•A-453/R-94-066-E
uiuicu ouu.es Office of Air Quality
Environmental Protection Planning and Standards
Agency Research Triangle Park, NC 27711
EPA-453/R-94-066-E
March 1995
Air.
& EPA
STUDY OF
VOLATILE ORGANIC COMPOUND EMISSIONS
FROM
CONSUMER AND COMMERCIAL PRODUCTS
Economic Incentives to Reduce VOC Emissions
from
Consumer and Commercial Products
-------
EPA-453/R-94-066-E
STUDY OF
VOLATILE ORGANIC COMPOUND EMISSIONS
FROM
CONSUMER AND COMMERCIAL PRODUCTS
V
•x
t
ECONOMIC INCENTIVES TO REDUCE
VOC EMISSIONS FROM
CONSUMER AND COMMERCIAL PRODUCTS
March 1995
-------
ACKNOWLEDGEMENT
The Study of VOC Emissions from Consumer and Commercial Products was
completed with a great degree of cooperation and assistance from the Chemical Specialties
Manufacturers Association, the Cosmetic, Toiletry, and Fragrance Association, the Soap and
Detergent Association, the Automotive Chemical Manufacturers Council, the National
Aerosol Association, and the Adhesive and Sealant Council.
111
-------
TABLE OF CONTENTS
Section Page
EXECUTIVE SUMMARY ix
1.0 INTRODUCTION 1-1
1.1 Significant Characteristics of the Pollutants 1-2
1.2 Policy Context 1-3
1.3 Consistency with Economic Incentive Program Rule 1-5
1.4 Design Considerations for Economic Incentives 1-11
1.5 Scope of the Report 1-12
1.6 References 1-13
2.0 CONSUMER AND COMMERCIAL PRODUCTS AND ECONOMIC
INCENTIVES 2-1
2.1 Information Needs .2-2
2.2 Sources of Emissions and Implications for Regulatory Design 2-3
2.3 Product Diversity, Product Innovation, and Advantages of Economic
Incentives 2-4
2.4 Product Performance and Advantages of Economic Incentives 2-7
2.5 Potential for Reducing a Large Proportion of Emissions with a Small
Universe of Sources 2-8
2.6 Existing Regulatory Requirements for Consumer and Commercial
Products and Implications for Regulation 2-13
2.7 References 2-14
in
-------
TABLE OF CONTENTS (CONTINUED)
Section Page
3.0 FEE PROGRAMS FOR REDUCING VOC EMISSIONS FROM
CONSUMER AND COMMERCIAL PRODUCTS 3-1
3.1 Purposes of a Fee Program 3-2
3.2 Economic Basis of the Fee 3-4
3.3 Market Relationships and Responses 3-6
3.4 Fee Rate Variability 3-8
3.4.1 Season 3-8
3.4.2 VOC Content 3-9
3.4.3 Location 3-11
3.4.4 Emission Reduction Progress 3-13
3.5 Initial Fee Rate 3-13
3.6 Fee Adjustment Mechanism 3-15
3.7 Universe of Sources 3-21
3.7.1 Selection of Sources by Type of Product: Sequence of Regulation 3-22
3.7.2 Selection of Sources by Type of Business 3-22
3.7.3 Selection of Sources by Size 3-26
3.8 Collection of Fee Payments 3-28
3.9 Monitoring and Enforcement 3-28
3.10 Use of Revenues Generated by Fee Programs 3-32
3.11 Information Needs 3-39
3.12 Summary of Design Options 3-42
3.13 References 3-46
IV
-------
TABLE OF CONTENTS (CONTINUED)
Section Page
4.0 EMISSIONS TRADING PROGRAMS FOR REDUCING VOC EMISSIONS
FROM CONSUMER AND COMMERCIAL PRODUCTS 4-1
4.1 Purposes of a Marketable Permit Program 4-4
4.2 Fundamental Attributes of the Permit 4-7
4.2.1 Detraction 4-7
4.2.2 Ambient Permit or Emission Permit 4-9
4.2.3 Emission Allowances or Emission Reduction Credits 4-10
4.2.4 Banking Provisions and Duration of Permits 4-11
4.3 Seasonal and Geographical Differentiation of Permits 4-12
4.3.1 Seasonal Differentiation 4-12
4.3.2 Geographical Differentiation 4-13
4.4 Universe of Sources 4-15
4.4.1 Selection of Sources by Type of Product and Sequence of
Regulation 4-16
4.4.2 Selection of Sources by Type of Business 4-17
4.4.3 Selection of Sources by Size 4-18
4.4.4 Opt-Ins 4-20
4.4.5 Participation in a Marketable Permit Program by Environmental
Organizations 4-21
4.5 Entry and Retirement of Sources 4-23
4.5.1 Entry of New Sources 4-24
4.5.2 Retiring Sources 4-27
4.6 Initial Allocation of Permits: Grants and Auctions 4-28
4.7 Monitoring and Enforcement 4-35
4.8 Information Needs for Program Design and Administration 4-37
4.9 Ancillary Projects to Increase Trading in a Permit Market 4-39
4.10 Emissions Averaging Programs 4-40
4.11 Summary of Design Options 4-43
4.12 References 4-49
-------
TABLE OF CONTENTS (CONTINUED)
Section Page
5 COMPARISON OF EMISSION FEE AND EMISSION TRADING
PROGRAMS 5-1
5.1 Program Costs and Uncertainty 5-2
5.2 Monitoring and Enforcement 5-8
5.3 Distributive Flexibility 5-9
5.4 Adaptation to Economic Growth 5-11
5.5 Incentives for Technological Innovation and Diffusion 5-12
5.6 Unintended Damages 5-15
5.6.1 Intermedia Transfers of Pollutants 5-16
5.6.2 Increased Health Risks 5-17
5.6.3 Locally Increased Air Pollution in the Near-Term from Nonlinear
Atmospheric Chemistry and Permit Trading 5-18
5.6.4 Increased Air Pollution in the Long-Term from Product Life-Cycle
Effects 5-20
5.7 Combination VOC Limit/Economic Incentive Regulatory Strategies 5-20
5.8 Summary 5-21
5.9 References 5-24
VI
-------
LIST OF FIGURES
Figure Page
1-1 Comprehensive Classification of Emissions from Consumer and Commercial
Products 1-2
1-2 Hypothesized Relation Between Economic Incentives and Economically Feasible
Environmental Goals 1-6
3-1 The Basic Principle of Emission Fee Programs 3-3
3-2 Flat and Variable Rate Fee Structures 3-10
3-3 Emission Fee Should Be Greater Where Benefits Are Greater 3-12
3-4 Markets Where Emission Fee May Be Imposed and Entities on Whom Fee May
Be Levied 3-23
3-5 Incidence of Fee 3-25
4-1 Emission Permit Market 4-5
5-1 Welfare Losses When Costs Are Uncertain—No Effects Threshold 5-4
5-2 Welfare Losses When Costs Are Uncertain—Effects Threshold 5-5
vn
-------
LIST OF TABLES
Table Page
1-1 Required Economic Incentive Program Elements 1-9
2-1 Concentration of Production in Consumer and Commercial Product Industries 2-10
2-2 Production of Primary and Secondary Products within Potentially Regulated
Industries 2-11
2-3 Primary Product Specialization in Potentially Regulated Industries 2-12
3-1 Fee System Without an Adjustment Mechanism 3-17
3-2 Fee System with an Adjustment Mechanism 3-18
3-3 Example of Proposed Rebate Scheme 3-35
3-4 Uses of Fee Revenues 3-37
3-5 Information Needs for the Design of Emission Fee Programs 3-41
3-6 Information Needs for the Administration of Emission Fee Programs 3-42
3-7 Summary of Fee Program Design Options 3-43
3-8 Example Fee Program Designed to Minimize Adverse Impacts on Industry 3-44
4-1 Information needs for the Design of Marketable Permit Programs 4-38
4-2 Information needs for the Administration of Marketable Permit Programs 4-38
4-3 Summary of Trading Program Design Options 4-44
4-4 Example Marketable Permit Program Designed to Maximize Incentives for
Technological Change 4-47
5-1 Comparison of Economic Incentives and Hypothetical VOC Content Standards
When the Environmental Goal is a Fixed Quantity of Emissions 5-23
viu
-------
EXECUTIVE SUMMARY
This report presents a preliminary assessment of the feasibility and desirability of
employing Federal economic incentive programs to reduce volatile organic compound (VOC)
emissions from the use of consumer and commercial products. This investigation of
economic incentives responds to the mandate to apprise Congress of the desirability of
different regulatory strategies that may be appropriate for consumer and commercial
products.
The principal tasks of the study are to examine alternative economic incentives and to
compare them to a hypothetical command-and-control program, VOC content standards,
which would consist of product-specific limitations on maximum VOC content (grams of
VOC per unit of product). It is the basis of comparison because the ultimate purpose of this
investigation is to search for the most desirable instrument in the set of potential instruments,
which obviously would include instruments based on command-and-control.
The purposes of comparison are to determine how well the instruments accomplish
certain policy objectives and to appraise their ability to cope with the complexities inherent
in the task of environmental regulation. The specific bases of comparison are the following:
program costs and initial information requirements when effects are uncertain,
• monitoring and enforcement,
• flexibility in distributing the economic impacts of regulation,
• adaptation to economic growth,
• incentives for technological innovation and diffusion, and
• unintended damages.
These criteria ensure a fairly comprehensive examination of the issues involved in designing
environmental policy instruments. Further, a broad basis of comparison more fully reflects
differences between economic incentives and command-and-control approaches.
IX
-------
This report does not serve as the final assessment of economic incentive strategies to
reduce VOCs from the many consumer and commercial products that may ultimately be
regulated. Rather, it serves as a general assessment of the most prominent economic
incentives for reducing air pollution, and it serves to set the stage for the future development
of more detailed and industry-specific economic incentives.
The Environmental Protection Agency's authority for the study and regulation of
VOC emissions from consumer and commercial products derives from Title I, Section 183(e)
of the Clean Air Act Amendments (CAAA) of 1990. Section 183 (e)2(A) instructs the EPA
to conduct a study of VOC emissions from consumer and commercial products to determine
their contribution to ozone levels, to establish criteria for their regulation, and to assist with
determining the sequence in which products shall be regulated. Section 183(e)4 explicitly
states that the EPA may consider economic incentive regulatory strategies.
Economic incentives may be particularly appropriate for the control of VOCs from
consumer and commercial products. The emissions source of concern is the use of the
product but not the manufacture. This aspect of the pollutant leads to the existence of a great
many small sources for which "end-of-pipe" control is practically infeasible. The products
are very diverse. The types of consumer and commercial products and product forms number
in the thousands. Product diversity reflects the adaptation of products to consumer demands,
which vary from region to region and change from one year to the next. The core of the
dynamics of consumer and commercial product industries is product innovation—each year,
manufacturers and importers introduce hundreds of new products. In these circumstances,
economic incentives may have great advantages over command-and-control strategies, such
as VOC content standards, because (1) the Agency may need to accumulate less information
on product technology and on performance and (2) manufacturers have more flexibility to
find the best balance between reducing VOC emissions and satisfying consumer demands.
-------
VOCs have specific characteristics that influence the design of a regulatory program.
VOCs are assimilative and uniformly mixed implying that emissions do not accumulate in
the atmosphere and that the concentration of ozone is independent of the location of sources
in the airshed. Therefore, efficient economic incentive programs need only to target current
total emissions of the pollutant in an airshed.
Emission fees and emissions trading are the economic incentives most commonly
proposed for the purposes of air pollution control. A comparison of these designs makes
explicit the requirement for the consideration of certain fundamental policy issues, including
the tradeoffs between: (1) certainty over the cost of emission reductions versus certainty over
the quantity of emissions reduced, (2) the cost to control emissions versus administrative,
monitoring, and enforcement costs, and (3) distributional flexibility versus the provision of
incentives to advance technology. This report examines these tradeoffs in some detail.
Emission fee programs can be used to obtain real and quantifiable reductions in the
emissions of VOCs from consumer and commercial products. The basic rationale for these
programs (and economic incentive programs in general) is to bring the full "social cost" of
using VOCs into the economic system. A fee on the emission of VOCs would increase the
cost of using the atmosphere as a waste sink. The fee would have the same effect as the
prices for the goods and services exchanged in conventional markets: for example,
manufacturers would economize on their use of VOCs because the "price" of using VOCs
would be higher, just as they would economize on the use of labor if wage rates were to
increase.
Designing a fee program to reduce VOC emissions from consumer and commercial
products presents many choices. Many of these design options are summarized in Table E-l.
XI
-------
Tf\
V/J
0
E
o
2
a
S5
u
a
<
a
U
o
f^
K
ft.
U
U
fc.
fe
o
0£
<
*
Z
=3
c«
in
c
^o
^*
Q.
O
•^f
e
0
'a
a.
O
«*^
JO
Q.
O
|>
"o
•<
w
rt>
•S
Product demai
Product supplier
Sources above size
cutoff
d) O
_ O
3 en
t ^
•a
cu
•o
IS I
* 5
0 »
0 u
b U
£ u
^ s
es O
ft. CA
c
p
_ o
2 3
-"S
s =
21 en
S § S
^•i g
•C e« go
i'es
!> a> a.
MI
3
cS
e15
With respect t<
nonattainment
O
O
0
o
8,
CO
a 81
'I Si
a ^8
e
CO
w
O
§
8 €
Z ^
8
w
0 -
^ E
£ €
« "S
= .•
a t
£ £
w<
o
e
0
Ifl
lo O
If
UE
Negotiation
&>
Value of Federal/Stat
marketable emission
rights
"3
Economic mod
8fi_
£^»
IS
u«
j£
"§1
«£
y
o
>
s<
5E
is
!Ǥ
O en
C_> On
C
tyj «5
_o
S
t*.
•^rf
1
8
S<
i^U
8
£
•3
e
_o
"S
o>
s
en
C
o
3
O.
O
V-
O
c
o
1
3
1
U
•8-S
•-u
|o
%>
51
a 13 a
•SgfM
fifi*
sa o.
EPA verifies reports
en
U
C s
Q fi/3
Sources self-re]
VOC content &
BO
e
|
°£
o
U_i
O
C
_o
«
f"
.z en
.E C
P o
C r*a
at
w c
fen O
C?3
Is
1_ ">
£'2
>.E
£1
8
Rebate revenu
Fund labeling and/or
education programs
i M
BJ>
0
>i
al
O 2
=y oo
otf e
=1
£l
i
1
4>
K
*s
o
CU
D
Xll
-------
Quantity-based, emissions trading programs are alternative ways to generate
economic incentives to obtain real and quantifiable reductions in the emissions of VOCs
from the use of consumer and commercial products. The two types of trading programs
studied in this report are marketable emission permit programs and emissions averaging
programs. Particular emphasis is given to emission permit programs because of their unique
advantages and other unique characteristics.
The distinguishing characteristics of marketable emission permit programs are inter-
source trading of emissions and the cap on the aggregate, absolute quantity of emissions.
Unlike emission fee programs, VOC content standards, and emissions averaging programs,
the absolute quantity of emissions is fixed in marketable permit programs. The restriction
ensures the achievement of the environmental goal regardless of changes in the economy
including the expansion of existing and entry of new sources (e.g., companies) of emissions.
Emissions averaging programs also rely on emissions trading and are most cost-effective
when trades occur between companies, but they characteristically place an upper limit on
emission rates instead of total emissions. The nature of the environmental goal is an essential
difference between marketable permit programs and emissions averaging programs.
Table E-2 summarizes many of the design options for marketable emission permit programs.
In all emission trading programs that allow external trades, companies have the
incentive to seek potential trading partners. Opportunities for trade exist when the
incremental or marginal cost of emission reductions differ in the industry. The consumer and
commercial products industry is diverse, and manufacturers very probably will reduce
emissions through unique combinations of product reformulation, packaging redesign, and
reducing production. Therefore the marginal cost of emission reductions is likely to vary
from company to company, and the company for which a reduction is more expensive will
xin
-------
c«
Z
O
E
o
Z
o
53
u
Q
i i
a U
Ed O
" 8s
« fc
< 0
H Z
M
Q
i
&-
o
>-
PS
^t
^,
*-<
2
S
c«
V)
5;
se
"a
O
T
e
^
"a
O
f>
c
o
O
n
.0
"Si.
O
»•«
e
1
0
*•*
1
U
S
Ml
•••
V)
V
Q
<4_
O
c
_o
S
.£ f=
•S §
S-2
0 Q.
U o
o
w
>
-a ~5
a c
u .2
.S 3 *
E .0 o
•= S g.
2^ E
«e5 ^ .2
Provide greatesi
incentives for
technological
change
Ut
Achieve VOC
reductions at lowe
cost
VI
1
£
S
0£
S
cu
Average rate of
emissions
i*.
Aggregate mass o
emissions
•*^
2
C/3
Emission
CS
Lmt
4J
"><
U
T3
es
"3
C t«
C u
U T3
£ 2
C/3
U
T3
2
13
c
_>^
O
M
C
^
2
H
*S
^
o
w
1/5
,*J
E
Uri
8.
"3
3
C
C
'e
By nonattainme
status
o
en
a
o
CO
>•>
CQ
C
JO
is
«^
e
Permit
Differe
Based on size
(emissions)
<<-.
0
8,
>,
*w»
§ S
"8 1
cS "^
CQ ^
•B
W
1
y
Sources 1
"^
o
't/3
cn
'E
U
5
sn
>,
03
U
o
S
i
o
W3
e
Phasing-i
U-
O
C
_o
"a
._ tn
J3 C
e 2
*• "^2
^^ *J
u &
« S.
C 3
u o
E 00
jj
O t«
S y
i|
ttJ .S
= S
« J
S2 o
a «*-
S O
5 2
£ £
3 5
§ 1
5 E
^^« ••*
•a
c
Other C&CP
manufacturers i
"Small" sources
13
c
u
•tari
X
U
& M
90 g.
if
ic
0
00
g
c
a
o
u
U -o
51
1^ O*,
importers
UM
o
u ^2
"a 'E
"5 £
u u
0 Q.
'fe.'O
"r g
•Q &
•u o
X M
t£ g
MM krt
4.^
Reserved permi
auction
Purchase permits
from existing
sources
£
|
o
C0
£
u
1
5
>*^
XIV
-------
«o
*^
1
H
K
T
I—I
0
iS
/•— *
Q
W
3
Z
M
H
Z
0
u
C/3
0
E
o
z
r h
U
53
u
Q
3
rf5
ROGRA
C.
O
Z
3
<<
«
tu
O
>•
SUMMAR
V}
c
_o
H.
O
•v
JO
•+*
0.
0
fO
gj
_o
"a.
O
^
s
JU>
"K
u
Q
5
on
1
on C
£ £
O O
Z3 *0
a =
. c ^«
d. r^ a »-
3 .^ ^"*
tfl £ «5 O
U b c^! • —
a g.-s % •«
= s E '= ^
c^ « 8. w §
D
U
J2 S Q
Si II
l« «r
«« ^ o .S
11 II
c8 o. m C
c £
1 1
<_ ^ <_ £
o *s o e
y is
H H
fft
tv
2
C
C
K!
I
O
1
U
C3
U x-N
= "«
« 2
« 1
oi v5
^>,
Zero-revenue
auction (annua
1
1 s
n w
= 'c
1 •§ Jf
o — <,
— = Q*
< D ia
o
o.
None
Limited
Sources self-re
.1 1 1
^fi C8 w
C U Bl
C5 = C
ca Q •£
«> -w O
i i -1
u u c
U W M
a, cu S
C/i
_O
^^
&
^^
W5 C«
•1 §
C3 «-^
^^
«« E
M0
IB
^ >
^
^\J
^i
u
0
u
U y.
0-s
> S
«rf
V
orcem
(M
K
i—
o
c
_o
ffS
.£ ™
£> c
E.2
O Q,
U o
1<-I
O
U U
3. a
•a si
U as
J= C
u u
vi a.
>,
"3
u
o.
"«
c
1
'C
U
.2 ««
o c3
^1
21
«>f
c • —
S S
is °-
O.S
>»
Financial pena
V
u
c
« va
ii
0 «
u C
c ^
o Cu
z
VM
r*
4O
^_<
ca
c ..
« c«
JD C
E-2
r9 0-
U o
S
C3
>
C.
O t.
>, °
1 3
Q£ M
2
^ c
o.2
u •*-•
'o1 «
is «"
O. c
-.*. CO
^-
— U
rs u
M-i
< s
c
_0
'•S
3
u
a
b
"S
^£
u
CJ
<5
&
o
c
o
J3 =
E-2
a&
u
c
u
i
&l
II
«- .S2
,0 c
Si
"
i|
ail
8
XV
-------
pay another company (i.e., one with lower costs) to reduce emissions. The benefit to the
former company is reduced costs, and the benefit to the latter, increased profits.
Table E-3 compares the performance of economic incentives and hypothetical VOC
content standards on the criteria listed above. This criteria, though not exhaustive, provides a
reasonable basis for comparing alternative strategies. Table E-3 assumes that every program
aims to achieve the same environmental goal, a predetermined reduction in VOC emissions.
The table includes the most important variants of fee and tradable permit programs: emission
fee programs with and without rebates, and permit programs in which the EPA allocates
permits to existing sources through an ordinary auction, zero-revenue auction, or grant. The
ranking of each program on a given criterion indicates relative performance and may not hold
in every circumstance.
Abatement Costs. In theory, marketable permit programs with unrestricted trading as
well as fee programs will lead to the least costly emission reductions. In practice, however,
none of these strategies will likely achieve the truly least cost solution that might be
obtainable in a world of zero transaction costs and perfect certainty. In marketable permit
programs with freely granted permits, transaction costs may restrict trading and hence reduce
potential abatement cost savings, though costs of obtaining the same level of emission
reductions would never exceed those resulting from content standards. Permit auctions may
reduce the need for trading, but periodic auctions may impose uncertainty costs. Similarly,
fee programs with adjustable fee rates may impose planning uncertainty or transition costs.
Nevertheless, economic incentive strategies will generally result in reduced abatement costs
compared to content standards because economic incentives provide the flexibility to obtain
emission reductions from the least costly sources.
xvi
-------
c0
Q
tf
^
<;
Q
Z
NTENT STA
O
U
u
o
j
<
u
H
[=2
BLE E-3
,ND HYPOTH]
H
Z
Cd
U
Z
M"N
U
i
o
z
o
u
H
tt,
o
z
o
50
»•*
»>
Hi
<<
£
s
o
u
EMISSIONS
F^
fa
O
UANTITY
O*
Q
U
X
S
<
C0
NM
«
_^(
<,
O
r ^
U
U
g
JS
z
o
OS
»M
^
H
K
B
H
Z
td
*^^^
£
•*rf
e „
a -3
c «-
o «
u-g
u 5
OCA
^
u
•3
2
H
Z
£
Emissions
Averaging
^g
n o
c ts
15 w
^J
O^
Free Permits
and Zero-
Revenue
Auction
xl
5-2
?jS
S6i
3 2
11
5fi
es
9)
«*
'C
u
^
I
T3
8
o
o
t.
2
6
•o
o
o
u
O
•o
8
o
(/I
'w
o
U
<*-*
ff\
{iuf
3
C3
p
•^j
<
'c
u
Ts
o
u
1
^
*/OC content «!
sales
^-
«««
«^ p4
C E U5
U U 00
c &£
8^2
U«3 2
ox;
O-«
> M
^.^
e E w
u u oo
c S..S
8*1
U S^
0-3
^ 5/5
K^
Js>
^
'c
u
*5 c«
8^
U W3
o
^J
s M
o-S
8 s
u
o
00
0
'c
o
s
•o
4J
'i
j
1
'E
j
T3
U
^^
'E
j
es
Ji
0
'S
H
o
Limited
>,
^3
is
'S
u
tu
^ 's;
^ c
ISL
CO X
u
.
Slow&
expensive
Automatic
Automatic
£ 0
.2 .S:
VI 't/5
>,§
_D O.
'is
^^
Bu ^^
^
"3 '55
>>£
s S-
S 5
o$J
ON ^
U
E
o
o
rH
w
0
I
ca *>
i- feA
t^t oy
&. c
CO K3
•o ^
u
VI
u
e
00
o
£
MM
ft "«
<2 .0
C/3 '5b
"o
s|
0 ij
£H
JU
IE
cs
XI
1
Possible
Possible
Possible
4)
33
'i«
-------
Monitoring and Enforcement. For the purpose of comparison, nothing is lost by
assuming that the penalty for noncompliance and probability of detecting noncompliance
ensure a reasonable level of deterrence. With equal levels of deterrence, the comparison of
policy instruments is a matter of differences in the activities and hence costs of detecting
noncompliance. The economic incentive programs all require the collection of companies'
reports of VOC content in the regulated consumer and commercial products and report
verification; verification involves product sampling and reviewing product sales information.
Marketable permit programs also require the Agency to track permit holdings. VOC content
standards are less demanding than economic incentives because the Agency does not
absolutely need to obtain sales information to monitor compliance nor does it need to track
permit holdings. Sales information, however, would be necessary to monitor program
performance.
Distributive Flexibility. At issue is not whether one pattern of distributive effects is
preferred to some other pattern but rather the dialogue over distributive considerations. The
ranking of policy instruments indicates the degree of flexibility to incorporate distributive
values into instrument design. Fee programs with rebates and marketable permit programs
with freely distributed permits or permits distributed through a zero-revenue auction may
serve to minimize adverse distributional impacts on industry.
Adaptation to Economic Growth. All instruments can adapt to the changes in the
economy that may strain the ability of the instrument to achieve the environmental goal, but
the quickness of the adaptation and the costs to the EPA and regulated companies differ
substantially. All permit programs adapt without action by the Agency because of the cap on
emissions and hence are ranked the highest.1 All fee programs, even those with formulas for
'Permit banking requires a qualification of that judgment because aggregate emissions will exceed the cap in a
year when previously banked permits are withdrawn and used, everything else equal. Nonetheless, without
action by the EPA, the annual average emissions equals the emissions cap.
xvm
-------
the automatic but potentially sluggish adjustment of the fee rate, are likely to require the EPA
to take corrective action when economic growth leads to an undesired increase in VOC
emissions in nonattainment areas. Emissions averaging programs and VOC content
standards would require the EPA to tighten the standards and, if new consumer and
commercial products have appeared, to issue new standards; adaptation is likely to be slowest
and Agency expenditures greatest.
Incentives for Technological Change. The rankings are sensitive to the incentives
that companies have to innovate, sell, and adopt cost-reducing technologies for reducing
VOC emissions from consumer and commercial products. All of the economic incentive
strategies promote technological innovation to a similar degree, but some strategies promote
technological diffusion more than others. As a class, economic incentives are superior to
standards with regard to promoting technological progress.
Unintended Damages. Narrowness in the design of traditional or economic incentive
regulatory programs may cause problems during implementation: intermedia transfers of
pollutants, increased VOC emissions due to inferior product performance, increased health
risks, and increased air pollution in some areas. These unintended consequences are possible
with all the policy instruments under review. Some of these potential problems, such as
those due to product performance, appear to be more probable and to have greater avoidance
costs with standards because companies have reduced flexibility to modify their products
under the restrictions imposed by standards.
Conclusion. Overall, economic incentive programs appear superior to VOC content
standards as strategies to reduce VOC emissions from consumer and commercial products.
However, content standards may still be preferable if monitoring and other implementation
costs are excessive.
XIX
-------
The choice between traditional and economic incentive regulatory strategies is not
necessarily an all or nothing choice. Combination content limit/economic incentive
approaches are possible and may provide significant advantages, though these advantages
need to be weighed against the shortcomings they may create.
The optimal selection of a regulatory strategy will depend on the specific
characteristics of the universe of sources being regulated. For example, potential abatement
cost savings, administrative and monitoring costs, and distributional implications of
employing economic incentive strategies to regulate different industries will vary
substantially.
The optimal selection of a regulatory strategy will also depend upon the program's
objectives. For example, if stimulating technological advancement is considered of most
importance, a marketable permit program with an ordinary auction might be most preferable.
Alternatively, if distributional considerations are considered more important, a permit
program with freely granted or zero-revenue auctioned permits, or perhaps a fee program
with rebates, might be most preferable. Similarly, marketable permit programs might be best
if certainty of emission reductions is of most importance (especially if product performance
problems exist), but might be the worst if protecting consumers from potential future product
price increases is of most importance. Tradeoffs are intrinsic to policy design, and the best
regulatory strategy will depend upon the particular universe of sources being regulated and
the priority of objectives.
xx
-------
SECTION 1.0
INTRODUCTION
This report studies the feasibility and desirability of using economic incentives to
reduce the emission of volatile organic compounds (VOCs) from the use of consumer and
commercial products. The report is not intended to serve as a final assessment of alternative
economic incentive strategies. Rather, it is intended to serve as a general, preliminary
assessment of the most prominent economic incentives and to set the stage for the future
development of more detailed, industry-specific economic incentive programs.
Section 183 (e)(l)(B) of the Clean Air Act Amendments of 1990 defines a consumer
or commercial product as "any substance, product (including paints, consumer and
commercial products, and solvents), or article (including any container or packaging) held by
any person, the use, consumption, storage, disposal, destruction, or decomposition of which
may result in the release of volatile organic compounds." Thus, generally speaking, the
environmental purpose of regulation is to reduce the flow of VOCs into the atmosphere from
consumption and disposal. Figure 1-1 shows the dissipative emissions and the disposal
emissions into the air that would be the target of regulation, distinguishing them from the
production-related emissions that would be beyond the scope of the regulation (as would
emissions to land and water).
The Environmental Protection Agency's mandate to regulate consumer and
commercial products channels the design of economic incentives in several important ways,
as do the characteristics of the pollutant. This section of the report describes the context of
the economic incentives that could constitute the regulatory strategy, or one component
thereof, to be promulgated under Section 183(e) of the Clean Air Act as amended.
1-1
-------
agro-chemicals
Industrial/energy
production
Imports
mm. m
Materials use & consumption
(pollutants embodied in commercial
and consumer goods)
Point- and area-source
production emissions
foodstuffs
Oissipative
emissions
Disposal emissions
Sewage
Agricultural
production
application to soils of
agro-chemicals
Area-source
agricultural
emissions
Incineration
Landfill
AIR, LAND, AND WATER
Note: Shaded area indicates consumption-related emissions.
Figure 1-1. Comprehensive Classification of Emissions from Consumer and
Commercial Products
Source: Adapted from Stigliani, William M. 1990. "Chemical Emissions from the Processing and Use of
Materials: the Need for an Integrated Emissions Accounting System." Ecological Economics
2(4):325-341 (Figure 2).
