0.62 0.23 0.19 0.07 0.05
2ป5 4.1 4.8 2.1 1.6 0.68 0.26 0.21 0.09 0.06
5.0 2.1 1.6 0.68 0.27 0.21 0.09 0.06
-------
EXHIBIT 4-20
CHANGE IN THE PERCENT OP TOTAL LOCATIONS NOT REQUIRED TO REPORT
PER A ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location need only be below one parameter to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions oE dollars)
PRODUCTION QUANTITY 0 0.1 0.5. 1.0 2.5 5.0 10 20 30 50
kilograms (pounds) to .1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0 (0)
2.5
4.1
5.0
2.1
1.6
0.70
0.26
0.22
0.10
0.07
454 (1,000)
2.0
3.2
4.2
2.1
1.6
0.66
0.23
0.22
0.09
0.07
41540 (10,000)
3.0
2.7
3.8
1.7
1.3
0.62
0.26
0.17
0.09
0.03
45,400 (100,000)
1.0
2.0
2.0
1.1
0.92
0.44
0.20
0.17
0.08
0.06
454,000 (1 million)
1.0
0.75
0.80
0.40
0.44
0.28
0.13
0.14
0.06
0.05
4.54 million (10 million)
1.0
0.50
0.20
0.27
0.08
0.16
0.08
0.09
0.04
0.03
22.7 million (50 million)
0.00
0.25
0.2
0.07
0.08
0.10
0.05
0.05
0.02
0.02
45.4 million (100 million)
0.00
0.25
0.00
0.13
0.08
0.06
0.04
0.03
0.02
0.02
227 million (500 million)
0.00
0.25
0.00
0.07
0.00
0.02
0.00
0.01
0.01
0.01
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.00
0.01
0.00
-------
EXHIBIT 4-21
CHANGE IN THE PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED
PER A ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location must be below both parameters to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds)
0
to .1
0.1
to 0.5
0.5
to 1.0
1.0
to 2.5
2.5
to 5.0
5.0
to 10
10
to 20
20
to 30
30
to 50
50
to 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1?000)
0.50
0.85
3.0
0.20
0.32
0.32
0.05
0.13
0.02
0.02
4,540 (10,000)
0.60
1.6
3.4
0.67
0.64
0.48
0.08
0.14
0.04
0.03
45,400 (100,000)
1.0
2.1
4.5
1.1
1.0
0.66
0.13
0.19
0.06
0.04
454,000 (1 million)
1.0
2.5
5.0
1.3
1.3
0.76
0.18
0.23
0.08
0.05
4.54 million (10 million)
1.0
2.7
5.2
1.4
1.4
0.82
0.20
0.26
0.09
0.06
22.7 million (50 million)
1.2
2.7
5.4
1.4
1.4
0.84
0.22
0.28
0.09
0.07
45.4 million (100 million)
1.2
2.7
5.4
1.5
1.4
0.88
0.23
0.28
0.09
0.07
227 million (500 million)
1.2
3.0
5.4
1.5
1.4
0.88
0.24
0.29
0.10
0.08
454 million (1 billion)
1.2
3.2
5.2
1.5
1.5
0.88
0.24
0.29
0.10
0.08
-------
EXHIBIT 4-22
CHANGE IN THE PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED
PER A ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(Location need only be below one parameter to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions of dollars)
PRODUCTION QUANTITY 0 0.1 0.5 1.0 2.5 5.0 10 20 30 50
kilograms (pounds) to .1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0 (0)
1.2
3.2
5.2
1.5
1.5
0.88
0.24
0.30
0.10
0.08
454 (1,000)
1.0
2.1
2.4
1.2
1.2
0.56
0.20
0.17
0.08
0.07
4,540 (10,000)
1.0
1.5
1.8
0.87
0.80
0.40
0.17
0.15
0.06
0.06
45,400 (100,000)
0.00
1.0
0. B0
0.33
0.44
0.24
0.11
0.11
0.04
0.05
454,000 (1 million)
0.00
0.50
0.40
0.13
0.16
0.12
0.06
0.07
0.02
0.04
4.54 million (10 million)
0.00
0.50
0.00
0.13
0.04
0.06
0.05
0.03
0.02
0.02
22.7 million (50 million)
0.00
0.25
0.20
0.00
0.04
0.04
0.02
0.02
0.01
0.02
45.4 million (100 million)
0.00
0.25
0.00
0.07
0.04
0.00
0.02
0.01
0.01
0.04
227 million (500 million)
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.01
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.00
-------
EXHIBIT 4-23
CHANGE IN PERCENT OF TOTAL PRODUCTION NOT COVERED PER A
ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(Location must be below both parameters to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions of dollars)
PRODUCTION QUANTITY 0 0.1 0.5 1.0 2.5 5.0 10 20 30 50
kilograms (pounds) to .1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4,540 (10,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
45,400 (100,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454,000 (1 million)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4.54 million (10 million)
0.00
0.00
0.00
0.01
0.01
0.00
0.00
0.00
0.00
0.00
22.7 million (50 million)
0.00
0.07
0.02
0.02
0.01
0.01
0.01
0.00
0.00
0.00
15.4 million (100 million)
0.00
0.15
0.02
H
O
O
0.03
0.02
0.02
0.01
0.00
0.01
227 million (500 million)
0.00
0.62
0.02
0.11
0.08
0.07
0.05
0.01
0.04
0.03
454 million (1 billion) 0.00 1.2 0.36 0.18 0.10 0.08 0.07 0.04 0.02 0.05
-------
EXHIBIT 4-24
PRODUCTION QUANTITY
kilograms (pounds)
0 (0)
454 (1,000)
4,540 (10,000)
45,400 (100,000)
454,000 (1 million)
4.54 million (10 million)
22.7 million (50 million)
45.4 million (100 million)
227 million (500 million)
454 million (1 billion)
CHANGE IN PERCENT OF TOTAL PRODUCTION NOT COVERED PER A
ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location need only be below one parameter to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions oฃ dollars)
0 0.1 0.5 1.0 2.5 5.0 10 20 30 50
to .1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0.00
1.2
0.40
0.13
0.12
0.12
0.07
0.08
.0.05
0.08
0.00
1.2
0.40
0.13
0.12
0.12
0.07
0.08
0.05
0.08
0.00
1.2
0.40
0.13
0.12
0.12
0.07
0.08
0.05
0.08
0.00
1.2
0.40
0.20
0.08
0.12
0.07
0.08
0.05
0.08
0.00
1.2
0.40
0.13
0.12
0.12
0.07
0.08
0.04
0.08
0.00
1.2
0.20
0.20
0.12
0.12
0.06
0.08
0.04
0.08
0.00 1.2 0.40 0.13 0.12 0.10 0.06 0.07 0.05 0.08
0.00 1.0 0.40 0.13 0.12 0.08 0.06 0.06 0.04 0.08
0.00
0.50
0.40
0.07
0.04
0.04
0.02
0.04
0.03
0.05
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.04
0.02
0.01
-------
EXHIBIT 4-25
CHANGE IN THE PERCENT OF TOTAL CHEMICALS MISSED PER A
ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location must be below both parameters to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions of dollars)
PRODUCTION QUANTITY 0 0.1 0.5 1.0 2.5 5.0 10 20 30 50
kilograms (pounds) to .1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.95
2.6
0.19
0.23
0.26
0.02
0.11
0.01
0.02
4,540 (10,000)
0.00
1.1
2.8
0.51
0.43
0.36
0.05
0.12
0.04
0.02
45,400 (100,000)
0.20
1.3
3.4
0.69
0.66
0.44
0.08
0.15
0.05
0.04
454,000 (1 million)
0.20
1.3
3.5
0.75
0.79
0.47
0.10
0.18
0.06
0.05
4.54 million (10 million)
0.20
1.3
3.6
0.77
0.83
0.50
0.11
0.19
0.06
0.06
22.7 million (50 million)
0.20
1.3
3.6
0.78
0.84
0.50
0.11
0.19
0.06
0.06
45.4 million (100 million)
0.20
1.3
3.6
0.78
0.84
0.50
0.11
0.19
0.06
0.06
227 million (500 million)
0.20
1.3
3.6
0.78
0.84
0.50
0.11
0.19
0.06
0.06
454 million (1 billion)
0.20
1.3
3.6
0.78
0.84
0.52
0.11
0.19
0.06
0.06
-------
EXHIBIT 4-26
CHANGE IN PERCENT REDUCTION IN TOTAL COST TO INDUSTRY PER A
ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location must be below both parameters to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions of dollars)
PRODUCTION QUANTITY 0 0.1 0.5 1.0 2.5 5.0 10 20 30 50
kilograms (pounds) to .1 to 0.5 to 1.0 to 2.5 to 5.0 to 10 to 20 to 30 to 50 to 100
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.50
0.85
2.5
0. 22
0.25
0.27
0.04
0.10
0.01
0.01
4,540 (10,000)
0.60
1.5
3.1
0.61
0.57
0.44
0.06
0.12
0.03
0.02
45,400 (100,000)
1.1
2.1
4.2
1.1
0.96
0.57
0.12
0.16
0.05
0.03
454,000 (1 million)
1.1
2.6
5.0
1.4
1.3
0.70
0.17
0.20
0.07
0.05
4.54 million (10 million)
1.1
2.8
5.2
1.5
1.4
0.77
0.20
0.24
0.08
0.06
22.7 million (50 million)
1.4
3.0
5.2
1.5
1.4
0.80
0.22
0.26
0.08
0.06
45.4 million (100 million)
1.4
3.0
5.2
1.5
1.5
0.81
0.23
0.27
0.09
0.07
227 million (500 million)
1.4
3.2
5.2
1.6
1.5
0.84
0.24
0.28
0.09
0.08
454 million (1 billion)
1.4
3.3
5.3
1.6
1.5
0.84
0.25
0.28
0.09
0.08
-------
EXHIBIT 4-27
CHANGE IN PERCENT REDUCTION IN TOTAL COST TO INDUSTRY PER A
ONE MILLION DOLLAR INCREASE IN THE SALES PARAMETER
(location need only be below one parameter to be exempted)
SHIFT OF ANNUAL SALES PARAMETER
(millions of dollars)
PRODUCTION QUANTITY
kilograms (pounds)
0
to . 1
0.1
to 0.5
0.5
to 1.0
1.0
to 2.5
2.5
to 5.0
5.0
to 10
10
to 20
20 30 50
to 30 to 50 to 100
0 (0)
1.4
3.3
5.3
1.6
1.5
0.84
0.25
0.29
0.10
0.08
454 (1,000)
2.0
2.25
2.8
1.4
1.2 .