1.1 SIGNIFICANT CHARACTERISTICS OF THE POLLUTANTS
The pollutants (VOCs) have specific characteristics that influence the design of an
effective economic incentive. VOCs mix uniformly with nitrogen oxides in the troposphere
to form ozone, one of the main components of urban smog. The characteristic of being
uniformly mixed implies that the concentration of ozone is generally independent of the
location in the airshed of sources emitting VOCs, but were dependent upon the total amount
of VOC emissions in an airshed. At current rates, VOC emissions do not accumulate in the
1-2
-------
atmosphere from year to year and hence the injury to human health and the environment is
due not to the historical mass of emissions, but from current emissions. Therefore, at least
within an airshed, cost-effective incentives do not need to differentiate among the locations
of sources and receptors and incentives may target the rate of current emissions of VOCs.
1.2 POLICY CONTEXT
In addition to the characteristics of the pollutant, the design of economic incentives to
reduce VOCs emitted from consumer and commercial products is shaped in specific ways by
the Clean Air Act as amended. The parameters established by legislation involve the
following important topical areas:
• sequence of regulations,
• regulated entities, and
• best available controls.
The last area refers to the environmental goal of regulation, and the other areas to the
approach to be taken to achieve the goal.
Section 183(e)3(A) directs the EPA to create a list of categories of consumer or
commercial products which account for at least 80 percent of VOC emissions from those
products in ozone nonattainment areas. EPA is to divide this list into 4 groupings,
establishing priorities for regulation. EPA shall regulate one of those groups no later than
every 2 years after promulgating the list until all four groups have been regulated. Thus the
regulation of consumer and commercial products involves a sequence of regulations.
An economic incentive program could apply to any single group or to all four groups
comprehensively. A comprehensive economic incentive program should be compatible with
the sequence of regulations. For example, architectural and industrial maintenance (AIM)
coatings may constitute the first group of products to be regulated. Therefore, a potentially
1-3
-------
important sequencing issue is how to mesh an economic incentive (or some other regulation)
applying to AIM coatings with economic incentives (or some other regulation) applying to
the remaining consumer and commercial products.
Regulations developed under Section 183(e) may be imposed only with respect to
"manufacturers, processors, wholesale distributors, or importers of consumer or commercial
products for sale or distribution in interstate commerce in the United States" or certain
entities which supply such products to the former [Sections 183(e)(l)(C) and 183(e)(3)(B)].
The definition of regulated entities excludes retailers and users.
The regulations affecting consumer and commercial products shall require best
available controls (Section 183(e)(3)(A)). The Administrator, on the basis of "technological
and economic feasibility, health, environmental, and energy impacts," shall determine the
desired degree of emissions reduction that "is achievable through the application of the most
effective equipment, measures, processes, methods, systems or techniques, including
chemical reformulation, product or feedstock substitution, repackaging, and directions for
use, consumption, storage, or disposal" (Section 183(e)(l)).
The requirement for best available controls establishes the general environmental goal
of regulation, not the means by which regulated entities will comply. The discussion in this
report of potential economic incentive programs will refer to an assumed environmental goal
without attempting to specify the magnitude of the goal. Further, it is assumed that the goal
1-4
-------
for each of the four groups of consumer or commercial products takes the form of an overall
goal, and that regulation does not mandate emission reductions for individual products.1
Although the goal itself is outside of the scope of the report, the significance of
"economic feasibility" to the determination of the desired degree of emission reduction
implies an interaction between the goal and the policy instrument. Cost-effective policy
instruments allow the greatest improvement in environmental quality: for any given cost, the
cost-effective instrument achieves the greatest emission reduction (and for any given
emission reduction, the cost-effective instrument is the least costly). If an economic
incentive were cost-effective, then it would achieve the most desirable environmental goal
that could be achieved at any given level of cost deemed to be economically feasible (see
Figure 1-2).
1.3 CONSISTENCY WITH ECONOMIC INCENTIVE PROGRAM RULE
Design of a Federal economic incentive program to reduce VOCs from consumer and
commercial products should be consistent with or, at least, not contradict the proposed
Economic Incentive Program Rule, which the EPA proposed on February 23,1993 (58 FR
11110). The proposed rule encourages creativity and flexibility while at the same time
1 Hypothetical goals for individual products are consistent with the assumption of an overall goal, and they may
be used in a "bottom-up" approach to determining the overall goal. For example, the first step in
determining the goal could be to determine hypothetical VOC content limits for each regulated product
(VOCj). Second, using the content limits and historical sales data, calculate the hypothetical implied total
allowable emissions for each manufacturer and importer (VOCsj). Third, sum allowable emissions over all
manufacturers and importers to determine the actual environmental goal for the regulation. The second and
third steps are expressed in the following equations (in which Qjj is the sales of coating i by source j in the
base year).
VOCj * Qjj = VOCsj
£ VOCSj = VOC Goal
Sj
1-5
-------
Total Cost
($/year)
Greatest
"economically
feasible" cost
Cost-effective
instruments
Emission Reduction
(tons/year)
Economic incentive "C" sets the greatest feasible emission reduction target.
Figure 1-2. Hypothesized Relation Between Economic Incentives and
Economically Feasible Environmental Goals
ensuring that the programs meet the same degree of accountability and enforceability as
traditional regulatory programs.
The proposed Economic Incentive Program Rule establishes criteria for economic
incentive programs (EIPs) that States develop to meet certain Clean Air Act requirements
that apply to the regulation of criteria pollutants in nonattainment areas.2 The Agency also
intends to use the rule to guide the development of EIPs included in any Federal
2In some cases the CAA mandates that states adopt EIPs, or that they consider them as one option to address
failures to meet statutory requirements for nonattainment areas. States also may voluntarily choose to adopt
an EIP to help meet progress requirements. The proposed EIP Rule applies to both statutorily-mandated and
discretionary programs. Specifically the EIP rule applies to (1) any statutory EIP submitted to EPA as a state
implementation revision plan revision to comply with sections 182(g)(3), 182(g)(5), 187(d)(3), or 187(g) of
the Clean Air Act, and (2) discretionary EIPs submitted as implementation plan revisions for any purpose
other than to comply with the statutory requirements specified in (1) (58 FR 11110).
1-6
-------
Implementation Plan or in other Federal measures promulgated under the authority of the
criteria pollutant provisions contained in Title I of the Act .3
Therefore, the Economic Incentive Program Rule will apply as guidelines for any
Federal economic incentive program for consumer and commercial products. Some specific
requirements may not literally apply to a Federal consumer and commercial products
program because this program is not subject to all requirements that apply to State
Implementation Plans.
The programs to which the EIP criteria apply include market creation (i.e., marketable
emission permits); direct market incentives (i.e., emission taxes and fees); and indirect
market incentives (i.e., tax code changes and subsidies). Further, depending on both whether
a program directly specifies an emission reduction and whether results are quantifiable, the
EIP rule establishes three regulatory categories: emission limiting, market response, and
directionally sound. Emission limiting programs set limits that must be met, but allow
flexibility in how sources reduce emissions. A marketable permit program with a cap on the
total mass of emissions is an example of an emission limiting program. Market response
programs, such as fees or taxes on emissions, create incentives to reduce emissions without
mandating emission reduction requirements for individual sources or groups of sources.
Directionally sound programs directly or indirectly rely on market mechanisms, but differ
from the other programs in that reliable quantification of emission reductions is infeasible.
The proposed EIP rule sets forth general requirements and principles. Programs shall
be designed such that they:
• do not discriminate in favor of intrastate commerce or against interstate
commerce;
•^Much of the following discussion in the text of the proposed EIP rule is taken from Martin, Richmond, and
Beal (1993).
1-7
-------
• have quantifiable emission reductions that may be credited to the program;
• are consistent with attainment and RFP demonstrations:
• limit emission reductions credited to the program to those that are surplus to
other SIP-credited programs;
• are enforceable at the Federal and State levels;
• achieve emission reductions that are permanent within the time frame
specified by the program;
• do not interfere with other requirements of the CAA (58 FR 11110; Martin,
Richmond, and Seal, 1993).
In addition to general principles, the proposed EIP rule requires specific program
elements for economic incentive programs. Table 1-1 lists specific requirements and
provides a brief description of each. A clear purpose and well-defined environmental and
economic goals are prerequisites to choosing an appropriate economic incentive strategy and
are necessary to ensure that the proper mechanisms needed to achieve these goals are
included in the program design. Without specific goals, the tradeoffs that are inherent in the
design of economic incentives will go unrecognized. Clarity in the goals of regulation
facilitates the design of the EIP that achieves the most important goal with the least sacrifice
of competing goals.
A well-defined program scope identifies the affected sources and defines provisions
for new sources and for sources exiting the program, whether they are closing operations or
have reduced emissions to zero. The scope must be defined such that requirements of the
EIP do not interfere with other CAA regulations, and procedures must be in place for solving
conflicts with other Federal regulations affecting the universe of sources. Compliance with a
Federal EIP must be monitored through enforceable source requirements. Many of these
1-8
-------
TABLE 1-1
REQUIRED ECONOMIC INCENTIVE PROGRAM ELEMENTS
EIP Requirement
Description
Statement of Purpose, Goals,
and Program Description
Program Baseline
Quantification Procedures
Source Requirements
Projected Results and
Audit/Reconciliation
Procedures
Implementation Schedule
Administrative Procedures
Enforcement Mechanisms
Clear statement of purpose, including environmental problem being
addressed and intended environmental and economic goals
Economic incentive strategy to be used
How program mechanisms will reach goals
Program Scope • Affected sources
• Procedures for dealing with sources entering or exiting affected source
categories
• Designed so the program will not interfere with Federal requirements of the
Clean Air Act
• Used as a basis for projecting program results
• Determined from information of sufficient quality to provide for at least as
high a degree of accountability as currently exists for traditional regulations
• Replicable and credible methods to quantify emissions and changes in
emissions for SIP credit
• Projection of program results from affected sources using specified
quantification methods and averaging times (directionally sound programs
exempt)
• Must include accounting for shutdowns and production curtailments; for
batch, seasonal, and cyclical operations
• Procedures for determining emissions from affected sources when data
monitoring is missing or inaccurate
• Actions of source necessary for compliance
• Emission limits, monitoring, recordkeeping, and reporting requirements
• Projected emission reductions associated with implementation of program
• Quantitatively projected results shall be adjusted by uncertainty factors for
incomplete compliance and programmatic uncertainty
• Audit procedures to evaluate program implementation and track program
results
• Reconciliation procedures (for market response strategies) to make up for
any shortfall between SIP credit predictions and actual results
• Dates for notifying potentially affected sources
• Program initialization and start-up procedures
* Submittal requirements from affected sources
• Procedures appropriate for the type of incentive strategy necessary to
implement the program
Requirements which are legally binding and State and Federally
enforceable
Penalties for noncompliance
Source: Federal Register.
Printing Office.
February 23, 1993. Vol. 58, No. 34, p. 11110. Washington, DC: U.S. Government
1-9
-------
reporting, monitoring, and recordkeeping requirements may be performed by the source, but
with adequate incentives to promote reporting integrity. Ideally an EIP should be designed
with administrative procedures which operate efficiently without placing a large burden upon
sources or regulatory staff, while at the same time providing incentives to comply with the
program through effective enforcement mechanisms.
The EIP rule also identifies guidelines for the use of any revenues generated by
mandatory economic incentive programs to which the rule applies.4 However, these
guidelines do not apply to this study since guidelines for the use of revenues generated from
the regulation of consumer and commercial products are specified in Section 183(e)(5). Any
use of revenues from a Federal economic incentive should consider the approved revenue
uses outlined in the EIP rule.
The economic incentives discussed in Sections 3 and 4 of this report follow the
principles of the proposed EIP rule. The program designs discussed below come into contact
with the EIP rule primarily at the points that pertain to program purpose, scope, source
requirements, reconciliation procedures, initialization, administrative procedures, and
enforcement mechanisms. The design of an actual incentive program for consumer and
commercial products will require additional attention to the more specific requirements
contained in the EIP rule, for example, quantification procedures. The program design
elements that are discussed in this report have been chosen with certain fundamental
tradeoffs in mind: certainty over the cost of emission reductions versus certainty over the
quantity of emissions reduced; cost to control emissions versus administrative, monitoring,
and enforcement costs; and distributional flexibility versus the provision of incentives to
advance technology.
4 Revenues generated from discretionary EFPs are exempt from these requirements.
1-10
-------
1.4 DESIGN CONSIDERATIONS FOR ECONOMIC INCENTIVES
The potential of economic incentives for improving ambient air quality in the nation
is great, and this potential is achieved with a program design that is sensitive to the
complexities inherent in environmental regulation. A program that meets the following
criteria will accomplish most policy objectives.5
• Does the program achieve the environmental goal?
• Is the environmental goal achieved with the least expenditure by regulated
entities on emission abatement and expenditure by the EPA on information?
• Will monitoring produce data that merit high confidence, and are monitoring
costs acceptable within the context of either the program itself or other
regulatory efforts that the program could assist?
• Are potential enforcement actions credible?
• Does the program incorporate distributive values in a way that enhances the
initial and continuing acceptability of the program? Does the design process
identify potential inequities in the distribution of programmatic benefits and
costs among regulated entities and users of commercial and consumer
products?
• Does the design process enhance public discussion of the objectives and
techniques of environmental policy?
• Will the program enhance the EPA's capacity for effective policy-making in
the future with respect to regulation of consumer and commercial products?
• Does the program minimize the cost of using inaccurate or incomplete
information on the costs of abating VOC emissions and on the benefits of the
program?
• Will the program adapt without unnecessary strain to future changes in the
economy and in values?
5The purposes of economic instruments for pollution control are evolving as program designers respond more
fully to the complexities of regulation. The given list of criteria reflects the insights of Stavins (1991), Hahn,
McRae, and Milford (1988), and Bohm and Russell (1985).
1-11
-------
• Does the program provide incentives for the invention and diffusion of
environment-saving technology or cost-saving abatement technology?
• Will the program avoid or limit unintended damages?
This report conducts an assessment of economic incentives based on the criteria listed
above, with particular emphasis on obtaining an environmental goal at least cost. This
assessment requires the development of prototypical economic incentives that are informed
by the relevant characteristics of the potentially regulated commodities, the companies that
manufacture or distribute these commodities, and users of the commodities.
1.5 SCOPE OF THE REPORT
The EPA is increasingly looking toward economic incentives to reduce the costs of
pollution reduction to industry, to government, and to consumers (U.S. EPA, 1991). The
Clean Air Act Amendments heralded the Agency's interest in economic incentives in at least
two ways. It provided specifically for the design and implementation of an economic
incentive program to reduce sulfur dioxide emissions from utilities. It also called, in some
sections, for the consideration of economic incentives as regulatory tools. Section 183(e)4
provides a list of regulatory systems which may be employed for the regulation of VOC
emissions from consumer and commercial products. This section explicitly states that such
systems may include economic incentives concerning the manufacture, processing,
distribution, use, consumption, or disposal of products.
This report examines three types of economic incentives:
• fees on VOC emissions,
• marketable VOC emission permits, and
• VOC emission averaging systems.
1-12
-------
The last two economic incentives may be considered alternative forms of emissions trading
strategies.
Although Federal policy appears to have favored emissions trading programs over fee
programs in the past, the study of both fees and emissions trading makes the greatest
contribution to the analysis of potential approaches to regulating consumer and commercial
products. A comparison of these designs makes explicit the requirement for the
consideration of certain fundamental policy issues, including the tradeoffs between: (1)
certainty over the cost of emission reductions versus certainty over the quantity of emissions
reduced, (2) the cost to control emissions versus administrative, monitoring, and enforcement
costs, and (3) distributional flexibility versus the provision of incentives to advance
technology. This report examines these tradeoffs in some detail.
1.6 REFERENCES
1. Bohm, Peter and Clifford S. Russell. 1985. "Comparative Analysis of
Alternative Policy Instruments." In Handbook of Natural Resource and Energy
Economics, Vol. I, A. V. Kneese and J. L. Sweeney, eds., pp. 395-460.
Amsterdam: Elsevier Science Publishers.
2. Federal Register. February 23, 1993. Vol. 58, No. 34, p. 11110. Washington,
DC: U.S. Government Printing Office.
3. Hahn, Robert W., Gregory J. McRae, and Jana B. Milford. 1988. "Coping with
Complexity in the Design of Environmental Policy." Journal of Environmental
Management 27:109-125.
4. Martin, Karen M., Harvey M. Richmond, and Willis P. Beal. 1993. Economic
Incentive Programs Under Title I of the Clean Air Act. Presented at the 86th
Annual Meeting of the Air and Waste Management Association, Denver, CO,
June 14-18.
5. Stavins, Robert N. 1991. Project 88—Round II. Washington, DC.
6. Stigliani, William M. 1990. "Chemical Emissions from the Processing and Use
of Materials: the Need for an Integrated Emissions Accounting System."
Ecological Economics 2(4):325-341
7. U.S. Environmental Protection Agency (U.S. EPA). 1991. Economic
Incentives: Options for Environmental Protection. 21P-2001. March.
1-13
-------
-------
SECTION 2.0
CONSUMER AND COMMERCIAL PRODUCTS AND ECONOMIC INCENTIVES
Information on the consumer and commercial product industries assists with
observing the special opportunities and challenges to using economic incentives to reduce
VOC emissions from the use of these products. The design of efficient and effective
economic incentives is adapted to certain characteristics of the industry, the products, and the
users. Product diversity and product innovation affect the ease and efficiency with which
consumer and commercial products can be regulated. Further, the preference for a
marketable emission permit program—a very prominent economic incentive—or a
command-and-control alternative depends upon expectations regarding expenditures on
program administration and pollution abatement, which are influenced by the universe of
sources and the implied structure of the permit market. This section therefore presents a
tightly focused overview of the commercial and consumer product industries, investigating
program design issues stemming from or involving product diversity, the universe of sources,
and conflict between environmental regulation and health and safety regulation.
This section draws on profiles that cover the potentially regulated products, the firms
that manufacture or distribute these products, and the product users. The companion
document, Economic Incentives to Reduce VOC Emissions from Consumer and Commercial
Products: Industry Profiles and Technical Appendixes (U.S. EPA, 1993), contains profiles
of the following industries:
• soaps and detergents,
• polishes and sanitation goods,
• cosmetics and toiletries,
• adhesives and sealants,
• pesticides,
• drugs and health use products,
2-1
-------
• automotive after-market products and miscellaneous household chemicals
(chemical preparations),
• allied paint products, and
• office supplies.
These profiles were developed to provide the basic information needed for assessing the
applicability and desirability of employing economic incentives in regulatory strategies for
consumer and commercial products.
2.1 INFORMATION NEEDS
Various information needs are relevant to the design of a regulatory strategy to reduce
VOC emissions from consumer and commercial products. It would be desirable to have
information regarding the:
• emissions reduction potential,
• ease of monitoring emissions,
• competition in product markets,
• industry dynamics—introduction of new and changed products,
• product distribution channels,
• potential cost-savings from emission fees or emissions trading,
• potential cost-savings accruing to small firms,
• potential VOC-reducing product or packaging reformulation (in advance of
Federal regulation),
• potential impacts on consumers (by income class), and
• potential unintended damages (e.g., increased use of toxic chemicals).
The above information would facilitate the design of economic incentives and better allow
the comparison of different programs on the specific criteria that are especially important:
attainment of the ozone standard, cost-effectiveness (including administrative costs), and an
equitable distribution of benefits and costs. Another important use of such information is to
raise issues that may pose difficulties later on when (and if) the Environmental Protection
Agency undertakes the actual design of an economic incentive. The Industry Profiles and
Technical Appendixes document (U.S. EPA, 1993) fills some of these information needs, but
2-2
-------
more information would be needed to optimally design specific economic incentive programs
for consumer and commercial products.
2.2 SOURCES OF EMISSIONS AND IMPLICATIONS FOR REGULATORY DESIGN
For consumer and commercial products, the emissions of concern occur during either
use of the product or disposal of the product (and, perhaps, container)—not during
manufacture of the product. This aspect of the pollutant leads to the existence of a great
many dispersed and individually small sources for which "end-of-pipe" controls are
infeasible. Hence, VOC content limits may generally be the only practical traditional
regulatory strategy.
Another conceivable approach to regulating consumer and commercial products is to
influence the user of a product by, for example, levying an emission fee on the product at the
point of retail sale. Many commodities sold in the United States are subject to this type of
fee (although they are not motivated by environmental concerns): cigarettes and alcoholic
beverages are obvious examples. By increasing the emissions fee (which is now zero) levied
on users of consumer and commercial products, users would respond by purchasing products
with reduced VOC content, by using such products less frequently, and by foregoing use of
some products.
The regulation of users could be an effective strategy for reducing emissions but not
necessarily the most efficient strategy, even if allowed under the Clean Air Act Amendments
of 1990 (which prohibits it). One source of inefficiency would be the higher administrative
costs involved in collecting fee payments from retail outlets. Every retail outlet in the
country would need to be included in the record keeping and monitoring programs. The
volume and hence costs of such activity would be far less if manufacturers or distributors and
importers were to be regulated instead. A regulatory strategy that operates at a higher level
2-3
-------
in the distribution channel for consumer and commercial products could be just as stringent
and could cost less than the regulation of users (or retailers).
The nature of the source of emissions—product users—is still relevant to the analysis
of economic incentives. Because the regulated entity will be either a manufacturer,
distributor, or importer but not the user, the pattern of trades in an emission permit program
(see Section 4) will not necessarily correspond to changes in the pattern of emissions. For
example, a manufacturer who purchases permits will not by that action necessarily retard
progress in his or her community toward attainment with the ozone air quality standard
because the company's products may be distributed outside of the community (as is typically
the case for large manufacturers). Therefore, problems associated with permit trades that go
in the "wrong direction" may be less likely under a marketable permit program for consumer
and commercial products than they might be for similar programs to reduce pollutant
emissions from other sources.
2.3 PRODUCT DIVERSITY, PRODUCT INNOVATION, AND ADVANTAGES OF
ECONOMIC INCENTIVES
Consumer and commercial products are very diverse and constantly change as
manufacturers market new products and new forms of existing products. Product diversity
reflects the tailoring of formulations and product forms to the technical requirements of
product usage, differences in required performance and ease of use, and users' aesthetics.
Even when the overall growth in sales volume in an industry is slight, change is occurring,
the outcome of product introductions determining companies' shares of the market. The
headline in an advertisement from a manufacturer of cosmetics chemicals states a truth about
industry dynamics: "If your product isn't to miss tomorrow's trends even by a hair's breadth,
we must develop today the ingredients for the day after tomorrow" (Household and Personal
Products Industry, 1991, p. 107).
2-4
-------
In one recent year alone, over fifteen thousand (15,879) new products were
introduced in drugstores and supermarkets (Household and Personal Products Industry,
1991, p. 10). Manufacturers of food, health and beauty aids, and other household products
sold in supermarkets introduced 520 "environmentally friendly" products in 1990 (Fucini,
1991, p. 22). Although these examples portray trends in a much broader group than
consumer and commercial products as defined in the Clean Air Act Amendments of 1990,
sales of all these products depend upon manufacturers' success in offering products that
satisfy many different demands for performance, ease of use, and aesthetics.
Several examples will illustrate current trends specifically in the consumer and
commercial products industries. Research and development in the soaps and detergents
industry are the key to creating improved and innovative products that are important for
increasing a manufacturer's market share (O'Reilly, 1991b). Manufacturers of polishes and
waxes are researching ways to formulate their products with reduced VOC content while
retaining product efficacy (Carson, 1990). Despite expectations that growth in the next few
years will be modest for the industry, manufacturers of cosmetics and toiletries will continue
to aggressively exploit product niches with the introduction of new products, allowing some
manufacturers to grow faster than the average (Standard and Poor's Corporation, 1991).
Growth in the domestic market for pesticides is expected to occur from the introduction of
more effective and environmentally safer products (O'Reilly, 199la).
Several new coating technologies may greatly reduce or eliminate VOC emissions.
Reactive diluent technology uses chemical reactions during film-forming to make organic
compounds an integral part of the coating, preventing release to the air. Radiation curing
technologies use ultraviolet light or electron beams to dry and harden a solventless coating.
Another recently introduced solventless coating is in the form of an electrically charged
powder that adheres to metallic surfaces. Although these coatings are technically appropriate
for only some of the applications for which conventional architectural and industrial
2-5
-------
maintenance coatings are commonly used, the new technologies illustrate research and
development in the paints and coatings industries.
Even a cursory survey shows that product innovation is widespread in the consumer
and commercial products industries. The pattern of funding for research and development
further makes the point. In the pharmaceutical industry, new product development is a key
component of expenditures on research and development; a survey of the members of the
Pharmaceutical Manufacturers Association found that 83 percent of R&D funds for human
and veterinary use spent in the United States were for research and development of new
products and that 17 percent went toward improvement of existing products (Pharmaceutical
Manufacturers Association, 1991). Therefore, information on new products and on R&D
funding establishes the importance of continually changing the product line.
Product diversity and product innovation are fundamental structural characteristics of
the consumer and commercial products industries. The attempt to fit a command-and-control
regulatory approach, such as a VOC content standards only strategy, into this structure may
generate unnecessary friction and costs because the EPA may have to revise its regulations to
keep abreast of the industry. A standard may have to be developed for each new product or
product form, or an agreement may need to be reached concerning which existing standard is
applicable before the product could legally be sold. Either effort is likely to be frequently
repeated and, because standard-setting takes time, to retard marketing new products.
Economic incentive programs may be more adaptive to the product diversity and
innovation characteristics of consumer products. An emission fee program (see Section 3)
smoothly adapts to industry dynamics because the fee would be applied to the emissions
from the use of new products, exactly as it applies to emissions from existing products. In a
marketable emissions permit program (see Section 4), the manufacturer or distributor of the
new product would increase his or her demand for permits, and the increase in permit price
2-6
-------
would induce other permit holders to sell additional permits. Friction is avoidable and costs
may be lower with economic incentives because adjustment occurs at the margin of the
program rather than in the basic components.
2.4 PRODUCT PERFORMANCE AND ADVANTAGES OF ECONOMIC INCENTIVES
A risk exists that the regulation of VOC content in consumer and commercial
products may jeopardize the performance of certain products. It is even possible that such
regulation, if imposed too strictly, could cause an increase in aggregate emissions from
certain products. For example, VOC content limits for paints could, at least theoretically,
make it necessary to apply more coats and to paint more often which could possibly lead to
more VOC emissions.
The issue of product performance may be less problematic under an economic
incentive regulatory strategy. Manufacturers would have the flexibility and some incentive
to allocate VOCs to products where performance would be threatened without sufficient
VOC content.
It may also be possible to design an economic incentive strategy that would ensure
that the desired emission reductions are obtained even when product performance is
problematic. For example, a VOC allowance strategy (see Section 4) could cap each
manufacturer's allowable current and future VOCs. In making reformulation decisions to
reduce VOCs, manufacturers would not want to sacrifice product durability since additional
future sales would require the purchase of additional VOC allowances. Generally durability
would be maintained, at the potential sacrifice of other quality attributes, and emission
reductions could be achieved with some certainty.
2-7
-------
2.5 POTENTIAL FOR REDUCING A LARGE PROPORTION OF EMISSIONS WITH A
SMALL UNIVERSE OF SOURCES
The products and companies to which regulation applies constitute the universe of
sources, a very important component of the design of any regulation. The universe of
sources has the right size and composition when the administrative and pollution abatement
costs associated with a marginal expansion of the universe just offset the value of the
environmental gains of the expansion. Although the difficulties of quantifying these costs
and benefits make the implementation of this rule problematic, the rule keeps in the
foreground the question of whether it may be sensible to regulate some but not all sources.
This question is relevant to all regulatory strategies, but an additional complexity
arises in constituting the universe of sources for a marketable emission permit program.1 A
large proportion of emissions may be reduced by constituting the universe of sources with a
small proportion of the companies in the industry. If the emissions are concentrated in the
products of a few manufacturers, then constituting the universe of sources with these
companies would reduce the EPA's administrative costs because fewer emission permit
transactions would need to be recorded and monitoring activities would be reduced.
Furthermore, exemptions may be used to lessen impacts on small businesses.
Some insight into emissions concentration can be obtained with statistics on the
concentration of production, which describe the proportion of production manufactured by
the largest companies in the industry. The 50-firm concentration ratio, for example, gives the
proportion of output manufactured by the 50 largest firms; the other concentration ratios are
similarly interpreted. If production and emissions were perfectly correlated, then the
concentration ratio would accurately measure the proportion of emissions attributable to the
'No special difficulty arises with emission fees. Imperfect competition in the product market should lead to an
adjustment of the optimal fee (Cropper and Oates, 1992, p. 684), but the fee program does not introduce a
market distortion as could the marketable emission permit program.
2-8
-------
firms' output. A perfect correlation requires that $1 worth of one product produces the same
emissions as $1 worth of any other product. The actual correlation is undoubtedly weaker.
Table 2-1 shows the concentration ratios and the number of companies in consumer
and commercial product industries for 1987. In every industry listed, the 50 largest
companies produced more than 60 percent of the total value of shipments, and the median
was 87 percent; in the pesticides industry, the 50 largest companies produced 94 percent of
the total. Despite the imprecision of inferences about emissions drawn from concentration
ratios, economies in the administration of economic incentives may be feasible without
compromising on the emissions reduction goal.
However, it is possible that a small universe of sources may inadvertently lead to
excessive expenditures on pollution abatement by the companies participating in a
marketable permit program because the conditions for imperfect competition are enhanced.
Given the large number of companies in the consumer and commercial products industry,
this possibility should not be a problem unless the universe of sources is substantially
restricted to reduce administrative costs or small business impacts. The likelihood and
potential significance of imperfect competition are more fully discussed in Section 4.
The potential for imperfect competition in the permit market could be diminished by
including in the universe of sources companies that do not compete in the product market.
Information on each industry's production can give a general idea of which companies do not
compete in the product market. Tables 2-2 and 2-3 summarize the data on production in each
of the potentially regulated industries. Table 2-2 gives the value of production, and Table 2-3
indicates the degree of specialization of production. For example, the manufacturers in the
paints and allied products industry (SIC 2851) made over $8 billion worth of commodities
that are classified as paints and allied products (i.e., primary products), and they made other
types of commodities (i.e., secondary products) valued at several hundred million dollars.
2-9
-------
TABLE 2-1.
CONCENTRATION OF PRODUCTION IN CONSUMER AND COMMERCIAL
PRODUCT INDUSTRIES
Concentration (Percentage) of Production in:a
Industry (SIC Code)
Soaps and detergents (2841)
Polishes and sanitation goods
(2842)
Cosmetics and toiletries (2844)
Adhesives and sealants (2891)
Pesticides (2879)
Drug and health use products
(2834)
Automotive and miscellaneous
chemicals (2899)
Paint and allied products (2851)
Pens and mechanical pencils
(3953)
Carbon paper and inked ribbons
(3955)
Four
Largest
Companies
65
44
32
23
49
22
23
27
49
37
Eight
Largest
Companies
76
60
49
34
69
36
33
40
65
56
20
Largest
Companies
84
70
72
52
88
65
47
53
84
80
50
Largest
Companies
89
79
86
70
94
88
62
66
97
95
Number of
Companies
683
669
648
537
233
640
1,366
1,121
106
108
aBased on value of shipments in 1987.
Source: U.S. Department of Commerce. 1992. 1987 Census of Manufactures, Concentration Ratios of
Manufacturers. Washington, DC: U.S. Government Printing Office.
The primary product specialization ratio for this industry was 96.6 percent. The
manufacturers of cosmetics and toiletries (SIC 2879) are less specialized, but they did not
produce any paints or allied products. Conversely, the manufacturers of paints and allied
products did not produce any cosmetics or toiletries nor any other product also manufactured
by cosmetics and toiletries companies. None of their products overlap, and the
manufacturers in these two industries would have nothing to gain from discriminating against
each other in a permit market.
2-10
-------
Cfl
3
2
£
•3
e
z
o
H
^J
J
^
O
BS
^
H
Z
O
z
Mt
^*
s
"^
H
u
Q
O
tf
ft.
>•
OS
^£
Q
Z
8
Ed
e
Cfi
^£
S
OS
A.
U.
O
z
0
H
U
O
o
OC
eu
a>>
*
§
(J5
t.
£
5
2?
9v
B
"S
1
I
£*
&•
75
IA
0
i
*g
s
.2
i
a
s
"m
£
S
a
if,
_
ao
<-)
Oi
"
Jg
«"<
<•*
g
S
g
l*>
<8
f^
1
vM
«£
r*
£
S
i
Ot
00
_
s
1
i
(s4
_
2
-j
e
/SIC Code
t
1
^
e
O <*> oor^OtN — *f m — * — «n in r-i
o^o ^••»r~^ — i— ov n n <~* OOO^ OOOrM
do odd>n ri rs "" ci o —
>0-
O •» — OOO -O OOOO OOOO
>ovi Kdgjci — oddd dddo
— — p.
(N
dd oo •» — od oddd
tn r^
(N
O— O^omoo « OOOO OOOO
n oddo do'dd
rJ ^ oo r^
•^ fs (N fn
oor^ i^v^i^v^ ^ — OOO OO^O
ri &• oo — •«• — o< oppp ppino
Wr^ K — on o- odddd ddodd
— ^
r~
o v-> r-ve — of o"
2 '5 S
|a- S I 1 -8 I
illJ3||l-3 -g i.
u 2 ~ T5 S "^ = a. .S "" •-
1 1 ; ; H i 1 1 M . t
Q •= S § 7; S «-y < ^ o S. ~ 2
I li I i ! 11 1 1 1 } ii 1 1 1 i!
-N S_0>* SJ, __„„ fN-OOv-
SS ooSoooo oo ooSSJov ooooo
(NCN r-irs
r~.
—
0
S
•»"
rs
§
Z
v>
»
g
oC
S
V)
O
S
oo
>n
oo
r-
m
oo
oo
i,
«rT
i
tN
Ov
P
•>»
'v
^_
d
P4
vb
tn
o"
jn
rn
r-
p-'
o
r~
0
R
p-
oo"
—
•3
y
1
&I
V
^
(9
^
3
H
|
^
e
C
-
I
u
3
wi
a
i
£
.S 'f
3*
"s *'-
i s
T"2
c 2f
ri
s^
Je four oil
ccounis o)
II
,s 9
II
•g «
3
« ^
U*
B S
•3,2:
Jc
g
>.e>
dustry we tggregaled b
menl of Commerce. 19
al
I*
5 W
w 2
Ha
S U
trt §
« CO
2-11
-------
8
2
3
c
z
IW
o
H
J
O
w
fi£
s
O
d
o
d
in
d
o
d
0
d
o
d
o
d
o
d
o
d
oo
S
1
c
Carbon Paper and
Ribbons
in
wi
—
O
d
o
d
i*^
00
oo
oo
o
d
o
d
0
d
o
d
o
o
o
d
o
d
o
d
o
d
o
d
0
d
r-
d
ON
1
CM
S
fM m
O 0
d d
O r-i
d m'
00
in p
fv O
00
n o
_ ci
o o
d d
o o
d d
O 0
o d
o o
d d
o o
d d
o o
d o
O 0
d d
o o
o d
o o
d d
o —
d d
O
-------
To summarize, the diversity and large number of companies in the consumer and
commercial products industries suggest that potentially substantial control cost savings may
be obtainable by employing flexible market-based regulatory strategies, although the
administrative, monitoring, and enforcement costs may be substantial unless smaller
companies are excluded. This discussion has observed a tradeoff between reducing the
universe of sources to reduce administrative, monitoring, and enforcement costs and small
business impacts, and increasing the likelihood of imperfect competition under marketable
permit programs, which may offset some of the (theoretically achievable) control cost
savings. However, if imperfect competition is not problematic, substantial control cost
savings may be possible while still achieving emission reductions by excluding, or perhaps
treating differently, small companies.
2.6 EXISTING REGULATORY REQUIREMENTS FOR CONSUMER AND
COMMERCIAL PRODUCTS AND IMPLICATIONS FOR REGULATION
Several categories of consumer and commercial products are already regulated under
Federal law. Pesticides are regulated under the Federal Insecticide, Fungicide, and
Rodenticide Act. Drugs and certain cosmetics are regulated under the Federal Food, Drug,
and Cosmetic Act. In this section, the implications of the existing regulatory requirements,
which are described in the Industry Profiles and Technical Appendixes document (U.S. EPA,
1993), for the design of economic incentives and the choice of regulatory strategy are briefly
discussed.
A manufacturer of pesticides, drugs, or drug-like cosmetics is likely to find that a
completely new product, or a reformulation of an old product, developed in response to an air
quality regulation will also be subject to either the Federal, Insecticide, Fungicide, and
Rodenticide Act or to the Federal Food, Drug, and Cosmetics Act. The process of registering
pesticides or of applying for approval of a drug probably will increase the total cost of
2-13
-------
complying with a consumer and commercial products regulation. The additional cost would
be incurred regardless of the regulatory strategy.
However, for products that are subject to these additional regulatory requirements,
economic incentive strategies may be preferable to command-and-control strategies because
of the flexibility that economic incentives provide. For example, VOC standards might make
it excessively costly or impossible to produce and sell certain drugs because of FDA
requirements. Economic incentives may allow some drugs to be produced at lesser cost than
with VOC standards and would not prohibit the manufacture or sale of any drug.
2.7 REFERENCES
1. Carson, Hamilton C. 1990. "The Wax and Polish Market." In Household and
Personal Products Industry. Ramsey, NJ: Rodman Publishing Corporation.
September, p. 45.
2. Cropper, Maureen L. and Wallace E. Dates. 1992. "Environmental Economics:
A Survey." Journal of Economic Literature 30(June):675-740.
3. Fucini, Suzy. 1991. "It's Not Easy Being Green." U.S. Distribution Journal
January:22-26.
4. Household and Personal Products Industry. 1991. Advertisement for Tego
Cosmetics. Ramsey, NJ: Rodman Publishing Corporation. February, p. 107.
5. Household and Personal Products Industry. 1991. "Newsfront." Ramsey, NJ:
Rodman Publishing Corporation. February, p. 10.
6. O'Reilly, Richard. 199 la. Standard and Poor's Industry Surveys: Chemicals
Current Analysis. "Agricultural Chemicals, Crop Acreage Should Continue
Up." New York: Standard and Poor's Corporation, pp. C38-C39.
7. O'Reilly, Richard. 1991b. Standard and Poor's Industry Surveys: Chemicals
Current Analysis. "Product Development Drives Market Growth." New York:
Standard and Poor's Corporation, pp. C40-C43.
8. Pharmaceutical Manufacturers Association. 1991. 1989-1991 Annual Survey
Report, U.S. Pharmaceutical Industry. Washington, DC: Pharmaceutical
Manufacturers Association.
9. Standard and Poor's Corporation. 1991. Standard and Poor's Industry Surveys:
Health Care Current Analysis. "Recession Slows Sales Growth." New York:
Standard and Poor's Corporation, pp. H40-H43.
2-14
-------
10. U.S. Department of Commerce. 1992. 1987 Census of Manufactures,
Concentration Ratios of Manufacturers. Washington, DC: U.S. Government
Printing Office.
11. U.S. Environmental Protection Agency. Office of Air Quality, Planning, and
Standards. 1993. Economic Incentives to Reduce VOC Emissions from
Consumer and Commercial Products: Industry Profiles and Technical
Appendixes.
2-15
-------
SECTION 3
FEE PROGRAMS FOR REDUCING VOC EMISSIONS FROM CONSUMER AND
COMMERCIAL PRODUCTS
A system of regulation that uses emission fees can obtain a real and quantifiable
reduction in the emission of VOCs from consumer and commercial products. The basic
rationale for emission fees is to bring an environmental service provided by the
troposphere—its capacity to assimilate VOCs—into the economic system.1 A fee would have
the same type of effect as the prices for the goods and services exchanged in actual markets:
for example, manufacturers would further economize on the use of VOCs just as they would
further economize on the use of labor if wage rates were to rise.
If the fee is charged in the input market, manufacturers have a direct motivation to
reformulate consumer and commercial products. Products with lower emissions will become
relatively less expensive than products with a higher VOC content. As the effect of the fee
on product price becomes clear to users, they have a motivation to substitute less polluting
products.
Input and product substitutions are only some of the environmentally beneficial
reactions that will occur with fee programs (wherever the fee is imposed). Fee programs
leave manufacturers and users free to choose their response and continually encourage
additional reductions in the emission of VOCs. These features may reduce the information
burden on the Environmental Protection Agency, reduce the cost of attaining the ozone
standard, and enhance the likelihood of environmentally beneficial technical change. The
latter effect is especially important: emission fee programs create rewards for the "green"
entrepreneur who includes environmental protection in the product design strategy.
'Freeman and Haveman (1972) simply and succinctly state the basic rationale for emission fees.
3-1
-------
Designing a fee program presents many choices. This section discusses the critical
elements of a fee program and explores several options for the most important elements. To
facilitate a comparison with alternative regulatory strategies, the section ends with an
example of a specific fee program for reducing VOCs from consumer and commercial
products.
3.1 PURPOSES OF A FEE PROGRAM
The purposes of a fee program influence the choice of specific options for the many
elements in the design of the program. A primary purpose of a fee program for consumer and
commercial products would be to provide incentives for actions leading to the attainment of a
specific reduction in total VOC emissions from the regulated products. It is important to note
that the achievement of the goal is not a one-time accomplishment, but must be met on a
continuous basis thus requiring a permanent incentive.
The environmental goal could be set in the following way. The Environmental
Protection Agency, independently or in consultation with affected parties, could determine a
hypothetical VOC content standard for each regulated product category; then the implied
emission reduction could be summed over all regulated products to determine the overall
goal. The Agency could impose a fee to provide regulated entities an incentive to reduce
their emissions by the desired amount.
Figure 3-1 illustrates the basic principle of an emissions fee program whose goal is
the attainment of a specific level of emissions at least expenditure on pollution abatement.
Two companies emit a total of 30 units of pollution, and the government intends to reduce
pollution to 15 units. A fee of $500 per unit of pollutant will induce Source 1 to reduce
emissions from 15 units to 5 units. Source 2 reduces emissions from 15 units to 10 units.
3-2
-------
Dollars/unit
of emissions
reduced
MC2
MCi
500 -
Source no. 1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Emissions reduced
Source no. 2
Figure 3-1. The Basic Principle of Emission Fee Programs
Source: Adapted from T.H. Tietenberg, Emissions Trading: An Exercise in Reforming Pollution Policy
(Washington: Resources for the Future, 1985), p. 20.
That allocation of responsibility for emission reduction minimizes costs. The figure readily
shows that any other pattern of emissions increases total costs; for example, Source 2 would
save the amount indicated by area A if it were to increase emissions by one unit, but Source 1
would spend A plus B. An emissions fee program leads to the cost-minimizing pattern of
emission reductions by levying a uniform fee on each unit of pollutant.
An alternative purpose of a fee program is to internalize the costs of pollution, leading
to an efficient allocation of resources. When individuals release pollutants into the air they
use a natural resource service (the assimilative capacity of the air) without paying for the use
3-3
-------
of that service. While there is no cost, or inadequate cost, to the polluter for using the
resource, there is a cost to all persons whose health or well-being is diminished by the air
pollution. One way to correct the misallocation of resources is to charge the polluter a fee
equal to the marginal or incremental damage caused by the pollution. A program with this
purpose would reduce VOC emissions although the determination of the desired reduction
explicitly would consider the health effects of exposure to ozone and the monetary equivalent
of the loss in well-being.2
Any fee program may raise revenue even if revenue-raising is not the purpose of the
program. The total revenue generated by a Federal program could be large enough to make
the uses of the revenue an important secondary purpose. In general, many activities are
suitable for funding from fee revenues: technological development and risk communication
are two examples. Alternatively, revenues may be rebated to reduce adverse impacts on
industry. A more specific discussion of these uses and the constraints imposed by the Clean
Air Act as amended appears later in this Section.
A fee program may serve more than one purpose. The multipurpose aspect of fees
does not mean that a fee program will serve each purpose equally well. In fact, as will be
shown, the designated purpose of the program determines important characteristics of the
program design.
3.2 ECONOMIC BASIS OF THE FEE
If the primary purpose of the fee is to achieve a specified quantity of emissions
reduction, the most accurate basis of the fee is a model of the markets that are directly and
indirectly affected by the fee program. Such a model would quantitatively describe a
complex set of relationships governing how users will change their purchases of the
2Those considerations would not necessarily be explicit in the prior approach to determining the environmental
goal of the program.
3-4
-------
commodity on which the fee is charged and how manufacturers will change the commodity
itself in response to the fee. The model would characterize the relationship between fee rate
and emissions reduction, but would not by itself indicate the desired fee rate or the desired
reduction in emissions. Once the environmental goal had been determined, the model would
indicate the initial fee that should be levied (see below for a discussion of revising the fee
rate in response to contingencies that arise during implementation).
If the primary purpose of the fee is to correct the market failure associated with the
absence of a cost to the product user for discarding VOCs to the atmosphere, the appropriate
basis of the fee is the damage resulting from air pollution. Ambient ozone is injurious to
human health, to materials, and to plants. The estimation of damages reflects the sequence of
processes that begins with the photochemical transformation in the troposphere of VOCs to
ozone and ends with the application of economics methodologies to determine the monetary
value of the injuries caused by exposure to ozone.
The most precise estimate of damages involves the detailed modeling of ozone
concentrations, exposures to ozone, concentration-response functions for human health,
materials damage, and ecosystem risks, and valuation. This effort would require a very large
amount of information, much of it difficult to obtain, for each nonattainment area and for
attainment areas as well, if the Federal program were to apply to the entire nation. The
manufacturers of consumer and commercial products would face greater fees on products
used in areas where the ozone problem is more severe. One strength of damage-based
emission fee rates is that manufacturers would be more likely to incorporate environmental
costs into strategies for product reformulation and packaging redesign.
Although better information on damages facilitates the more precise estimation of fee
rates, full information is not a prerequisite for designing fees that reflect regional differences
in the environmental costs of VOC emissions. An approximate, but more implementable
3-5
-------
approach to setting fee rates is to scale fees to the severity of each region's nonattainment
status and to the size of the population in the region.3 Still, the ease of movement of
consumer and commercial products across regions may make differentiated fee rates too
difficult to implement effectively.4
Given the difficulties with estimating the benefits of improving air quality in different
regions, this Section emphasizes the primary purpose of achieving emission reductions at
least cost. Fee programs may also achieve other purposes, for example, stimulating
technological change, and thus these objectives are also discussed.
3.3 MARKET RELATIONSHIPS AND RESPONSES
Input suppliers, product manufacturers and importers, and users of consumer and
commercial products form a complex system of related markets.5 Changes in one market
directly and indirectly affect the other markets in the system. For example, a fee on VOCs
levied on purchasers of solvents (i.e., product manufacturers) increases the price paid by
manufacturers for a solvent, which changes the relative prices of the inputs to consumer and
commercial products. The changes in input prices motivate manufacturers to substitute the
now lesser expensive inputs, in particular, materials formulated with a lower concentration of
VOCs, for the more expensive inputs (solvents formulated with a higher concentration of
VOCs); actual reformulation depends upon consumer and commercial products technology.
The cost of manufacturing consumer and commercial products increases, and the increase
will be greater for consumer and commercial products which are higher in VOCs. Users of
consumer and commercial products react to the higher prices by reducing purchases of some
3 A theoretical examination of regionally differentiated emission fee programs shows that this approach sets fee
rates that accurately reflect relative differences in environmental costs, although the determination of the
absolute levels of the fees would require additional analysis and, most likely, trial-and-error (Tietenberg,
1974).
4This point is discussed more fully in Section 3.4.3.
5The industry profiles (see U.S. EPA, 1993) describe the basic materials used to manufacture consumer and
commercial products, types of products, and substitution possibilities available to users of those products.
3-6
-------
consumer and commercial products and increasing purchases of relatively less expensive
consumer and commercial products. Overall VOC emissions from consumer and commercial
products decline; however, emissions from some types of consumer and commercial products
may actually increase despite reformulation because of increased demand.
The final state of the markets—the new prices, sales volume, and emissions—
depends upon technological possibilities, the behavior of product users, and the secondary
effects of the initial changes in the markets. One cannot a priori determine the magnitude of
change and the direction of change for all prices, sales volume, and emissions. Therefore,
without further investigation, program designers cannot predict the gainers and losers from
the fee program or the social costs. An empirical model of the markets is necessary to
completely investigate the economic and environmental consequences of the program.
To illustrate market interrelationships, a model has been developed to investigate the
effects of imposing a VOC-based fee on the purchase of solvents by manufacturers of
architectural and industrial maintenance coatings. The Industry Profiles and Technical
Appendixes document, Appendix B describes the model and identifies the data which are
required to operate the model.6 This document also presents the results of one simulation of
a fee program. Because the data were contrived, the simulation illustrates but does not
project the precise effects that one would really expect to occur following the imposition of
the indicated fee. The model represents coating technology with a two input production
function and represents coating usage as allowing substitution between a high- and a low-
VOC formulation. The emission reductions attributable to an initial fee level cannot be
rigorously estimated without a model like the one described.7
^The Industry Profiles and Technical Appendixes document, Appendix B also presents a graphical analysis of
the effects of imposing the fee in the coating market (rather than the solvent market).
7Multiple models (for different coatings) may be required for an investigation of the fee program for
architectural and industrial maintenance coatings; further investigation is required to determine the
appropriate level of aggregation for analyzing this very heterogeneous set of commodities.
3-7
-------
For fee programs, the uncertainty in projecting emission reductions is due to both the
intrinsic nature of fee programs (i.e., emissions will increase if the industry expands
production) and limits on modeling markets. One significant limitation is the uncertainty in
estimating the parameters of the model (see the Industry Profiles and Technical Appendixes
document, Appendix B for a description of the parameters), which causes uncertainty in
estimating the total emission reduction which any particular fee will elicit. A sensitivity
analysis can determine the frequency distribution of emission reductions that results from
uncertainty about the parameter estimates. Thus market models could help determine the
degree of "reasonable certainty" (U.S. EPA, OAQPS, 1991b: p. 9) of projected emission
reductions attributable to an initial fee rate and the consequences for progress toward
compliance with the statutory deadlines for attainment with the ozone standard.
3.4 FEE RATE VARIABILITY
A fundamental design choice for a fee program is the uniformity or variability of the
rate of the fee. This design choice arises with respect to several factors: season, VOC
content, location, and progress in reducing emissions.
3.4.1
The EPA could make the fee effective only during a particular time of the year. Since
the presence of ambient ozone is strongly related to heat and humidity, the Agency could
make the fee effective (or greater) only during the seasons characterized by that type of
weather. Seasonally differentiated fees may lower the aggregate cost of emission reductions
without reducing the decrement in ozone concentration.
However, opportunistic behavior appears to make seasonally differentiated fees
inappropriate for consumer and commercial products. It is possible and even likely that
3-8
-------
users would purchase consumer and commercial products during the season when the fee was
ineffective (or less) and store the product until the time of use. For example, because many
paints last in storage and are applied outdoors, users are likely to buy them in the winter and
to apply them in warm weather, thus negating the intended effects of a seasonal fee on VOCs.
Agricultural chemicals (insecticides and pesticides) also are typically applied during warm
weather, and users would have the incentive to circumvent seasonally differentiated fees.
Many other consumer and commercial products are used for purposes such as sanitation,
personal hygiene, and grooming that are largely independent of the ozone season. These
products are also stable and are sold in packages of which the volume is sufficient for an
extended period of time. Therefore it is doubtful that seasonally differentiated fees would
alter the pattern of use because circumvention would be easy.
3.4.2 VOC Content
The uniformity-variability consideration also arises in designing the fee structure
which relates the fee rate to VOC content. A uniform or flat rate fee structure requires the
regulated entity to pay the same amount for each unit of VOC potentially emitted regardless
of the total emissions potential of a unit of product. A variable rate fee structure also requires
the regulated entity to pay a fee for each unit of VOC; however, the per-unit amount of the
fee depends on the total amount of VOCs potentially emitted. Figure 3-2 shows a flat rate
and a variable rate fee structure plotted against the amount of VOCs potentially emitted from
a unit of product.
Under the flat-rate fee structure, the total fee is computed by multiplying the amount
of VOCs potentially emitted by the constant fee, fc. Under the variable rate fee structure, the
fee is computed by multiplying the amount of VOCs (x units) by the fee rate specific to that
3-9
-------
$Fee/
VOC
fx - Variable Rate Fee
fc -Flat Rate Fee
VOCs/Unrt of
Product
Figure 3-2. Flat and Variable Rate Fee Structures
"concentration" of VOCs (fx). Under the variable rate fee structure, a manufacturer can
reduce the fee rate applied to his or her product by reducing the VOCs potentially emitted by
a unit of product.
In a static analysis of the efficiency of fee structures, the variable rate fee does not
minimize the total cost of achieving an emissions reduction goal. This result occurs because
the marginal costs of control would not be equal across all regulated consumer and
commercial products. However, an advantage of the variable rate is that the fee burden on a
manufacturer decreases as its products approach very low VOC content and it encounters
possibly insurmountable barriers to further reducing VOCs. The possibility of a barrier to
reducing VOCs may make a modification of the continuously variable rate fee structure
attractive: the fee is zero below a threshold VOC content. For example, the fee could be
imposed on every unit of VOC only when the VOC content of a regulated product exceeds a
threshold value. For example, if the threshold value for VOCs in a particular type of coating
were 100 grams per liter of coating then there would be no fee on a coating containing 99
3-10
-------
grams of VOC, but a per unit fee would be imposed on a coating containing 125 grams of
VOC (i.e., the fee would apply to the 25 grams above the threshold). This type of fee
structure, however, has the disadvantage of eliminating the incentive to invent a formula in
which the VOC content is substantially less than the threshold.
3.4.3 Location
In nonattainment areas, excess concentrations of ozone pervade the area. High ozone
values registered at the monitors reflect air quality in a large part of the metropolitan area.
In contrast, the ozone problem differs remarkably across nonattainment areas. Peak
concentrations in the Los Angeles South Coast Air Basin are extremely high; in Kansas City,
they are marginally excessive. Therefore, in general, air quality differences within any one
nonattainment area are not significant for the design of economic incentives, but the
differences between nonattainment areas may justify a tailoring of the program fo air quality.
The fee rate is the most important element in a fee program that could be used to
modify the incentive to reduce emissions, increasing the incentive in areas where the greatest
reduction is necessary to comply with the air quality standard. An emission fee that increases
with the severity of the nonattainment status is the simplest way to tailor the fee to air
quality.
A greater fee provides both a stronger incentive to reduce emissions and comes closer
than a uniform fee to indicating the environmental significance of decrements in emissions.
Figure 3-3 illustrates the principle that the emission fee should be greater where the benefits
of emission reductions are greater. Because the risk to human health increases with exposure
to ozone, the benefits of emission reduction—avoided damage—increase. The curve labeled
MBo could correspond to a severe nonattainment area, and curve MB i, to an extreme
nonattainment area. The fee rate should be higher in the extreme nonattainment area.
3-11
-------
Dollars per
quantity of
emissions
reduced
Pi
PO
MC
MC
Qo
Q
Zero
pollution
Emissions reduced
Figure 3-3. Emission Fee Should Be Greater Where Benefits Are Greater
If the Federal program were to use regionally differentiated fees, then combining
approaches to setting the fee may be desirable. The damage-based approach, if feasible,
would indicate the relative magnitude of the fees, and the market model would inform the
selection of the absolute level of fees. The administrative costs of a fee program with
differentiated fees would be somewhat higher than the costs of a program with a uniform fee,
but the least attractive feature of the former program is a difficulty in enforcement. Users of
consumer and commercial products would pay higher prices for products sold in more
polluted areas. They would have an incentive to buy products where they are less expensive,
3-12
-------
in some cases transporting products across attainment area boundaries to take advantage of
fee differentials. Opportunistic behavior would undercut the program without extremely
intrusive monitoring. Nonetheless, geographically differentiated fees may be practical for
products the value of which is small enough to render uneconomical a trip that would be
made to take advantage of price differences.
3.4.4 Emission Reduction Progress
It is unlikely that the EPA will have enough information to set the initial fee such that
the program exactly achieves the environmental goal. The realistic expectation is that actual
emissions will overshoot or undershoot the target. Consequently, it is advisable to build
sufficient flexibility into the fee system so that the fee may be adjusted. A schedule of
adjustments could be laid out in advance or the adjustment could be made mechanically at
predetermined time periods according to a formula. Adjusting the fee rate is discussed more
thoroughly in Section 3.6.
3.5 INITIAL FEE RATE
A critical task in the initialization of a fee program is the selection of the fee rate.
The rate is important because it influences the effectiveness of the fee program in two ways.
First, the fee is the instrument for motivating regulated entities to search for new product
formulas and packaging. Second, the acceptability—technical reasonableness, fairness—of
the process by which the initial fee is set affects peoples' intentions to delay the start of the
program or to circumvent the program once implemented. Either result would be costly: the
anticipated emission reductions would occur later than desired or would require more
resources devoted to monitoring. The following discussion of the method for setting the
initial fee rate begins with a market model, but, recognizing potential problems with
acceptability, alternatives are also considered.
3-13
-------
The market model described in Section 3.3 and the Industry Profiles and Technical
Appendixes document, Appendix B can be used to estimate the initial fee rate required to
achieve any aggregate emission reduction goal. However, the strength of the model's results
depends on the accuracy of the data used in the model and assumptions about elasticities and
opportunities for input substitution.
The difficulty associated with rigorously justifying the appropriate fee rate may be a
weakness of the fee program. This difficulty derives from the necessity for information
about the current technology of manufacturing consumer and commercial products, users'
and manufacturers' likely responses to the fee, manufacturers' costs of reducing VOCs in
consumer and commercial products, and the likely success of research and development in
reformulation, repackaging, and new product development. In light of the expectation that
not all parties who are affected by regulation may agree on the accuracy of the data used to
estimate the market model or the estimated parameters, it is germane to consider alternative
methods for setting the initial fee.
One alternative method for setting the initial fee rate is negotiation between the EPA,
representatives of the consumer and commercial products industry, and representatives from
interested citizen and environmental groups. On the table for negotiation could be the initial
fee rate, the length of the period in which the emissions reduction goal would be achieved, a
set of interim goals, and a backstop measure (such as fee rate increases) that would become
effective if progress toward the goal were judged to be insufficient.
A second alternative method for determining the initial fee rate is to make a fee on the
VOC content of consumer and commercial products comparable to the cost to firms of
buying VOC credits or the cost of obtaining VOC reductions from other industries. Thus the
fee rate could be determined by the market value of VOC emission rights. One such market
already exists: the market for offsets created under the aegis of the EPA's emissions trading
3-14
-------
policy (U.S. EPA, 1986). State and regional authorities are forming other markets now.
Alternatively, the fee could be based on estimates of the marginal cost to achieve additional
VOC reductions from other industries.
A third alternative is to combine the initial fee rates computed by different methods or
to combine the methods themselves. One possibility would be to obtain estimates of
elasticities of demand and supply from experts inside or outside of the consumer and
commercial products industry. These estimates could then be used in the market model. One
other approach is to use the initial fee rate calculated by the model or the value of traded
rights as a starting point for negotiations.
A number of states and air quality management districts have already promulgated
VOC content standards, and other regulatory agencies may promulgate standards prior to the
promulgation of Federal regulations for consumer and commercial products.8 Such standards
require an adjustment in the fee rate to make the economic incentive effective. The necessary
adjustment is to set the fee rate above the rate that would be desired in the absence of content
standards (see the Industry Profiles and Technical Appendixes document, Appendix A).
Considering the uncertainty in the regulated entities' and product users' responses to
the fee, concern with the initial fee rate should not overshadow provisions for adjusting the
rate. A properly designed adjustment mechanism will ensure eventual progress toward the
emissions reduction goal.
3.6 FEE ADJUSTMENT MECHANISM
It is not possible to perfectly forecast the responses of users and regulated entities to
the fee because the modeling of response involves economic parameters that, when formally
^Consumer and commercial products are currently regulated in several states, including California, New York,
New Jersey, Texas, and Arizona.
3-15
-------
estimated with statistics, are estimated imprecisely or are informally estimated with far less
precision. Consequently, the fee rate that will meet an emissions reduction goal cannot be
determined with precision. By including an adjustment mechanism in the fee program, the
fee can be made responsive to unsatisfactory or too rapid progress toward achieving the
emissions reduction goal.9 If the objective is to achieve an emissions reduction goal, a fee
adjustment mechanism is an essential element in the design of an effective fee program for
consumer and commercial products.
There are three steps in designing an adjustment mechanism. The first step is to set
interim emissions reduction goals. For example, if the overall emissions reduction goal is to
be achieved over 10 periods, then one could set a schedule of interim goals such that ten
percent of the overall goal would be achieved in each period.10 Column 3 of Table 3-1
shows an example of a set of interim goals. Note that the interim goals can be constant over
time, increasing, or decreasing as long as the sum of the interim goals equals the overall
emissions reduction goal (151,000 units in the example).
The second step is to project the fee rates necessary to achieve the interim goals. One
could use any of the methods described in Section 3.5 to determine these fee levels. Column
2 in Table 3-1 shows a fictitious set of projected fee levels to meet the interim goals.
'This reason for adjusting the fee is independent of the increase in the fee that may be necessary to obtain the
desired emission reductions in a program whose universe of sources expands according to the sequence of
increasing cost of emission reductions, which could occur in the regulation of consumer and commercial
products.
choice of the length of a period might reflect both statutory requirements for air quality improvement and
the time that regulated entities must take to redesign products and packaging. Early adjustment can lead to
an unnecessarily high fee because apparently slow progress can reflect the time sources need to respond (as
well as lack of incentive).