0.56
0.21
0.17
0.08
0.07
4,540 (10,000)
2.0
1.5
2.2
1.0
0.92
0.44
0.18
0.16
0.07
0.06
45,400 (100,000)
1.0
1.0
1.2
0.47
0.52
0.28
0.12
0.13
0.04
0.05
454,000 (1 million)
0.00
0.75
0.20
0.20
0.24
0.14
0.08
0.08
0.03
0.04
4.54 million (10 million)
1.0
0.25
0.20
0.13
0.04
0.08
0.05
0.05
0.02
0.03
22.7 million (50 million)
0.00
0.25
0.00
0.07
0.04
0.06
0.02
0.03
0.01
0.02
45.4 million (100 million)
0.00
0.25
0.00
0.07
0.04
0.02
0.02
0.01
0.01
0.01
227 million (500 million)
0.00
0.00
0.02
0.00
0.00
0.00
0.01
0.00
0.00
0.01
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-------
EXHIBIT 4-28
LOWER BOUND ON THE PERCENT OF CHEMICALS THAT WOULD
BE MISSED AS A FUNCTION OF THE PRODUCTION QUANTITY PARAMETER
WHEN LOCATIONS NEED ONLY BE BELOW ONE PARAMETER TO BE EXEMPTED
Production Quantity Percent of Chemicals
kg. (lbs.) That Would Be Missed (%)
0 (0)
0.0
454 (1,000)
12.5
4,540 (10,000)
18.8
45,400 (100,000)
31.6
454,000 (1 million)
49.3
4.54 million (10 million)
67.9
22.7 million (50 million)
80.0
45.4 million (100 million)
84.5
227 million (500 million)
92.8
454 million (1 billion)
95.4
ICF INCORPORATED
-------
EXHIBIT 4-29
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
TOTAL EMPLOYMENT
PRODUCTION QUANTITY
kilograms (pounds) 1 4 6 11 21 51 101
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.08
0.17
0.54
0.83
1.0
1.2
1.3
4,540 (10,000)
0.00
0.08
0.25
0.92
1.4
2.5
3.0
3.2
45,400 (100,000)
0.00
0.13
0.50
1.9
3.4
6.2
7.2
7.8
454,000 (1 million)
0.00
0.34
1.0
3.0
5.4
9.4
11.3
12.5
4.54 million (10 million)
0.00
0.38
1.3
3.6
6.4
11.2
13.8
15.5
22.7 million (50 million)
0.00
0.46
1.4
3.9
6.9
11.9
14.8
17.0
45.4 million (100 million)
0.00
0.46
1.4
3.9
6.9
12.2
15.1
17.5
227 million (500 million)
0.00
0.46
1.4
3.9
6.9
12.3
15.6
18.7
454 million (1 billion) 0.00 0.46 1.4 3.9 6.9 12.4 15.8 18.9
-------
EXHIBIT 4-30
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(location need only be below one parameter to be exempted)
TOTAL EMPLOYMENT
PRODUCTION QUANTITY
kilograms (pounds) 1 4 6 11 21 51 101 201
0 (0)
0.00
0.46
1.4
3.9
7.0
12.4
15.8
19.0
454 (1,000)
2.8
3.1
4.0
6.1
8.9
14.1
17.4
20.5
4,540 (10,000)
6.7
7.1
7.9
9.7
12.3
16.6
19.6
22.5
45,400 (100,000)
16.1
16.4
16.9
18.0
19.6
22.3
24.7
27.3
454,000 (1 million)
31.5
31.6
31.9
32.4
33.1
34.5
36.0
38.0
4.54 million (10 million)
47.9
48.0
48.0
48.2
48.5
49.2
50.0
51.4
22.7 million (50 million)
64.0
64.0
64.0
64.0
64.1
64.5
65.0
66.0
45.4 million (100 million)
71.3
71.3
71.3
71.4
71.4
71.6
72.0
72.8
227 million (500 million)
89.9
89.9
89.9
90.0
90.0
90.1
90.1
90.3
454 million (1 billion)
95.1
95.1
95.1
95.1
95.1
95.1
95.1
95.2
-------
EXHIBIT 4-31
PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(location must be below both parameters to be exempted)
TOTAL EMPLOYMENT
PRODUCTION QUANTITY
kilograms (pounds) 1 4 6 11 21 51
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.07
0.35
1.2
2.3
4.0
5.7
6.4
4,540 (10,000)
0.00
0.08
0.54
1.6
3.2
6.4
8.7
9.8
45,400 (100,000)
0.00
0.18
0.77
2.2
4.7
9.4
12.2
13.9
454,000 (1 million)
0.00
0.23
1.0
2.7
5.6
10.9
14.4
16.7
4.54 million (10 million)
0.00
0.26
1.1
3.0
5.9
11.7
15.4
18.0
22.7 million (50 million)
0.00
0.28
1.1
3.0
6.1
11.9
15.8
18.7
45.4 million (100 million)
0.00
0.28
1.1
3.0
6.1
12.0
16.0
19.1
227 million (500 million)
0.00
0.28
1.1
3.1
6.1
12.1
16.2
19.5
454 million (1 billion) 0.00 0.28 1.1 3.1 6.2 12.2 16.2 19.6
-------
EXHIBIT 4-32
PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(location need only be below one parameter to be exempted)
TOTAL EMPLOYMENT
PRODUCTION QUANTITY
kilograms (pounds)
11
21
51
454 million (1 billion) 98.0
98.0
98.0
98.0
98.0
101
201
0 (0)
0.00
0.28
1.6
3.1
6.2
12.2
16.2
19.7
454 (1,000)
12.7
13.0
13.5
14.7
16.6
20.9
23.3
26.0
4,540 (10,000)
20.2
20.4
20.8
21.8
23.2
26.0
27.8
30.1
45,400 (100,000)
31.1
31.2
31.5
32.0
32.5
33.9
35.1
36.9
.454,000 (1 million)
44.7
44.7
44.8
45.1
45.3
45.9
46.5
47.6
4.54 million (10 million)
60.0
60.0
60.0
60.2
60.3
60.5
60.8
61.6
22.7 million (50 million)
74.3
74.3
74.3
74.4
74.4
74.6
74.8
75.3
45.4 million (100 million)
81.5
81.5
81.5
81.6
81.6
81.7
81.8
82.1
227 million (500 million)
95.5
95.5
95.5
95.5
95.5
95.6
95.5
95.6
98.0
98.0
98.1
-------
EXHIBIT 4-33
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Locations must be below both parameters to be exempted)
TOTAL EMPLOYMENT
PRODUCTION QUANTITY
kilograms (pounds) 1 4 6 11 21 51 101 201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4,540 (10,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
45,400 (100,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454,000 (1 million)
0.00
0.00
0.00
0.00
0.00
0.01
0.02
0.02
4.54 million (10 million)
0.00
0.00
0.00
0.01
0.02
0.04
0.06
0.07
22.7 million (50 million)
0.00
0.00
0.00
0.01
0.06
0.12
0.20
0.27
45.4 million (100 million)
0.00
0.00
0.00
0.04
0.07
0.19
0.30
0.60
227 million (500 million)
0.00
0.00
0.00
0.21
0.24
0.50
0.93
1.9
454 million (1 billion)
0.00
0.00
0.00
0.47
0.59
1.0
1.4
2.8
-------
EXHIBIT 4-34
PRODUCTION QUANTITY-
kilograms (pounds)
0 (0)
454 (1,000)
4,540 (10,000)
45,400 (100,000)
454,000 (1 million)
4.54 million (10 million)
22.7 million (50 million)
45.4 million (100 million)
227 million (500 million)
454 million (1 billion)
PERCENT OP TOTAL PRODUCTION NOT COVERED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(location need only be below one parameter to be exempted)
TOTAL EMPLOYMENT
ซ 11 21 51 101 201
0.00
0.00
0.00
0.50
0.60
1.0
1.7
3.3
0.00
0.00
0.00
0.50
0.60
1.0
1.7
3.3
0.00 0.00 0.00 0.50 0.60 1.0 1.7 3.3
0.00 0.00 0.00 0.50 0.60 1.0 1.7 3.3
0.10
0.10
0.10
0.50
0.70
1.1
1.7
3.3
0.80
0.80
0.80
1.2
1.3
1.7
2.4
4.0
5.0
5.0
5.0
5.5
5.5
5.9
6.5
8.0
11.4
11.4
11.4
11.8
11.9
12.2
12.8
14.0
52.8
52.8
52.8
53.0
53.1
53.3
53.5
54.1
75.4
75.4
75.4
75.4
75.4
75.4
75.6
75.9
-------
EXHIBIT 4-35
PERCENT OP TOTAL CHEMICALS MISSED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(location must be below both parameters to be exempted)
PRODUCTION QUANTITY
kilograms (pounds)
TOTAL EMPLOYMENT
6 11
21
51
454 million (1 billion) 0.00
101
201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.00
0.38
1.66
1.95
2.53
3.83
4.08
4,540 (10,000)
0.00
0.00
0.46
1.87
2.63
3.70
5.48
5.95
45,400 (100,000)
0.00
0.02
0.53
2.22
3.26
4.92
7.14
7.95
454,000 (1 million)
0.00
0.02
0.55
2.30
3.43
5.40
7.75
8.76
4.54 million (10 million)
0.00
0.02
0.55
2.34
3.48
5.57
8.06
9.13
22.7 million (50 million)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.23
45.4 million (100 million)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.23
227 million (500 million)
0.00
0.02
0.55
2.35
3.52
5.61
8.12
9.23
0.02
0.55
2.35
3.52
5.61
8.12
9.23
-------
EXHIBIT 4-36
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS t EMPIX)YMENT AND PRODUCTION QUANTITY PARAMETERSฎ/
(Location must be below both parameters to be .exempted)
PRODUCTION QUANTITY
kilograms (pounds)
TOTAL EMPLOYMENT
6 11
21
51
101
201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.07
0.32
1.04
2.01
3.47
4.84
5.48
4,540 (10,000)
0.00
0.08
0.49
1.48
2.90
5.71
7.60
8.58
45,400 (100,000)
0.00
0.17
0.72
2.17
4.49
8.77
11.3
12.73
454,000 (1 million)
0.00
0.25
1.02
2.80
5.52
10.6
13.8
15.9
4.54 million (10 million)
0.00
0.28
1.16
3.09
6.00
11.6
15.1
17.6
22.7 million (50 million)
0.00
0.32
1.19
3.20
6.23
11.9
15.6
18.4
45.4 million (100 million)
0.00
0.32
1.19
3.20
6.23
12.0
15.8
18.8
227 million (500 million)
0.00
0.32
1.19
3.25
6.28
12.2
16.1
19.4
454 million (1 billion)
0.00
0.32
1.19
3.28
6.32
12.2
16.1
19.5
a/cost = (Number of reports required to be prepared) + (Number of locations rpaiilred to reiiort)
-------
EXHIBIT 4-37
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS: EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERSฎ/
(Location need only be below one parameter to be exempted)
TOTAL EMPLOYMENT
PRODUCTION QUANTITY
kilograms (pounds) 1 4 6
0 (0)
0.00
0.32
1.19
3.28
6.32
12.2
16.2
19.6
454 (1,000)
10.8
11.1
11.7
13.1
15.2
19.6
22.2
24.9
4,540 (10,000)
17.7
17.9
18.4
19.5
21.1
24.2
26.3
28.7
45,400 (100,000)
28.2
28.4
28.7
29.3
30.1
31.7
33.1
35.1
454,000 (1 million)
42.2
42.2
42.4
42.7
43.0
43.8
44.5
45.8
4.54 million (10 million)
57.7
57.8
57.8
57.9
58.0
58.4
58.8
59.7
22.7 million (50 million)
72.4
72.4
72.4
72.5
72.5
72.7
72.9
73.5
45.4 million (100 million)
79.6
79.6
79.6
79.7
79.7
79.8
80.0
80.4
227 million (500 million)
94.4
94.4
94.4
94.4
94.5
94.5
94.5
94.6
454 million (1 billion)
97.5
97.5
97.5
97.5
97.5
97.5
97.5
97.5
ง/cost = (Number of reports required to be prepared) + (Number of locations required to report)
-------
EXHIBIT 4-38
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location tnust be below both parameters to beexempted.)
Percent of
Information
Missing
o
3
0
a
s
1
o
40
30
20
10
t ^ ^ %
10 32 ioo
Total Employees Parameter (log scale)
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,360 kg.
* (100,000 lb.)
454 kg.
(1000 lb.)
I
316
-------
EXHIBIT 4-39
100
80
Percent of
Information
Missing
60
40
20
PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT FOR VARIOUS
EMPLOYMENT AND PRODUCTION OUANTITY PARAMETERS
(Location need only be below one parameter to be -exempted}
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
o
0
3D
S
1
45,360 kg.
(100r000 lb.)
454 kg.
(1000 lb.)
X
10 32
Total Employees Parameter
: I
100
(log scale)
316
-------
EXHIBIT 4-40
PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
40
Percent of
Information
Missing
30
20
10
o
0
a
s
1
o
45.4 million kg.
(100 million lbi)
4.54-million kg.
(10 million lb.)
45,360 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
10 32 100
Total Employees Parameter (log scale)
316
-------
EXHIBIT 4-41
100
80
Percent of
Information
Missing
60
40
20
o
0
3D
s
1
PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location need only be below one parameter to be exempted)
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,360 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
ฑ
10 32 100
Total Employees Parameter (log scale)
316
-------
EXHIBIT 4-42
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be .exempted)
* upper bound
/
454 million kg.
(1 billion lb.)
227 million kg.
(500 million lb.)
22.7 million kg.
(50 million lb.)
3
10
32
100
316
Total Employees Parameter (log sealei
-------
EXHIBIT 4-43
100
80
Percent of
Information
Missing
60
40
20
o
PERCENT OF TOTAL PRODUCTION NOT COVERED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location need only be below one parameter to be exempted)
upper bound
454 million kg.
(1 billion lb.)
lower bound
upper bound
227 million kg.
(500 million lb.)
lower bound
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
o
o
30
s
s
o
10 32 100
Total Employees Parameter (log scale)
316
-------
EXHIBIT 4-44
PERCENT OF TOTAL CHEMICALS MISSED FOR VARIOUS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted
20
Percent of
Chemicals
Missed
15
o
n
0
30
s
1
o
10
45.4 million kg.
(100 million lb.)
J L
10 32 loo
Total Employees Parameter (log scale)
4.54 million kg.
(10 million lb.)
45,400 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
316
-------
EXHIBIT 4-45
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS: EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERSฎ/
(Location must be below both parameters to be exempted)
40
30
Percent
Reduction in
Total Cost
20
10
ฆ - t rSsz.
10 32 100
Total Employees Parameter (iog scale)
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,400 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
316
a/Cost = (Number of reports required to be prepared) + (Number of locations required to report)
-------
EXHIBIT 4-46
PERCENT REDUCTION IN TOTAL COST TO INDUSTRY FOR VARIOUS SMALL MANUFACTURER
EXEMPTIONS: EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERSฎ/
(Location need only be below one parameter to be exempted)
Percent
Porlnrtion in
Total Cost
ฎ/ Cost
80
60
40
20
* 0 -
(
10 32 100
Total Employees Parameter (log scale}
45.4 million kg.
(100 million lb.)
4.54 million kg.
(10 million lb.)
45,400 kg.
(100,000 lb.)
454 kg.
(1000 lb.)