3-16
-------
TABLE 3-1
FEE SYSTEM WITHOUT AN ADJUSTMENT MECHANISM
(1)
Period
1
2
3
4
5
6
7
8
9
10
(2)
Projected
Fee
0.05
0.06
0.07
0.08
0.09
0.10
0.12
0.14
0.16
0.18
(3)
Interim
Goals
16,500
16,500
16,500
16,500
16,500
16,500
16,500
16,500
16,500
16,500
(4)
Actual
Reductions
16,500
16,200
15,900
15,600
15,300
15,000
14,700
14,400
14,100
13,800
(5)
Cumulative
Goals
16,500
33,000
49,500
66,000
82,500
99,000
115,500
132,000
148,500
165,000
(6)
Cumulative
Reductions
16,500
32,700
48,600
64,200
79,500
94,500
109,200
123,600
137,700
151,500
Columns 4, 5, and 6 illustrate what might occur if no adjustment mechanism were used. The
cumulative actual reduction falls further behind the cumulative interim goal. The cumulative
emissions reduction achieved by the end of the tenth period is less than the overall emissions
reduction goal. The projected fees were insufficient to obtain the emissions reductions
necessary to meet the goal. The example in Table 3-1 represents just one possible scenario.
It is also possible that the projected fees are set too high, leading to the achievement of the
overall emissions reduction goal several periods early. Because one cannot accurately
project the fee rate necessary to meet the overall goal, and because this will result in
undershooting or overshooting the goal, a mechanism that automatically adjusts the fee rate
in each period is desirable.
The third step in designing the adjustment mechanism is to develop a system of
formulas that relates the fee to progress toward the overall goal, changes the interim goals,
3-17
-------
and further refines the fee to reflect the new interim goals. Therefore adjustment requires a
triple mechanism, each serving a separate but complementary purpose. The first adjusts the
projected fee rate by the percentage of the interim goal achieved in the previous period. For
example, if the emissions reduction achieved in period 3 falls short of the goal for period 3
by 15 percent, the projected fee level in period 4 would" increase by 15 percent. The reverse
operation is executed when actual reductions exceed the interim goal in the previous period,
decreasing the projected fee level by the appropriate percent.
A hypothetical case is constructed in Table 3-2 based on the scenario depicted in
Table 3-1. The numbers in column 6, labeled Rate Adjustment Factor #1, are calculated by
the following formula, in which ERA is the emissions reduction achieved and ERG is the
interim emissions reduction goal:
r> . AJ- *T? * .ui i ERA in period t
Rate Adjustment Factor #l(t+1) = 2 - ERG in j,eriod t
TABLE 3-2
FEE SYSTEM WITH AN ADJUSTMENT MECHANISM
(1)
Period
1
2
3
4
5
6
7
8
9
10
(2)
Projected
Fee
0.05
0.06
0.07
0.08
0.09
0.10
0.12
0.14
0.16
0.18
(3)
Interim
Goal
16,500
16,649
16,612
16,870
16,853
17,033
17,081
17,072
16,771
16,639
(4)
Actual
Reductions
15,000
17,000
14,000
17,000
15,000
16,500
17,100
20,000
18,000
17,000
(5)
Interim
Goal
Adjustment
149
-37
258
-17
180
47
-9
-301
-132
-46
(6)
Rate
Adjustment
Factor #1
1.00
1.09
0.98
1.16
0.99
1.11
1.03
1.00
0.83
0.93
(7)
Rate
Adjustment
Factor #2
1.000
1.009
1.007
1.022
1.021
1.032
1.035
1.035
1.016
1.008
(8)
Adjusted
Fee
0.066
0.069
0.095
0.091
0.115
0.128
0.145
0.135
0.168
3-18
-------
To obtain the adjusted fee rate, one simply multiplies the projected fee rate (in
column 2) by the first adjustment factor. For example, if the emissions reduction achieved in
period t were only 91 percent of the emissions reduction goal for period t, then the
adjustment factor would be 1.09. If the fee rate scheduled for the next period was $0.06, then
the new fee rate, without any further adjustments, would be $0.06*1.09, or $0.065. This
adjusted fee rate is higher than the fee rate initially computed for period 2, and should
provide a greater incentive in period 2.
The second adjustment mechanism changes the interim goals to make up for
shortfalls or excesses in the actual reductions in previous periods. It allocates the difference
between the goal in period t and the actual reduction in period t evenly to all future interim
goals. For example, if in period 5, the actual reduction fell short of the goal by 20,000 units,
then each of the goals for periods 6 through 10 would be increased by 4,000 units. In
contrast, if in the following period, the actual reduction exceeded the goal by 12,000, then the
adjustment mechanism would decrease the interim goals for periods 7 through 10 by 3,000.
Column 5 in Table 3-2 shows the number of units added to or subtracted from the interim
goals in future periods.
Interim ERG adjustments are calculated by the following equation:
ERG — ERA
Interim ERG Adjustment = io_t—~ •
assuming that the total number of periods allotted to achieve the overall emissions reduction
goal is 10. Each remaining interim goal (i.e., ERGt+i, ERGt+2, —) is adjusted by adding the
interim ERG adjustment calculated above. Hence the interim goal in any period will be a
function of the progress in meeting the past interim goals.
3-19
-------
In addition to adjusting the interim goals, it is necessary to further adjust the fee rate
to account for changes in the interim goals. For example, if one increases the interim goal
without increasing the fee, then actual reductions that period are likely to still fall short of the
interim goal. Consequently, both the interim goal and the fee rate need to be adjusted. This
second adjustment to the fee level is shown in column 7 of Table 3-2 and is computed by the
following equation.
„ _ . ,. . c . un Adjusted Interim ERGt
Rate Adjustment Factor #2(t) = ''Original ERGt
The final adjusted fee rate (shown in column 8) is computed by multiplying the
projected fee rate by the two adjustment factors. In this manner, progress toward the overall
goal feeds back to the program to ensure effectiveness.
The use of such a mechanism to adjust the fee sends a clear message to regulated
entities that future fee rates are dependent upon progress in meeting emission reduction goals.
Manufacturers play an especially important part in meeting emission reduction goals through
the development of new products and packaging. Hence the adjustment mechanism
reinforces the incentive to invest in VOC-reducing research and development.
However beneficial fee adjustment, it is not costless. A balanced discussion of fee
adjustment involves the costs that regulated entities will incur as they respond to changing
but uncertain fees (Walker and Storey, 1977).u After fees change, manufacturers, for
example, may find that they incur greater manufacturing costs and fee payments than they
would incur if they had known the fee in advance and adopted the most appropriate response.
Some manufacturers will adjust their technology, but others will prefer to stay with the
existing technology (and pay higher operating expenses and fees). The expectation of those
1 Ipee adjustment might also involve additional costs to be borne by the EPA or the general population: the cost
of information used to determine whether unsatisfactory progress is due to a lack of incentive or the reaction
lag, and the cost of polluted air.
3-20
-------
capital adjustment and lock-in costs will motivate some manufacturers to plan for change by
adopting more flexible strategies for reducing emissions, which could entail adaptations of
the manufacturing process as well as product and package redesign. Flexibility is not free—
regulated entities will incur costs to reduce the costs of change. Therefore the costs of an
emission fee program with a fee adjustment mechanism include capital adjustment and lock-
in costs and the costs of flexibility, although the extent to which such costs are significant is
uncertain. The alternative to fee adjustment is to gamble on the initial fee, at the price of
increased risk of failure to achieve the environmental goal.
3.7 UNIVERSE OF SOURCES
The universe of sources comprises the regulated entities (companies) on whom the fee
would be levied. Regulated entities are manufacturers, processors, wholesale distributors,
and importers of consumer and commercial products (Clean Air Act Amendments of 1990,
Section 183 (e)(l)(C)). The definition of regulated entity excludes suppliers of the raw
materials for the manufacture of consumer and commercial products and, at the opposite end
of the product distribution channel, retailers and users. Inclusion in the universe of sources
implies that the company is legally responsible for paying the fee and must comply with, at a
minimum, monitoring and record keeping requirements.
The following characteristics of regulated entities are relevant to the composition of
the universe of sources:
• type of product,
• type of business, and
• size.
Each characteristic implies a criterion for determining whether a regulated entity will be in
the universe of sources. Inclusion may also depend upon the simultaneous satisfaction of
multiple criteria. Further, pursuant, to Section 183(e)(3) of the Clean Air Act Amendments of
3-21
-------
1990, the universe of sources may progressively expand over time as regulation encompasses
more groups of consumer and commercial products.
3.7.1 Selection of Sources by Type of Product: Sequence of Regulation
The most general issue in the selection of regulated entities for inclusion in the
universe of sources is the grouping of products by priority of regulation. Although the
identification of those groups is beyond the scope of this report, it is germane to consider
whether the sequence of regulation should influence other elements of a design of fee
program that encompasses multiple groups of consumer and commercial products.
Cost-effectiveness suggests that the order in which products are regulated should
correspond to the projected costs of emission reduction, the least costly reductions being
obtained first. Under this regulatory sequence, if continual progress toward an overall
emissions reduction goal is desired, the magnitude of the fee may need to increase over time
to elicit the desired emission reductions.12 In this scenario, the fee increases whenever a
group of products is added to the universe of sources. Otherwise, with a constant fee rate,
more emission reductions would be forthcoming from the earliest product groups regulated
than from product groups regulated subsequently.
3.7.2 Selection of Sources by Type of Business
Figure 3-4 portrays a generalization of the manufacture-distribution system for
consumer and commercial products. It shows the markets where the fee may be imposed
and the entities on whom the fee may be levied. Manufacture comprises several activities:
specification of the product-package system by the marketer, preparation of the product
(formulation), and filling the package. Distribution refers to the shipment of finished goods
'2This increase in the fee is in addition to any adjustment for inflation.
3-22
-------
0)
3
•g
i
"a.
Q.
Q.
•o
o
m
o
£
i
o
C
UJ
TS
CO
•o
Q.
0)
CD
>.
(0
1
o
To
^2
UJ
I
w
i
re
CO
O)
1
ffl
CO
11
M
O 3
5 £
o o
•s U
IS
^"8
fl
II
0 1 J
1 a I
•si
V)
u
I
o
3-23
-------
but excludes retail sales. In any actual consumer or commercial product industry, the course
of a product involves a shorter or longer path depending upon whether formulation and
filling occur on the site of the marketer's operations and the number of distributors as well.
A fee may be imposed on either the VOCs in the products or the product itself, and
this choice has implications for the selection of source by type of business. If the fee were to
be imposed on the product itself, then a fee rate would be chosen to be applied to each unit of
product. The next choice would be the node in the distribution channel and the side of the
market—supplier or demander—on whom to levy the fee.
If the fee were charged in the wholesale product market, the manufacturer or product
distributor could be made responsible. Alternatively, the fee could be levied at an
intermediate node in the distribution channel, and either the selling or buying distributor
could be responsible for paying the fee. (For complete coverage, all importers should be
included in the universe of sources.)
The incidence of the fee and the ease of collection may be thought to be criteria for
the choice of entity on whom to levy the fee. The incidence of a fee is the weight of the fee
borne after markets adjust. As shown in Figure 3-5, the incidence of the fee is independent
of the entity on whom the fee is levied. In general, the portion of the fee paid by suppliers or
demanders of a commodity will be determined by the market supply and demand curves, but
the portion is unaffected by the attribution of legal responsibility for paying the fee. Whether
levied on the suppliers or demanders in a given market, the fee reduces total emissions from
consumer and commercial products by the same amount and otherwise has identical effects.
Therefore, the ease of collection should be the deciding factor once the node in the
distribution channel has been chosen.
3-24
-------
Price
($/unit)
Price paid by user
(includes fee)
PO
Price received by
supplier (excludes fee)
Demand curve
before fee on users
Demand curve
after fee on users
Supply curve after
fee on producers
Supply curve before
fee on producers
Q1 Quantity (units of input or
product per year)
The fee is imposed on either the supplier or user.
Figure 3-5. Incidence of Fee
One rationale for the product fee is that every consumer and commercial product
contains and releases VOCs and, therefore, every product should be subject to a fee. Also,
this type of fee might be easiest to implement, involving only the total quantity of consumer
and commercial products sold in the United States by each manufacturer or distributor and
importer and each company's remittance.
However easy to administer, a fee on the product itself may not be the most efficient
approach to reducing VOC emissions. Cost-effectiveness in the allocation of effort to reduce
VOC emissions requires equality of the marginal cost of emission reductions for all sources;
a uniform fee on VOCs will lead to cost-effectiveness (Fisher, 1981: p. 188). The problem
with a product fee is that the fee is not uniform with respect to VOCs. Therefore, a uniform
fee on consumer and commercial products implies a nonuniform fee on VOCs.
3-25
-------
A fee on the VOC-containing inputs to the manufacture of consumer and commercial
products could be cost-effective.13 If the fee were charged in the input market, for example,
propellant market, the user of the input (i.e., the product manufacturer) would be made
responsible for paying the fee but not the supplier of the input because the definition of
regulated entity excludes input suppliers. Again, importers should be included in the
universe of sources for complete coverage of emissions.
The input fee would induce manufacturers to redesign consumer and commercial
products but would not prohibit the manufacture of any product. Product prices would
increase, and some product users would switch from high-VOC products to low-VOC
products in response to relative price increases. Total VOC emissions from consumer and
commercial products would decrease, and the cost of achieving the desired reduction in
emissions would be minimized.
The method for quantifying emissions in an input-based program assumes that
emissions occur in the year in which the input is purchased by a product manufacturer.
Similarly, the quantification method in a product-based program assumes that emissions
occur in the year in which the product is manufactured in the United States or imported.
These assumptions may be imprecise; for example, the purchase and use of a product may
occur in different years. The resulting discrepancy between assumed and actual emissions
may be negligible and might not justify changing the methodology for quantifying emissions.
3.7.3 Selection of Sources by Size
The EPA must consider the economic impact of proposed regulations on small
businesses and develop alternatives when the impacts are onerous. Regardless of this
requirement, the exemption of "small" companies from a fee program may benefit such
13The assumption that the mass of emissions equals the mass of VOCs in the inputs may not be not entirely
accurate.
3-26
-------
companies without substantially affecting the effectiveness or efficiency of the program,
especially if size is defined in terms of annual emissions.
The companies on whom the fee is levied will search for ways to reduce their fee
payments and comply with reporting and record keeping requirements. These activities may
be costly, and although the costs correlate with the volume of trade, costs may be
proportionately greater for small companies than for large companies. The EPA will also
incur costs in identifying and communicating with small companies and keeping the
appropriate records. The value of the emission reductions obtained from small companies
may not outweigh the total costs (including the companies' costs and the EPA's costs) of
including these companies in a fee program.
In most potentially regulated industries, a small number of manufacturers is
responsible for a large percentage of sales (see U.S. EPA, 1993). Therefore it may be
possible to control a large percentage of VOC emissions from consumer and commercial
products by including only a small number of companies in the fee program. Small
companies could be excluded either permanently or temporarily from the requirements of a
fee program.
However desirable it is to lessen onerous impacts on small businesses, a selection
criterion based on size could have undesirable side-effects. First, after the division of large
and small companies was determined (where large companies are included in the program
and small companies are not), a mechanism would have to be designed to handle small
companies which become large companies. Second, basing inclusion on company size could
provide an incentive for large companies to divest themselves of some product lines. As
new, smaller manufacturing companies were established, they and their products could
conceivably be small enough to be excluded from the program.
3-27
-------
3.8 COLLECTION OF FEE PAYMENTS
The method of collecting fee payments is important for both program effectiveness
and administrative efficiency. The following discussion of collection methods assumes that
the regulated entity is a manufacturer, but these methods could easily be applied to other
sources.
One method for collecting fees would require manufacturers of consumer and
commercial products to compute the fees owed to the government for each product and to
remit this amount to the government, along with the data used to compute the fee, on a
periodic basis (annually, for example). The manufacturers' computations would be based on
the VOC content and the sales volume of each product. This strategy takes advantage of the
manufacturers' wealth of information concerning their own production and sales, and also
places the burden of collection/remittance on the manufacturers. This approach has the
additional advantage of continually supplying the government with detailed information for a
complete emissions inventory.14
An alternative collection method is to require manufacturers to purchase "VOC
stamps" from the government. The manufacturer would affix a stamp onto each regulated
product indicating that the emissions fee had been paid, and the stamp could indicate VOC
content as well. One possible implementation problem with this collection method is that the
EPA would have to concern itself with possible counterfeiting of stamps.
3.9 MONITORING AND ENFORCEMENT
Compliance is a necessary condition for the successful operation of any economic
incentive program. If noncompliance is tolerated, the regulatory program fails to achieve the
14Thus fee programs also provide an incentive to manufacturers to adapt their management information systems
to the requirements of cost-effective emissions control.
3-28
-------
environmental goal and serves as an instrument to disadvantage complying companies vis-a-
vis noncomplying companies (through changes in costs and in market share). In designing a
fee program, it is important to take into account the feasibility of ensuring compliance.
Because the overall gains of a regulatory program are dependent upon compliance and
administrative costs, it may be necessary to make tradeoffs between simple and
sophisticated designs.15
The operating principle behind a strategy to ensure compliance is deterrence. An
effective strategy will lead regulated entities to perceive that the expected costs of
noncompliance outweigh the expected gains of noncompliance. Effective deterrence has two
primary components: sufficient monitoring and credible enforcement.
Under cost-effective fee programs, manufacturers (importers) could be required to
monitor and periodically report the VOC content of each formulation of regulated consumer
and commercial products the company manufactures (imports). The states or EPA could
check for falsification of companies' reports through a random audit of products on shelves
in retail establishments or at wholesale distribution centers. Incorrect reports (either through
falsification or incompleteness) could be a form of noncompliance.
In practice, monitoring compliance in a fee system involves determining whether or
not regulated entities are paying the correct amount of fees. In a fee system based on the
VOC content of consumer and commercial products, the correct fee payment has three
components:
1. VOC content per unit of product,
2. total sales volume (in units) for each product, and
3. correct fee rate.
^Noncomplying companies who undcrreport emissions may still reduce emissions to the desired level
(Harford, 1978). Therefore, air quality may improve as expected, although activities funded by fee payments
may suffer.
3-29
-------
If the fee is imposed in the input market, the correct amount of fees to be remitted by
regulated companies is based on the amount and type of input the companies use in regulated
consumer and commercial products. Because the same inputs may be used to manufacture
regulated and unregulated products, the correct calculation of the fee payment cannot be
based solely on the use of inputs. Inputs used in regulated products, for the purposes of
compliance with the fee program, must be tracked separately. The use of inputs can be
monitored with auditing of company records and product sampling.
Auditing and product sampling would also be followed if the fee were to be levied on
the product itself. The purpose of sampling would be only to verify that a purportedly VOC-
free product is in fact not a source of emissions. Therefore, monitoring compliance with a
fee imposed in the solvent market requires the same types of activity as a fee imposed in the
product market, but the latter might be less expensive.
In one of the collection schemes described in the previous section, regulated
companies would be required to remit the fees owed and to submit reports of VOC content
and quantities sold. To determine that the regulated companies were reporting true sales of
regulated products, the states or EPA could conduct record keeping audits. The subject of the
audit would be required to show how sales volume (and destination, if necessary) of
regulated products is tracked for the purpose of compliance with the consumer and
commercial products regulation, and to demonstrate consistency with reports prepared for
other purposes.
Under the collection scheme involving the VOC stamp, compliance could be
monitored in three ways. First, random audits could verify the VOC content of the coating as
stated on the stamp. The second step in monitoring compliance would require the
government to verify that a stamp is affixed to every regulated product. The government
could alert retailers of regulated products to the new regulations making it illegal to sell
3-30
-------
regulated products without a stamp. Third, the government would have to ensure that the
stamp is genuine. Checking for counterfeiting of stamps could be accomplished by
government examination of stamps on the containers at the retail level and unannounced
inspections of plants where consumer and commercial products are packaged.
Both monitoring programs rely heavily on self-reporting by the manufacturers and
random audits. To the extent that self-reported sales volume information may be difficult to
corroborate, compliance may be more difficult to monitor under the scheme relying solely on
manufacturers' self-reports. However, identifying counterfeited stamps, and even designing
a stamp that is difficult to counterfeit are disadvantages of the alternative scheme.
Geographically differentiated fees entail a unique monitoring problem. It is
reasonable to expect that the volume of consumer and commercial products transported
across air quality regions will depend upon the fee differential, the cost of transportation, and
any product quality differences. Users of large volumes of regulated products are most likely
to circumvent the fee program because they are best able to take advantage of economies of
scale in transportation. An effective program may require some form of monitoring of large-
volume users of regulated products.
The enforcement component of an effective compliance strategy must have two
qualities. First, penalties must be severe enough to deter unacceptable levels of
underreported emissions. Second, regulated companies must perceive the EPA as willing
and able to determine and apply the appropriate penalty in a timely manner. If a company
believes that the EPA will not follow through on an enforcement action, it may not comply
regardless of the severity of the stated penalties.
Section 113 of the CAA outlines the Federal enforcement procedures which apply to
persons in violation of the Act or any regulations promulgated under the Act's provisions.
3-31
-------
Section 113(b) authorizes EPA to assess civil penalties of up to $25,000 per day of violation
and/or to obtain a temporary or permanent injunction against a non-complying source. In
assessing the amount of the penalty, the court is to consider such factors as the size of the
business, the economic impact of the penalty on the business, the violator's full compliance
history and good faith efforts to comply, the duration of the violation, previous penalties
assessed for the same violation, the economic benefit of noncompliance, and the seriousness
of the violation. Under § 113(d) administrative penalties may be assessed for the same
violations for which civil penalties may be sought and as with §113(b) penalties, may be
assessed in an amount of up to $25,000 per day of violation and established by the factors
listed above. In addition to the civil and administrative penalties, under § 113(c) criminal
penalties may be enforced in cases where it is determined that a person has knowingly
violated a requirement. Thus, the CAA provides clear authority to impose substantial
noncompliance penalties.
3.10 USE OF REVENUES GENERATED BY FEE PROGRAMS
An efficient and effective fee program achieves its purpose by guiding manufacturers
and product users to change behavior in specific ways. The use of fee revenues can have a
positive, negative, or neutral effect on these changes in behavior. Consequently, the use of
fee revenues is an important design consideration of any fee program.
Although the Clean Air Act Amendments of 1990 constrain the uses of fee revenues
(see Section 183 (e) (5)), a variety of uses may be consistent with the amendments. These
uses include: product labeling, public education, subsidizing technological development, and
rebates.
Using fee revenues to fund a labeling and public education program could enhance
the effectiveness of a fee program by promoting awareness of the air quality impacts of
3-32
-------
consumer and commercial products. Informing users about the VOC content of their
purchases and the consequences to the environment may lead to additional changes in
behavior and in additional emission reductions.
Creating an organization to fund technological development in the consumer and
commercial products industry is another potential use of fee revenues. Applications for
funding could be accepted solely from companies manufacturing consumer and commercial
products, or could be accepted from any R&D organization. Alternatively, the fees paid by
each company could be held in a trust account that could be accessed only by the
contributing company, and only after the company had submitted a reasonable plan for
research and development to reduce emissions. Ideally, technology developed with money
from the technological development fund could be made available to all consumer and
commercial products manufacturers. This feature may have added benefits for small
manufacturers who are unable to conduct aggressive R&D on their own. However, requiring
companies to give away the newly developed technology, and with it a competitive
advantage, will also be a disincentive to apply for and use those R&D funds.
Employing a rebate scheme to refund fee revenues back to manufacturers could have
a negative, positive, or neutral effect on the changes in behavior originally intended by the
fee program. The most preferable scheme returns the fee revenues to the contributing
sources without nullifying the original incentive to reduce VOC emissions. Four desirable
qualities for the rebate scheme might include the following:
• The rebate scheme should be budget-neutral in each time period: there should
neither be a surplus of unrebated fee revenues nor a deficit of excess rebates.
• The rebate that each company receives should be insensitive to any strategic
action that company might take (such as increasing VOC content in any
period).
• The rebate scheme should be insensitive to mergers, acquisitions, or
divestitures in the industry. Furthermore, the rebate scheme should not
encourage or discourage such actions.
3-33
-------
• The rebate scheme should be able to accommodate and not discriminate
against companies entering or retiring from the industry.
One rebate scheme that possesses most of these qualities is illustrated in Table 3-3.
The first step in computing rebates is to determine a historical baseline that quantifies
companies' shares of total VOC emissions. For example, sulfur dioxide allowances in the
acid rain emissions trading program were initially distributed according to each utility unit's
share of emissions in the base year. The same principle could be used to determine a
historical baseline for consumer and commercial products, possibly using VOC emissions
from consumer and commercial products in 1990. If no recent year by itself is appropriate
for establishing an historical baseline, one could average shares of emissions over the past
three or four years.
The rebate given to any company in the universe of sources is calculated by
multiplying the total fee revenues collected in the year by the company's historical share of
emissions. For example, in Table 3-3 Company A historically accounts for 50 percent of
VOC emissions. The rebate given to Company A in period t is computed by multiplying the
total amount of revenues collected in period t ($204,000 charged at $1 per unit of VOC) by
the company's baseline percentage (50 percent), giving the Company a rebate of $102,000.
Rebates in every year of the fee program for every company in the universe of sources are
calculated on the same basis. The rebate scheme is budget-neutral in each period.
As shown in the table, this rebate scheme is insensitive to the acquisition of one
company by another. When Company E acquires Company B in period t+1, Company E also
acquires Company B's baseline share of emissions. The rebate scheme is also insulated
against strategic behavior of companies trying to manipulate the rebate scheme to their
advantage. For example, when Company F increases VOC emissions in period t, the
company's net position after the rebate (indicated by the rightmost column in Table 3-3) is
3-34
-------
TABLE 3-3
EXAMPLE OF PROPOSED REBATE SCHEME
Time: T
Company
A
B
C
D
E
F
G
Ha
Total
Time: T+l
Company
A
B
C
D
E
F
G
Ha
Total
Time: T+2
Company
A
B
C
D
E
F
G
Ha
Total
Time: T+3
Company
A
B
C
D
E
F
G
H
Total
Historical
Baseline (%)
50
5
6
3
12
20
4
100
Historical
Baseline (%)
50
0
6
3
17
20
4
100
Historical
Baseline (%)
50
0
6
3
17
20
4
100
Historical
Baseline (%)
50
0
6
3
17
20
4
0
100
VOCs
(T)
95,000
9,000
8,000
7,500
22,000
55,000
7,500
204,000
VOCs
(T+l)
90,000
0
8,000
7,500
30,000
41,000
7,000
183,500
VOCs,
(T+2)
90,000
0
7,500
7,500
30,000
41,000
6,500
182,500
VOCs,
(T+3)
85,000
0
7,500
7,000
29,000
41,000
6,000
25,000
200,500
Percentage of
Current
Emissions
47
• 4
4
4
11
27
4
100
Percentage of
Current
Emissions
49
0
4
4
16
22
4
100
Percentage of
Current
Emissions
49
0
4
4
16
22
4
100
Percentage of
Current
Emissions
42
0
4
3
14
20
3
12
100
Rebate
(T)
102,000
10,200
12,240
6,120
24,480
40,800
8,160
204,000
Rebate
(T+l)
91,750
0
11,010
5,505
31,195
36,700
7,340
183,500
Rebate
(T+2)
91,250
0
10,950
5,475
31,025
36,500
7,300
182,500
Rebate
(T+3)
87,750
0
10,530
5,265
29,835
35,100
7,020
0
175,500
Net to
Company
7,000
1,200
4,240
-1,380
2,480
-14,200
660
0
Net to
Company
1,750
0
3,010
-1,995
1,195
-4,300
340
0
Net to
Company
1,250
0
3,450
-2,025
1,025
-4,500
800
0
Net to
Company
2,750
0
3,030
-1,735
835
-5,900
1,020
-25,000
-25,000
a Company H enters the industry in period T + 3.
3-35
-------
less in period t and in subsequent periods than it would have been if Company F had not
increased emissions. This rebate scheme deters companies from increasing their emissions to
obtain a greater share of the rebates.
Some form of discrimination against entering companies is difficult to avoid with this
rebate scheme.16 The fee payment made by a company during its first year of operation
cannot be rebated to the company in a future year because the incentive for emission
reduction would be nullified. Budget-neutrality would require that the fees be returned to the
industry in some manner, but if the entering company's fees were distributed to existing
companies, then it would be placed at a competitive disadvantage. An alternative is to
channel new companies' fee payments into a fund that would subsidize technological
development or some other purpose related to environmental improvement. New (as well as
old) companies would be eligible for the subsidy, making the alternative less
disadvantageous to them.
Rebates could also be combined with one of the other options for the use of program
revenues. For example, the total amount of fees rebated could be limited to 50 percent of all
fees collected in that year, and the remainder could be used to promote technological
development or a product labeling program. An alternative is to set aside a percentage of
total fee revenues for administration of the fee program. Still another alternative would be to
use revenues to fund award programs for manufacturers who develop lower no-VOC
technologies.
Promoting technological development could encompass subsidies for both research
and development on low-VOC formulations of consumer and commercial products and
research and development on VOC-free products that are substitutes for existing consumer
is difficulty is shared by marketable emission permit programs in which the EPA freely grants permits only
to sources who existed at the time of program initiation (see Section 4).
3-36
-------
and commercial products. Subsidies for the latter are in accord with an approach to
regulation that aims at reducing not just emissions from coatings but more generally the
environmental impact of the means that consumers use to obtain the services provided by
consumer and commercial products (Morton and Madariaga, 1993).17 For example, the
regulation of VOC emissions from paints and coatings could both increase the cost to coating
manufacturers of using VOCs and subsidize the development or use of materials that do not
require surface coatings.
Table 3-4 identifies several specific uses for fee revenues. Other uses are of course
possible. Fee revenues could support any single use or a combination of uses.
TABLE 3-4
USES OF FEE REVENUES
• Fund public education program
• Fund labeling program for regulated products
• Subsidize research and development on low-VOC formulations
• Subsidize research and development on VOC-free substitutes for regulated
products
• Direct revenues to U.S. Treasury for general use
• Direct revenues to the EPA to administer fee program
• Rebate revenues to companies in universe of sources
17An analogous objective guides demand-side management in the electricity market. Electricity is valuable
because it is a source of energy for lighting and heating, which are directly consumed, not because electricity
itself directly provides services that people value. The distinction between electricity and the services that it
provides leads to a recognition that society may be better served by reducing electricity demand through
increasing energy efficiency rather than by increasing the degree of pollution control at power plants.
3-37
-------
Another consideration which bears on the desirability of rebates is their effect on the
incentives for companies to create new products (i.e., technological innovation) and to adopt
technologies invented by other companies (i.e., technological diffusion). The willingness of
a company to promote innovation and diffusion—the two steps in technological change—
depends on the company's potential private gain. A company's gain depends on the net
change in profits from product manufacture and sale, gross fee payments, and rebates (if
any). Different regulatory strategies imply different incentives to promote technological
change (Milliman and Prince, 1989).