316
(Number of reports required to be prepared) + (Number of locations required tb report)
-------
EXHIBIT 4-47
CHANGE IN PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT
PER A TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(location must be below both parameters to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 ซ to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.27
0.45
0.74
0.29
0.06
0.04
.01
4,540 (10,000)
0.27
0.85
1.3
0.48
0.37
0.10
0.02
45,400 (100,000)
0.43
1.8
2.8
1.5
0.93
0.20
0.06
454,000 (1 million)
1.1
3.3
4.0
2.4
1.3
0.38
0.12
4.54 million (10 million)
1.3
4.6
4.6
2.8
1.6
0.52
0.17
22.7 million (50 million)
1.5
4.7
5.0
3.0
1.7
0.58
0.22
45.4 million (100 million)
1.5
4.7
5.0
3.0
1.8
0.58
0.24
227 million (500 million)
1.5
4.7
5.0
3.0
1.8
0.66
0.31
454 million (1 billion)
1.5
4.7
5.0
3.0
1.8
0.68
0.31
-------
EXHIBIT 4-48
CHANGE IN PERCENT OF TOTAL LOCATIONS NOT REQUIRED TO REPORT
PER A TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location need only be below one parameter to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
1.5
4.7
5.0
3.1
1.8
0.68
0.32
454 (1,000)
1.0
4.5
4.2
2.8
1.7
0.66
0.31
4,540 (10,000)
1.3
4.0
3.6
2.6
1.4
0.60
0.29
45,400 (100,000)
1.0
2.5
2.2
1.6
0.90
0.48
0.26
454,000 (1 million)
0.33
1.5
1.0
0.70
0.47
0.30
0.20
4.54 million (10 million)
0.33
0.00
0.40
0.30
0.23
0.16
0.14
22.7 million (50 million)
0.00
0.00
0.00
0.10
0.13
0.10
0.10
45.4 million (100 million)
0.00
0.00
0.20
0.00
0.07
0.08
0.08
227 million (500 million)
0.00
0.00
0.20
0.00
0.03
0.00
0.02
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.00
0.01
-------
EXHIBIT 4-49
CHANGE IN PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED
PER A TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location must be below both parameters to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 10L to 201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.23
1.4
1.7
1.1
0.57
0.34
0.07
4,540 (10,000)
0.27
2.3
2.1
1.6
1.1
0.46
0.11
45,400 (100,000)
0.60
3.0
2.9
2.5
1.6
0.56
0.17
454,000 (1 million)
0.77
3.8
3.4
2.9
1.8
0.70
0.17
4.54 million (10 million)
0.87
4.2
3.8
2.9
1.9
0.74
0.26
22.7 million (50 million)
0.93
4.1
3.8
3.1
1.9
0.78
0.29
45.4 million (100 million)
0.93
4.1
3.8
3.1
2.0
0.80
0.31
227 million (500 million)
0.93
4.1
4.0
3.0
2.0
0.82
0.33
454 million (1 billion)
0.93
4.1
4.0
3.1
2.0
0.80
0.34
-------
EXHIBIT 4-50
CHANGE IN PERCENT OF TOTAL REPORTS NOT REQUIRED TO BE PREPARED
PER A TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location need only be below one parameter to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
0.93
6.6
3.0
3.1
2.0
0.80
0.35
454 (1,000)
1.0
2.5
2.4
1.9
1.4
0.48
0.27
4,540 (10,000)
0.67
2.0
2.0
1.4
0.93
0.36
0.23
45,400 (100,000)
0.33
1.5
1.0
0.50
0.47
0.24
0.18
454,000 (1 million)
0.00
0.50
0.60
0.20
0.20
0.12
0.11
4.54 million (10 million)
0.00
0.00
0.40
0.10
0.07
0.06
0.08
22.7 million (50 million)
0.00
0.00
0.20
0.00
0.07
0.04
0.05
45.4 million (100 million)
0.00
0.00
0.20
0.00
0.03
0.02
0.03
227 million (500 million)
0.00
0.00
0.00
0.00
0.00
0.00
0.01
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.00
0.01
-------
EXHIBIT 4-51
CHANGE IN PERCENT OF TOTAL PRODUCTION NOT COVERED
PER A TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location must be below both parameters to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4,540 (10,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
45,400 (100,000)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454,000 (1 million)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
4.54 million (10 million)
0.00
0.00
0.02
0.01
0.01
0.00
0.00
22.7 million (50 million)
0.00
0.00
0.04
0.04
0.02
0.02
0.01
45.4 million (100 million)
0.00
0.00
0.08
0.03
0.04
0.02
0.03
227 million (500 million)
0.00
0.00
0.42
0.03
0.09
0.09
0.10
454 million (1 billion)
0.00
0.00
0.94
0.12
0.14
0.08
0.14
-------
EXHIBIT 4-52
CHANGE IN PERCENT OF TOTAL PRODUCTION NOT COVERS)
PER A TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location need only be below one parameter to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
0.00
0.00
1.0
0.10
0.13
0.14
0.16
454 (1,000)
0.00
0.00
1.0
0.10
0.13
0.14
0.16
4,540 (10,000)
0.00
0.00
1.0
0.10
0.13
0.14
0.16
45,400 (100,000)
0.00
0.00
1.0
0.10
0.13
0.14
0.16
454,000 (1 million)
0.00
0.00
0.80
0.20
0.13
0.12
0.16
4.54 million (10 million)
0.00
0.00
0.80
0.10
0.13
0.14
0.16
22.7 million (50 million)
0.00
0.00
1.0
0.00
0.13
0.12
0.15
45.4 million (100 million)
0.00
0.00
0.80
0.10
0.10
0.12
0.12
227 million (500 million)
0.00
0.00
0.40
0.10
0.07
0.04
0.06
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.04
0.03
-------
EXHIBIT 4-53
CHANGE IN THE PERCENT OF TOTAL CHEMICALS MISSED PER A
TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location must be below both parameters to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.00
1.9
2.6
0.29
0.20
0.26
0.01
4,540 (10,000)
0.00
2.3
2.8
0.76
0.36
0.36
0.05
45,400 (100,000)
0.07
2.5
3.4
1.0
0.55
0.44
0.08
454,000 (1 million)
0.07
2.6
3.5
1.1
0.66
0.47
0.10
4.54 million (10 million)
0.07
2.6
3.6
1.1
0.70
0.50
0.11
22.7 million (50 million)
0.07
2.6
3.6
1.2
0.70
0.50
0.11
45.4 million (100 million)
0.07
2.6
3.6
1.2
0.70
0.50
0.11
227 million (500 million)
0.07
2.6
3.6
1.2
0.70
0.50
0.11
454 million (1 billion)
0.07
2.6
3.6
1.2
0.70
0.50
0.11
-------
EXHIBIT 4-54
CHANGE IN PERCENT REDUCTION IN TOTAL COST TO INDUSTRY PER A
TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location must be below both parameters to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
454 (1,000)
0.23
1.2
1.4
0.97
0.49
0.27
0.06
4,540 (10,000)
0.27
2.0
2.0
1.4
0.94
0.38
0.10
45,400 (100,000)
0.57
2.7
2.9
2.3
1.4
0.51
0.14
454,000 (1 million)
0.83
3.8
3.6
2.7
1.7
0.64
0.21
4.54 million (10 million)
0.93
4.4
3.9
2.9
1.9
0.70
0.25
22.7 million (50 million)
1.1
4.3
4.0
3.0
1.9
0.74
0.28
45.4 million (100 million)
1.1
4.3
4.0
3.0
1.9
0.74
0.30
227 million (500 million)
1.1
4.3
4.1
3.0
2.0
0.78
0.33
454 million (1 billion)
1.1
4.3
4.2
3.0
2.0
0.78
0.33
-------
EXHIBIT 4-55
CHANGE IN PERCENT REDUCTION IN TOTAL COST TO INDUSTRY PER A
TEN PERSON INCREASE IN THE EMPLOYMENT PARAMETER
(Location need only be below one parameter to be exempted)
SHIFT OF TOTAL EMPLOYMENT PARAMETER
PRODUCTION QUANTITY
kilograms (pounds) 1 to 4 4 to 6 6 to 11 11 to 21 21 to 51 51 to 101 101 to 201
0 (0)
1.1
4.3
4.2
3.0
2.0
0.80
0.34
454 (1,000)
1.0
3.0
2.8
2.1
1.5
0.52
0.27
4,540 (10,000)
0.67
2.5
2.2
1.6
1.0
0.42
0.24
45,400 (100,000)
0.67
1.5
1.2
0.80
0.53
0.28
0.20
454,000 (1 million)
0.00
1.0
0.60
0*30
0.27
0.14
0.13
4.54 million (10 million)
0.33
0.00
0.20
0.10
0.13
0.08
0.09
22.7 million (50 million)
0.00
0.00
0.20
0.00
0.07
0.04
0.06
15.4 million (100 million)
0.00
0.00
0.20
0.00
0.03
0.04
0.04
227 million (500 million)
0.00
0.00
0.00 .
0.10
0.00
0.00
0.01
454 million (1 billion)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-------
EXHIBIT 4-56
Annual Sales
(millions of dollars)
(log, scale)
100
ISOQUANTS SHOWING THE PERCENT OF TOTAL LOCATIONS
NOT REQUIRED TO REPORT AS A FUNCTION OF THE
SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to' be exempted)
10
\
\
\
\
\
5%
\
\
X
10%^.
\
.01
.0011
0.0
(0.0)
X
45.4
(100)
1-
454
(1000)
4-
4.
4-
J.
4-
4-
4540 45,400 454,000 4.54 million 45.4 million 454 million
(10,000) (100,000) (1 million)(10 million) (100 million)(1 biJiinn)
-------
EXHIBIT 4-57
Annual Sales
nillions of dollars)
(log scale)
100
ISOQUANTS SHOWING THE PERCENT REDUCTION IN TOTAL COST
TO INDUSTRY AS A FUNCTION OF THE SALES AND
PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
10
1 -
.1
.01-
0011 1 1 I ซ ' I t 1 ' ! 1 1 ฆi -.. I... r, ; , i ป ฆ . -t. . i
ฐ*ฐ 45-4 454 4540 45,400 454*000 4.54 million 45.4 million 454 million
(0.0) (100) (1000) (10,000) (100,000) (1 million)(10 million) (100 million)(1 billion)
Production Quantity kg. (lb.)
(log scale)
-------
EXHIBIT 4-58
LOCATION AND COST ISOQUANTS GRAPHED TOGETHER
AS FUNCTIONS OF THE SALES AND PRODUCTION
QUANTITY PARAMETERS
(Locations must be below both parameters to be exempted)
Percent of Total Locations
Not Required to Report
Percent Reduction in Total
Cost to Industry
-I i-'-1 ป | i ii t } i i r.... I ฆฆ ป.... t i. 11 i.
454 4540 45,400 454,000 4.54 million 45.4 million 454 million
(1000) (10,000) (100,000) (1 million) (10 million) (100 million) (1 bill^ni
prooiirf i on Hiianf i fw Im fiu\ - -
-------
EXHIBIT 4-59
Annual Sales
millions of dollars)
(loq scale)
100
REPORTS AND COST ISOQUANTS GRAPHED TOGETHER
AS FUNCTIONS OF THE SALES AND PRODUCTION
QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
10
1 -
.1 -
.01-
.001
Percent of Total Reports
Not Required to be Prepared
Percent Reduction in Total
Cost to Industry
X
4-
+
-4.
O-
4-
4-
0,0 45,4 454 4540 45,400 454,000 4.54 million 45.4 million 454 million
(0*0) (100) <1000> (10,000) (100,000) (1 million)(10 million) (100 million)(1 billion)
Production Quantity kg. (lb.)
(loa scale)
-------
EXHIBIT 4-60
Annual Sales
(millions of dollars)
(log, scale)
100
10
PERCENT OF CHEMICALS HISSED AND COST IOSQUANTS GRAPHED
TOGETHER AS FUNCTIONS OF THE SALES AND PRODUCTION
QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
15%
5%
; Percent of Total Chemicals Missed
Percent Reduction in Total
Cost to Industry
,01
.0011
0.0
(0.0)
X
4-
-I-
45.4
(100)
J-
4-
X
4-
4-
4540 45,400 454,000 4.54 million 45.4 million 454 million
(1000) (10,000) (100,000) (1 million) (10 million) (100 million) (1 bi"''"n)
-------
EXHIBIT 4-61
Annual Sales
(millions of dollars)
(log scale)
100 -
REPORTS AND LOCATIONS ISOQUANTS GRAPHED TOGETHER
AS FUNCTIONS OF THE SALES AND PRODUCTION
QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
10 -
1 -
1 -
.01-
Percent of Total Reports Not
Required to be Prepared
Percent of Total Locations
Not Required to Report
.001
0.0
(0.0)
4-
4-
+
-l-
4.
4~
X
.U
45-4 454 4540 45,400 454,000 4.54 million 45.4 million 454 million
(100) (1000) (10,000) (100,000) (1 million)(10 million) (100 million)(1 billion)
Production Quantity kg. (lb.)
(log scale)
-------
EXHIBIT 4-62
Annual Sales
(millions of dollars)
(log scale)
100 -
10 -
1 -
,1 -
.01-
.OOlL
-L-
REPORTS AND LOCATION ISOQUANTS GRAPHED TOGETHER TO
DEMONSTRATE THE INSENSITIVITY OF THE COST-EFFECTIVENESS
CONCLUSION TO ASSUMPTIONS REGARDING THE COEFFICIENTS
IN THE COST EQUATION
(Location must be below both parameters to be exempted)
C
Percent of Total Reports Not
Required to be Prepared
Percent of Total Locations
Not Required to Report
4-
4-
-4-
JL
ฆ t i ฆฆ I ฆ i l i I., i. i .. i... ,iu
0.0 45.4 454 4540 45,400 454,000 4.54 million 45.4 million 454 million
(0.0) (100) (1000) (10,000) (100,000) (1 million)(10 million) (100 million)(1 billion)
4-
Pr< cic.. w-an-w-w, kq. . lb. <
(luu bcalei
-------
EXHIBIT 4-63
STATISTICS SHOWING THE SIMILARITY OF PRODUCTION REPORTS FOR THE LOCATIONS USED IN THE
ANALYSIS AS COMPARED WITH THE TOTAL NUMBER OF LOCATIONS
All Locations a/
Production Percent of Percent of
Category All Records Records in 0-9
yj (2)
Locations in Analysis b/
Percent of Percent of (3)xl5,416
All Records Records in 0-9 (l)xl9,485
(3) 14] (5)
BLANK
N
0
1
2
3
4
5
6
7
8
9
22.1
11.1
7.8
5.1
7.7
9.6
10.8
9.7
4.6
8.5
1.6
1.4
11.6
7.7
11.5
14.4
16.1
14.6
6.9
12.7
2.4
2.1
22.6
10.8
8.5
5.0
7.2
9.1
10.2
9.5
4.8
9.3
1.7
1.3
12.7
7.5
10.8
13.6
15.3
14.3
7.2
13.9
2.6
2.0
81.2
76.7
86.2
77.1
74.0
70.6
75.0
77.5
81.8
86.3
85.7
76.9
5/Total number of locations ฆ 3778
Total number of reports = 19,485
^/"Tbtal number of locations = 3044
-------
EXHIBIT 4-64
COMPARISON OF THE DISTRIBUTION OF REPORTS IN PRODUCTION CATEGORIES N AND BLANK
TO THE DISTRIBUTION OF REPORTS IN PRODUCTION CATEGORIES 0 THROUGH 9
Sales Category
(million of dollars)
Percent of Total Reports
of Production in
Categories 0 through 9^/
ID
Percent of Reports
of Production
in Category N\i/
ill
Percent of Reports
of Production
in Category BLANKฎ/
-------
5. ANALYSIS OF OPTIONS
5.1 INTRODUCTION
EPA can set the small manufacturer exemption for Section 8(a) in a
variety of ways, including:
set the exemption on a case-by-case basis for each
rule promulgated under Section 8(a);
set one exemption level for all future Section
8(a) rules;
set a base exemption level below which locations
are always exempt, but retain the option of exempting
additional locations on a case-by-case basis;
set a ceiling exemption level above which
locations must always report, but retain the option
of requiring additional locations to report on a
case-by-case basis; and
set a series of tiered exemption levels that would
apply to different types of rules promulgated under
Section 8(a).
The choice of the number and type of parameters to use must also be made.
This chapter analyzes these options within the framework discussed in
Chapter 2,. and the information requirements of future Section 8(a) rules
discussed in Chapter 3. The results of the data analyses presented in Chapter
4 are used to evaluate the options and to highlight the implications of
setting the exemption(s) in various ranges.
The remainder of this chapter is divided into three sections, Section
5.2, Choice of Number and Type of Parameters, analyzes whether it is preferred
to set an exemption for which locations need to be below both parameters of
the exemption or just one parameter in order to be exempted. This section
also analyzes whether total employment can be substituted for annual sales as
a parameter to indicate corporations' abilities to absorb reporting costs.
Section 5.3, One-Level Exemption, analyzes the implications of setting a
single small manufacturer exemption for all future Section 8(a) reporting
rules. A one-level exemption provides the most certainty to industry;
locations will either always have to report, or will always be exempt from
reporting.
ICF INCORPORATED
-------
5-2
Section 5.4, Multi-Level Exemption, analyzes the implications of allowing
the exemption level to vary. We do not draw sharp distinctions between the
case-by-case, base, ceiling, and tiered exemption approaches. These options
are similar in that they provide the Agency increased flexibility over the
one-level exemption, but they also reduce the certainty to industry.
5.2 CHOICE OF NUMBER AND TYPE OF PARAMETERS
EPA can use a variety of parameters to define small manufacturers for the
purposes of the Section 8(a) small manufacturer exemption. We examined
production quantity at the manufacturing location, annual sales of the parent
corporation, and total employment of the parent corporation as parameters.
This section is directed toward the choice of how many and which parameters to
use.
The production quantity of a specific chemical does not necessarily
indicate the "size" of a manufacturer, because a small manufacturer may
produce as much or more of a particular chemical as a large diversified
corporation. However, production quantity is assumed to be an important
parameter of the small manufacturer exemption because it may be an indicator
of the potential for exposure to the chemical. It may be important not to
exempt reporting on large production quantity chemicals, regardless of the
annual sales or total employment of manufacturers, in order to protect human
health and the environment. Therefore, production quantity is the first
parameter of the exemption.
The choice of whether to add a second parameter can be analyzed using the
results presented in Chapter 4. From Exhibit 4-9, one sees that if only
production quantity were used as a parameter for the definition, the total
cost to industry would be reduced approximately 10.8 percent if it were set at
454 kilograms (1000 pounds). Under this exemption, all manufacturers who
produced the chemicals included in the rule in quantities greater than 454
kilograms would have to report.
In Exhibit 4-8, one sees that a 10 percent reduction in total cost to
industry can also be accomplished with any number of combinatipns of two
parameters; for example, the production quantity parameter could be set at
45,400 kilograms (100,000 pounds) with the sales parameter at $10 million. In
this case, the manufacturing locations must be below both parameters in order
to be exempted.
Although the total cost to industry is approximately the same for these
two exemption levels, the exemptions are not equivalent. To evaluate which
type of exemption is preferred, we examined the proxies for benefits and the
distribution of the reporting burdens among the smallest manufacturers in our
data base. The manufacturers among which the distribution of the reporting
burden was examined were the 600 manufacturing locations reported as having
annual sales under $100 million. These represent approximately twenty-five
percent of the locations used in the analysis.
ICF INCORPORATED
-------
5-3
With the 454 kilogram exemption, 2.8 percent of the locations and 12.7
percent of the reports would be exempted; and 12.5 percent of the chemicals
would be missed completely. With the two-parameter exemption set at 45,400
kilograms and $10 million, 7.1 percent of the locations and 10.7 percent of
the reports would be exempted; and 7.1 percent of the chemicals would be
missed completely. Production not covered is virtually zero for both
exemptions. The one-parameter exemption may be preferred only on the basis of
a larger number of different locations having to report for a given level of
total cost to industry. However, the cost of these additional locations being
required to report is borne almost entirely by the smallest manufacturing
locations.
The one-parameter exemption only exempts 7 percent of the locations and
29 percent of the reports of manufacturers with under $100 million in annual
sales. The two parameter exemption targets more of the costs toward
manufacturers with larger annual sales by exempting 28 percent of the
locations and 41 percent of the reports of those under $100 million. The
total cost to the 600 locations under $100 million is reduced by over 18
percent.
The two-parameter exemption also redistributes some cost among the 600
locations. The 360 locations with annual sales under $10 million all
experience reductions or no change in reporting requirements. Of the 240
locations with annual sales between $10 million and $100 million, thirteen
additional locations are required to report and the average number of reports
per location increases from approximately four to five. The manufacturers in
this range will experience no change or a small increase in their reporting
requirements.
The two-parameter exemption therefore allows one to more accurately
target the reporting burden toward those manufacturers most able to absorb
reporting costs, i.e., those with larger annual sales. The level of the
two-parameter exemption can be set to maximize any of the benefit proxies for
a given cost constraint as described in Chapter 4. The level of the
exemption depends on balancing the value of additional information against the
additional cost of obtaining it. Regardless of the level chosen, however, two
parameters enables the burden on small manufacturers to be as small as
possible given the amount of information desired. A production parameter
alone does not sufficiently target the reduction in reporting costs, and a
sales parameter alone does not satisfy EPA's information needs.
This analysis also holds for the choice between requiring that locations
be below either or both parameters to be exempted under a two-parameter
exemption. A one-parameter exemption is actually a special case of the
two-parameter exemption when locations need only be below one parameter to be
exempted. Therefore, a two-parameter exemption requiring that locations be
below both parameters in order to be exempted is preferred.
Annual sales and total employment of the parent company were two
parameters examined to accompany the production parameter for the small
manufacturer exemption. The sales quantity was chosen as a measure reflecting
ICF INCORPORATED
-------
the ability of corporations to absorb reporting costs. A measure of this sort
enables the reporting exemption to be targeted toward those manufacturers that
would be most adversely affected by increased costs imposed by government
reporting requirements.
Annual sales was chosen as an indicator of manufacturers' profits, and
its use implicitly assumes that profit margins are constant throughout the
industry and that corporations with larger profits are better able to afford
reporting costs. Neither assumption is completely true; profit margins will
vary across manufacturers and over time, and profits are not the only
indicator of the financial state of a corporation. However, given constraints
on the availability of verifiable information, these assumptions and these
data probably provide the most reasonable method of assessing the ability of
corporations to absorb reporting costs. To the extent that these assumptions
are correct, the annual sales parameter is useful in minimizing the adverse
effects that the reporting requirements will have on industry.
Annual sales as an exemption parameter may require indexing to account
for inflation because the Section 8(a) exemption is expected to be used over a
period of years. Two important issues regarding indexing are the choice of
the standard index and the timing of indexing. The Bureau of Labor Statistics
(BLS) and the Bureau of Economic Analysis both publish price indexes.
Seasonally adjusted producer price indexes for various industries (i.e.,
commodity groupings) and stages of production (i.e., finished goods,
intermediate goods, and crude materials) are available from BLS on a regular
basis along with the familiar producer and consumer price indexes. For
analysis of price trends, stage-of-processing indexes may be more useful than
commodity-grouping indexes.1 Commodity-grouping indexes can exaggerate
signals of price changes by reflecting the same price movement through various
stages of processing. For example, a price increase in a raw material could
be counted in the overall producer price index three times:
1) as the raw material price increase; 2) as the price increase in the
intermediate good resulting from the raw material price increase; and 3) as
the commensurate final good price increase. The stage-of-processing indexes
would each show the increase once, and may avoid any potential exaggeration
of price movements. Therefore, although a producer price index for
"chemicals" is available, the intermediate-products (less foods and feeds)
producer price index published by the BLS may be preferred. This assumes that
the chemicals about which information will be collected under Section 8(a) are
intermediate, as opposed to finished or crude, materials.
Because the sales parameter investigated here is based on annual sales,
there is no need to index more than once a year. In fact, the parameter could
be updated less frequently by using a base year of sales, e.g., 1980, for two
years and then indexing to a 1982 sales cut-off level. Less frequent indexing
fails to exempt firms that shrink below the cut-off during the period between
1Bureau of Labor Statistics, Producer Prices and Price Indexes Data
for April 1981 (Washington, D.C.: June 1981 [ISSN 0161-7311]), p. 88.
ICF INCORPORATED
-------
5-5
indexing, and fails to require reporting from the appropriate growing firms.
The opportunity for such errors may be small for a period as short as two
years.
Total employment was investigated as a substitute for annual sales as a
parameter of the small manufacturer exemption. Employment is not related
conceptually to corporations' abilities to absorb reporting costs. However,
if the same locations can be exempted using either parameter, then total
employment could be used and indexing would be avoided.
The results of the analysis are that although total employment and annual
sales are generally well correlated, in the relevant range of definitions
under consideration total employment is not a good predictor of annual
sales. Therefore, total employment cannot be used to exempt the same
locations as the annual sales parameter. The analysis below demonstrates this
result.
A random sample of 250 locations was drawn from the total number of 3044
locations for which all employment and sales data are known. Exhibit 5-1
displays a plot of sales (in millions of dollars) versus employment for these
locations. As Exhibit 5-1 shows, employment and sales are generally well
correlated. Regressing sales on employment results in:
2
(1) Sale = 186,600,000 + 81,900 (Employment); R = 0.541; and
2
(2) Sale = 85,200 (Employment); R = 0.536.
Regression equation (2) is constrained to go through the origin. Both
equations are shown in Exhibit 5-1.
The relevant range of exemptions under consideration is in the bottom
left-hand corner of Exhibit 5-1. Exhibit 5-2 enlarges that corner by plotting
sales versus employment for only those locations with annual sales under $100
million. Again, sales and employment appear to be generally well correlated,
and the regression equations are:
2
(3) Sale = 5,900,000 + 45,000 (Employment); R = 0.681; and
(4) Sale = 50,900 (Employment); R2 = 0.621.
Regression equation (4) is constrained to go through the origin.
For locations with annual sales under $100 million, total employment
varies to as high as approximately 1800 employees. More importantly, the
locations with annual sales between $30 million and $100 million also have
total employment figures that vary widely. Employment does not appear to
correlate well with sales for those locations in this sample with annual sales
in this range. Exhibit 5-2 also shows that locations with sales between $30
million and $100 million generally have employment figures of greater than 500
employees. To exempt these locations with an employment parameter, the
employment parameter would have to be set so high that locations with annual
ICF INCORPORATED
-------
EXHIBIT 5-1
PLOT OF TOTAL ANNUAL SALES VERSUS TOTAL EMPLOYMENT
(Sample of 250 locations)
Annual Sales
(millions of $)
-
a j J > j 10 -
I ft 1. HO -
-
5 y ^ ^ o 1 i -
3Jr>o.V.l -
-
"tit jlu -
-
21'b.iil -
l-
17 io> idi -
1 t ' I. 1st.
o
T1
- 'j I'j.i1*/
5
o
a
J -U. Jtl)
i
o
SALE = 85,200(EMPLOYMENT)
Xts-v
SALE = 186,000,000 + 81,900(EMPLOYMENT)
a = 25 observations
b =137 observations
c = 10 observations
.EMPLOYMENT
I I I I I I I I I I I
*100 Ms/ooui Ob7^r..010 09373.00J 11199U.OOO
lOlDJ.olJ ' i2fc0?.3'ปf> 1 7ซ. 7JC60.500 1 00t>B 5
-------
EXHIBIT S-2
Annual Sales
(millions of $)
f/5. J.lj -
8 2 u5. -
6Cซ ilu "
7 6..J.,, -
PLOT OF TOTAL ANNUAL SALES VERSUS TOTAL EMPLOYMENT
FOR LOCATIONS WITH TOTAL ANNUAL SALES BELOW $100 MILLION
(Taken from sample of 250 locations)
o
3
0
a
s
1
o
Cv.aTj -
61 j
5 / C'Ot, -
S0.^-t -
* -
)8.Ull -
3 1 ป j ^ j
2 ? J'f a -
I*.jls -
l.'.jt', -
:ฆ -
J. if >J -
SALE = 50,900(EMPLOYMENT)
SALE = 5,900,000 + 45,000(EMPLOYMENT)
a = 27 observations
I
I I I
l3r . 170
1 .-> 3 . o I 4 *; 31 5 J i
EMPLOYMENT
, ' > ( I I I I
7^'-,.j7h lG'ซi.:i77 |4',6*2'IU I 8 3o . 9:3 3
'Ij.tZs I2f!u.9>9 loSl.bll
-------
5-8
sales much greater than $100 million would also be exempt. This result can be
seen from Exhibit 5-1. Therefore, in the sales parameter range of $30 million
to $100 million, employment does not reflect corporations' abilities to absorb
the costs of reporting.