A fee program without rebates and a budget-neutral fee program imply different gains
to companies from the overall sequence of technological change. Without rebates, the
innovator's gain from creating a less polluting product equals the sum of the change in profits
and the reduction in gross fee payments. The changes in profits and gross fee payments will
be same with or without rebates because rebates do not affect product supply or demand.
With rebates, the innovator's rebate from the government also could change, but a more
realistic assumption is that his or her rebate stays constant because of this one company's
small contribution to aggregate emissions. Therefore, both fee programs provide an equal
incentive for innovation.
The effect of diffusion on the innovator comes from two sources: the change in its
rebate and the change in its competitive advantage. Diffusion does not change the
innovator's gross fee payments. Diffusion reduces the industry's emissions, thus reducing
the size of the pool from which the innovator's rebate is drawn, and the innovator's rebate
declines. Because the innovator also could lose its competitive advantage by providing its
technology to competitors, the total incentive for diffusion is especially low in a fee program
with rebates. Any loss in competitive advantage also occurs in an ordinary fee program.
Therefore, the rebate provides an additional negative incentive to the innovator to promote
3-38
-------
diffusion. The fee/rebate program is more disadvantageous to an innovator than an ordinary
fee program, although the additional disadvantage is difficult to measure.
In summary, a budget-neutral fee program and an ordinary fee program equally
promote technological innovation, but a budget-neutral, fee program may slow the pace of
diffusion of a new technology. This investigation shows that a tradeoff exists between
enhancing technological diffusion and redistributing economic impacts on manufacturers
through rebates.
The assessment of budget-neutral fee programs should also consider the implications
of the special fund provision contained in Section 183 of the Clean Air Act.18 Fee revenues
would be deposited in a special fund in the United States Treasury. Withdrawals, including
rebates, could be made only if authorized by annual appropriation Acts. The special fund
provision may limit the usefulness of fee and rebate programs.19
3.11 INFORMATION NEEDS
Information needs fall into two categories: program design and program
administration. The most important uses of information for program design are to determine
the initial fee rate and to establish the baseline (VOC emissions) against which progress will
18"Any amounts collected by the Administrator under such regulations [i.e., Federal consumer and commercial
products regulations] shall be deposited in a special fund in the United States Treasury for licensing and
other services, which thereafter shall be available until expended, subject to annual appropriation Acts,
solely to carry out the activities of the Administrator for which such fees, charges, or collections are
established or made" [Section 183(e)(5)].
19Nonetheless, important functions of budget-neutral fee programs can be performed by a certain type of
emissions trading program. An emissions trading program in which the rate of emissions per unit of product
(X grams VOC/unit of product) is capped is in important respects equivalent to an emission fee program
with rebates. In a ratio-based emissions trading program, total emissions from the regulated sources are not
capped, and each source creates permits when the VOC content of a product is below a particular threshold.
Manufacturers of low-VOC products may sell permits to manufacturers of high-VOC products. The
program is budget-neutral, the mass of emissions may increase as industry output expands, and the program
avoids new-source bias. The absence of new source bias is an advantage that ratio-based trading programs
have over fee/rebate programs. In addition, administrative costs might be lower because the government
does not need to develop and operate an accounting system for rebates. Ratio-based trading programs are
discussed more fully in Section 5.7 of this report.
3-39
-------
be measured. The most important use of information for program administration is to
monitor compliance. For concreteness and comprehensiveness, the discussion in this section
assumes that the specific fee program involves a fee imposed in the input market,
manufacturers and importers are legally responsible for paying the fee, and fees are rebated.
Table 3-5 lists information needs for program design. Some of these needs are more
critical than others. The first part of the table lists the information needs when the design
team sets the initial fee rate without using economic modeling. As discussed in Section 3.6
above, several means for doing so are possible, and the information requirements will differ
accordingly. Consequently the table indicates that some of the requirements will be specified
by the design team. The other requirements specified in the first part of the table allow the
creation of the current emissions inventory, form the parameters of the rebate formula, and
enhance the credibility of the expectation that manufacturers and users of consumer and
commercial products can adapt to the increased cost of emissions.
The second part of Table 3-5 lists the information needs when the initial fee rate is set
using economic modeling (for example, the model described in the Industry Profiles and
Technical Appendixes document, Appendix B). The economic model could represent
markets for solvents, propellants, other inputs, and consumer and commercial products.
Information on the state of the markets, supply, demand, and technology would be employed.
The specific data to be obtained and parameters to be estimated will vary with the consumer
and commercial products being modeled. The Industry Profiles and Technical Appendixes
document, Appendix B describes the information used to model a market system involving
two coatings, two solvents, and two types of other inputs. This model uses 30 sales, price,
and parameter values. Increasing the richness of the model by working with more specific
manufacturing inputs or additional coatings would increase the information needs.
3-40
-------
TABLE 3-5
INFORMATION NEEDS FOR THE DESIGN OF EMISSION FEE PROGRAMS
Initial Fee Rate Set Without Using Economic Modeling
• As required by design team
• EPA's administrative costs for the program
• Annual VOC emissions from regulated consumer and commercial products in a
recent period (for example, 1989-1991) by manufacturer and importer
• Qualitative information on alternative formulations of existing consumer and
commercial products
• Qualitative information on product substitution by product users
Initial Fee Rate Set Using Economic Modeling
• Annual VOC emissions from regulated consumer and commercial products in a
recent period (for example, 1989-1991) by manufacturer and importer
• EPA's administrative costs for the program
• Qualitative information on alternative formulas and packaging of existing consumer
and commercial products
• Qualitative information on product substitution by product users
• Purchases and prices of VOC-containing inputs
• Purchases and prices of other inputs
• Sales and prices of consumer and commercial products
• Elasticities of supply of VOC-containing inputs
• Elasticities of supply of consumer and commercial products
• Own-price elasticities of demand for consumer and commercial products
• Cross-price elasticities of demand for consumer and commercial products
• Elasticities of substitution between VOC-containing inputs and other inputs
• Cost-share of VOC-containing inputs in manufacture of consumer and commercial
products
• Cost-share of other inputs in manufacture of consumer and commercial products
• VOC content of consumer and commercial products
3-41
-------
Table 3-6 lists the information needs for program administration. The information
would allow the EPA to determine the accuracy of fee payments, make rebates, calculate
VOC emissions, and communicate changes in the fee rate.
TABLE 3-6
INFORMATION NEEDS FOR THE ADMINISTRATION OF EMISSION FEE
PROGRAMS
• Name and address of every company in the universe of sources
• Fee rate adjustment mechanism
• Economic model to determine emission reductions due to program
• Remittance from every manufacturer and importer
• VOC content of each regulated product
• VOC content of sampled products
• Sales of consumer and commercial products by product and manufacturer and
importer
• Rebate formula
• Reliable method for communicating with manufacturers and importers
3.12 SUMMARY OF DESIGN OPTIONS
The preceding sections describe various design options for a fee program to reduce
VOC emissions from consumer and commercial products. These design options are
summarized in Table 3-7 for programs of which the primary but, perhaps, not only purpose is
to achieve VOC reductions at lower cost than is achievable with command-and-control
regulatory strategies.
One specific example of a fee program is described in Table 3-8. The selection of
options for this example is based on: (1) the administrative cost of implementing the
program and (2) the ability to incorporate values related to the distribution of the cost burden
3-42
-------
IY
ri
H
M
PQ
S
t/3
O
E
0
z
C3
w
^•q
S
a
M
S
5
O
O
tf
0.
K
U
ta
fe
O
#
S
S
cc
V)
e
^2
"•S
D,
O
^T
e
^o
'•S
a
O
f^
" 4
e
.2
V*
a
O
s
u o
.a'is
11!
i
i_
&
«
cc
£
ft.
C M.
• ^ ^
-S i
O P3
— C J=
« -
El
•o
w
tf
85
y
>>
••i
•M
U
<
b.
u
T3
C
U
•o
o
1
Cu
Product supplier
!_
U
Cfi
O.
0)
£
(k>
O
V
1
Sources above size
cutoff
iurces
g
<
•o
u
•o
"S
a
|
e
cc
.§
'S5
>,
ca
u
o
Cfl
S
o
U3
_c
M
*5
a
a.
c
_o
«p<
_ U
S 3
•** "^
O JJ
fil *"
Q. M
U! C CA
O O <"
C •— u
1/5 ill
JS !/3 tJJQ
•|ES
>• U Cu
in
«
Q "
W ^
0 U
8-1
s|
js "55
.•a c
> 0
> c
U
O
o
t>
8,
i«
He
•£ -
> °
!> o
e
o
V3
C3
U
srt
o
1
CQ
£
j=
S
.-^
15
R
"u
ea
>•
I
u-
o
c
o
3-s
JS 0
E-S
a§
_o
—
o
bO
u
u
Value of Federal/Stat
marketable emission
rights
u
1
o
o
£
M
e ~
'5 «
*• W
W J
u«
|£
"§^
*.«
i-
en
o
Cu
O
(*•
o
c
_o
«
c
15
o
U
«• ^
^ C
W •—
IS
S O
">
en «j
j»! tn
u c
u 'rt <"
_p* P3 *^
U M 0
w es 3
< «fl"2
C1- S 2
Wc£ D.
y
5 S
s;< 8.
Jw K
Or-
S
cu
fc,
(W
O Q£
e .=
.2 C
•S 2
i 'S
3 I
*o
c
_o
H
e „,
. „ 5/J
•° S
E.2
_°. Cu
U o
E
2 5
M °
° «
S.S
t-.^
£ E
>.'E
« -o
Cu «
en
U
3
C
U
K
V
13
x>
u
Q£
Fund labeling and/or
education programs
CA
"•a
o §»
•*"" o
£1
O «
c^j e*
BS.S
•g =
il
2
3
W
-------
TABLE 3-8
EXAMPLE FEE PROGRAM DESIGNED TO MINIMIZE ADVERSE IMPACTS ON
INDUSTRY
Design Element
Recommendation
Program Purpose
Activity Charged
Payer of Fee
Sources Included
Phasing-in Sources
Fee Variability
Method for Setting Initial Fee Level
Collection of Fees
Monitoring
Use of Revenues
Achieve VOC reductions at lower cost
Impose fee in input markets
Input user (product manufacturer and importer)
All "large" manufacturers and importers
By size
With respect to emissions reduction progress
Combination of methods
Manufacturers and importers remit fees to EPA
Combination of options
Rebate revenues to industry
(distributive flexibility). In addition, the program outlined in Table 3-8 intends to achieve a
well-defined and predetermined environmental goal at a lesser cost than command-and-
control regulatory strategies such as VOC content standards. The goal takes the form of a
definite reduction (compared to the baseline), over a period of several years, in the quantity
of VOC emissions from regulated consumer and commercial products.
The basis of the fee in this example program is the VOC content of consumer and
commercial products rather than the product itself because the former type of fee leads to the
least costly allocation of emission reductions. That consideration implies the desirability of
an input fee.
Product manufacturers and importers could be legally responsible for paying the fee.
Of all regulated entities involved in the manufacture and distribution of consumer and
commercial products, manufacturers have the most knowledge about product formulas and
packaging. A fee on solvents, propellants, and other VOC-containing inputs is a charge on
the materials that are responsible for the air pollution from consumer and commercial
3-44
-------
products. The emission fee would stimulate manufacturers to economize on the VOCs that
are released during use of the product.
In the example program, initially only large companies would pay the fee. After
some period of time, small companies might also be required to pay the fee or perhaps pay a
lesser fee (which would interfere with achieving the least costly allocation of emission
reductions).
Due to inflexibility in the timing of the use of consumer and commercial products, the
fee may need to be uniform with respect to season. Geographical differentiation of fee rates
is attractive, but the possibility of opportunistic transportation of regulated products across
the boundaries of air quality regions makes this option contingent upon the demonstration of
satisfactory program performance without intrusive monitoring. A fee with a variable rate
structure may increase administrative and record keeping costs and would not lead to the
least costly allocation of emission reductions. However, the environmental purpose of the
program entails varying the fee with respect to emissions reduction progress in accordance
with an adjustment mechanism similar to the one described in Section 3.6.
A combination of methods for establishing the initial fee rate may best address the
affected parties' concerns over the difficulty of determining the desired fee. Economic
modeling is the preferable method but may raise concerns over cost, or it may not be possible
if sufficient information is unavailable.
To collect the fees and monitor companies' compliance with the program, the
example program uses the self-reporting system in which manufacturers and importers send
to EPA the total amount of fees owed and the data on which the fee payment was calculated.
The data would provide valuable information for evaluating the success of the fee program
and for compiling a complete emissions inventory. This collection method implies that the
3-45
-------
states or EPA would have to check for companies' compliance with the regulations by
random audits of companies' records and laboratory tests of the VOC content of regulated
products.
Rebates can make the program budget neutral. In this way, adverse industry impacts
may be minimized.
The example fee program suggests that distributional flexibility is of primary
importance. If instead providing incentives for technological advancement is deemed of
utmost importance, rebates would be less attractive. Rebates decrease incentives for
technological diffusion. Further, revenues raised from fees could be used to fund research
and development efforts to reduce VOCs. This discussion illustrates a tradeoff between
distributive flexibility and promoting technological advancement.
3.13 REFERENCES
1. Air Credit Advisor. 1992a. "SCAQMD Marketable Permit Program Moves
Closer to Reality." Vol. 2, No. 1, p. 1.
2. Air Credit Advisor. 1992b. "SCAQMD Board Gives Nod to RECLAIM
Marketable Permit Program." Vol. 2, No. 2, pp. 1-2.
3. Clean Air Weekly. 1992. "South Coast Votes Historic Market for NOX, VOCs,
SO2." March 9, pp. 2-3.
4. Fisher, Anthony, C. 1981. Resource and Environmental Economics.
Cambridge: Cambridge University Press.
5. Freeman, A. Myrick ffl, and Robert H. Haveman. 1972. "Residuals Charges
for Pollution Control: A Policy Evaluation." Science 177:322-329.
6. Gibbs, Michael. August 1990. Study of Economic Incentives to Control
Photochemically Reactive Organic Compound Emissions from Consumer
Products, report prepared by ICF Consulting Associates, Inc. for the California
Air Resources Board, Sacramento, California.
7. Harford, Jon D. 1978. "Firm Behavior Under Imperfectly Enforceable
Pollution Standards and Taxes." Journal of Environmental Economics and
Management 5(l):26-43.
3-46
-------
8. Milliman, Scott R., and Raymond Prince. 1989. "Firm Incentives to Promote
Technological Change in Pollution Control." Journal of Environmental
Economics and Management 17:247-265.
9. Morton, Brian J. and Bruce Madariaga. 1993. "Economic Incentives to
Stimulate the Development and Diffusion of Low- and No-VOC Coating
Technologies." Paper presented at the Pollution Prevention Conference on
Low- and No-VOC Coating Technologies, San Diego, CA, May 2-27.
10. South Coast Air Quality Management District. 1990. Interim Report on Market
Incentives. December.
11. Tietenberg, Thomas, H. 1974. "Derived Decision Rules for Pollution Control
in a General Equilibrium Space Economy." Journal of Environmental
Economics and Management. 1:3-16.
12. Tietenberg, Thomas, H. 1985. Emissions Trading: An Exercise in Reforming
Pollution Policy. Washington, DC: Resources for the Future.
13. U.S. Environmental Protection Agency (U.S. EPA) 1986. "Emissions Trading
Policy Statement; General Principles for Creation, Banking and Use of
Emission Reduction Credits; Final Policy Statement and Accompanying
Technical Issues Document." Federal Register 51(233):43814-43860.
14. U.S. Environmental Protection Agency (U.S. EPA), Office of Air Quality
Planning and Standards (OAQPS). 199 Ib. "Economic Incentive Program
Rules: Background and Issues: Clean Air Act Section 182(g)(4)." Research
Triangle Park, NC. September.
15. U.S. Environmental Protection Agency. Office of Air Quality, Planning, and
Standards. 1993. Economic Incentives to Reduce VOC Emissions from
Consumer and Commercial Products: Industry Profiles and Technical
Appendixes.
16. Walker, Martin and D. J. Storey. 1977. "The 'Standards and Price' Approach
to Pollution Control: Problems of Iteration." Scandinavian Journal of
Economics 79:99-109.
3-47
-------
SECTION 4
EMISSIONS TRADING PROGRAMS FOR REDUCING
VOC EMISSIONS FROM CONSUMER AND COMMERCIAL PRODUCTS
A marketable emissions permit program or an emissions averaging program is an
alternate way to obtain real and quantifiable reductions in the emissions of VOCs from the use of
consumer and commercial products. A restriction on the aggregate quantity of emissions is
essential to all marketable permit programs. In contrast, a VOC content standard—a maximum
proportion of VOCs—is essential to emissions averaging programs. Marketable permit
programs and emissions averaging programs have in common the "trading" of emissions; the
trades may occur within companies and between companies (although emissions averaging
programs do not generally entail external trading).
The most important difference between the two types of programs is that marketable
permit programs place a cap on emissions from the universe of sources, but emissions averaging
programs allow aggregate emissions to increase. Their most important similarity is that they
allow companies much more flexibility in choosing strategies to reduce emissions compared to
product-specific VOC content standards, which are the command-and-control analog to
economic incentives for consumer and commercial products.
This section emphasizes marketable permit programs because the relative advantages of
these programs are substantial and because these programs are most consistent with the goal of
achieving a given emissions reduction at least cost. Section 4.10 considers emissions averaging
programs.
For the most commonly discussed types of marketable permit programs, the quantity
restriction is on either pollutant concentration measured at specified air quality monitors or
4-1
-------
aggregate emissions of a pollutant.1 Thus the restriction ensures the achievement of the
environmental goal regardless of changes in the economy, including the expansion of production
of existing products and the introduction of new products.
/
Equally important, companies may "trade" emission reductions in a marketable permit
program, and they have the incentive to seek potential trading partners.2 Opportunities for trade
exist when the incremental or marginal cost of emission reductions differ in the industry.
Emission reductions are costly to a company because it undertakes research and development,
changes the manufacturing process or inputs, or experiences losses due to product substitution
and other adaptations by product users. The consumer and commercial products industry is
diverse, and manufacturers very probably will reduce emissions through unique combinations of
the options just sketched. Therefore the marginal cost of emission reductions is likely to vary
from company to company, and the company for which a reduction is more expensive will pay
another company (i.e., one with lower costs) to reduce emissions. The benefit to the former
company is reduced costs, and the benefit to the latter, increased profits. If the market for
permits functions well, permit trades will occur whenever marginal costs differ.
A marketable permit program allows companies the flexibility to determine the best (for
them) method of reducing emissions. Such a program reduces the information burden on the
EPA, promotes search for the least expensive emission reductions, and continually stimulates
technological change (to reduce the marginal and hence total costs of emission reductions).
If the States or EPA were to regulate (by product) the VOC content of consumer and
commercial products, then these products would be more homogeneous within a product
category because the maximum VOC content of each product would be lower than it is now.
1Tbe first type of program is commonly referred to as an ambient permit program, and the second and more
common, an emission permit program.
2Program participants literally trade permits.
4-2
-------
Companies might face a narrower range of options for further reducing VOC emissions. The
marginal costs of emission reductions may be higher than under a marketable permit program.
The SO2 allowance trading program mandated by the 1990 amendments to the Clean Air
Act has strongly influenced the following discussion of marketable permit programs for reducing
VOC emissions from consumer and commercial products. The SC«2 program is relevant because
the characteristics of the pollutant that are significant for designing economic incentives are
similar to the significant characteristics of VOCs: uniform mixing in the atmosphere and
assimilation.3 Also, the SC»2 program has proven to be politically acceptable. It therefore
provides valuable insights into designing an effective and politically acceptable marketable
permit program for consumer and commercial products.
Just as Congress established an annual cap on SO2 emissions from power plants, the EPA
could establish an absolute maximum (per defined time period) for the quantity of VOCs
contained in and hence emitted by all or some subset of the regulated consumer and commercial
products. To start the program, EPA could distribute this aggregate level of emissions in the
form of VOC emission permits to the manufacturers and importers of the regulated products
included in the universe of sources. Each company in the universe of sources would own permits
(denominated, for example, in tons of VOC per year) which it could retain or sell to other
participants in the program. Companies would be required to possess a sufficient number of
permits to cover the total quantity of VOCs contained in their domestically sold products.
Companies could trade permits according to considerations of profitability.
The design of an implementable marketable permit program is complex because of the
many details and options for these programs. The most important are examined below.
3See Section 1.1 above.
4-3
-------
4.1 PURPOSES OF A MARKETABLE PERMIT PROGRAM
A primary purpose of a marketable permit program is to achieve a reduction in VOC
emissions at lower overall expenditure on pollution abatement than the most likely command-
and-control regulatory strategy, product-specific VOC content standards. Another important
purpose is to provide a continual incentive for cost-reducing technological change that is
potentially beneficial to the environment.
Figure 4-1 illustrates the basic operation of a marketable permit program. Overall,
Figure 4-1 describes the derivation of each source's demand for permits and in turn the aggregate
demand curve for permits, which, in conjunction with the fixed supply of permits, determines the
price of an emission permit. The figure also shows how the permit market leads to the least
expensive allocation of responsibility for emission reductions. Lastly, the figure illustrates that,
in a competitive market, any initial allocation of permits is consistent with the achievement of
least cost.
Each source emits 15 units of pollutant in the absence of emission control. Panels (a) and
(b) in Figure 4-1 show the marginal cost of emission reduction for Sources 1 and 2, respectively.
The marginal cost curves are the same as in Figure 3-1, which illustrated the basic operation of
an emission fee program. The marginal cost curve implies the emission permit demand curve
because, for any quantity of emissions reduction, the source will demand permits for the
remaining emissions at the price indicated by the marginal cost curve. A source has found its
optimal level of control when it is indifferent between increasing emissions (purchasing a permit)
or decreasing emissions. For example, when Source 1 has reduced emissions by 10 units, its
marginal cost of emissions reduction is $500, and the source is willing to pay $500 for the last
permit purchased. Similarly, when the first source has reduced emissions by 12 units and
demands 3 permits, the marginal cost of emissions reduction and the source's willingness-to-pay
4-4
-------
-------
for a permit equal PI, as indicated by panel (a). The rationale for the second source's demand
curve is the same. Panels (c) and (d) portray each source's permit demand curve.
The aggregate permit demand curve is obtained by the horizontal summation of each
source's demand curve. Panel (e) shows the aggregate demand curve and the aggregate supply
curve, which is vertical at 15 units because in this example the regulator has set the
environmental goal at 15 units of emissions. The price which clears the market is $500 per
permit.
To initiate the permit program, the regulator could allocate 3 permits to Source 1 and 12
permits to Source 2. Initially, Source 1 would be responsible for reducing 12 units of emissions,
and Source 2, 3 units. Assuming negligible transaction costs, trading would occur: Source 1
would prefer to buy permits from Source 2 at a price less than PI, and Source 2 would be willing
to sell them at a price above ?2, which is less than PI. The marginal costs of emissions reduction
are different (Pi > ?2), giving both sources the incentive to trade. Source 1 will buy a permit if it
obtains a net reduction in costs (including transfer payments), and Source 2 will sell a permit if it
obtains a profit In the example, Source 1 buys just two permits. Source 1 reduces emissions
from 15 units to 5, and Source 2, from 15 units to 10. Trading establishes the permit price of
$500.
As in the example of the emission fee program (see Figure 3-1), that allocation of
responsibility between the sources for emission reductions requires the least expenditure on
pollution abatement. Another similarity is that the correct fee rate and market-clearing permit
price are both $500 (and must be equal for the programs to achieve the same environmental
goal).
To summarize, the equilibrium allocation of responsibility for emission reductions clears
the permit market and equates the marginal costs of emissions reduction across sources. The
4-6
-------
permit program, by allowing trading, achieves the environmental goal while requiring the
minimum total expenditure on emission reductions. In a competitive permit market, any initial
allocation of permits is consistent with cost-minimization in the short-run, although the pattern of
permit buying and selling and the consequent inter-company transfers will vary across
allocations.
4.2 FUNDAMENTAL ATTRIBUTES OF THE PERMIT
Under a marketable permits program, the emission permit is the tradable commodity.
The environmental purpose of the program is incorporated into the type of permit. Sources base
their valuation of a permit on the specific attributes of the permit. This section discusses the
attributes of the permit that are fundamental in the sense that they are essential to the definition
of the permit regardless of any further specification of the permit.
4.2.1 Detraction
The possible mechanisms for allocating marketable emission permits lie on a spectrum:
at one extreme, the government would own the entitlement to emit air pollutants and sources
would pay to pollute; at the other extreme, sources would own the entitlement to pollute and the
government would pay to improve the environment (Tietenberg, 1980: p. 399). The acid rain
program shows that a third mechanism is also possible: the EPA, even though it owns the
entitlement, will freely grant SO2 allowances to utility units. Therefore the issue of who pays to
obtain an initial allocation of permits is conceptually separate from the issue of who initially
owns the entitlement. Although the Federal government is the initial owner of the entitlement,
an issue related to ownership that must be resolved is the right to rescind or diminish emission
permits from their current owner without compensation.
4-7
-------
The acid rain program assigns this right—referred to as detraction—to the EPA. At any
point in time, the EPA may reduce the number of SO2 allowances to conform emission
reductions to the emissions cap.
Detraction is important because the EPA, sometime after the start of a marketable permit
program, may find that air quality has not improved sufficiently. The complexity of the program
may have inadvertently resulted in an excessive level of emissions, or societal goals for
environmental quality may have increased. The only backstop measure that preserves the
integrity of a marketable permit program is a reduction in the number (or lower denomination) of
permits and hence emissions from the participants. If the EPA were to possess the right of
detraction, it would be able to act unilaterally to enhance the stringency of the program without
committing an unconstitutional taking of the detracted permits.
Although a discussion of the EPA's authority is beyond the scope of this report, it is
appropriate to observe that Title IV of the Clean Air Act Amendments of 1990 appears to give an
unlimited right of detraction to the Federal government.
An allowance allocated under this title is a limited authorization to emit sulfur dioxide in
accordance with the provisions of this title. Such allowance does not constitute a
property right. Nothing in this title or in any other provision of law shall be construed to
limit the authority of the United States to terminate or limit such authorization (Section
403(f)).
The acid rain program may provide a precedent for assigning the right of detraction to the EPA
for the purposes of implementing a marketable permits program for consumer and commercial
products.
Possession of the right is not costless. If a company is uncertain over whether it will
continue to own an excess permit, then the value of future use or sale is less than when the
government does not possesses the right of detraction. One potential cost of this assignment of
rights is that permit owners may be less certain of the future value of permit holdings, increasing
4-8
-------
reluctance to bank permits and to engage in intertemporal trades. Therefore, the efficiency of the
permit program may be somewhat reduced.
One criticism of marketable permit programs asserts that detraction, viewed as a virtual
necessity due to the "extreme uncertainties involved in calculating sustainable levels of
pollution" (Colby, 1991: p. 206), is made costly by permit owners' demand for compensation.
This evaluation overlooks the possibility of creating a marketable permit program without
establishing new private property rights.
4.2.2 Ambient Permit or Emission Permit
Depending upon the environmental problem, the tradable commodity could be either an
ambient permit or an emission permit. Emission permits allow emissions per se, but ambient
permits allow air quality impacts at specific monitoring sites. Ambient permit programs are
more efficient when emissions from different sources impose different air quality impacts.
However, ambient permit programs are more complex, requiring detailed air quality modeling
information.
Volatile organic compounds are precursors to the formation of tropospheric ozone, and
they mix uniformly in the atmosphere with the other precursors. The total quantity of precursor
emissions in an airshed but not necessarily the location of the sources influences the
concentration of ozone. For a uniformly mixed, assimilative pollutant, an emission permit
program is cost-effective because control of the total quantity of emissions is sufficient to
achieve an environmental goal at least cost (Tietenberg, 1985: pp. 16-22). Therefore, a
marketable permit program for consumer and commercial products that employs emission
permits would not require air quality modeling information and may be as effective as an
ambient permit program.
4-9
-------
For purposes of quantification, it would be assumed that emissions occur in the year in
which the product is manufactured in the United States or imported. The assumption would be
imprecise if either manufacture and distribution occur in different years or purchase and use
occur in different years. Although some manufacturers, distributors, and users probably behave
in ways that have those consequences, the resulting discrepancy between assumed and actual
emissions is probably negligible and would not justify changing the methodology for quantifying
emissions. Most important, any discrepancy would not affect the incentive to reduce emissions.
The emission quantification method treats as equivalent the emission reductions that
occur from redesign of the product or package and the reductions from operational contingencies
(i.e., production curtailments and shut-downs).4 This aspect of quantification is appropriate
because neither the particular means which sources employ to reduce emissions nor the
motivations for the emission reduction are significant to the achievement of the environmental
goal of a marketable permit program.
4.2.3 Emission Allowances or Emission Reduction Credits
In the acid rain trading program, the Phase I emission limitation for each included utility
unit is 2.5 pounds of SOi per million BTUs of fuel combusted during the base year, and each
unit receives as a free grant from the EPA the implied number of SO2 allowances less a fraction
retained by the EPA for special reserves. (Therefore, the emission limitation exceeds the
allocation of allowances.) Each unit's control requirement equals the excess of actual emissions
over the allocation of allowances.5 The key point is that the commodity in the program is an
emission allowance rather than an emission reduction credit, which is valid only after
4 Add-on controls (for example, paint booths with carbon absorbers) are generally infeasible for consumer and
commercial products. Reductions in VOC emissions can come from reformulation, packaging redesign, and
reduced production.
^Trading allows utilities to comply with the control requirement by achieving on-site emission reductions or
acquiring allowances from other utility units.
4-10
-------
government certification, prior to trading or banking, of over-control of emissions and which can
be used only in trades approved by the government.
With emission allowances, the EPA's certification of a participant's compliance with the
program occurs at an aggregate level instead of the level involving individual emission
reductions and trades. At the end of the year (or other defined period), the EPA would compare
each participant's actual emissions to its permit holdings for the year. A participant is in
compliance with the permit program if its emissions do not exceed its holdings. Emission
allowances may involve less administrative costs, may provide more certainty to participants,
and may allow trades to occur sooner.
4.2.4 Banking Provisions and Duration of Permits
A very important objective of establishing a permit market is cost-effectiveness: in a
competitive market, buyers and sellers of emission permits will exhaust all mutually beneficial
trades and reduce emissions with the least total expenditure on abatement. To achieve these
gains, different sources will trade with each other. Some sources will assess current and future
demands for permits and conclude that their best interest is served by reducing emissions now
and selling or using permits in the future—banking permits. Restrictions on current or
intertemporal trading reduce the potential gains from trade and increase the aggregate
expenditure on abatement.
Allowing sources to bank permits for an indefinite period conduces to the minimization
of emission reductions cost but may increase administrative costs.6 Banking also makes possible
an increase in emissions in some future but predictable year. Thus banking implies a tradeoff
between cost-minimization and certainty about future emissions in any particular year.