In the sales range below $30 million, the same problem exists. To exempt
all the appropriate locations in this sales range, the employment parameter
must be set so high that locations with sales over $30 million are exempted.
By lowering the employment parameter to correct for this, many locations with
low sales are not exempted.
For a specific set of production quantity and sales parameters, a more
detailed analysis can explore what employment level options can be substituted
for the given sales level, and can examine the exact effects each option has
on the locations required to report. In general, employment is not a good
substitute for sales as a parameter of the small manufacturer exemption.
Therefore, the remainder of this chapter analyses the one-level and
multi-level options for exemptions set with production quantity and annual
sales parameters.
5.3 ONE-LEVEL EXEMPTION
A one-level exemption for all future Section 8(a) reporting rules is the
least flexible option evaluated here. Regardless of the varying information
requirements of the expected rules discussed in Chapter 3, manufacturers below
the exemption would always be exempt, and those above would always be required
to report. However, a one-level exemption also provides the greatest
certainty to industry. This increased certainty to industry can be balanced
off against the opportunities to reduce reporting costs to industry in some
cases by requiring less reporting; and to increase reporting costs in other
cases by requiring more reporting.
Industry's relative valuation of the certainty of reporting versus
unknown future reporting requirements is not assessed here. Instead, this
option is evaluated in terms of its implications for setting an exemption
level that enables EPA to obtain all the information required for future 8(a)
rules. The cost that setting this level could be expected to impose on
industry is also examined.
A one-level exemption is most appropriate if all future rules will
require approximately the same coverage of manufacturers. To the extent that
the desired coverage varies significantly across rules, a one-level exemption
may include too many manufacturers for one rule or too few for another. For
purposes of the analysis that follows, we assume that a one-level exemption
should be set so as to meet the strictest coverage requirements of potential
future Section 8(a) rules. Just how much coverage is sufficient is a judgment
for the Agency to make. In the following example we demonstrate how the
ICF INCORPORATED
-------
5-9
reports and locations proxies can be used to set an exemption level for a
group chemical rule.2
As shown in Chapter 4, to maximize the number of locations for a given
cost constraint, "low" production quantity and "high" sales parameters should
be used. To maximize the number of reports, the opposite is true. Because
both proxies may be important in the evaluation of a rule, a "middle" approach
may at first glance seem appropriate. However, the importance of obtaining
information representative of the wide range of sales categories, be it in
terms of the number of reports prepared or the number of locations reporting,
also influences the choice of the appropriate range of parameters to consider.
Exhibit 5-3 shows that at point A, a "low" production quantity parameter
And "high" sales parameter combination ensures more representative information
for a given cost. The points A and B in Exhibit 5-3 are assumed to impose the
same total cost on industry. The locations in the shaded region are exempt
under either definition. Point A additionally exempts locations in area
SjS^AX; and Point B additionally exempts locations in area PjPjBX.
If the exemption were set at B, nearly all the locations with total firm
sales under Sj would not have to report because is set "high". Short
of setting close to zero, which would impose a large burden on industry
as a whole including small manufacturers, very little or no information would
be received from the small locations in the industry, even if they have
moderately high production (between and ?^). By setting the exemption
at A, additional information on production above P^ is received for a loss
of some information on production below P^. The additional information both
increases the representativeness of the information to a wider range of sales
categories, and focuses on moderate to high production quantities. Setting
the exemption at A spreads the reduction in reporting costs among more sales
categories. Also, the reporting that is exempted is primarily on very low
production chemicals that may be of less importance. (The way the data
support this analysis is discussed below.)
Point A may also be preferred on equity grounds. Locations below
may produce a fair number of moderate quantity chemicals (between P^ and
P^) that compete with moderate quantity chemicals produced by medium sized
firms (between and Sฃ) Point B could give the smallest firms an
unwarranted advantage over the medium sized firms. Therefore, Point A is
again preferred.
aThe example analyzed here is for a hypothetical reporting rule that
requires reporting on a group of chemicals similar to the group used as data
for this report. As discussed in Chapter 4, the results of the data analyses
presented here may not hold for groups of chemicals or individual chemicals
that are not representative of the data examined here.
ICF INCORPORATED
-------
5-10
EXHIBIT 5-3
COMPARISON OF THE REPRESENTATIVENESS OF INFORMATION RECEIVED
UNDER A "LOW PRODUCTION/HIGH SALES" EXEMPTION (POINT A)
VERSUS A "HIGH PRODUCTION/LOW SALES" EXEMPTION (POINT B)
Sales
Parameter
-B
Production Quantity Parameter
ICF INCORPORATED
-------
5-11
If production processes vary systematically across the chemical industry
in such a way so as to make representativeness an important criterion to
consider, then a "low" production/"high" sales range seems most appropriate
for the one-level exemption. The reports proxy still holds, but for the
information to be valuable it may have to be representative.3
Just how low and high the parameters should be set depends on the value
of the information relative to the cost of obtaining it. We have not assigned
values to the information. However, by choosing an exemption level in the
appropriate range, and examining how the information obtained and the
reporting burden vary as one varies the exemption level, we highlight the
information/cost trade-offs required.
Starting (arbitrarily) with an exemption level of $50 million and 45,400
kilograms (100,000 pounds), total cost to industry is reduced nearly 14
percent; 10.5 percent of the chemicals may be missed entirely; approximately
15 percent of the total possible reports are exempt; and 8.6 percent of all
locations are exempt. The total cost to manufacturers with annual sales under
$50 million is reduced by 64 percent of the total possible cost to those
manufacturers. These manufacturers account for nearly 22 percent of the total
number of manufacturers used in the analysis. Reports are received from a
fairly representative cross-section of sales ranges. Except for the under
$100,000 ranges, over 40 percent of the locations in each range must report.
To obtain additional information, the sales parameter and/or the
production parameter can be reduced. Reducing the sales parameter to $20
million picks up additional information from those locations with annual sales
between $20 million and $50 million. No addition information is obtained from
locations below $20 million or above $50 million. The additional information
is obtained only on those chemicals produced in quantities under 45,400
kilograms (100,000 pounds). Of the 96 locations within this sales range,
twenty additional locations must report when the sales parameter is lowered,
and the average number of reports per location increases from 2.5 to 5.3. The
percent of locations exempted drops 0.8 percent to 7.8 and the percent of
reports exempted is reduced from 15 percent to 12 percent. Also, the percent
of chemicals missed drops to under 8 percent from 10.5 percent. The
additional cost of the information is approximately 2.5 percent of the total
possible cost to industry.
The parameter should be lowered if the value of the additional
information outweighs the additional burden placed on manufacturers in that
sales range. The actual burden will depend on the complexity of the reporting
'This result is the same as the hypothesis put forward in a previous
1CF report on this topic, that marginal increases in the number of locations
reporting were valued more highly than marginal increases in the number of
reports prepared. See "Analysis of TSCA Section 8(a) Generic Small
Manufacturer Exemption" (February 1981), p. 7 and 37.
ICF INCORPORATED
-------
5-12
rule flpH the time it takes to complete the needed forms. The value of the
information depends on whether the information that would be received
regarding the production of small volume chemicals at locations in this sales
range varies from the information received regarding:
the production of other larger quantity chemicals
by the same locations; or
the production of the same or similar very small
quantity chemicals by locations with high sales
figures.
If the production process of the chemicals covered in this incremental
information differ significantly from the production processes reported
elsewhere, then this information may be sufficiently valuable to warrant a
reduction in the sales parameter. However, of all the reports of production
below 45,400 kilograms (100,000 pounds), nearly one third come from locations
with annual sales greater than $50 million; and of all the locations that
produce only in quantities under 45,400 kilograms, nearly one-half have
annual sales greater than $50 million. Therefore, in the absence of
information demonstrating that the production processes of small quantity
chemicals produced by the locations in the $20 million to $50 million sales
range are different from the processes of locations in the over $50 million
sales range, it appears that the additional information obtained by lowering
the sales parameter to $20 million may not outweigh the cost of obtaining the
information.
The only remaining reason for which the value of the additional
information in this example may outweigh the costs is that fewer chemicals are
missed with the sales parameter at the lower level. Alternative methods of
gathering information for the chemicals missed in' reporting may be costly for
EPA. However, as described in Chapter 3, chemicals in such small production
quantities may be too unimportant to warrant detailed examination if missed in
reporting. The need for Agency follow up, of course, depends on the purpose
of the individual rules.
This method of analysis holds for lowering the sales parameter even below
$20 million. The production processes of locations in the range under say, $5
million, may be more likely to differ from other production processes about
which information is collected elsewhere. The marginal benefit of obtaining
this information may be higher than the marginal benefit of obtaining the
information analyzed above. It is not clear whether the benefit could be so
great as to warrant a small manufacturer definition set below $5 million; in
fact, it appears unlikely.
Beginning again at the $50 million, 45,400 kilogram exemption level, one
can examine the benefits and costs of obtaining more information by lowering
the production parameter to 4,540 kilograms (10,000 pounds). The total cost
to industry increases significantly, 5 percent of the total possible cost.
This entire increase is borne by locations with annual sales below $50
million. The additional information obtained is on chemicals produced in
ICF INCORPORATED
-------
5-13
quantities between 4,540 and 45,400 kilograms. As above, if the information
regarding the production of these chemicals does not differ significantly from
the other information which is received, then the incremental value of this
information is not very high and is unlikely to outweigh the additional
costs. In other words, the primary benefit of lowering the production
parameter in this way is that it increases the representativeness of the
respondents; over 80 percent of the locations in each sales range would be
required to report. Even if representativeness is deemed by EPA to be an
important criterion, it is unlikely that it will be required to this extent
even for a rule requiring a wide coverage of manufacturers.
Less information would be received if the parameters were increased.
Raising the sales parameter to $100 million would reduce the total cost to
industry by approximately 1.7 percent of the total possible cost of
reporting. This reduction in cost would be received entirely by the locations
with annual sales between $50 and $100 million. The information lost would be
on those chemicals produced in quantities under 45,400 kilograms at the
locations in this sales range. This information may not be very valuable for
the same reasons cited above for not lowering the sales parameter to $20
million. However, the implication that the sales parameter can be increased
indefinitely without the loss of any important information is not correct. As
the sales parameter is increased, the information reported about chemicals in
the lowest production categories (by the locations with high annual sales) is
likely to become less and less representative of the information on chemicals
produced in small quantities by locations with lower annual sales. Therefore,
if representativeness is an important criterion, the sales parameter must be
kept sufficiently low to ensure that the full range of information on low
production quantity chemicals is obtained.
Just how low is "low enough" is necessarily a judgment for EPA to make.
Additional information on how the production processes of the small quantity
chemicals of interest vary across sales categories could be helpful. Clearly,
however, there is an upper limit on the most appropriate level for the sales
parameter.
Because there may always be insufficient information to choose this level
with a great degree of confidence, one can examine the importance of reducing
the burden on locations in the $50 to $100 million annual sales range.
Increasing the sales parameter would reduce the average number of reports per
location in this range by 2.0 from 5.4 to 3.4. The importance of this cost
reduction to firms of this size would likely be minimal, even if the firms had
as many as ten or more locations.u Therefore, to be conservative, it may be
best to leave the sales parameter at (or below) $50 million. This will ensure
sufficient information is obtained, but will probably not impose an
unreasonable burden on locations with annual sales greater than this amount.
fcThis would be more than twice the expected number of locations per
firm in this sales range. See Analysis of Options for Definition of Small
Business, and Estimated Cost of the Initial Section 8(a) Reporting
Requirements, (EPA-561/1-77-001: November 1977), p. 45.
ICF INCORPORATED
-------
Less information could also be received if the production quantity
parameter were increased. Increasing this parameter to 454,000 kilograms (1
million pounds) would reduce the reporting burden on those locations with
annual sales under $50 million by approximately 3.6 percent of the total
possible cost to industry. Only ten to thirty percent of the locations in the
sales categories below $5 million would be required to report. (Over fifty
percent of the locations in the other sales categories would still have to
report.) This fair to low representation in the smaller sales ranges may be
unacceptable for some potential future rules. This is particularly true
because the incremental information foregone would be on chemicals produced in
quantities up to 454,000 kilograms annually. Therefore, although the cost
reduction would be received exclusively by the smaller locations (in terms of
sales), it may be unwise to increase the production parameter of the one-level
exemption to 454,000 kilograms.
This analysis can be repeated for additional levels of the one-level
exemption. The conclusions drawn in this section reflect judgments on the
relative value of obtaining different types of information from various
segments of manufacturers. The next section discusses multi-level exemptions.
5.4 MULTI-LEVEL EXEMPTION
The analysis of a multi-level exemption is identical to that performed
for a one-level exemption. The results will only vary insofar as the coverage
required for the different reporting rules varies.
For example, a 454,000 kilogram (1 million pound) production quantity
parameter may yield sufficient information for a rule that does not require
broad coverage of manufacturers. With the sales parameter at $50 million,
total cost to industry is reduced by approximately 17.5 percent. The entire
cost reduction is received by locations with annual sales under $50 million.
As mentioned in the previous section, there is only fair to low representation
in the lower sales ranges at this production quantity. However, unlike the
one-level exemption that must meet the constraints of the the strictest
coverage requirements of potential future rules, parts of a multi-level
exemption can be set higher; even a 4.54 million kilogram production quantity
parameter may be acceptable.
As above, the choices of the exact levels depend on various judgments and
valuations of information. Rather than repeat the lengthy example of the
previous section, we may conclude that:
The upper tiers of a multi-level exemption will be
higher than the one-level exemption if the coverage
required for the future rules is expected to vary;
The total cost to industry of a series of different
Section 8(a) reporting rules can be less under a
multi-level exemption than under a one-level
exemption; and
ICF INCORPORATED
-------
5-15
If representativeness is desired, the preferred
combinations of parameters at all levels is "low"
production and "high" sales.