6If detraction applies to active permits, then it should also apply to banked permits.
4-11
-------
4.3 SEASONAL AND GEOGRAPHICAL DIFFERENTIATION OF PERMITS
4.3.1 Seasonal Differentiation
The national ambient air quality standard for ozone specifies that the daily maximum
hourly average concentration (0.12 ppm) shall not be exceeded more than once a year. Because
of the chemistry of ambient ozone formation, peak concentrations "typically occur during hot,
dry, stagnant summertime conditions (high temperature and strong solar insolation)" (U.S., EPA,
OAQPS, 1991: p. 3-29). Emissions of VOCs during warm weather are more likely to result in a
high peak hourly concentration than emissions during cool weather.
For that reason seasonal differentiation of permits could be desirable. Two types of
permits could be created: one for VOC emissions during the ozone season and the second for the
rest of the year (the "off-peak season"). The price of the ozone season permit would be higher
because greater emission control would be required to reduce the formation of ozone. Sources
would economize on emission control by switching emissions from the ozone season to the off-
peak season. Therefore, the most cost-effective—with respect to the cost to reduce ozone rather
than VOCs—marketable permit program may involve seasonally differentiated permits.
However desirable seasonal differentiation is in theory, a major problem would arise
when applying this idea to a marketable permit program for consumer and commercial products.
The primary purpose of the program would be to reduce emissions of VOCs from the use—not
manufacture—of consumer and commercial products. Therefore, for the seasonal differentiation
of permits to result in a seasonal shift of emissions, the program must shift the use of consumer
and commercial products to the off-peak season.
The problem arises because, in a program whose universe of sources conforms to the
statutory definition of regulated entity, the owners of permits generally will not be the users of
consumer and commercial products, and the incentives for switching to off-peak use would be
4-12
-------
weak. The price differential of emission permits may induce a price differential in consumer and
commercial products: within a product category, a product made from "ozone season" VOCs
would be more expensive than a product made from off-peak season VOCs. Product users would
respond by shifting the timing of purchase but not the timing of use. For example, users of
architectural and industrial maintenance coatings may stillfavor application of the coating during
the ozone season because of superior adhesion to the substrate and shorter drying time. They
may respond to the coating price differential by buying coatings when they are cheap (i.e., during
the off-peak season) and applying them when the weather is warm (i.e., the ozone season). The
response would defeat the purpose of the program (and bid up the price of the cheaper coating,
nullifying the impact on coating users).
One can imagine an enforcement scheme that would guide behavior in the preferred
direction, but the requisite monitoring of many thousands of users would be extremely intrusive
and expensive. Seasonal differentiation of permits appears to be inappropriate for reducing VOC
emissions from using consumer and commercial products—unless a more acceptable approach to
monitoring is found.7
4.3.2 Geographical Differentiation
Tropospheric ozone is a regional pollutant Emission reductions in one nonattainment
area or transport zone, for example, Los Angeles, improve air quality in Los Angeles but do not
affect air quality in San Francisco, Houston, or any other nonattainment area.
The regional nature of the ozone problem is an argument for geographically differentiated
permits: every regulated entity in the universe of sources could hold emission permits for each
nonattainment area or transport zone in which the entity's consumer and commercial products
7Tne temporal differentiation of permits also could include diurnal and episodic dimensions (Tietenberg, 1985:
pp. 149-167)) but the monitoring problem is similar.
4-13
-------
emit VOCs. Theoretically, geographical differentiation would allow the EPA to exactly limit
emissions from the consumer and commercial products used in an area.
For example, in the program with the greatest geographic differentiation, if the universe
of sources consists of manufacturers, then every manufacturer would be required to hold the
requisite number of permits to cover the emissions from products used in Los Angeles, Houston,
and every other nonattainment area in which its products are used; if its products are used in
every nonattainment area, then the manufacturer would be required to hold as many types of
permits. Because this program strictly incorporates the principle that emissions in different areas
have different and incommensurable effects, permits could not be traded across geographical
boundaries. If instead it is decided that the effect of emissions in one nonattainment area is the
same as the effect of emissions in another area, then both areas should be treated as one for the
purposes of structuring the permit market.8 The comparative effect of emissions in different
areas is a crucial parameter for geographically differentiated marketable permit programs.9
An economic consideration that is relevant to the determination of the number of permit
markets is the impact of increasing geographical differentiation on the competitiveness of the
permit market. It is reasonable to conjecture that the smaller the nonattainment area—in which
size is probably most usefully and simply defined as population—the smaller the number of
different manufacturers of the same type of consumer or commercial product. As the number of
potential trading partners decreases, fewer trades occur because the cost of finding a suitable
partner increases, and dominance of the market by one buyer or seller becomes more likely. The
8 Between these extremes is the intermediate case in which the effect of emissions in one area is proportional to the
effect of emissions in a different area. The corresponding permit program creates as many types of permits as
areas but allows trading across areas, the relative exchange ratio being determined by the "transfer coefficient" or
constant of proportionality.
9 A permit program with geographically differentiated permits requires a greater expenditure on administration by
the EPA and permit holders, reducing somewhat the gain from a more desirable spatial pattern of emissions. A
compromise between differentiation and ease of administration can be achieved with a hybrid program in which
small sources participate in a national system, but large sources participate in multiple regional systems
(Tietenberg, 1980: p. 407).
4-14
-------
reduction in competitiveness reduces the abatement cost savings achievable with the permit
program.10
It is also appropriate to repeat the conjecture that the incentive to product users to
transport products across boundaries to take advantage of price differentials (reflecting
differences in permit price) leads to an ineffective program without extremely intrusive
monitoring. Therefore, the benefits of geographic differentiation come at the price of increasing
the monitoring of users and decreasing the cost-effectiveness of marketable permit programs.
4.4 UNIVERSE OF SOURCES
The universe of sources refers to the regulated entities who participate in the marketable
permit program. The potentially regulated entities are manufacturers, processors, wholesale
distributors, and importers of consumer and commercial products (Clean Air Act Amendments of
1990, Section 183 (e)(l)(Q). The definition of potentially regulated entity excludes suppliers of
the raw materials for the manufacture of consumer and commercial products and, at the opposite
end of the product distribution channel, retailers and users.
The following characteristics of regulated entities are relevant to the selection of the
universe of sources:
• type of product,
• type of business, and
• size.
Each characteristic implies a criterion for determining whether a potential participant will be in
the universe of sources; inclusion may also depend upon the simultaneous satisfaction of
multiple criteria. Much of the discussion of the significance of type of product, type of business,
and size for the universe of sources in fee programs is relevant to marketable permit programs;
1(>The deterioration of competitiveness could be offset by increasing the scope of the universe of sources to include
additional products.
4-15
-------
that material is repeated below. Permit programs do present unique issues such as the
participation of environmental organizations and the elective participation by regulated entities
that opt-in. These issues are also discussed below.
4.4.1 Selection of Sources by Type of Product and Sequence of Regulation
The greater the variety of consumer and commercial products that are included, the
greater the opportunities for permit trading. A corollary is that a permit buyer is more likely to
find a permit seller who does not compete in the product market Increasing the variety of
included products increases emissions reduction cost savings.
The universe of sources and the calculation of the emissions limitation are linked in the
most efficient permit programs. If the universe of sources is planned to expand in the future via
the inclusion of additional product categories, then determination of the emissions limitation
formula at the outset enhances the ability of future program participants to undertake strategic
planning and hence reduces the aggregate cost of emissions reduction.
The SO2 trading program allows the universe of sources to change over time, and
provides a model for the consumer and commercial products program. The 862 program has
two phases. During the first phase, the universe of sources includes only large utility units that
also have a high emissions rate. During the second phase, nearly all other utility units are added
to the universe of sources. All sources know the formulas to be used to determine their
emissions limitation.
This feature of the SO2 trading program required the Congress to choose an overall cap
on emissions. Similarly, the EPA may need to choose an overall cap on VOC emissions to
minimize costs in an expanding permit program for consumer and commercial products.
4-16
-------
Another issue in the selection of regulated entities for inclusion in the universe of sources
is the grouping of products by priority of regulation. Although the identification of those groups
is beyond the scope of this report, it is germane to consider whether the sequence of regulation
should influence other elements of the design of marketable permit programs.
4.4.2 Selection of Sources by Type of Business
Product manufacturers control VOC emissions from the use of consumer and commercial
products by changing solvents, changing propellants, redesigning packaging (for example,
substituting a pump sprayer for an aerosol can), reducing production, and reformulating paints
and coatings to reduce emissions during film-forming. Manufacturers are good candidates for
participation in a permit program because the incentives provided by potential sale of emission
permits motivates appropriate responses. To extend the coverage of the permit program to
imported products, importers also should be included in the universe of sources.
However desirable the inclusion of manufacturers and importers, distributors are also
potential participants. A practical problem with operating the permit program at the level of the
distributor is ensuring complete coverage of emissions because some but not all consumer and
commercial products pass through a distributor on their way to a retail outlet. Complete
coverage could be ensured by defining "distributor" sufficiently broadly.
Distributors can discontinue products, carry other products than what they have been
carrying, and increase wholesale prices. These responses to the permit program would motivate
users of consumer and commercial products to change purchasing patterns and would provide
information to manufacturers on the salability of products with different VOC content.
Manufacturers would then be prompted to reformulate products, to redesign packaging, and to
discontinue some products. Therefore manufacturers would receive the same incentives as if
they participated in the permit program, but the incentive would be indirect.
4-17
-------
A distributor-based program achieves the same environmental goal as a manufacturer-
based program, but increases administrative costs for both the EPA and sources. Administrative
costs are higher because the number of participants is greater: there are typically more
distributors than manufacturers and, in addition to distributors, manufacturers who distribute
products directly would need to be included.
4.4.3 Selection of Sources by Size
The size (e.g., quantity of emissions, value of sales, etc.) of a source is relevant to the
construction of the universe of sources for two reasons. First, the EPA must consider the
economic impact of proposed regulations on small businesses and develop alternatives when the
impacts are onerous. Second, the exemption of "small" sources from mandatory participation in
the marketable permit program may benefit both the source and the EPA without substantially
affecting the effectiveness or efficiency of the program. Although this report does not
investigate the magnitude of small business impacts of marketable permit programs, it is
desirable to briefly investigate the conditions under which they might arise and to consider
adaptations of the permit program.
Any source which participates in a marketable permit program evaluates its own demand
for permits, searches for buyers or sellers of permits, judges offers, and complies with reporting
and record keeping requirements. Those activities may be costly, and although the costs
correlate with the volume of trade, it is reasonable to expect that costs will be proportionately
greater for small sources than for large sources. The EPA will also incur costs in identifying and
communicating with small companies and keeping the appropriate records. The value of the
emission reductions obtained from small companies as a result of their participation in a permit
program may not outweigh the total costs (including the companies' costs and the EPA's costs)
of including these companies in the universe of sources.
4-18
-------
A small number of product manufacturers appear to be responsible for a large percentage
of sales (see Table 2-2). Therefore it may be possible to control a large percentage of VOC
emissions from consumer and commercial products by including only a small number of
companies in the permit program.
A selection criterion based on size, however desirable from the standpoint of reducing
costs, could have some unintended consequences. First, after the division of large and small
sources was determined (where large sources are included in the universe and small sources are
not), a mechanism would have to be designed to handle small sources which grow and become
equal in size to large sources. Second, basing inclusion in the universe of sources on company
size could provide an incentive for large companies to divest themselves of some product lines.
As new, smaller companies were set up to manufacture the products, they could conceivably be
small enough to be excluded from the universe of sources. (Exclusion may not be problematic if
these small companies were to be subject to VOC content standards rather than exempt from
regulation.) Third, including only a small number of large manufacturers in the universe may not
provide for active trading in the market for permits because of too few buyers and sellers and
could enhance the likelihood of permit price manipulation.
Independently of the sequence of regulations, sources included in the permit trading
program could be phased-in. For example, on the basis of volume of emissions, the largest
manufacturers or distributors and importers could be selected for inclusion in Phase I of the
program.11 All other manufacturers, distributors, and importers then could be added to the
universe of sources in Phase II.
universe of sources should also be sufficiently large to create an active and competitive permit market.
4-19
-------
4.4.4 Opt-Ins
Another way in which the universe of sources could expand over time is to allow
manufacturers and importers to "opt-in" to the permit program. Opting-in could encompass 1)
the voluntary inclusion of a company who would not otherwise be included in the universe of
sources and 2) a company already included in the universe of sources could elect to have
additional product lines (other than its regulated consumer and commercial products) included in
the permit program.
Depending upon the sequence of regulations and the universe of sources to be regulated,
additional consumer and commercial products releasing VOCs would be outside of the reach of a
permit program. Opting-in would allow the permit program to increase the VOC emissions
brought under control and enhance trading opportunities, thus providing additional compliance
flexibility to all participating sources and reducing abatement costs. Because opting-in would be
voluntary, companies electing to participate would only do so if they anticipated making profits
from permit sales or if they were subject to more costly alternative regulatory requirements (e.g.,
content standards). No company would have anything to lose from having the freedom to opt-in
to the permit program.
Opting-in requires a procedure for allocating permits to opting-in sources. The SO2
trading program provides an example. Sources that opt-in to the SO2 program receive an
allocation of emission allowances from the EPA based on the following formula (Lock and
Harkawick, 1991, p. 21):
TT . , , ,. f ., c , „/ lesser of its actual or \
Unit s baseline fossil fuel use x (aUowable 1985 SO2 emissions rate)
In the consumer and commercial products program, opting-in sources could receive permits
according to a formula that similarly takes into account either actual or allowable emissions.
4-20
-------
4.4.5 Participation in a Marketable Permit Program by Environmental Organizations;
The EPA could extend participation in the program to groups outside the consumer and
commercial products industry, specifically, environmental interest groups. The value that
environmental interest groups attach to marketable permits, is based on the emission reductions to
be obtained by permanently removing the permits from circulation and use by manufacturers,
distributors, or importers. The discussion also applies to other external groups such as local
environmental agencies and city governments.
There are two opportunities for environmental organizations to participate in a
marketable permit program: during the initial allocation of permits and in the trading of permits.
If the EPA employed an auction to distribute a finite number of permits, environmental
organizations could participate in the auction and purchase as many permits as they desired and
could afford. Their interest in participating in the auction presumably is due to the ability to
reduce the aggregate emissions by taking permits out of circulation.
The second possible opportunity for environmental organizations to actively participate
in the program is to buy and sell permits in the market. If an environmental organization wanted
to reduce the number of permits that could be used during a particular time period, it could offer
to purchase permits from existing sources or the EPA if it has a permit reserve. Likewise,
existing sources that had generated excess emission reductions could potentially find buyers for
permits in environmental organizations.
There are advantages and disadvantages to allowing environmental organizations to
participate in a marketable permit program. One potential advantage is direct, continuing public
participation in the determination of the amount of environmental protection. If, in setting the
overall emission cap, the EPA underestimates the amount of environmental protection society
wants, then public participation in the permit market may be desirable.
4-21
-------
A related argument for including environmental organizations in permit trading is that the
organizations' purchase of permits could make the permit program more responsive to changing
environmental conditions and technological change. For example, if the introduction of a new
technology reduced the marginal costs of abatement, or if new information substantiated an
increase in the estimated marginal damage of emissions, then environmental organizations could
purchase permits and hold them indefinitely. In this situation, adaptation of the program to
changed circumstances (new technology or new information on damages) may occur faster than
if the government were to change the program (by reducing the emissions cap).
One potential disadvantage of allowing participation by environmental organizations is
that the additional demand for permits that they create could make it less costly for existing
sources to engage in anti-competitive behavior. For example, an existing source that generates
excess emission reductions can either sell or bank the extra permits. There is an opportunity cost
to the source in hoarding permits (from competitors) in the form of foregone revenues from the
sale of the permits. By selling to an environmental organization, the source can reduce its
opportunity cost of taking permits out of circulation by the amount which it receives from the
environmental organization for the sale of the permits. A second, though probably minor,
disadvantage to allowing environmental organizations to participate in the permit program is that
it may raise the costs to administer the program.
One frequently encountered objection to environmental organization participation is that
environmental organizations have had their opportunity to participate at the stage of program
design and formulation of the environmental goal. The objection expresses the concern that
environmental organizations, if allowed to buy and sell permits, would buy enough permits to
restrict industry growth. However, existing sources could voluntarily decide whether to sell
permits to environmental organizations. If sources legitimately were to feel that environmental
organization purchases of permits would restrict their growth, they could simply refuse to sell
permits to environmental organizations.
4-22
-------
The issue of participation by environmental organizations in permit markets may be of
minor importance given the limited budgets of these organizations. Oppenheimer and Russell
(1981) suggest that environmental organizations are not likely to have sufficient funds to
purchase a significant number of permits and that their resources might be more effectively spent
lobbying the legislature in an attempt to influence the environmental goal, initial distribution of
permits, and trading rules. They conclude that allowing environmental organizations to
participate in the permit market would not significantly reduce the supply of permits available to
existing sources and hence would not threaten the continuing operation of these sources.
4.5 ENTRY AND RETIREMENT OF SOURCES
The previous section described how the universe of sources may change when sources are
phased-in over a period of time or when companies not originally included in the universe opt-in.
The universe of sources may also change when new companies enter the consumer and
commercial products industry and when existing companies cease to exist This section
discusses program design issues that arise in connection with entering and retiring companies,
including the expansion of existing companies.
The entry of new sources is potentially important to the total cost of emission reductions
obtained through the permit trading program and is relevant to the maintenance of a competitive,
dynamic consumer and commercial products industry. When new manufacturers enter an
industry, they are likely to enter with modem technology and equipment Modem technology
can include new and possibly cheaper products which result in lower VOC emissions. In
addition to contributing less expensive emission reductions, new sources may also lower
production costs, increase competition, and offer new products to consumers.
The retirement of inefficient existing companies is also important in a productive
economy. Less efficient consumer and commercial products manufacturers with older
4-23
-------
equipment and technology are likely to face higher costs to obtain emission reductions than new
manufacturers. In addition, older factories are likely to have higher production costs. A permit
program should not impede the retirement of inefficient plant and equipment Therefore a permit
program should develop a method for handling the entry and retirement of sources.
4.5.1 Entry of New Sources
If new sources are required to participate in a permit program, and hence required to hold
permits for their emissions, there are two potential sources for them to obtain permits: existing
sources (and brokers) and the EPA. If new sources can obtain permits only from existing
sources, then, if the permit market is thin, existing sources could act strategically to deny new
sources the permits they need to begin operation. Theoretically, this treatment of new sources
could lead to a decrease in competition in the product market
Note that an existing source expanding operations is analogous to a new source. Existing
sources could act to prohibit another existing source from expanding operations by refusing to
sell permits or charging discriminatory prices. Therefore similar concerns about the ability to
obtain additional permits also arise for expanding sources.
Tietenberg (1985, p. 140) observes that the likelihood of predatory behavior—
discriminatory permit pricing—by existing sources declines with the heterogeneity and number
of sources. The potentially regulated consumer and commercial products are very diverse, and
they are made by thousands of manufacturers. More specifically, as Table 2-3 (above) shows,
some potentially regulated industries produce few commodities made by some of the other
potentially regulated industries. For example, in the year covered by the table, the manufacturers
of paints and allied products (SIC 2851) did not produce any commodity also produced by the
manufacturers of cosmetics and toiletries (SC 2879) and vice versa. The manufacturers in these
two industries are unlikely to compete in the product market, and would have nothing to gain
4-24
-------
from discriminating against each other in the permit market Other similar examples can be
found. Information on the products made by each manufacturer could be used to evaluate the
potential for permit trades to occur between companies that do not compete in the product
market In addition, allowing manufacturers and importers of unregulated products to opt-in
would enhance the breadth of the universe of sources. A judicious construction of the universe
of sources—its size and composition—appears to be able to protect new and expanding sources.
To provide additional protection to new and expanding sources, the EPA could reserve a
number of permits from the initial allocation to existing sources. There are two basic
implementation issues that need to be addressed if permits are to be reserved. One is how many
permits ought to be reserved from the initial allocation and from whom the permits should be
obtained. Because existing sources may refuse to sell a new source any permits at all, the
minimum number of permits to be reserved should equal the number of permits new sources
would need to begin operation. Equity may require that the burden of providing permits should
not be concentrated on a subset of existing sources and hence the responsibility for creating the
permit reserve should be shared equally by existing sources. This sharing could be accomplished
by reserving an equal percentage of each existing source's initial allocation, the sum of which
would equal the total number of permits to be reserved.
The second implementation issue is the manner in which new and expanding sources
obtain the reserved permits and what price, if any, do they pay. One possibility is for the EPA to
use an auction to distribute permits to new and existing sources and to require bids at least equal
to the market price of permits. The purpose of the reservation price would be to enhance
competition when few sources are bidding; the administrative burden on the EPA would thus
increase because it would need to stay current with the permit market. Another possibility is for
4-25
-------
the EPA to offer permits for sale to new and existing sources at a fixed price, giving priority to
new sources.12
The SO2 program provides a precedent for reducing any bias toward new sources in a
consumer and commercial products permit program. Approximately three percent of each
existing source's initial permit allocation is withheld each year. These reserved permits are
auctioned to both new and existing sources and the revenues paid out to the existing sources
from whom the reserved permits were withheld. Thus the price for auctioned permits is set
competitively, not by the EPA. Because existing sources could possibly obtain all of the permits
at auction, to the exclusion of new sources, the EPA also offers for direct sale a limited number
of permits. New sources have priority to purchase these permits over existing sources. Thus the
EPA serves as the supplier of last resort (through direct sales of permits) should new sources be
unable to purchase permits from existing sources. The price of these direct-sale permits has been
set two to three times higher than the expected trading price of the permits in the open market
Regardless of the method by which the price of permits to new sources is determined, if
existing sources receive an initial allocation of permits for free, there is an implicit disadvantage
in general to new sources in requiring them to pay for their initial allocation of permits. If
permits are valid for only one period, for example, a year, then the source must incur the same
costs or similar costs in the following and all later years because it does not receive a free initial
allocation from the government. The purchase of permits in this case would not be a one time
expense, and would to some degree be an operating expense that existing sources would avoid on
average because of the free grant of permits to them.
The extent to which new sources would have difficulty obtaining permits is uncertain.
Elman, Braine, and Stuebi (1990) analyzed the availability of allowances for new coal-fired
12It may be that new sources deserve more favorable treatment than expanding sources because the former do not
have the capacity to internally supply permits as do expanding sources (for example, by reducing the VOC
content of another product).
4-26
-------
generating plants in the SO2 program. Their analysis indicates that there should be sufficient
allowances available for new sources and that additional allowance costs relative to annual
operating costs for new sources should be small (ranging from 1 to 5 percent). However, there
are significant and obvious differences in production, demand, and structure between the utility
industry and the consumer and commercial products industry. Those differences prevent one
from easily extrapolating the results from Elman's study to the consumer and commercial
products industry.
The diversity of manufacturers and importers in the consumer and commercial products
industry can provide a sufficient safeguard against discriminatory permit pricing targeted against
new sources. If the universe is broad, additional measures such as a permit reserve and
preferential treatment of new sources may not be necessary.
4.5.2 Retiring Sources
As stated in the introduction to this section, the primary concern about exiting sources is
that the permit program could result in an incentive for inefficient companies to remain in
operation. One way to avoid this problem is to allow retiring sources to sell their permits to
other (new or existing) sources.
The EPA could detract the permits of retiring sources and thus permanently reduce
emissions through the attrition of existing sources. However, such action is likely to provide an
incentive for existing sources, who might otherwise find it profitable to sell their permits and go
out of business, to continue operation of potentially inefficient plant and equipment. When
making the shutdown decision, the owners of a company compare the salvage value of the
company's assets to the value of continuing operation. Emission permits are assets because they
can be sold to other companies and hence increase the salvage value of a company. If the EPA
allows retiring companies to sell permits, then the salvage value will be more likely to exceed the
4-27
-------
value of continuing operation and a company facing a shut down decision will more likely close.
The EPA's decision to allow retiring companies to sell their permits will accelerate rather than
retard the retirement of inefficient plant and equipment
Other existing or new companies may employ better technology which could
manufacture more product than the exiting company with the same number of permits. By
allowing retiring companies to sell permit holdings (including its annual initial allocation of
permits), the EPA could allow for expansion of production without increasing emissions and at
the same time provide another way for new sources to obtain permits.13
4.6 INITIAL ALLOCATION OF PERMITS: GRANTS AND AUCTIONS
To start the marketable permit program, the EPA establishes: 1) the total allowable
emissions of VOCs from all program participants and 2) the allocation of emission permits to
each participant.14 A source's allocation and its actual emissions when the program starts
implies a source's initial emission reduction requirement. Thus the EPA assigns the initial
responsibility for emission reductions and influences the distribution of the burden of the
program.
A participant's initial control responsibility may take any value: positive when
uncontrolled or actual emissions exceed the permitted amount; zero when uncontrolled and
permitted emissions are equal; negative when permitted emissions exceed actual emissions. If
the government possessed the entitlement to emit, the use of an auction to allocate permits would
reflect the judgment that each participant's control responsibility is equivalent to total control
because the participant must either reduce emissions entirely or purchase permits to cover
emissions. If instead the source possessed the entitlement to emit, the control responsibility must
13Production can increase while holding emissions constant if the exiting source sells its permits to a new or existing
source which manufactures products with lower VOC content
14For a program in which the EPA periodically issues permits, the "initial" allocation refers not to a single event but
to regularly repeated allocations.
4-28
-------
be presumed to be zero, and permitted emissions would equal actual emissions.15 It is only when
the EPA authorizes a partial control responsibility that it encounters a choice with any latitude
for the handling of individual companies.
If the EPA were to authorize partial control responsibility, then it would have two choices
for allocating permits. It could freely grant permits according to a distributional rule, or it could
auction them, collecting revenues for various purposes.
Granting emission permits is attractive because it allows the EPA the greatest latitude to
allocate permits and it may enhance the ability of the EPA in an uncertain world to obtain
emission reductions (detractions) beyond those contemplated at the start of the marketable permit
program. The EPA can make the allocation without regard to static efficiency: in a competitive
permit market, regardless of the initial allocation of permits, trading will result in the least
aggregate expenditure of resources on emission reductions (Montgomery, 1972).16 This result
assumes that transaction costs are minimal; in general, transaction costs otherwise will preclude
trades from occurring, increasing expenditures on emission reductions above the theoretical
minimum.
The following steps illustrate the implementation of this approach for existing companies
and assumes that the total cap on emissions has been selected by aggregating over hypothetical
VOC-content standards for each regulated product.
1. Calculate each company's emission limitation (VOCsj) using this formula:17
I VOCi*Qij=VOCsj
i
2. Grant the implied number (VOCsj) of allowances to each company in the first period.
^Tbe EPA could purchase permits thus subsidizing emission reductions.
16In general, the starting point of a company does not affect its desired level of emission reductions, but the initial
allocation does affect whether a company is a net buyer or seller of permits.
17VOQ is the hypothetical VOC content limit for product i and Qy is the company's sales of product i in the base
year.
4-29
-------
3. If the aggregate emissions cap declines in the future, grant proportionately fewer
allowances to each company.
In the example, the EPA grants emissions allowances to companies. Other allocative
mechanisms are feasible and may be more desirable.
Although distributive flexibility is important, grants do have disadvantages. The areas of
concern are competition in the permit and product markets and technological change. It is most
informative to evaluate the desirability of grants by comparing them to auctions, which are the
most probable alternatives.
Permit markets may be imperfect in various ways (Tietenberg, 1985, pp. 125-148). One
or more companies may attempt to manipulate the auction permit market to reduce the permit
price and its financial burden (permit payments plus abatement expenditures). Alternatively,
companies may attempt to sequester permits (thus raising the price) to injure competitors in the
product or input markets. In each case, imperfect competition increases aggregate control costs
above the optimum.
A heterogeneous and large universe of sources thwarts the company intent on injuring
competitors because the permit market contains many other potential trading partners. They
reduce the aggressive source's leverage over the permit price (because its demand is less
significant the larger is the total demand18), and competition in the product or input markets is
not a consideration, at least for some sources. Thus the tools for limiting discrimination in the
permit market are to be found in the construction of the universe of sources, just as they were
when the problem was reducing the potential bias against new sources. Special types of auctions
provide another means for enhancing competition and are discussed more fully below.
18Specifically, the price-making source's excess demand for permits—the difference between its initial allocation of
permits and permit use—but not the source's use of permits per se (i.e., equilibrium emissions) is the basis for
market power (Hahn, 1984).
4-30
-------
An artificially low permit price is also undesirable because it induces an inefficient
allocation of control responsibility. Simulation studies of permit auctions and permit grants
provide insights into the possible quantitative effects of this type of distortion.
After reviewing simulated auctions in which the dominant source reduces the permit
price, Tietenberg comes to the "strong conclusion" that control costs would increase by a very
small amount (by one or two percent in the simulations) even when circumstances are conducive
to permit price manipulation.
We have examined a very congenial setting for market manipulation, one involving few
sources, low marginal control costs for the price-setting source, and high control costs for
the competitive fringe. Despite finding circumstances where prices and total financial
burden were dramatically affected, regional control costs were remarkably insensitive to
market manipulation in all simulations (1985: p. 132).
Further, "incentive-compatible" auctions, which cleverly frustrate intentions to manipulate
permit prices, may be available for use in auction markets which are prone to manipulation.19
The free distribution of permits also creates opportunities for companies to manipulate
the permit market Tietenberg reviewed simulation studies of controlling SC>2 in Los Angeles
and hydrocarbons in the United States. In these studies, the dominant source chooses the permit
price that minimizes its own expenditures on pollution abatement and permits. He concluded
that the studies "support the notion that high degrees of cartelization are necessary before control
costs are affected to any appreciable degree and that even high degrees of cartelization [e.g., 90
percent] do not significantly erode the large savings to be achieved from permit markets"
compared to command-and-control regulation (1985: p. 136).
The special type of auction known as the "zero-revenue auction" is one defense against
attempts to exercise power in the permit market, and it has other advantages as well (Hahn, 1983;
19The efficacy of incentive-compatible auctions in the context of marketable permits is controversial. Tietenberg
expresses optimism (1985: p. 133), but Marin, doubt because the theoretical investigations of auctions have not
examined the implications of the type of bid that is characteristic of permit auctions (i.e., a demand schedule) or
the existence of the post-auction permit market (Marin, 1991: p. 298).
4-31
-------
Hahn and Noll, 1983). Several steps are involved in allocating permits through a zero-revenue
auction. The government first gives permits to each participant, following a method or formula
that reflects diverse considerations, including equity. Then the auction itself begins. Every
participant offers for sale all permits that it received, and every participant submits a bid for the
permits that it wants to buy. A bid describes the number of permits that are desired at various
prices.20 The government uses these bids to redistribute permits. The participants who are net
buyers of permits make a payment to the government; the participants who are net sellers
receive a payment from the government. The government's receipts equal disbursements.