Chapter 6 summarizes the results of the analyses.
ICF INCORPORATED
-------
CHAPTER 6
CONCLUSIONS
In developing a general small manufacturers exemption for Section 8(a)
reporting, EPA must consider a number of factors. This study was designed to
help identify and analyze those factors and generally facilitate EPA's
decisions regarding setting an appropriate exemption. We believe the
following conclusions, important to EPA's decision, can be drawn from the
analyses conducted in this study:
The appropriate proxies to use to evaluate the
amount of information received under the various
Section 8(a) rules currently foreseen are the number
of reports prepared and the number of locations
reporting. Additionally, a measure of the
representativeness of the information received from
firms in the low sales ranges may be desirable
(Chapter 3).
To maximize the number of reports prepared for a
given total cost to industry, "high production/low
sales" parameters are preferred. To maximize the
number of locations reporting for a given total cost
to industry, "low production/high sales" parameters
are preferred. These results are not sensitive to the
relative magnitudes of the fixed and variable costs of
reporting. However, these results may not hold for
reporting rules for which the mix of chemicals and
manufacturers are not similar to those examined here
(Chapter 4).
Nearly all production is covered in the exemption
ranges examined here (Chapter 4).
To minimize the number of chemicals missed for a
given cost constraint, "high production/low sales"
parameters are preferred. This result may not hold
for reporting rules for which the mix of chemicals and
manufacturers are not similar to those examined here
(Chapter 4).
A two-parameter exemption can more accurately target
reductions in reporting costs to firms with less
ability to absorb these costs (as measured by annual
sales) than a one parameter exemption based on
production volume. Employment is not a good
substitute for annual sales as a parameter to indicate
ICF INCORPORATED
-------
6-2
firms' abilities to absorb reporting costs.
Therefore, the preferred exemption parameters are
production quantity and annual sales. The preferred
structure of the exemption is to require that
locations be below both parameters in order to be
exempted (Chapter 5).
If production processes vary across segments of the
chemical industry such that representativeness becomes
an important criterion to consider, then the preferred
range of exemption parameters to consider is the low
production/high sales" range This range ensures that
the information obtained will be as representative as
possible given the total cost to industry, and is also
the cost effective range for the locations proxy.
Other ranges may be equally appropriate if
representativeness is deemed unimportant (Chapter 5).
A one-level exemption is appropriate if all future
rules will require approximately the same coverage of
manufacturers. If coverage of manufacturers is
expected to vary, then a multi-level exemption may be
preferred. Given this uncertainty over future
coverage requirements, the added flexibility of the
multi-level exemption structure may be preferred, even
if only a single level is contemplated at this time
(Chapter 5).
The choice of the exemption level requires judgments
regarding the value of obtaining different types of
information from various segments of the chemical
industry. Additionally, the choice of exemption level
distributes the total cost among the manufacturers.
The actual cost imposed (in dollars) will be a
function of the information requested by the reporting
rule. The value of the incremental amount of
information received must be balanced against the
incremental cost imposed on certain segments of the
industry to obtain it. Exhibit 6-1 lists six levels
of the exemption in order of decreasing total cost to
industry. Columns (2) and (3) show the exemption
parameters. Columns (4) through (7) show the values
for the cost and benefit proxies; columns (8) and (9)
indicate both the representativeness of the
information received and the cost imposed on the
smallest manufacturers (in terms of sales). The
higher the values in columns (8) and (9), the more
information is required from manufacturers with annual
sales below $50 million. In Chapter 5, we discuss how
one can use information of the type shown in Exhibit
6-1 to identify the most appropriate level for the
exemption. (Chapter 5).
ICF INCORPORATED
-------
EXHIBIT 6-1
COMPARISON OF THE COSTS AND BENEFITS OF SIX SMALL MANUFACTURER EXEMPTION LEVELS
(location must be below both parameters to be.exempted)
Average of
Average of
Percent of
Percent of
Locations in
Total Possible
Sales Cate-
Reports from
Below $50
Sales Categories
Million
Below $50 Million
Exemption
Parameters
Percent
Reduction
in Proxies
(%)
Required
Required to
Option
Sales ($)
Production (kg.)
Total Cost
Locations
Reports
Chemicals
to Report (%)
be Prepared (%)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
1
50
million
4,540
9.1
3.5
10.4
7.9
91.0
52.0
2
20
million
45,400
11.2
7.8
12.0
7.9
52.0^
27.0^
3
50
million
45,400
13.9
8.6
15.1
10.5
58.0
29.0
4
100
million
45,400
15.6
9.1
17.1
12.7
58.0
29.0
5
50
million
454,000
17.5
14.1
18.3
11.7
36.0
16.0
6
50
million
4.54 million
19.4
17.6
19.8
12.3
22.0
10.0
3/Average is for categories below $20 million in annual sales. All locations must submit all possible reports in
categories between $20 million and $50 million.
-------
APPENDIX A
DESCRIPTION OF DATA BASE
EPA provided on computer tape three files of information for this project,
the Sales File, the Production File, and the Chemical File. The information
was provided on a computer tape which was read into ICF s computer. The
content of each of the files is described in this appendix.
Sales File
The Sales File contained ten pieces of information on 3778 manufacturing
locations of 1459 chemicals. The information for each location was obtained
by EPA from the Dun and Bradstreet Corporation and is as follows:
A. Manufacturer's Identification Number (MID): seven
alpha-numeric characters uniquely identifying each
manufacturing location. This numbering system is used
by EPA.
B. DUNS Number: nine digit number uniquely identifying
each manufacturing location on which the Dun and
Bradstreet Corporation maintains information. Dun and
Bradstreet maintains records on over 4 million
businesses nationwide.
C. Headquarters DUNS (HQDUNS): if the manufacturing
location is a branch that reports to a headquarters,
the DUNS Number of the headquarters is reported.
Also, if the location is itself a headquarters, its
DUNS Number is repeated here.
D. Parent DUNS Number (PAR DUNS): if the manufacturing
location is a subsidiary, the DUNS Number of its
parent firm is reported here.
E. Ultimate DUNS Number (ULT DUNS): number identifying
the top firm in the location s corporate structure.
F. Status: designates whether the location is a
headquarters, branch, or single location.
G. Sales: the annual sales for the manufacturing
location. This is not reported for branches. If the
location is the top company in the corporate
structure, the sales for the entire corporation is
reported.
ICF INCORPORATED
-------
A-2
H. Subsidiary: designates whether the location is a
subsidiary. All subsidiaries must have a Parent DUNS
Number designated. Branch locations cannot be
subsidiaries.
I. Total Employment (EMPTOTAL): total employment for
the entire corporation.
J. Employment (EMPHERE); total employment for that
location. This is not reported for branch locations.
The Sale File is used to associate the appropriate total employment and
annual sales figures with each manufacturing location. These data are then
used along with the production data to assess the extent to which each
location is required to report under various small manufacturer exemptions. A
significant portion of the data were missing, however. In the original data
provided by EPA, there were:
912 locations for which the sales and employment
data were completely missing because Dim and
Bradstreet information was not provided;
1357 locations for which the sales and employment
data were not provided at the corporate level; and
117 locations for which the sales figure was
reported as zero.
This left only 1392 locations or 37 percent of the original data with
sufficient information to be included in the analysis. To improve the data
base, we used Standard Corporate Descriptions published by Standard and
Poor's to find the annual sales and total employment of the corporations to
which the locations belonged. Because the Standard and Poor's data are
arranged by name, EPA provided the names of the 3778 locations as they were
reported in the 1977 Inventory. Many of the names were not sufficiently
specific to identify the corporation to which the location belonged (for
example, many locations were listed similarly to "Plant No. 2"). We were
successful in obtaining the needed sales and employment data for 455 of the
912 locations for which no Dun and Bradstreet data were provided and 55 of the
117 locations for which the sales figure was reported as zero.
For the other 1357 locations, we used a list provided by EPA of sales and
employment data for approximately 925 corporations identified by Ultimate Duns
Number. We successfully matched up the Ultimate Duns Numbers for 1142 of the
1357 locations which left only 215 of these locations without sales and
employment data. Therefore, following our data search, the final Sales File
included:
1392 locations: provided originally by EPA;
455 locations: sales and employment data found by
ICF Incoroporated in Standard and Poor's;
ICF INCORPORATED
-------
A-3
55 locations: sales data found by ICF Incorporated
in Standard and Poor's; and
1142 locations: sales and employment data found by
ICF Incorporated in a printout provided by EPA.
A total of 3044 locations or 80.6 percent of the original data were used
in the analysis.
Production File
The Production File contained six pieces of information on 3778
manufacturing locations. The information was provided by EPA and was
developed by matching the reports of production of 1459 chemicals in the 1977
Chemical Inventory to the locations at which the production was reported as
taking place. The list of 1459 chemicals chosen by EPA includes:
all chemicals on both the ITC Initial Listing of
chemicals and the 1977 Chemical Inventory; and
chemicals being considered by EPA for the TSCA
Section 8(a) Level A rule which were not chosen
solely on the basis of high production quantity.
The information for each location is as follows:
A. Manufacturer's Identification Number (MID): seven
digits of numbers and letters uniquely identifying
each manufacturing location. This numbering system is
used by EPA.
B. Mean, Standard Deviation, Low and High: derived
production statistics for the manufacturing location.
, (These data were not used in the analyses.)
C. Blank: the number of chemicals for which that
manufacturing location reported the production
quantity confidential.
D. N: the number of chemicals for which that
manufacturing location reported no current production,
but expected future production.
E. 0, 1, 2, 3, 4, 5, 6, 7. 8, 9: the number of
chemicals for which that manufacturing location had
reported production in each of ten possible production
categories. The categories are as follows:
ICF INCORPORATED
-------
A-4
0
0 -
1,000 pounds
1
1,000 -
10,000 pounds
2
10,000 -
100,000 pounds
3
100,000 -
1 million pounds
4
1
million -
10 million pounds
5
10
million -
50 million pounds
6
50
million -
100 million pounds
7
100
million -
500 million pounds
8
500
million -
1 billion pounds
9
greater than
1 billion pounds
F. Total: total number of Inventory chemicals produced
at that manufacturing location.
Chemical File
The Chemical File contains the same information as in the Production File,
except the Chemical File is arranged by chemical instead of manufacturing
location. The information obtained for each chemical from the 1977 Chemical
Inventory is as follows:
A. CAS Number: nine digit number uniquely identifying
each chemical.
B. Mean, Standard Deviation, Low and High: derived
production statistics for the chemical. (These data
were not used in the analyses.)
C. Blank: the number of reports of that chemical for
which the production quantity was kept confidential by
the manufacturer.
D. N: the number of reports of that chemical for which
the manufacturing location reported no current
production, but expected future production.
E. 0, 1, 2, 3. 4, 5, 6, 7, 8, 9: the number of reports
of production of that chemical in each of ten possible
production categories. The ten categories are defined
in the previous section on the Production File.
The Chemical File only included data on 1243 of the 1459 chemicals chosen
by EPA to examine in this study. The list of 1243 was developed by EPA by
choosing all those chemicals on both the 1977 Chemical Inventory and on the
Interagency Testing Committee (ITC) Initial Listing of Chemicals. Data on
these chemicals were provided by EPA to ICF Incorporated.
EPA then expanded the list of chemicals to include 216 additional
chemicals being considered for the TSCA Section 8(a) Level A rule which were
not chosen solely on the basis of high production quantity. Data on these
additional chemicals were not provided. Nevertheless, the data on the 1243
ICF INCORPORATED
-------
A-5
chemicals are very likely to be representative of the larger list of 1459
chemicals. Additionally, the production data for these 1459 chemicals have
been updated since the performance of this analysis. The updated information
was also not available for the Sales and Production Files described above.
The newer data would probably not affect any conclusions drawn in this
report. Exhibit A-l displays the similarity in the distributions of the
reports of production across the various production categories for the two
lists of chemicals, and the updated data. The distribution of reports for the
expanded list was developed using the Production File data. The larger list
has a slightly higher percentage of reports of production in which the
production quantity was kept confidential by the manufacturer (category
BLANK). The list of 1243 appears to be well representative of the larger list
and is therefore used in the analysis. The distribution of reports for the
updated data was provided by EPA. Again the distributions are sufficiently
similar such that errors introduced by using the "old" list are likely to be
insignificant. Therefore, EPA concluded that no further revisions to the
analysis were necessary.
The Inventory Small Manufacturer Exemption
The initial reporting rule for the Inventory included a small manufacturer
exemption of $5 million, except that no manufacturer was considered small with
respect to any chemical manufactured at a location in an amount exceeding
100,000 pounds. Small manufacturers were exempted from reporting production
quantities and were permitted to aggregate data from more than one
manufacturing location. This small manufacturer exemption may bias the
Inventory data (upon which the anayses in this report are based), and
consequently may affect the results presented in the chapters above.
The data and results regarding the marginal benefits and costs of setting
the exemption below both $5 million and 100,000 pounds are suspect. The
greater the bias in the Inventory data (which is unknown), the more inaccurate
is the information reported above on small manufacturer exemptions set below
both of these levels.
The effects of a small manufacturer exemption induced bias in the
Inventory data on the data and results regarding the marginal benefits and
costs of setting the exemption with at least one parameter above $5 million or
100,000 pounds are negligible. The data and analysis in the report are based
on the number of locations that would have to report, which was obtained
directly from the data. The small manufacturer exemption in the Inventory
reporting rule did not affect the accuracy of this information. In the
presentation of the results, however, we used the percentage of locations not
required to report. This was estimated by dividing the difference between the
total number of locations and the number of locations required to report by
the total. The resulting percentages are in error only to the extent that the
total number of locations (which may be biased downward due to the small
manufacturer exemption) is in error. All the percentages would be
innaccurate, but the shape of the graphs shown in the report would remain
unchanged. Therefore, given that the marginal costs and benefits are the
primary important decision variables, the usefulness of the data and results
in the report are not sensitive to a small manufacturer exemption induced bias
in the Inventory data.