Because each participant is on the demand side of the market, the zero-revenue auction precludes
some of the distortions that could arise if the EPA were to simply grant permits, specifically, the
emergence of aggressive sellers—at least initially when information on the value of permits is
most uncertain.
The zero-revenue auction quickly establishes an equilibrium price using the maximum
amount of information that is available at the start of the economic incentive program, and it
allows the EPA to freely grant permits. Experiments show that even when the initial allocation
of permits is very different from the least-cost allocation, such that some participants may be in a
position to dominate the market, the participants do not manipulate the permit price (Hahn,
1983). The zero-revenue auction therefore promotes efficiency without requiring participants to
transfer wealth to the government (Harrison and Portney, 1982, p. 175).
This review of permit market imperfections implies that any potential imperfections in
the consumer and commercial products permit market are probably insignificant (i.e., will have
negligible effect on aggregate control costs) regardless of the permit allocation mechanism.
Further, should evidence on the potential for price manipulation be adduced, incentive-
compatible or zero-revenue auctions may be used to lessen this potential.
20In effect, each participant submits its demand curve to the government, which acts as the market clearing agent.
4-32
-------
Permit allocation methods can affect the static efficiency of the permit market, but this
effect is not the only attribute which is relevant to program design. Permit allocation methods
should also encourage new ways of making low-VOC consumer and commercial products.
From a macroeconomic perspective, new technology can decrease the conflict between
environmental quality and economic growth. Society continually redefines the terms of
engagement for those conflicting goals, and technological change adapts the economy to
heightened demands for environmental quality.21 "Over the long haul, perhaps the most
important single criterion on which to judge environmental policies is the extent to which they
spur new technology toward the efficient conservation of environmental quality" (Kneese and
Schultze, 1975: p. 82).
Theoretical analysis shows that the initial allocation of permits influences the dynamic
efficiency of marketable permit programs. Milliman and Prince (1989) have shown that the
method of allocating permits influences the willingness of program participants to promote
innovation in pollution control and diffusion of the innovation—the two steps in technological
change. Their willingness to promote technological change depends upon their potential private
gain. Specifically, a company's gain depends on the net change in profits from product
manufacture and sale, profit from permit sales, and reduction in expenditures on permits.
Different allocation methods imply different incentives to promote technological change.
Freely granted permits and auctioned permits provide equal incentives to innovate, but
free permits provide less incentive to promote diffusion.22 When permits are auctioned, all
regulated companies (permit buyers) gain from the diffusion of low-VOC technologies because
2 Economists who equate economic growth with economic progress would make the point with a different
emphasis: technological change permits demands for environmental quality to be tolerable. "Unless we want to
blight the fruits of economic growth, we shall have to be continuously pressing on the technological frontier in
the search for means of accommodating economic progress within a clean environment" (Kneese and Schultze,
1975: p. 83.)
22The auction analyzed creates revenue for the government as permit seller.
4-33
-------
the aggregate demand for permits falls, reducing permit prices. When permits are given away,
this gain from diffusion is absent The innovator's gain from diffusion is greatest with auctioned
permits because, after diffusion, he or she buys permits from the government at a reduced price,
but with freely granted permits, the innovator loses the permit sales revenue that was obtainable
prior to diffusion. As in emission fee/rebate programs, because the innovator also could lose its
competitive advantage by providing its technology to competitors, the total incentive for
diffusion is especially low in freely granted permit programs. Any loss in competitive advantage
also occurs in auctioned permit programs. The freely granted permit program is more
disadvantageous to an innovator than an auctioned permit program, although the additional
disadvantage is difficult to measure.
Therefore, one disadvantage of freely granting permits in a consumer and commercial
products program would be that the pace of diffusion of technological change would not be as
great as with auctioning permits.23 One cannot be specific because the Milliman and Prince
analysis only compares incentives in relative terms and does not provide a methodology for
correlating potential gains from diffusion into actual rates of diffusion.
In forming an overall comparison of different permit allocation methods, the certain
advantages of freely granting permits (including the homologous zero-revenue auction)—
distributional flexibility and demonstrated acceptability to the EPA, industry, and
environmentalists—are strongly influential but may not be decisive. The counterpoint to these
advantages may be somewhat slower diffusion of cost-reducing technology.
23Zero-revenue auctions are equally disadvantageous. Zero-revenue auctions and freely granted permits provide the
same incentives for technological change because the distributional impacts of permit price changes are
subsumed within program participants in both allocations.
4-34
-------
4.7 MONITORING AND ENFORCEMENT
Given the design of a marketable permit program, a profit maximizing company
determines its level of emissions based on the costs of compliance and the expected costs of
noncompliance. The expected cost of noncompliance is the product of the penalty for
noncompliance and the probability of detection of noncompliance. Consequently the efficiency
of the permit program is intimately tied to monitoring and enforcement.
Compliance with a consumer and commercial products permit program occurs when
participants hold a sufficient number of permits to cover the total amount of VOCs which their
products emit. Consequently, the objectives of monitoring must be to observe the quantity of
VOCs contained in each company's regulated products and permit holdings.
To reduce the EPA's monitoring costs, participants will be required to provide
information on VOC content and product sales. Then monitoring compliance in the permit
program involves three separate steps.
1) Verifying accuracy of participants' reports of VOC content.
2) Verifying participants reports of the volume of each regulated product they
manufacture, distribute, or import for domestic sale.
3) Ensuring that each participant's permit holdings cover total VOC emissions from its
regulated products.
The states or EPA could verify reports of VOC content by sampling and testing products
from retail establishments and wholesale distribution centers. To determine if participants'
reports of sales volumes are correct, the states or EPA could randomly audit their sales records
and compare the records with similar reports prepared for other purposes. In addition to ensuring
compliance with the permit program, this information (VOC content and sales volume) is
valuable to the EPA as a means to evaluate the success of the permit program in reducing VOC
emissions.
4-35
-------
The final step in monitoring compliance with the permit program is to determine if
participants hold enough permits to cover the total number of VOCs contained in their consumer
and commercial products. In order to make this determination, the states or EPA would have to
track each participant's holding of permits, which would change from the initial allocation if it
bought or sold permits.
As discussed in the fee program section, the enforcement instrument of a pollution
control strategy must have two qualities. First, penalties must be severe enough that, when
combined with the chosen monitoring strategy, they deter noncompliance. Second, companies
must perceive the enforcement agency as willing and able to determine and apply the appropriate
penalty in a timely manner.
The following actions could constitute noncompliance under a consumer and commercial
products permit program:
1) Failure to submit VOC content and sales reports to the EPA in a timely manner.
2) Falsification of any information contained in the required reports.
3) Failure to hold a sufficient number of permits.
4) Failure to remit financial penalties to the EPA in a timely manner.
Each failure could be penalized to the extent permitted under the CAA.
The resources spent on enforcement are linked to the resources spent on monitoring. A
permit program which combines accurate and reliable monitoring with appropriate penalties
would effectively deter noncompliance. As a result, fewer resources (administrative and legal)
would be required to impose penalties and the overall program would be more efficient
4-36
-------
4.8 INFORMATION NEEDS FOR PROGRAM DESIGN AND ADMINISTRATION
The discussion in this section conveys the categories of information needed to design and
administer marketable permit programs. An important design parameter is the aggregate
emissions cap because the basic principle of the program is. the creation of a market for a fixed
supply of VOC emission permits.24 An important use of information for program design is to
establish the baseline of VOC emissions against which progress will be measured. An important
use of information for program administration is to monitor compliance. For concreteness, this
section assumes that the specific marketable permit program involves an emission allowance
auction, although the section also describes additional information that could allow the EPA to
develop a distributional rule for granting allowances.
Table 4-1 lists information needs for program design. The aggregate emissions cap can
be determined in several ways; therefore, the specific information requirements for this design
element are uncertain. The other needs specified in the table allow the creation of the current
emissions inventory, enhance the credibility of the expectation that product manufacturers and
users can adapt to the increased cost of emissions, and allow the EPA to set a fee to cover its
administrative costs, if desired. Some of this information is valuable but not absolutely
necessary. The parameters of a reasonable distributional rule for granting allowances include
VOC content standards and product sales in a base year or years; this information is the same as
that which could be used to set the aggregate emissions cap.25
Economic modeling may play a role in program design. For example, if the EPA wished
to project permit prices, then a model like the one described in the Industry Profiles and
Technical Appendixes document, Appendix B would be useful. Information needs would then
24For convenience, the discussion assumes a single permit market. Multiple permit markets (for differentiated
permits) require the corresponding number of aggregate emissions caps.
25Section 1.2 describes the use of this information in a "bottom-up" procedure to determine the aggregate emissions
cap.
4-37
-------
increase beyond those indicated in Table 4-1 and would include baseline economic conditions
and model parameters (see Table 3-6).
TABLE 4-1
INFORMATION NEEDS FOR THE DESIGN OF
MARKETABLE PERMIT PROGRAMS
• Aggregate emissions cap
• Name and address of every manufacturer or distributor and importer included in the
program
• Annual net VOC usage in regulated consumer and commercial products in a recent year
(or years) by company
• Qualitative information on alternative formulations of existing consumer and
commercial products
• Qualitative information on product substitution by users
• EPA's administrative costs for the program
Table 4-2 lists the information needs for program administration. This information would
be necessary to allow the EPA to track permit holdings and to calculate VOC emissions for each
company in the program.
TABLE 4-2
INFORMATION NEEDS FOR THE ADMINISTRATION OF
MARKETABLE PERMIT PROGRAMS
• Number of permits owned at any given moment by each program participant
• VOC content of each product
• Sales of coating by product for each manufacturer or distributor and importer
4-38
-------
4.9 ANCILLARY PROJECTS TO INCREASE TRADING IN A PERMIT MARKET
In general, markets are most effective when neither one actor (a buyer or seller) nor a
colluding group dominate and when the trades which would be beneficial to buyer and seller
actually do occur. The costs of determining what one wants, finding a potential trading partner,
and evaluating offers—transaction costs—reduce the gains to trade and the effectiveness of the
market. The market for permits to emit VOCs from consumer and commercial products would
have few precedents, and the participation of manufacturers, distributors, or importers in this
type of market may be a new experience for them. The transaction costs in this market might be
higher, at least initially, than in markets for more familiar goods.
The EPA could undertake or support projects to reduce transaction costs in a consumer
and commercial products marketable permit program. The general objective of these projects
would be to remove obstacles to trade thus reducing the aggregate expenditure on pollution
abatement.
Such projects could provide assistance in different forms. They could include one or
more of the following:
• guidebooks,
• instructional seminars,
• computer programs, and
• computer bulletin boards.
A guidebook could discuss the strategic issues facing participants and provide information on the
record keeping and reporting requirements. Instructional seminars could cover all aspects of the
marketable permit program and provide the EPA with opportunities to clarify the operation of the
program. The computer programs could assist participants with the evaluation of input
substitution and packaging redesign, the determination of their demand for permits, and the
appraisal of current offers to buy or sell permits.
4-39
-------
These forms of assistance are already being developed for electric utilities who will
participate in the SO2 allowance trading program. For example, the "Clean Air Response Kit,"
developed by the Electric Power Research Institute, in addition to including a strategic
guidebook and seminars, features several computer programs which estimate the costs of several
means of reducing SO2 emissions and allow utility planners to simulate their participation in the
allowance market (Lamarre, 1991).
As the utility example shows, the private sector may sponsor projects that reduce the
transaction costs of a marketable permit program. In fact, private sector brokers already exist
who could facilitate permit trading. It may be that private sector brokers or trade associations
can most efficiently reduce transaction costs. However, if, for example, the trade associations in
the consumer and commercial products industry are less prepared to provide sophisticated
technical assistance or provide assistance only to a small group within the industry, then the EPA
may wish to sponsor projects to increase the efficiency of the permit market.
4.10 EMISSIONS AVERAGING PROGRAMS
The characteristic feature of emissions averaging programs is a limit on emissions that is
expressed in terms of a product-specific rate (for example, pounds/gallon) instead of an absolute
rate (pounds/year). The emissions limit in an averaging program is thus production-based.
Manufacturers are free to change the VOC content and sales of individual products in any way
provided that the rate of aggregate emissions does not exceed the production-based limit
An emissions averaging program for consumer and commercial products was recently
discussed at the recent California Air Resources Board's workshop on alternatives to VOC
content standards (CARB, 1992). The proposal allows internal trades but not external trades.26
26Other emissions averaging programs may allow internal and external trades (see below).
4-40
-------
For example, under this program a company that manufactures two forms of the same product
would need to conform its products to the following rule:
(VOCi x Salesi) + (VOC2 x Sales2>£ VOCp X (Salesi+Salesz)
Thus the sum of actual emissions cannot exceed the emissions that would be allowed if the sales-
weighted average VOC content of both product forms met the standard (VOCp) for the product
category. The manufacturer is free to adjust either sales or VOC content (or both) to comply
with the regulation—assuming that he or she makes multiple forms of the same product, or
multiple regulated products.
Emissions averaging programs are similar to the marketable permit programs previously
discussed in this section. Both types of incentive are quantity-based instruments, and emissions
trading is the activity that is responsible for the cost-effective allocation of responsibility for
emissions reduction. Both programs may allow only intra-plant trades, or they may allow any
inter-plant or inter-company trade as well. Internal trading can achieve a locally (either within
the plant or the company) cost-effective allocation of responsibility for emissions reduction but
not a globally cost-effective allocation (assuming the marginal cost of emissions reduction vary
across companies). Emissions averaging programs that allow internal and external trades can
achieve global cost-effectiveness with respect to the goal of limiting average emissions.
Emissions averaging programs also differ from marketable permit programs in one
fundamental way. The environmental goal in averaging programs is quite different: a limit on
average emissions at either the plant, company, or universe of sources rather than a limit on the
mass of emissions from all regulated sources. Consequently, aggregate emissions may increase
automatically as the sales of consumer and commercial products increase.
The incentive for environmentally-beneficial technological change with emissions
averaging is weaker than with either emission fees or marketable permit programs that limit the
4-41
-------
aggregate mass of emissions because a manufacturer may "shield" high-VOC content products
by increasing the sales of lower-VOC content products. This strategy is uneconomical with
emission fees because total fee payments increase. The cost of increasing permit holdings makes
the strategy uneconomical with a marketable permit program.
Information on the actual performance of emissions averaging is available in a recent
review conducted by EPA's Region IX and the California Air Resources Board. They studied
the Alternative Emission Control Plans that were applicable to coatings used by the aerospace
industry in the South Coast Air Quality Management District and the San Diego Air Pollution
Control District (U.S. EPA Region DC and CARB, 1990).
The study reports that the emissions averaging programs implemented in the Alternative
Emission Control Plans had not achieved emission reductions equivalent to the VOC content
standards for which the Plans are substitutes. In addition, because each plan included multiple
coatings, places of application, and methods of generating excess reductions, and because the
emissions calculation method is complex and requires specialized information, the review of a
Plan is lengthy and can be undertaken only by very experienced engineers. The records kept by
sources were typically incomplete, inadequate, and erroneous. In practice, the reviews actually
undertaken by district staff were insufficiently probing to reveal that the assumed equivalence of
the Plans to the standards was unjustified.
The procedures for approving Alternative Emission Control Plans have subsequently
been made more stringent, and it appears that the Plans now will achieve equivalent emission
reductions. Therefore, effective emissions averaging programs can be designed and
implemented, but the experience with the Alternative Emission Control Plans raises the issue of
the distribution and size of administrative costs.
4-42
-------
The tradeoff between increased administrative costs for the regulatory agency and
decreased abatement costs for sources may be unnecessary. The regulator's administrative costs
may actually be less with an emissions averaging program that allows external trades; an
example is the lead credit trading program operated by the EPA during the reduction in the lead
content of gasoline (U.S. Library of Congress, CRS, 1989). A potential buyer will give greater
value to emission credits of which the verifiability is higher, and the seller, in the most cost-
effective way, will consequently keep records and undertake other actions to ensure verification.
The buyer and the regulator have the same interest in verification of emission credits. Therefore
the seller of credits, in satisfying the information demands of buyers, simultaneously satisfies the
demands of the regulator, reducing the latter's administrative costs.
One possible advantage of emission averaging versus marketable permit strategies
involves industry expansion. New companies and expanding existing companies are not
disadvantaged in averaging programs as they could be in marketable permit programs because
the former offer an unlimited supply of emission permits at no cost. Industry expansion,
however, could lead to increased emissions under emissions averaging programs.
4.11 SUMMARY OF DESIGN OPTIONS
This section primarily described design options for a marketable permit program to
reduce VOC emissions from consumer and commercial products. The distinguishing
characteristics of this policy instrument are inter-company trading of emissions and the cap on
the aggregate, absolute quantity of emissions. Emissions averaging programs also rely on
emissions trading, and can be made most cost-effective when trades are allowed between
companies, but they characteristically place an upper limit on average emissions. The nature of
the environmental goal is the essential difference between marketable permit programs and
emissions averaging programs. For purposes of comparison with emission fees and VOC
content standards per se, Table 4-3 summarizes permit program design options.
4-43
-------
a,
O
S
o
c
. "5
fi
Z
O
O
Z
O
c.
O
«
Z
••
Q
o.
O
o
c
_o
«
.£ *
£ c
E-2
58-
OJ
•O "«
cs c
'E jo o
'S is S.
is i I
11
0) O
I DO
I 1
1 s
w.S
fa
O
2 «
o c
M.2
?S 'm
CO
£
^
X
> x e E
ffl O E .£
« c a
00 O C
a
en
j:
1
H
*s
iL
o
u
en
(A
O
cS
-------
f*J
4
U
.J
N"4
a
<*<
s
X^
c
U
N*<
H
Z
o
U
\^0f
Cfl
Z
o
E
Z
U
••«
W5
U
C
S
<
£
TRADING PROG
fa
O
SUMMARY
IT)
e
.0
"a.
0
TT
o
«^
cu
O
f>
.0
^0*
c.
O
r*
,0
"o,
O
|
Q.
O
Design Element
U.
flj
EPA allows transfi
at shutdown
e
£
o
"3
2%
" »
g «
•8.3
CU §
U a.
Retiring Sources
Sources possess
temporary
permission to emil
Sources possess
property rights
Type of
Entitlement
e
Emission reductioi
credit
U
g
is
15
J.s
e/5 c
'- 1
w o
Type of
Commodity
s-**
"«
3
c
C3
C
2
o
Revenue auction
(annual)
Zero-revenue
auction (annual)
Initial Allocation
Mechanism
Allowed
U
o
Z
Permit Banking
Unlimited
Limited
Permit Duration
*o
^o
ca
.£ c*
•° S
£-2
O r^
r \ C"
U o
««»
'£ -
tt SO «
o. £ §
•8§a
o *^ »^^
ci _-, cs
(^ vj >••
*~* Mr\ ••
•= ^0
<:sy
a. "5 O
U ^ >
a
o
a.
E
8
c;
•n
(U
<
c-
u
g
O
cu^>
s?
U- C
IS
»3 O
(U U
iXs
«?1
Monitoring and
Enforcement
t—
o
o
'i
c —
• M ^
£ C
£-2
0 Q.
U o
5J
«
o.
0 u.
>%2
"" «
o£ d
o
«> e
n Ancillary project
reduce transactio
costs
Market Facilitatioi
"o
e
o
e
• «5 &
JO C
ff O
C •—
r9 0.
U 0
CO
0>
3
1
U
C
o
2
£
2 S
ob.2
£ «
o.is
. »s
•^ B^*
t2 «=
t*^ .^^
>»E
eo -a
ft. rt
Fund labelling
and/or education
programs
Fund R&D for
VOC-reducing
technologies
Use of Revenues
4-45
-------
The optimal design of a marketable permit program depends upon the objective. One
specific marketable permit program is described in Table 4-4. It illustrates a program designed
to obtain air quality improvements that will not be eroded by expansion of the consumer and
commercial product industries, reduce VOC emissions from consumer and commercial products
with the least aggregate expenditure on pollution abatement, and provide the greatest incentives
for technological innovation and diffusion while reducing potential impacts on small businesses.
Basically the example program accomplishes these purposes by:
• creating a market for a fixed quantity of emission allowances,
• providing maximum flexibility to manufacturers and importers to find the least
expensive sources of emission reductions,
• periodically auctioning allowances,
• using revenues to promote research and development, and
• requiring strict monitoring and enforcement.
The permit market will reflect equity considerations as they apply to the initial allocation
of control responsibility, initiate movement toward the least-costly control of emissions, and,
through auctions, publicly reveal the current demand for permits. The market may therefore
perform all of its basic functions (Hahn, 1983: p. 87) and minimize barriers to entry into the
industry (consistent with the environmental goal).
The justification for the mandatory participation of product manufacturers and importers
is the same as that for making these companies legally responsible for paying the emission fee.
Of all regulated entities involved in the manufacture and distribution of consumer and
commercial products, manufacturers have the most knowledge about product formulas and
packaging. Emission permits increase the opportunity cost of the solvents, propellants, other
4-46
-------
TABLE 4-4
EXAMPLE MARKETABLE PERMIT PROGRAM DESIGNED TO MAXIMIZE
INCENTIVES FOR TECHNOLOGICAL CHANGE
Design Element
Recommendation
Program Purpose
Permit Differentiation
Sources Included
Phasing-in Sources
Opt-ins & External Groups
New Sources
Retiring Sources
Type of Entitlement
Type of Commodity
Initial Allocation Mechanism
Permit Banking
Permit Duration
Monitoring and Enforcement
Market Facilitation
Use of Revenues
Achieve VOC reductions at lower cost and
maximize incentives for technological change
Annual permit
All "large" manufacturers and importers
By size (emissions)
Combination of options & EIGs
Purchase permits from existing sources
EPA allows transfer at shutdown
Sources possess temporary permission to emit
Emission allowance (1 ton in 1 year)
Revenue auction (annual)
Allowed
Unlimited
Combination of options
Combination of options
R & D for VOC-reducing technologies
inputs, and formulations that are responsible for the air pollution from consumer and commercial
products.
The phasing-in of small manufacturers combined with providing the option to participate
attempts to lessen small business impacts while eventually including all VOC emissions from the
regulated products in the program. Nonetheless, the costs of phasing-in and the uncertainty in its
4-47
-------
desirability would need to be recognized. The postponement of mandatory participation may
reduce the emissions which the program will control and constrict the permit market. A
quantitative study of potential economic impacts would be helpful in determining the desirability
of special treatment of small companies.
Inflexibility in the timing of the use of consumer and commercial products argues against
seasonally differentiated permits. Geographically differentiated permits provide the EPA with
the ability to fine tune the emissions caps in nonattainment areas. However desirable on
environmental grounds, differentiated permits increase administrative costs. Again, it is
appropriate to recognize the possibility of opportunistic transportation of regulated products
across the boundaries of nonattainment and ozone transport areas—and to make this option
contingent upon the demonstration of satisfactory program performance without intrusive
monitoring.
If the program should provide the greatest incentives for technological change, then
auctions are the preferred mechanism for initially allocating permits because auctions increase
incentives to spread technologies once they have been developed. Auctions may also reduce
transaction costs that could prevent optimal permit trading from occurring, and they may
generate revenue that could be used to promote technological advance even more quickly.
The above example program design would be less desirable if minimizing adverse
economic impacts on manufacturers and importers was deemed more important than providing
the greatest incentives for technological change. For example, freely granting permits would
allow companies to avoid having to pay for their initial allocation of permits and may increase
their certainty regarding future permit availability. Hence, this discussion illustrates the tradeoffs
associated with alternative marketable permit design options.
4-48
-------
4.12 REFERENCES
1. Breger, Marshall J., Richard B. Stewart, E. Donald Elliott, and David Hawkins.
1991. "Providing Economic Incentives in Environmental Regulation." Yale
Journal on Regulation 8:463-495.
2. Colby, Michael E. 1991. "Environmental Management in Development: The
Evolution of Paradigms." Ecological Economics 3:193-213.
3. Elman, Barry, Bruce Braine, and Richard Stuebi. 1990. "Acid Rain Emission
Allowances and Future Capacity Growth in the Electric Utility Industry." Journal
of the Air and Waste Management Association 40:979-986.
4. Hahn, Robert W. 1983. "Designing Markets in Transferable Property Rights: A
Practitioner's Guide." In Buying a Better Environment: Cost-Effective Regulation
Through Permit Trading, Erhard F. Joeres and Martin H. David, eds., pp. 83-97.
Madison, WI: University of Wisconsin Press.
5. Hahn, Robert W. 1984. "Market Power and Transferable Property Rights."
Quarterly Journal of Economics November:753-765.
6. Hahn, Robert W., Gregory J. McRae, and Jana B. Milford. 1988. "Coping with
Complexity in the Design of Environmental Policy." Journal of Environmental
Management 27:109-125.
7. Harrison, David Jr. and Paul R. Portney. 1982. "Who Loses from Reform of
Environmental Regulation." In Reform of Environmental Regulation, pp. 147-179,
Wesley A. Magat, ed. Cambridge: Ballinger.
8. Kneese Allen V., and Charles L. Schultze. 1975. Pollution, Prices, and Public
Policy. Washington, DC: Brookings Institution.
9. Lamarre, Leslie. 1991. "Responding to the Clean Air Challenge." EPR1 Journal
April/May, pp. 21-29.
10. Marin, Alan. 1991. "Firm Incentives to Promote Technological Change in
Pollution Control: Comment." Journal of Environmental Economics and
Management 21:297-300.
11. Milliman, Scott R., and Raymond Prince. 1989. "Firm Incentives to Promote
Technological Change in Pollution Control." Journal of Environmental Economics
and Management 17:247-265.
12. Montgomery, W. David. 1972. "Markets in Licenses and Efficient Pollution
Control Programs." Journal of Economic Theory 5:395-418.
13. Oppenheimer, Joe A., and Clifford Russell. 1981. "A Tempest in a Teapot: The
Analysis and Evaluation of Environmental Groups Trading in Markets for Pollution
Permits." In Buying a Better Environment: Cost-Effective Regulation through
Permit Trading, E. Joeres and M. David, eds., pp. 131-148. Madison, WI:
University of Wisconsin Press.
4-49
-------
14. Tietenberg, Thomas H. 1980. 'Transferable Discharge Permits and the Control of
Stationary Source Air Pollution: A Survey and Synthesis." Land Economics
56:391-416.
15. Tietenberg, Thomas. H. 1985. Emissions Trading: An Exercise in Reforming
Pollution Policy. Washington, DC: Resources for the Future.
16. U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards. 1991. National Air Quality and Emissions Trends Report, 1989.
EPA-450/4-91-003.
17. U.S. Environmental Protection Agency Region DC and California Air Resources
Board. 1990. "Phase in Rule Effectiveness Study of the Aerospace Coating
Industry." May.
18. U.S. Library of Congress, Congressional Research Service. 1989. Using Incentives
for Environmental Protection: An Overview. Report No. 89-360 ENR. By John L.
Moore, Larry Parker, John E. Blodgett, James E. McCarthy, and David E. Gushee.
Washington.
4-50
-------
SECTION 5
COMPARISON OF EMISSION FEE AND EMISSION TRADING PROGRAMS
The fee-based economic incentive programs described in Section 3 and the emission
trading programs described in Section 4 are alternative instruments for reducing VOC
emissions from consumer and commercial products. Because one instrument achieves its
purposes by regulating a price variable, and the other instruments by regulating a quantity
variable, a comparison of fee and trading programs reveals the essential differences of the
two most commonly proposed economic incentives for pollution control. This section
compares the following specific variants of each instrument:
• fees—without rebates,
• fees—with rebates,
• permits—freely granted permits,
• permits—zero revenue permit auction,
• permits—ordinary permit auction, and
• emissions averaging.
It is assumed that the same environmental goal has been set for all programs: a particular
quantity of VOC emissions from the use of consumer and commercial products.1
The purposes of comparison are to determine how well the instruments accomplish
certain policy objectives and to appraise their ability to cope with the complexities inherent
in the task of environmental regulation. The specific bases of comparison are the following:
• program costs and initial information requirements when effects are uncertain,
• monitoring and enforcement,
• distributive flexibility,
• adaptation to economic growth,
• incentives for technological innovation and diffusion, and
• unintended damages.
1The comparison may be undertaken without specifying the goal.
5-1
-------
Those criteria are standard for the comparative analysis of environmental policy instruments
(Bohm and Russell, 1985; Hahn, McRae, and Milford, 1988).
The comparisons made in this section will also include a third type of instrument: a
hypothetical VOC-content standard. The content standard (described below) would be a
command-and-control analog to the economic incentives for reducing emissions. It is
included because the ultimate purpose of this investigation is to inform the search for the
most desirable instrument in the set of potential instruments, which obviously would include
instruments based on command-and-control.
5.1 PROGRAM COSTS AND UNCERTAINTY
This comparison of policy instruments begins with a simple theoretical examination
of the costs and efficiency associated with employing alternative regulatory approaches under
conditions of uncertainty. In a certain world where specific emission reduction technologies
and costs for each source are known to the EPA and in which only total emissions from
consumer and commercial products matter, the question of efficiency is trivial because the
Agency could design a fee-based program, marketable emission permit program, or set of
VOC content standards for which compliance costs are equal. Efficiency really is an issue
when uncertainty in cost or benefit estimates is present.
The interest in uncertainty arises because uncertainty makes it easier to be wrong
when designing an environmental policy instrument. In terms of social efficiency, being
wrong means that any source, in response to regulation, has either not gone far enough or
gone too far in reducing emissions. In the context of a static but uncertain world, and in
which monitoring and enforcement are equally effective and expensive, is the cost of being
wrong less with an economic incentive than with VOC content standards?
5-2
-------
Consider first the efficiency loss of being wrong when the control cost of emissions
reduction is uncertain. Figure 5-1 shows a marginal benefits curve, the best-estimate or
expected marginal cost curve MQ), and, to reflect uncertainty, high and low estimates of
marginal control costs.2 Either an emission charge equal to P*, an emissions quota of Qs, or
the equivalent emission standard is optimal when expected marginal costs equal the actual
costs. When emissions are limited to Qs, area 1 indicates the loss from overcontrol when
control costs are underestimated. For the same limit, area 2 indicates the loss from
undercontrol when control costs are overestimated. When the regulator charges P*, area 3
indicates the loss from overcontrol (i.e., P* exceeds the optimal tax when actual control costs
are less than expected), and area 4 indicates the loss from undercontrol. Losses with the
emission charge are least in this example: area 3 is less than area 1, and area 4 is less than
area 2.
The constancy of the fee rate is very important in establishing the preference for
emission fees. If the regulator were to change the rate to achieve the target level of
emissions, the rate would fall when observed emissions (Q) exceed the emissions target Qs,
and the rate would rise when observed emissions are less than Qs. In the first case, area 3
gradually would be reduced then eliminated, and area 2 would grow, reaching a maximum
when Q = Qs. In the second case, area 4 gradually would be reduced then eliminated, and
area 1 would grow, reaching a maximum when Q = Qs. The welfare losses that are unique to
an emissions fee (i.e., areas 3 and 4) are the short-run losses which occur until the program
achieves the fixed environmental goal. Adjustment of the fee basically transforms an
emissions fee program into a quantity control.
Therefore, when marginal benefits do not decrease especially rapidly as emissions are
reduced and marginal control costs rise steeply, the efficiency loss or social costs of being
2SeeCrandall(1983).
5-3
-------
Dollars per
quantity of
pollution
removed
MC1
. MCO
/ MC2
P*
Pollution removed
Figure 5-1. Welfare Losses When Costs Are Uncertain—No Effects Threshold
Source: Robert W. Crandall, Controlling Industrial Pollution: The Economics and Politics of Clean Air,
Washington, DC: Brookings, 1983, p. 62.
wrong are least with a constant emissions fee when control costs are uncertain. A similar
analysis shows that constant emission fees are preferable when benefits as well as control
costs are uncertain (Crandall, 1983, p. 65). The advantages of an emissions fee are transitory
and diminishing when the regulator adjusts the charge in pursuit of a fixed emissions target.
Given a comparatively steep marginal cost curve, the assumption that marginal
benefits decline gradually is crucial to the analysis underlying the preference for emission
5-4
-------
fees. When marginal benefits instead decline rapidly, emission standards or marketable
permits are preferable.3 Figure 5-2 shows the same marginal cost curves as in the first figure,
but the marginal benefit curve is steeper. The steeper curve decreases areas 1 and 2 but
increases areas 3 and 4. Losses with the emissions fee are now greater than losses with
policies that establish a limit on emissions. Adjustment of the fee rate diminishes the
disadvantage of the emission fee.
Dollars per
quantity of
pollution
removed
P*
Pollution removed
Figure 5-2. Welfare Losses When Costs Are Uncertain—Effects Threshold
Source: Robert W. Crandall, Controlling Industrial Pollution: The Economics and Politics of Clean Air,
Washington, DC: Brookings, 1983, p. 62.
3 Again, uncertainty in the benefits does not affect policy preferences.
5-5
-------
Therefore, in a static but uncertain world, the preference for price-based or quantity-
based instruments depends upon the relationship between the marginal benefit and cost
curves, although the preference weakens somewhat when the regulator adjusts the fee rate.
With rising marginal control costs, the specific factors that will determine the preferred
policy are the scientific and economic facts about the transformation of VOCs into ozone and
effects of exposure on human health. Effect thresholds imply a marginal benefits curve
similar to the one in Figure 5-2 because total benefits increase most rapidly for emission
reductions in the most polluted environments, but the increase is slight in very clean
environments in which pollutant concentrations are below the (unspecified) threshold
(Crandall, 1983, p. 61).
The preceding analysis featured a mild form of uncertainty that may not accurately
reflect the state of the EPA's knowledge concerning the technology of consumer and
commercial products and the costs of reformulation and repackaging to reduce VOC
emissions. It assumes that the EPA has data on a substantial number of formulations of a
particular product. A more accurate description of the actual data may be that the EPA can
observe only one or two forms of each product because they are the only ones that are
available in the market.
In the presence of minimal information on compliance costs, VOC content standards
are most unattractive with respect to efficiency, but emissions averaging may perform better
according to this criterion. Without usable knowledge about the marginal cost of emission
reductions, it is highly unlikely that the EPA could find or come close to the least costly
allocation of emission reductions. Expanding the scope of resource costs to include the
losses from reduced product performance would probably amplify the range of uncertainty,
further strengthening the expectation of inefficiency from VOC content standards.
Abatement costs are likely to be less with emissions averaging than with VOC content
5-6
-------
standards because internal trading and, especially, external trading allows sources flexibility
in determining how to comply with emission limits. However, the potential for basically
inefficient standards remains with emissions averaging.
Emission fees may also be unattractive in these circumstances because of the
prolonged process of trial and error which may be necessary to achieve the predetermined
emission goal. Trial and error is predictably costly because product manufacturers will make
investments which either turn out to be "wrong" after the fee has been revised or are more
flexible (and hence costly) than necessary if the correct fee had been known from the outset.
The argument for a marketable permit program is especially strong when the EPA has
very little information on marginal control costs and achieving an emissions goal at least cost
is the objective. To start the program, the EPA would determine the desired quantity of
emissions and could freely grant the corresponding quantity of emission permits. Provided
the permit market is competitive, monitoring and enforcement are effective, and transaction
costs are negligible, exchanges between program participants will lead to the least costly
allocation of emission reductions. To the extent that transaction costs are large, the
efficiency of a marketable emission permit program will be lessened but will never be
inferior to that of VOC content standards.
By efficiently distributing permits initially, permit auctions may serve to lessen the
potential problem that trades will not occur due to high transaction costs. An auction itself,
however, would involve some transaction or administrative costs. Furthermore, periodic
auctions may increase costs associated with uncertainty regarding the future availability of
permits.
Though generally reducing costs, none of the economic incentive strategies will likely
yield the truly least cost solution that might be obtainable in a world of perfect certainty. A
5-7
-------
practicable least cost strategy will depend on particular circumstances and will have to reflect
tradeoffs among transaction or administrative costs, costs associated with uncertainties, and
control costs.
5.2 MONITORING AND ENFORCEMENT
The purpose of monitoring is to observe actions antagonistic to the basic principle of
the policy instrument—noncompliance. Enforcement involves the selection and imposition
of a penalty for noncompliance. The EPA, regardless of the instrument, will seek penalties
for noncompliance as allowed under the CAA. Therefore, because enforcement actions may
be similar under fees, permits, VOC content standards, and emissions averaging, this
subsection focuses on monitoring.
The EPA's essential task in monitoring compliance with a fee program is to
determine the correctness of each source's gross payment of fees levied on the total VOC
content of its consumer and commercial products used in the United States. For a marketable
permit program, the Agency must determine any insufficiency hi the number of permits held
by each source. For a VOC content standard, the Agency must only determine the
acceptability of the composition of each regulated product.
Monitoring compliance with fee programs is similar to monitoring compliance with
marketable permit programs because each program requires the Agency to observe the VOC
content of regulated products and the quantity of every regulated product manufactured or
imported for domestic use. For example, to judge the adequacy of a domestic manufacturer's
remittance, the Agency could compare actual receipts to required receipts, which would be
calculated (for a single product) as the product of the emission fee ($/VOC), VOC content,
and annual production. To monitor accuracy, the Agency could cross check reported
production, obtain information on solvent and propellant usage, and test the VOC content of
5-8
-------
product samples.4 One important difference is that the Agency also must track permit
holdings under marketable permit programs.
Monitoring compliance with VOC content standards would require product testing,
only because the basis of a standard would be the product formula. In an emissions
averaging program, the basic requirement for sources is compliance with a VOC content
standard that may be met by averaging emissions over multiple products. Therefore, the
EPA would need to obtain information on VOC content and on product sales, and to test
products.
The expected monitoring activities and expenses are the greatest for marketable
permit programs, less for fee programs and emissions averaging (when external trading is
prohibited), and least for VOC content standards. It should be noted, however, that in order
to assess emission reductions progress with VOC content standards, annual sales data would
need to be obtained.
5.3 DISTRIBUTIVE FLEXIBILITY
Economic incentives for pollution control generate transfer payments and may also
create valuable assets, for example, emission permits, which are transferred from the
government to program participants. Economic incentives are intrinsically political because
they create and redistribute wealth, although they are not dissimilar from command-and-
control programs which may also create wealth.5 The pattern of transfer payments reflects
the distributive values that are incorporated into instrument design.
4Solvent and propellant usage would help to establish a lower bound on the total required payment because
some but not all VOCs are contained in those inputs.
^For example, conventional (i.e., untradeable) emission permits are valuable assets. Their value equals the
shadow price of emissions, which reflects the aggregate marginal cost of emission reductions (assuming that
sources are free to choose the least costly means for reducing emissions) and the quantity of allowable
emissions.
5-9
-------
Transfer payments take the form of either emission fees or expenditures on
marketable emission permits. The fee programs in Section 3 require manufacturers or
distributors and importers to pay the government for their customers' use of the assimilative
capacity of the atmosphere, although, in the version with a rebate program, the EPA rebates
all fees to the sources. Marketable emission permit programs also require sources to incur an
opportunity cost—the permit price—for emissions, and they create a market in which buyers
of permits will pay sellers; depending on the program version, the EPA will freely distribute
or auction permits. All fee and permit programs create winners (those obtaining net receipts)
and losers (those making net payments), but it is theoretically possible to design fee and
permit programs with the same pattern of transfer payments. The program element that
creates equivalence is the rebate formula or the mechanism for initially allocating permits.
The comparison between VOC standards and budget-neutral economic incentives is
uncertain. Some companies will prefer standards because they avoid paying fees or buying
permits. Other companies will prefer budget-neutral economic incentives because they will
receive rebates in excess of fee payments or they will sell permits.
The primary criterion that differentiates among policy instruments is the flexibility of
incorporating distributive values into instrument design. Distributive values underlie the
desired composition of the universe of sources and the desired pattern of transfer payments.
Distributive flexibility is greatest with free permits, permits allocated by a zero-revenue
auction, and rebated fees because the regulator selects the universe of sources and is
completely free to choose—and must choose—the formula for initially allocating permits or
calculating rebates. Distributive flexibility is limited and less focused with ordinary auctions,
fee programs without rebates, VOC content standards, and emissions averaging programs
because the regulator only controls one variable, the universe of sources, that determines
distributive effects.
5-10
-------
5.4 ADAPTATION TO ECONOMIC GROWTH
This section studies some of the implications of dropping the assumption of a static
world, and considers each policy instrument in a setting that allows for changes in the
economy. The specific type of change of interest is exogenous to the policy: an increase in
demand for consumer and commercial products, which could be due to population growth or
an increase in per capita income, although the specific cause is not crucial. The factors
responsible for increased demand are among the developments usually associated with
economic growth and considered essential to any realistic description of the American
economy. What changes (if any) in the policy instruments under review are necessary to
adapt them to the increased demand for consumer and commercial products, and are the
changes costly in terms of information, analysis, or trial and error?
Emission fees ensure that the marginal cost of emissions reduction will be constant
and equal (to the fee rate) for all sources irrespective of changes in the economy or
environment. Fee-based programs can neither impose nor induce an invariant limit on VOC
emissions from consumer and commercial products. As the demand for consumer and
commercial products increases, VOC emissions will increase unless the EPA increases the
fee. All of the information and analysis that is relevant to establishing the initial fee is
relevant to increasing it. Therefore, some of the effort made to start a fee-based program will
be useful in the process of adjustment, but it is reasonable to expect that substantial effort
will be necessary to update models of the input and product markets to reduce uncertainty
over the proper value of the new fee. If the new fee is set through a process of trial and error,
then sources will incur adjustment costs in addition to expenditures on emission reductions.
VOC content standards and emissions averaging programs must become more
stringent to maintain the achievement of the environmental goal because they impose a limit
on the quantity of VOCs potentially emitted by a unit of product or product group, but they
5-11
-------
do not limit aggregate emissions. The EPA may incur significant adjustment costs if it must
reexamine content limits in a growing economy.
Emission permits succeed in automatically constraining emissions when the demand
for consumer and commercial products increases and hence sales increase. The information
and analysis supporting the limit on emissions does not need to be revised as the economy
changes. A marketable permit program also spares sources the trial-and-error costs which are
unavoidable in an evolving fee-based program. Of course, they will spend more to achieve
emission reductions. Thus, under emission permit programs, uncertainty regarding emissions
reductions is replaced with uncertainty regarding the costs of achieving those reductions.
5.5 INCENTIVES FOR TECHNOLOGICAL INNOVATION AND DIFFUSION
One desirable type of technological change which environmental policy could foster
is the innovation and diffusion of less expensive methods of pollution abatement.6
Reductions in compliance costs, as Milliman and Prince observe, "can significantly expand
the boundaries of environmental policy design" (1989, p. 247). The specific example which
they give is new acid rain abatement technology which "may greatly spur efforts to regulate"
the precursors of acid rain (Ibid.). Technological change benefits the innovator, adopting
companies, and potentially the people and other organisms exposed to the pollutants.
The incentives of a company for innovation and diffusion come from three sources of
savings: compliance costs, transfer payments, and income. Specifically, technological
change can reduce expenditures on machinery, facility operations, emission fees, and
emission permits as well as provide income from the sale of emission permits and patent
royalties.
^Pollution abatement may include the capture of pollutants, reductions in the toxicity of emitted pollutants, and
pollution prevention.
5-12
-------
This section summarizes the results of Milliman and Prince's (1989) theoretical
analysis of the incentives for cost-reducing technological change in pollution abatement.
They assume a positive correlation between cost-savings and the probability that a company
will undertake innovation or diffusion. Thus they are able to rank the incentive under
different environmental policy instruments for each step and the entire process of
technological change.
•
The incentive for innovation is the greatest with auctioned marketable emission
permits, freely granted permits, and emission fees—all of which provide an equal incentive—
and least with emission standards. In general, economic incentives provide a continual
incentive for regulated companies to reduce emissions further.
The incentive for diffusion varies with the policy instrument. The incentive is
greatest with auctioned permits, least and negative with free permits (see Section 4), and
intermediate with emission fees and emission standards.
Sections 3 and 4 considered variants on fee-based and permit programs which differ
from the pure forms examined by Milliman and Prince. The variant on emission fees
featured rebates, making the program budget-neutral from the perspective of the government.
Similarly, the variant on permits featured a zero-revenue auction of permits. Milliman and
Prince's method is also applicable to the budget-neutral versions of the emission fee and
marketable permit programs for consumer and commercial products.
The introduction of rebates or a zero-revenue auction only changes the distribution of
the transfer payments for the environmental services provided by the atmosphere. The
government collects payments but disperses them to the program participants. Provided that
the rule for determining rebates is equivalent to the rule for initially distributing permits, the
net financial position of the program participants will be the same under the fee program with
5-13
-------
rebates, permit program with a zero-revenue auction, and permit program with freely given
permits.
The gains to an innovator from the innovation step are unchanged. Regardless of the
variant, fee and permit programs provide an equal incentive: the changes in expenditures on
pollution abatement and transfer payments (i.e., reduction in gross fees or increase in permit
sales) sum to the same amount under all policy instruments.
Zero-revenue auctions and rebates change the gains from diffusion. The alteration in
the permit program changes the distribution of the economic surplus generated by changes in
permit prices. When permits are auctioned in the ordinary way, all regulated companies gain
from diffusion because the aggregate demand for permits falls, reducing permit prices. With
free permits or a zero-revenue auction, this gain from diffusion is absent. A rebate has the
effect of making a fee program distributionally equivalent to free permits. Therefore,
assuming identical initial allocations of permits (when relevant to comparison), the fee
program with rebates and marketable permit program with a zero-revenue auction provide
equal incentives for diffusion and are equivalent to the marketable permit program with free
permits. The variations of the pure programs reduce the incentives for diffusion.
The reduction of the incentives for diffusion has a negative consequence for the
overall process of technological change. If a potential innovator plans expenditures on
research and development on the basis of the total gain from both innovation per se and
diffusion, then reducing the gains from diffusion will depress the overall incentive to
undertake research and development. However, because an innovator could lose its
competitive advantage by providing its technology to competitors, the significance of the
additional disincentive for diffusion attributable to some incentive strategies is difficult to
assess. The alteration of economic incentive programs to incorporate distributive values does
5-14
-------
lead to unintended consequences for dynamic incentives, but the effect may be of minor
importance.
Even so, the modified fee and permit programs are superior to standards because the
incentive for innovation is stronger. The incentive for technological change with emissions
averaging depends to some degree on whether interfirm trading of permits is allowed.
Without interfirm trading, this incentive is weaker than with emission fee and marketable
permit programs because the gains from innovation comprise only the value of "shielding"
high-VOC products with low-VOC products—a difficult value to quantify—and the
reduction in costs of pollution abatement. With interfirm trading, the incentive is greater
because an innovator will generate and then may sell permits to manufacturers with high-
VOC products (provided that these sales do not diminish the innovator's competitive
advantage in the product market). Emissions averaging programs offer a stronger incentive
for innovation than VOC content standards because the latter provides no incentive to
achieve reductions below the standards.
In conclusion, theoretical considerations of technological change favor economic
incentives, and the most productive are auctioned permits with an ordinary auction, followed
by fees without rebates. VOC content standards provide the weakest stimulus to
technological change.
5.6 UNINTENDED DAMAGES
A general observation on the nature of ameliorating environmental problems such as
ambient ozone will introduce the concern with unintended damages. Narrowly defined
solutions may be counterproductive when imposed on complex problems. Speaking of
ecosystems, whether natural or urban, Rolling and Goldberg remark that "[i]t is because of
the complexity of the interactions that it is so dangerous to take a fragmented view, to look at
5-15
-------
an isolated piece of the system. By concentrating on one fragment and trying to optimize the
performance of that fragment, we find that the rest of the system responds in unsuspected
ways" (1971: p. 224).
The designs of the economic incentive programs optimize the cost of reducing VOC
emissions from the use of consumer and commercial products, and the narrowness of the
objective may cause problems during implementation. VOC content standards also reflect a
narrow objective. This section compares the potential of each policy instrument to generate
these important types of unintended damages: movement of pollutants from one
environmental medium to another, increased health risks, and, perhaps most unexpected,
increased air pollution.
5.6.1 Intermedia Transfers of Pollutants
Intermedia transfers of pollutants could occur if reformulated consumer and
commercial products are more prone to "off-specification" manufacture or have a shorter
shelf life. The increasing importance of intermedia transfers and pollution prevention
justifies concern. For example, the Science Advisory Board specifically refers to intermedia
transfers in its report on priorities and strategies for environmental protection (U.S. EPA,
1990: p. 22).
The problem with off-specification manufacture and reduced shelf life is that the
product is more likely to become a waste rather than a usable product. Disposal would
increase the generation of hazardous or municipal waste. In attempting to solve an air
pollution problem, regulation may worsen the hazardous waste problem and, if the spoiled
products are deposited in landfills, may increase groundwater contamination. Manufacturers
of consumer and commercial products probably would find and implement solutions to the
5-16
-------
production and spoilage problems to increase profits, but the search for solutions would
require expenditures on research and development.7
Intermedia transfers of pollutants are undesirable in general and may frustrate the
intent of regulating consumer and commercial products. These problems may be more
prevalent and avoidance costs greater with VOC content limits than with economic
incentives because the former by definition reduce the flexibility of companies to offer
consumer and commercial products which satisfy internal and external demands for ease of
manufacture and product stability.
5.6.2 Increased Health Risks
Manufacturers may reduce VOC content by substituting a noxious solvent or
propellant. The substitution of 1,1,1-trichloroethane or methylene chloride for mineral spirits
and other aliphatic hydrocarbon solvents has already occurred in response to measures
implemented by the states of California, New York, Texas, and Arizona and in response to
previous regulation of air emissions from the coatings industry (Rauch Associates, 1990:
p. 120). Methylene chloride is a designated hazardous air pollutant (Clean Air Act as
amended, Section 112 (b) (1)). Such input substitution justifies concern for the safety of
workers and users.8
The Federal regulation of consumer and commercial products may similarly heighten
human health risk. Economic incentives and VOC content standards may be equally prone to
this liability, but the solutions to any such problem may be easier to achieve with the former.
A flexible economic incentive program to control VOCs may be more compatible with and
reduce the costs of future regulations to reduce hazardous air pollutants.
7The provision of information, say, a label stating the expiration date, would reduce the waste from stale but
unused products.
8The change in health risk is the combined effect of the change in exposure of a substance and the intrinsic
potential for harm of the substance.
5-17
-------
5.6.3 Locally Increased Air Pollution in the Near-Term from Nonlinear Atmospheric
Chemistry and Permit Trading
A third potential unintended consequence is an increase in ambient air pollution, in
particular, ozone and sulfate particulates, in some airsheds. This effect theoretically could
occur because of either the nonlinearity of atmospheric chemical transformations or the
geographic pattern of emissions in a marketable permit program.
Models of atmospheric chemistry reveal nonlinearities in the transformation of VOCs
and NOX into ozone. Specifically, depending upon both the ratio of NOX to VOCs and ozone
concentration, it is theoretically possible that a reduction in VOCs may increase ozone
concentration in certain localities, and the tendency for this effect increases with ozone
concentration (Hahn, McRae, and Milford, 1988, p. 116). The gas phase reactions that form
sulfate particulates also involve complex interactions with VOCs and NOX, and the starting
point concentrations of these compounds significantly influence the production of sulfate
from SO2 (Ibid., p, 115). Because of the nonlinearities in the formation of ozone and sulfate
particulates, air quality starting points matter.
Regionally undifferentiated economic incentives and VOC content standards are
equally likely to produce this side-effect because neither instrument is adapted to local
conditions. One solution is to match the desired emission reduction to the locale; either
regionally differentiated emission fees or multiple emission permit markets are appropriate.
Another solution is a multiple-objective marketable permit program for NOX and VOCs that
would achieve simultaneous reductions in both pollutants.
It is also possible that air quality in a particular airshed may decline after
implementation of a national marketable permit program because permit trading could go in
the "wrong direction." This result is less likely to occur under a national marketable permit
program to reduce VOCs from consumer and commercial products because, unlike the acid
5-18
-------
rain trading program, permit traders would not also be the direct sources of emissions. For
example, Southern California manufacturers that buy emission permits and sell their products
nationwide, need not cause emissions to increase in Southern California. Thus, even if
manufacturers in the worst ozone areas bought more permits than they sold, air quality in
these areas would not necessarily decline.
Furthermore, the possibility of a localized increase in VOC emissions lessens as the
environmental goal becomes more stringent. The probability that companies will reduce
emissions from all regulated products increases as the quantity of allowed emissions declines.
More important, a company would increase the VOC content of a product only if the current
formulation were less profitable than a feasible alternative, but the contingency is unlikely.
Companies now offer the most profitable, feasible consumer and commercial products.
Under a marketable permit program, increasing the VOC content would be more costly than
it is now because a company would incur the opportunity cost of using emission permits in
addition to the costs of reformulation per se. Therefore, even if a company were to possess a
sufficient quantity of permits to allow an increase in the VOC content of a product, it would
lack the incentive. Thus, the possibility of environmentally adverse permit trading is very
unlikely.
The more likely contingency is that emissions in any particular airshed will decrease,
but not sufficiently in some ozone nonattainment areas. The most efficient adaptation of the
marketable permit program might be to create separate subnational permit markets and to set
more stringent environmental goals for the areas requiring the greatest reduction in VOC
emissions to comply with the ozone national ambient air quality standard.9
h markets would not be without potential drawbacks: they would require increased administration, markets
might be thin, and product users might circumvent the program by purchasing products from vendors outside
of the boundaries of the user's permit market.
5-19
-------
5.6.4 Increased Air Pollution in the Long-Term from Product Life-Cvcle Effects
Paints and coatings provide good examples of the unintended effects on product life-
cycle characteristics that regulations can cause. A paint or coating with reduced VOC
content may not perform as well as the higher-VOC formulation. For example, a
reformulated paint may form a film that is less durable, and users might reapply the coating
more frequently. The emissions from any one application of the coating will decline, but
emissions over the life of the substrate may increase.
As with intermedia transfers of pollutants, product-life cycle impacts are likely to be
worse with VOC content limits than with economic incentives. Standards are more likely to
constrain manufacturers' choices, thus preventing them from balancing the tradeoffs of
reduced emissions and product performance.
5.7 COMBINATION VOC LIMIT/ECONOMIC INCENTIVE REGULATORY
STRATEGIES
This section has focused on a comparison of fee, emissions trading, and content limit
regulatory strategies to reduce VOCs from consumer and commercial products. It is also
possible, and in some cases may be desirable, to combine VOC content limits with economic
incentive strategies.
An example where a combination approach might be desirable would be as follows.
Suppose that a VOC fee strategy is deemed appropriate, except that the economic burden to
industry of paying a fee based on all VOC content is expected to be too great. A better
strategy under this scenario might be to employ VOC content limits, but allow firms to
exceed these limits if they pay a fee based on the difference between their actual VOC
content and the VOC content limits. Such a strategy might also reduce administrative costs if
only a small number of sources exceed their content limits. The drawbacks of this
5-20
-------
combination VOC limit/fee strategy compared to a comprehensive VOC content fee program
would be somewhat reduced abatement cost savings and reduced incentives for technological
innovation to reduce VOC content below the limits.
Another example where a combination approach might be desirable would be as
follows. Suppose that a marketable permits program is deemed appropriate, except that
administrative costs are expected to be too high because the number of sources is great. The
best strategy under this scenario might be to include only the largest sources in a marketable
permits program and require VOC content limits for other sources. One option would be to
allow smaller sources to opt in to the marketable permits program. The drawbacks of this
combination VOC limit/marketable permits strategy compared to a comprehensive
marketable permits program would be reduced permit trading and hence reduced abatement
cost savings, and potential problems associated with imperfect competition in the permit
market if the number of sources included in the marketable permits program is too small.
Thus, the choice between traditional and economic incentive approaches is not
necessarily an all or nothing choice. Combination approaches may have advantages which
need to be weighed against the shortcomings they create. The best regulatory strategy will
depend on the objectives for the program and the particular characteristics of the industries
being regulated.
5.8 SUMMARY
This section has described the criteria for a comparison of policy instruments
designed to reduce VOC emissions from consumer and commercial products and has
evaluated the alternatives under consideration on each criterion. The last step in this
appraisal of the instruments is to integrate the separate rankings.
5-21
-------
Table 5-1 compares the performance of the economic incentives and hypothetical
VOC content standards on the criteria discussed above. It is assumed that every program
aims to achieve the same environmental goal, a predetermined reduction in VOC emissions.
The table includes the most important variants of both types of economic incentives:
emission fee programs with and without rebates, and permit programs in which the EPA
allocates permits to existing sources through an ordinary auction, zero-revenue auction, or
grant. The ranking of each program on a given criterion indicates relative performance.
Overall, economic incentives appear superior to VOC content standards. Only when
monitoring and other implementation costs for incentive strategies are excessive might
content standards be preferable. If these costs are substantial, combination content
limits/economic incentive strategies may still be preferable to employing only limits.
The optimal selection of a regulatory strategy will depend on the specific
characteristics of the universe of sources under consideration for regulation. For example,
potential abatement cost savings, monitoring and administrative costs, and distributional
implications of employing economic incentive strategies for different industries will vary.
The selection of the most preferable regulatory strategy depends upon the program's
objectives. For example, if stimulating technological advance is most important, a
marketable permit program with an ordinary auction might be the most preferable.
Alternatively, if distributional considerations are more important, a permit program with
freely-granted or zero-revenue auctioned permits, or perhaps a fee program with rebates,
might be the most preferable. Thus, the best regulatory strategy will depend upon the
regulated industry and the program's objectives.
5-22
-------
?•*
1
V)
H
tt
.<
H
1/3
Q
OS
^;
Q
Z
£
C/3
H
Z
U
H
z
O
u
u
o
-
<
u
M
fe
U
=
H
g
><
S
Q
Z
•<
g
fa
t^^
M
H
W
U
z
!•*
u
S
0
§
u
u
b
o
z
o
Cfl
^•^
PS
$
ON
s
o
u
MISSIONS
w
fa
O
UANTITY
a
a
k.ji
X
fa
<
C/3
NX
J
«<
O
ri
V
>
M
s
z
u
S
o
ft.
*•*
Z
K
U
&
H
Z
w
E
^
^^
c
a;
«^
e
a
u
o
^^
^.
1^^
en
^e
•S
2
H
V
fa
Emissions
Ordinary
V)
Free Permil
and Zero-
Revenue
•M
^
^
C
X
•—
^
^
V)
"P
u
CQ
•o
n
r/5
>^J
Averaging
e
.0
««
u
<
Auction
jg
"«
ja
^
U5
•w
58
A
£
as
.S
*2
5
u
(•M
I
T3
O
O
O
O
l_l
2
T3
8
o
6
"§
o
O
^
o
O
-irf
=<}
C
5J .
c u
o"«
«-> 00
U
O
>
*!.ts
"S E to
U u 00
c °".S
8=«2
r i t« 2
y u js
O13
> S
,A> ^,
°<3 .tS
- ^ .
8J 5 W>
e 0-.S
8<82
u S^
0-5
> *
wj
=«
e
u
— C/3
S Ji
R 13
0 S
U
O
£*•
=3
c
u
c S
8"3
° 60
U
O
>
bo
_c
'i«i
o
'c
o
2
T3
_U
1
J
Limited
Limited
"S
^
o
"n
8
O
•o
32
'J
J*
>,
^T^
15
'S
u
tt.
.1
5
..n
*k«
t/3
Q
u
eW >
^ "Z
ll
00 X
u
SIow&
expensive
Automatic
Automatic
1 S
_o _>
U2 '3?
si
s 5
£-8
> M*
> U
_o .>
"33 «5
>,§
S B-
• v> M*
W5 «i
W5 U
£^
o
*p
o
o
u
tu
o
_o
'i «
*; OJ3
a. c
•§ J-S
.
E?
>
CA
on
H
oo
o
ex
.23
^H .M
en 00
U O
.>-o
1^
0 &
£H
_O
<*^
r*
"S
c.
Possible
Possible
Possible
flj
3
°io
c/;
c2
u
.0
(/5
to
o
D.
CO
U
00
C3
E
C3
Q
T3
O
•a
D
15
'c
3
5-23
-------
5.9 REFERENCES
1. Bohm, Peter, and Clifford S. Russell. 1985. "Comparative Analysis of
Alternative Policy Instruments." Handbook of Natural Resource and Energy
Economics. Vol. LAV. Kneese and J. L. Sweeney, eds., Amsterdam: Elsevier
Science Publishers.
2. Crandall, Robert W. 1983. Controlling Industrial Pollution: The Economics
and Politics of Clean Air. Washington, DC: Brookings Institution.
3. Hahn, Robert W., Gregory J. McRae, and JanaB. Milford. 1988. "Coping with
Complexity in the Design of Environmental Policy." Journal of Environmental
Management 27:109-125.
4. Rolling, C. S., and M. A. Goldberg. 1971. "Ecology and Planning." AIP
Journal July:221-230.
5. Milliman, Scott R., and Raymond Prince. 1989. "Firm Incentives to Promote
Technological Change in Pollution Control." Journal of Environmental
Economics and Management 17:247-265.
6. Rauch Associates, Inc. 1990. The Rauch Guide to the U.S. Paint Industry.
Bridgewater, New Jersey: Rauch Associates.
7. U.S. Environmental Protection Agency (U.S. EPA), Science Advisory Board.
1990. Reducing Risk: Setting Priorities and Strategies for Environmental
Protection. SAB-EC-90-021. September.
5-24
------- |