ICF INCORPORATED
-------
A-6
EXHIBIT A-l
COMPARISON OF THE DISTRIBUTION OF REPORTS OF PRODUCTION ACROSS
THE PRODUCTION CATEGORIES FOR THE 1243 CHEMICALS IN THE
CHEMICAL FILE VERSUS THE 1459 CHEMICALS INCLUDED IN
THE PRODUCTION AND SALES FILES
Category
Percentage of Reports
List of
1243 Chemicals a/
List of
1459 Chemicals b/
Updated List of
1459 Chemicals c/
BLANK
21.1
22.1
21.7
N
10.7
11.2
11.4
0
7.6
7.8
7.7
1
5.0
5.1
5.2
2
7.6
7.7
7.5
3
9.7
9.6
9.6
4
11.0
10.8
10.6
5
10.3
9.7
9.9
6
4.9
4.6
4.7
7
9.0
8.5
8.7
8
1.7
1.6
1.6
9
1.4
1.4
1.4
a/ Total Number of Reports = 18,363
b/ Total Number of Reports = 19,485
c/ Total Number of Reports = 19,374
ICF INCORPORATED
-------
APPENDIX B
DATA ANALYSES
The data analyses performed for this report are described in this
appendix. Appendix A describes the Sales, Production, and Chemical data files
used in the analysis. The information found by ICF Incorporated on 1652
manufacturing locations was entered manually into the computer. Following the
addition of these data, the following analyses were performed:
Merge the Sales and Production Files;
Calculate values for the locations, reports, and
production proxies from the Merged File;
" Calculate values for the cost proxy from the
locations and reports proxies; and
Estimate the percentage of chemicals missed using
the Chemical File.
Each of these st6ps is described below.
B.l MERGE THE SALES AND PRODUCTION FILES
The information on the Sales and Production Files was merged to create a
single record for each of the 3778 manufacturing loctions in the data base.
Associated with each location in the Merged File was its Manufacturer's
identification number, the annual sales and total employment of its parent
corporation, and the number of chemicals produced at that location in each of
twelve production categories. Of the 3778 locations, 734 did not have
complete sales and employment information. These locations were dropped from
the data base, and were not used in the following analyses. A total of 3044
locations remained.
B.2 CALCULATE VALUES FOR THE LOCATIONS, REPORTS. AND PRODUCTION PROXIES
The Merged File was used to calculate values for the locations, reports,
and production proxies for various small manufacturer exemptions. Eleven
sales intervals, eight employment intervals, and the twelve available
production categories were used to set various small manufacturer exemptions.
These intervals are shown in Exhibit B-l. The calculations for the locations
and reports proxies are described, and then the estimation of the production
proxy is described.
Locations and Reports Proxies. The following procedure was used to
calculate the number of locations that would have to report and the number of
reports that would have to be prepared for various small manufacturer
exemptions.
ICF INCORPORATED
-------
B-2
EXHIBIT B-l
INTERVALS OF THE PARAMETERS USED TO SET VARIOUS
SMALL MANUFACTURER DEFINITIONS
Production
Quantity
kg. (lb.)
Category Interval
Annual Sales
(millions of dollars)
Category Interval
Total Employment
Category Interval
0:
0-454 (0-1000)
1:
0-.10
1-4
1: 454-4,540 2:
. (1000-10,000)
2: 4,540-45,400 3:
(10,000-100,000)
3: 45,400-454,000 4:
(100,000-1 million)
4: 454,000-4.54 million 5:
(1 million-10 million)
5: 4.54 million-22.7 million 6:
(10 million-50 million)
6: 22.7 million-45.4 million 7:
(50 million-100 million)
7: 45.4 million-227 million 8:
(100 million-500 million)
8: 227 million-454 million 9:
(500 million-1 billion)
9: Greater than 454 million 10:
(greater than 1 billion)
BLANK: Quantity Held 11:
Confidential
.10-.50
.50-1.0
1.0-2.5
2.5-5.0
5.0-10
10-20
20-30
30-50
50-100
Greater
than 100
2: 5-6
3: 7-10
4: 11-20
5: 21-50
6: 51-100
7: 101-200
8: greater than
200
N:
None Produced in 1977
ICF INCORPORATED
-------
B-3
1. Examine the sales and employment figures for the
location and assign the location to the proper sales
and employment intervals.
2. Each sales and employment interval is a vector with
twelve variables. Each of the twelve variables is
assigned to one of the twelve production categories.
The number of reports that the location would have to
prepare if the definition were set at production
category 9 (i.e., the highest production category), is
added to the tenth variable in the appropriate sales
and employment vectors. The number of reports that
the site would have to prepare if the definition were
set at production category 8 is added to the ninth
variable in the appropriate sales and employment
vectors, and so on. The result is a matrix that
identifies how many reports would have to be prepared
for all the possible levels of sales and employment
when production varies from categories 0 through 9
(see Exhibit B-l).
3. The same procedure was used to calculate the number of
locations that would have to report for various
definitions.
4. A significant number of reports of production were
reported in production categories BLANK and N (33
percent of the total). It is not possible to
determine whether these reports would be prepared for
a given small manufacturer exemption. Therefore, they
were excluded from the analysis at this stage. The
total number of reports dropped from 19,485 to
15,416. Also, a significant number of locations had
reported only in categories N and BLANK. It is also
not possible to determine whether these locations
would have to report for a given small manufacturer
exemption. Therefore, these locations were dropped
from the analysis at this stage. The total number of
locations dropped from 3044 to 2386.
The matrices developed in the preceeding four steps are sufficient to
count up the number of reports prepared and the number of locations reporting
for small manufacturer exemptions with sales and production parameters,
employment and production parameters, and when locations need be below both or
only one parameter to be exempted. The "percent of information missing" was
calculated by dividing the numbers by their appropriate totals, subtracting
from one and multiplying by 100.
The values for the locations and reports proxies are reported in Chapter 4.
Production Proxy. Unlike the number of locations reporting and the
number of reports prepared, the percentage of total production exempted is an
ICF INCORPORATED
-------
B-4
estimate, because the actual production at each site is not known; only the
production range is known, liie range is identified by the production category
in which the chemical was reported. Therefore, this proxy is estimated
differently than the other proxies which are actual counts of locations and
reports.
The percentage of total production exempted from reporting was be
constructed in three steps:
* Estimate total production, call this quantity T;
* Estimate production exempted under a given small
manufacturer definition, call this quantity E; and
* Use the above to estimate the percentage of total
production exempted, call this quantity P which is a
function of T and E.
Each of the three steps is discussed below.
Estimate the Total Production, T. The estimate of total production
includes only those reports of production in production categories 0 through 9
(see Exhibit B-l). The production quantities of chemicals whose production
quantities were kept confidential by the manufacturing locations and the
production quantities of chemicals reported as not produced in 1977 when the
inventory data were collected (but were produced prior to 1977 and were
expected to be produced in the future) cannot be used because there are no
production statistics available for them.
Both gross total production and total production were estimated. The
gross total production, call it GT, includes all the reports of production (in
categories 0 through 9) of all the manufacturing sites on which production
information was provided by EPA. The total production, T, includes only those
manufacturing sites for which all necessary sales and employment data are
available. Total production, T, as a percentage of gross total production,
GT, was estimated to indicate the extent to which missing sales and employment
data reduce the coverage of the data base. (The method for estimating this
percentage is described below.)
Total production, T, and gross total production, GT, were estimated in the
same way, except that fewer sites were used in estimating total production.
Total production, T, was used to estimate the percentage of production
exempted by various definitions. The following explains how total production
was estimated.
All manufacturing locations for which sales and/or employment information
was missing were removed from the Merged File. For the remaining locations,
the number of reports of production in each of the ten production categories
were counted. The result, the total number of chemicals produced in each
category, was called COUNT; COUNTq equals the number of reports of
product^n in production category 0, COUNT^ equals the number of reports of
production in production category 1, and so on.
ICF INCORPORATED
-------
B-5
For each production category, EPA estimated the mean and standard
deviation of production. The list of means for the production categories was
put in a vector called MEAN, and the list of standard deviations was used to
construct a vector called VAR, for variance.
Each production category also has high and low bounds of production. For
example, production category 1 goes from 1000 pounds to 10,000 pounds. These
values were put in two vectors called MAX and MIN.
The following operations were done to estimate total production:
9
(1) Total Production = T = I COUNT x MEAN
i=0 i i
9
(2) Variance of T = VT = I COUNT x VAR
i=0 i i
9
(3) Minimum possible value for T = MINT = ฃ COUNT x MIN
1=0 i i
9
(4) Maximum possible value for T MAXT - 1 COUNT x MAX
i=0 i i
1/2
(5) Likely lowest value for T given VT = LOWT = T - 1.96 x [VT]
1/2
(6) Likely highest value for T given VT = HIGHT e t + 1.96 x [VT]
Total production is estimated by T with a standard deviation equal to the
square root of VT. The likely range of T goes from the greater of MINT and
LOWT to the lesser of MAXT and HIGHT.
Estimate Exempted Production, E. Exempted production was estimated the
same way as total production. Only the reports of production of those
manufacturing sites exempted under the definition being examined were
included. Only the COUNT vector changes; and for exempted production, this
vector will be renamed to COUNTC. The vectors MEAN, VAR, MIN, and MAX remain
unchanged.
ICF INCORPORATED
-------
B-6
Exempted production is estimated by E with a standard deviation equal to
the square root of VE. The likely range of E goes from the greater of MINE
and LOVE to the lesser of MAXE and HIGHE. These values are estimated as shown
in equations (1) through (6); T changes to E, VT .changes to VE, MINT changes
to MINE, MAXT changes to MAXE, LOWT changes to LOWE, HIGHT changes to HIGHE,
and COUNT changes to COUNTE.
Estimate the Percentage of Total Production Exempted. There are no
simple exact formulae for the mean and variance of the ratio of two random
variables. The following approximations are taken from Mood, Graybill, and
Boes:1
(7)
_X_
y
vol 1 yx
- 2 COV [X,Y] + 3 VAR [Y],
yy yy yy
(8)
VAR
_X_
Y
Var fXl
VARfYl
+ 2
vy
2 covrx.Yi
(ux) Cvy)
In our application, E = yx, VE = VAR[X], T = vy, and VT = VAR[Y]. If
T and E are independent, then COV[E,T] = 0 and all the terms in equations (7)
and (8) are specified. However, T and E are not independent because T is
actually comprised of the sum of an estimate of E, the exempted production,
and an estimate of the non-exempted production, call it N. By this
definition, the expected value of T equals the expected value of E + N, where
E and N are independent. Therefore, the COV[E,T] can be written as:
(9) COV[E,T] = COV(E, E+N), and simplifying:
COV[E,T] = t(E2) + e(EN) - [e(E)]2 - c(E) e(N)
COV[E,T] = t(E2) - [e(E)]2
(10) C0V[E,T] = VAR(E) = VE
This result in equation (10) can be substituted into equations (7) and (8)
to complete the formulae:
(11)
P =
_E_
T
VE + 3 VT, and
T
1Alexander Mood, Franklin Graybill, and Duane Boes, Introduction to the
Theory of Statistics (New York: McGraw/Hill Publishing Company, 1974), pp.
180-181.
ICF INCORPORATED
-------
B-7
(12) VP = VAR
T
VE VT 2 VE
2 + 2 -
E T ET
, where
P equals the percentage of total production exempted and VP equals the
variance of the estimate P. The likely range for P can be estimated by:
(13) LOWP = P - 1.96 [VP]
1/2
(14) HIGHP = P + 1.96 [VP]
1/2
The range calculated in equations (13) and (14) were used unless the values
fall outside the minimum and maximum possible values which are defined by:
Lower Bound for E
(15) MINP = Upper Bound for T
Upper Bound for E
(16) MAXP = Lower Bound for T
The upper and lower bounds for E and T are found with equations (3)
through (6) as explained above. The greater of LOWP and MINP and the lesser
of HIGHP and MAXP is the appropriate range.
These calculations result in an estimate of the percentage of total
production exempted by a given small manufacturer definition and estimates of
the likely bounds on the percentage. Values for this proxy are reported in
Chapter 4.
B.3 CALCULATE VALUES FOR THE COST PROXY
The cost proxy is the sum of the number of reports prepared and the number
of locations required to report. These values were calculated for the reports
and locations proxies. To calculate the percent reduction in total cost to
industry for a given exemption, we divided the sum of the number of reports
required to be prepared plus the number of locations reporting by the total
possible number of locations and reports. This fraction was subtracted from
one and multiplied by 100 to get the percent reduction in total cost to
industry. These values are reported in Chapter 4.
B.4 ESTIMATE PERCENTAGE OF CHEMICALS MISSED
To identify those chemicals about which no information would be reported
under various small manufacturer exemptions, the list of chemicals that each
manufacturer produces, and the quantities in which each manufacturer produces
them is required. With these data, one could keep track of which locations
must report which chemicals under each exemption level, and for each exemption
ICF INCORPORATED
-------
B-8
level, the chemicals for which no reports are required could be identified.
However, these data were not available for this report. Therefore, a
simulation was run to estimate, on an expected value basis, the percentage
of chemicals that would be missed under various small manufacturer exemptions.
A simulation is an attempt to reproduce a complex system of relationships
in what is the equivalent of a laboratory setting.2 The system, generally
expressed as a series of equations, is analyzed in a systematic and
reproducible manner. A simulation was developed using the Chemical File and
some statistics from the Merged File to examine the relationship between the
number of different chemicals not reported and the levels of the small
manufacturer exemption.
The Merged File was used to estimte the distribution of reports of
production across the sales and employment categories. For example, of the
770 reports of production in production category 7, 10 percent come from
locations in sales category 6. (See Exhibit B-l for the definitions of the
categories.) Therefore, knowing nothing about the report of production other
than its production category, one can assign probabilities for the sales and
employment categories of the locations from which the report may have come
from. Given these probabilities, the Chemical File was used to estimate
whether each of a random selection of chemicals would be missed for a certain
small manufacturer exemption. This was determined as follows:
1. Examine the reports of production for that chemical.
For each report of production listed, generate a
random number. Apply each -of the random numbers to
the appropriate lists of probabilities to see which
"locations" most likely produce that chemical. The
actual locations at which the chemical is produced
are not identified, but the sales and employment
ranges in which the producers of that chemical are
likely to fall are estimated. The highest production
category in which the chemical is produced can be read
directly from the data.
2. The highest sales and employment categories simulated
for that chemical are stored. These values can be
used to determine under which exemption the chemical
would not be reported at all. The chemical is not
reported if both the sales and production quantity
parameters of the exemption exceed the highest sales
and production quantity values stored for that
chemical. (If total employment is substituted for
sales, then the chemical is not reported if both the
employment and production quantity parameters of the
exemption exceed the highest employment and production
quantity values stored for that chemical.)
*Edith Stokey and Richard Zeckhauser, A Primer for Policy Analysis (New
York: Jtf.W. Norton and Company, Incorporated, 1978), p. 89-90.
ICF INCORPORATED
-------
B-9
Once this is done for all the chemicals, the number of chemicals missed
under each exemption level can be counted. The results will vary, however,
depending on how the randoms number come up. Therefore, we ran the simulation
thirty times, each time randomly choosing 250 different chemicals from the
list of 1243. Thirty trials are statistically sufficient to ensure unbiased
expected values for the average number of chemicals missed under the various
small manufacturer exemptions. Samples of 250 chemicals for each trial were
believed to be sufficiently large to ensure representative results for the
entire list of chemicals.
The average percent of chemicals missed, of those reported din production
ranges 0 through 9, is reported for each level of the exemption in Chapter 4.
Exhibits B-2 and B-3 display not only the average percent of chemicals missed,
but also the highest and lowest percent missed in the thirty trials of the
simulation. The range is fairly large, demonstrating the wide range of
possible outcomes.
Estimates of the lower bound of the percentage of chemicals that would be
missed for the case where the location need only be below one parameter to be
exempted are presented in Exhibit 4-28 in Chapter 4. The lower bounds were
constructed by calculating the percentage of chemicals whose highest report
of production is in each of the ten production categories 0 through 9. All
the chemicals whose highest report of production is in category 0 would be
missed if the production parameter were set at or above production category
1. Similarly, all the chemicals whose highest report of production is in
category 1 would be missed if the production parameter were set at or above
production category 2, and so on. Exhibit 4-28 shows that even without taking
into account locations exempted due to sales, a large portion of the chemicals
would be missed, even for very low exemption levels, when locations need only
be below one parameter to be exempted. Therefore, a simulation was not
developed to "fine tune" the estimates of chemicals missed for this type of
exemption.
Approximately 18 percent of the chemicals were reported only in production
categories BLANK and/or N. It was not possible to determine whether they
would be missed under various small manufacturer exemptions, and they were
therefore excluded from this analysis. If these chemicals are distributed
across the production and sales categories in the same manner as the other
chemicals, then the results are generalizable to them. The resits may not be
applicable for groups of chemicals unlike those in the data base, however.
ICF INCORPORATED
-------
EXHIBIT B-2
LOWEST, AVERAGE, AND HIGHEST PERCENTAGE OF CHEMICALS MISSED FOR THIRTY SIMULATION TRIALS
SALES AND PRODUCTION QUANTITY PARAMETERS
(Location must be below both parameters to be exempted)
o
5
0
9
3
1
o
Production
Categories
1
2
3
Sales Categories
4 5
6
7
0
9
10
11
PVO
0.0
0.0
0.0
J .0
0.0
U . 0
0.0
0.0
0.0
0.0
0.0
0.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
O.J
0.0
C.O
0.0
O.C
PVl
J.O
0.0
U .0
0.0
0.0
0.0
0.0
0.379
1.571
O.C
1.65 8
3 .3d 2
0.0
1.954
4 S 3 1
0.0
2.533
5.797
1.005
3.335
9.662
1 .005
4.010
10.145
1.005
5.165
12.077
1.503
5.44 7
12."360
3.015
6.297
13.527
p vz
3.0
J.O
0.0
o .u
0.0
0.3
0.0
0.463
1.S71
0.0
1.874
3.590
e.4A5
2.629
4 .831
0.971
3.700
7.000
2.434
5.479
10.62 B
3.315
5.952
11.111
3.015
7.203
13.04 3
4.523
7 . 0ฐ 7
14.s93
6.030
9.054
15.942
P V 3
0.0
u .0
0.0
0.0
0.01c
0.474
0.0
0.529
1.571
0.0
2.221
*.103
0.971
3.25b
S.64 1
2.414
4.921
3.500
3.922
7.142
12.077
3.922
7.954
13.043
4.523
9.499
15.459
5.312
10.521
17.391
6.995
12.740
19.607
P V4
0.0
J.O
0.0
0.0
0.015
0.474
0.0
0 . t
1 .571
0.0
2.334
<..615
0.971
3.435
6.154
2.427
5 .3ฐ0
9.500
3.922
7.750
12.550
3.922
a.759
13.527
5.0J2
10.519
15.942
5.882
11.724
17.374
9.524
1 4 . 399
20.773
p v*
0.0
0.0
0.0
0.0
0 .01b
0.474
0.0
C.S46
1.571
U.O
2.337
4.615
0.971
J.485
6.154
2.451
5.575
9. *00
3.922
8.057
13.527
3.922
9. 132
14.493
5.862
10.975
17.391
5.892
12.259
19.324
9.524
15.049
22.222
P V 6
J.C
0.0
O.J
0.0
O.Uli
0 47 4
0.0
0 . 5ซ b
1.371
0.431
2.353
<..615
1.442
3.510
6.154
2.451
5.508
9.500
3.922
6.123
13.527
3.922
7.214
14.493
5.8e2
11.073
17.391
5 d 9 2
12.356
I9.32t
9.524
15.152
22.222
P V7
0.0
0.0
0.0
0.0
0.016
C.474
0.0
0.5*6
1.571
0.481
2.353
4.615
1.442
3.5ie
6. 154
2.451
5.608
9.500
3.922
6.123
13.527
3.922
9.231
14.403
5.662
11.090
17.371
5.382
12.373
19.324
9.524
15.227
22.222
PV3
0 .0
0.0
0.0
0.0
0.01b
0.474
0.0
0.54b
1.571
0.24
15.227
22.222
-------
EXHIBIT B-3
LOWEST, AVERAGE, AND HIGHEST PERCENTAGE OF CHEMICALS MISSED FOR THIRTY SIMULATION TRIALS
EMPLOYMENT AND PRODUCTION QUANTITY PARAMETERS
(location must be below both parameters to be exempted)
Employment Categories
Production
Categories
o
3
o
I
o
PVO
PV1
PV2
P V 3
PV4
PV5
PV6
PV7
p ve
? V">
0.0
O.C
0.0
0.0
0.0
0.0
0.0
0.0
U .0
o.o
0.0
o.o
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
J.O
O.G
0.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.016
0.474
0.0
J.016
0.474
0.0
0.016
0. 47ซt
0.0
0.016
U . 4 T 4
0.0
0.016
U.'tM
0.0
0. U16
0.474
0.0
0.01b
0.474
0.0
0.0
0.0
0.0
0. 37?
1.571
0.0
0.463
1.571
0.3
0.52 9
1.571
0.0
O. 546
1.571
0.0
0.546
1.571
0.0
0. 546
1.571
J.O
0. 5'ป6
1.5/1
0.0
0. 546
1.571
0.0
0. 546
1.571
0.0
0.0
0.0
0.0
1.653
3.302
0.0.1.b74
3.59C
0.0
2.22 1
4.103
0.0
2.334
4.615
0.0
2. J 37
4.615
0.4? 1
2.353
4.615
o.-*ni
2.353
4.615
0.401
2.353
4.615
G.'Sl
2.333
<>.615
0.0
0.0
0.0
0.0
1 .954
4 .331
0.465
2.629
4.831
0.971
3.256
5.641
0.971
3.435
6.154
0.971
3.485
6.154
1.442
3.518
6.154
1 .442
J. 518
6.154
1 .442
3.518
6.154
1. *4 2
3.518
6.15ซ.
0.0
0.0
0.0
0.0.1. 533
5.777
0.971
3.700
7.000
i .404
4.) 21
8 . bOO
2.427
5.378
9.50C
2.451
5.575
9.500
2.451
5.608
9.50U
2.451
5.60C
9.500
2. .51
S.oOa
9.50C
2.4M
J. b09
V .500
O.C
U.O
O.C
1.005
3.d35
9.662
2.404
5.47?
10.628
3.922
7.142
12.077
3.922
7.750
12.560
3.922
G.057
13.527
3.922
8.123
13.527
3 V 2 2
8.123
13.527
3.922
8.123
13.527
3.922
3.123
13.577
1.005
4 . 0"0
10.145
3.015
5. 952
11.111
3.922
7.954
I 3.043
3. 922
d.759
13.527
3.922
9. 132
14.*93
3.922
9.214
1 4 9 3
3.922
9.231
14.493
3.922
9.231
14.493
3.922
9.231
14.493
-------
APPENDIX C
ANALYSIS OF A HYPOTHETICAL REPORTING RULE
The purpose of this appendix is to demonstrate that the conclusions
regarding the most cost-effective small manufacturer definition developed in
this report are not necessarily generalizable to all reporting rules. We
demonstrate this through the construction of a hypothetical reporting rule in
which the conclusions would not hold, thereby proving that they do not hold
for all reporting rules.
Our hypothetical reporting rule covers 7 locations, which generate IB
reports among them. Each point on Exhibit C-1 represents one of those
reports. The location at sales level A must submit reports on 5 substances,
whose production volumes are distributed as shown, and the other reports are
distributed among the six remaining firms. Assuming that firms must meet both
sales and production criteria to qualify as a small manufacturer, two location
isoquants and one report isoquant are shown.
In order to interpret the effect of a particular pair of sales and
production criteria, form an ordered pair of the two criteria and locate the
point on the graph. If it is below and to the left of the L = 7 isoquant, all
seven locations are included under those criteria. If the point lies between
the L = 7 and L = 6 isoquants, six locations are included under those
criteria. If the point is below and to the left of the R = 14 isoquant, at
least 14 reports are received under these criteria.
In our study, we found the cost-effective solution, when the firm must
meet both criteria and the number of reports is the measure of information, to
be in the "southeast" corner of the graph, no matter what the relative
coefficient values. But this is not the case here. Setting a = 0 in the
cost equation, as b increases in size relative to c, the cost isoquants
approximate the location isoquants more and more closely. If the cost
isoquants were equal to the location isoquants, and at least 14 reports were
required, the cost minimizing solution would fall in the shaded area in the
center, rather than in the "southeast" corner. Therefore, for some values of
b sufficiently large in relation to c, the cost-effective solution is in the
center.
ICF INCORPORATED
-------
HYPOTHETICAL EXAMPLE OF REPORTING RULE IN
WHICH COST-EFFECTIVE SOLUTION DOES NOT
LIE IN THE "SOUTHEAST" CORNER WHEN
REPORTS ARE USED AS A PROXY FOR
INFORMATION
(Firm must be below both parameters to be exempt)
L=6
R = Reports Isoquant
L ซ= Locations Isoquants
ooo
Production
-------
TECI-iMICAL REPORT DATA
'Picssc reju Jixniciiuiu on me rei cnc ocior; comnUtbtj/
1 Ri'OnT NC
4.
J. -z.CP.cHTS ACCESi.O.VNO.
ฆi TITLE AND SUBTITLE
Analysis of TSCA Section 8(a) Small Manufacturer
Exenptim
5 REPORT DATE
April 1982
o. PERFORMING ORGANIZATION CODE
7. AUTHOftlSI
Michael J. Gibbs, Joseph L. Kirk, Kenneth B. Kolsky,
Prank Lerman
3. PERFORMING ORGANIZATION '.EPORT n:
i. PERFORMING CRGANI.ZATION NA.\iฃ AND AODRESS
ICF Incorporated
1850 K Street, NW
Washington, D.C. 20006
10. PROGRAM ELEMENT NO.
B2CL2S
11. CONTRACT/GRAJvI't no.
1ฃsk Order No, 3
No. 68-01-6287
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 ANO PERIOD COVERS:
14. SPONSORING AfffeNitY CODE
'.s. supplementary -notes
EPA Project Officer: Scott Peters/Michael Shapiro
This report presents analyses in support of EPA's development of a email
manufacturer definition required for Section 8(a) of the Toxic Substances Control
Act. Under Section 8(a), manufacturers designated as "small" are exempt from
certain reporting requirements. The report develops a framework for balancing off
reporting costs with the benefits of receiving information from manufacturers.
Production data are used in examples of how this balancing of costs and benefits can
be performed for various types of small manufacturer definitions.
KEY WORDS AND DOCUMENT ANALYSIS
r3 DESCRIPTORS
b.lOENTlFlERS/OPEN ENDED TERMS
c. COSATi Field 'Crour
Toxic Substances Control Act, Chemical
Industry, Reporting Exemptions
1 i. DIBTHI3U7 'JN STATEMENT
Release Unlimited
*9 5-Cufll T v CLASS ; rms Reportj
Unclassified
21. NO. OF PAGES
126
SO. SECURITY CLASS (Thupugef
Unclassified
22. PRICE
EPA Form 2220-1 19.73)
------- |