POLICY RESEARCH WORKING PAPERE
T~~~~~~~~~~~~~~~~- "--  -:~- -eY.::
7's -~~~~~~~~~~~"-. 
.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. , '    -  .
The Cost of Air Pollution
;-pcilitionlabatementcastIr
Abatement
b-vienchmaorks tbeniefitcot-
Raymond S. Hfazrtman                                              f     i
David Wheeler                                                  ::    t,
Manisda Singh                                                     g
.~ ~~~ ~~~~~~~~~~~~~~~~~~~~~ ,              .r
Policy, Resurch Depomentcconmsztsqahougar
Envirornnenr, Inftastrucmre,controfArdailtuonDhasn
Demrnber 1994 ~ ~ ~ ~   ~    ~ ~ ~ "rdue-arpolflnar~
.~~~~~~~~~~~~~~~~~~~~~~~~- '-
- <t -.~unnecesnblahigh coL-'-- 
.,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -               'C'  - 0'''
-- PoliqReseardI)eparanent ---r ..i                  'i   -i
Envircm~~~~~~~~~~~~~~~~~~~~ 16tct'¾ndgiuleo  S?~5                 5



POLCY RESEARCH WORKING PAPER 1398
Summary findings
Using data from the U.S. Census Bureau, Hartrnan,      countries. Regulatory options that are judged to have
Wheelr, and Singh have developed comprehensive          high net benefits using these numbers would probably
estimates of pollution abatement costs by industry scctor    look even better if local abatement cost data were
for several major air pollutants. Their results provide  available.
conservative benchmarks for benefit-cost analysis of     The estimates in this paper can provide useful
pollution control strategies in developing countries. They   information for pollution charges. They can also help
also provide striking evidence of inefficiency in U.S.  make targeted regulation morc cost-effective. With
command-and-control regulation.                        scarce resources for monitoring and enforcement, new
The cost estimates reflect the experience of about    regulatory institutions in developing countries will want
100,000 U.S. manufacturing facilitics under actual     to focus initiaUy on industry sectors that are the main
operating conditions. They are based on a complete     sources of locally-dangerous polluants. After those
accounting of costs -including capital, labor, energy,  sectors arc identified, targeted regulation should be
material, and services. So, they should be more useful  informed by sectoral differences in abatemcnt coSL The
for benefit-cost analysis than idealized engineering   estimates suggest, for cxample, that cost-effective control
estimates But rhey also reflect stricr pollution control  of suspended particulatc enissions will focus on wood
regulation and inpur prices which are probably         pulping rather than steclmaking when both are major
somewhat higher, on average, than those in developing  sources of suspended particulames. The reason: avenge
countrics. They should be interpreted as conservative  rarticulate abatement costs are four times higher in
estimates for environmental planning in developing     steelmaking.
This paper- a product of the Environment, Infiastructure, and Agriculture Division, Policy Kesearch Department- is part of
a larger effort in the department to support more ost-effective environmenal regulation in developing counties. The study was
funded by the Bank's Research Support Budget under the research project 'Econometric Analysis of Pollution Abatement Costs"
(RPO 677-8 1). Copies of this paper are available free from the World Bank, 1818 H StreetNW, Washington, DC 20433 Please
contact Elizabeth Schaper, room N1O-037, extension 33457 (30 pages). December 1994.
The Policy ResartS Workig Paper Saies dsemnats the fJd    of iwk in prore to ascuage the _kchage of ideas abou
deelopmnt issuesAn objeaie of th series is to gSethe fiins out quicky, men ifthe presentations are lss than flly polished. The
papers cany the names of the auhws and should be usedands edco y. The findig, tprains, and concdsions are the
auhrs'own andshould io be attributed to the World Bank its Eeuv Board ofDeaos. orazy of its member countrim
Produced by the Policy Research Dissemination Center



THE COST OF AIR POLLUTION ABATEMENT
by
Raymond S. Eartman
David Wheeler
and
Manjula Singh
Abstract
Using data from the U.S. Census Bureau, we have developed
comprehensive abatement cost estimates by industry sector for
several major air pollutants. Our results provide conservative
estimates for benefit/cost analysis of pollution control strategies
in developing countries. They also reveal very high intersectoral
variances in marginal and average abatement costs: maximumnimum
ratios are frequently near ten, aud occasionally near one hundred.
They suggest that conmmand-and-control regulation in the U.S. has
reduced air pollution at unnecessarily high cost.
'The authors are, respectively, Principal, Law and Economics
Consulting Group; Principal Economist, Environment, Infrastructure
and Agriculture Division, Policy Research Dept., World Bank; and
Ph.D. Candidate, Boston University. The econometric research
reported in this paper was undertaken in collaboration with the
Center for Economic Studies, U.S. Bureau of the Census. The
research was supported by World Bank Research Grant #67781. Our
thanks to David Shaman for invaluable assistance with preparation
of final text and tables. Thanks also to Ken Chomitz, Gunnar
Eskeland and Shakeb Afsah for many valuable comments on previous
drafts.



Executive Summary
This project began, and ends, on a practical note. Using a
massive U.S. database, we have developed comprehensive abatement
cost estimates by industry sector for several major air
pollutants. Our results are summarized in Table 1 (p. E-5),
which can be employed for benefit-cost analysis of air pollution
regulation in client countries. For our colleagues in Operations
and elsewhere who need only "the bottom line," we have designed
the Executive Summary and Table 1 as a detachable, five-page
working document.
Our paper provides a full explanation of the methodology and
results. Comprehensive abatement cost estimates were our primary
goal, but our results also provide striking evidence of
inefficiency in U.S. command-and-control (CAC) regulation.' We
therefore hope that they will serve two ends simultaneously: A
caution against overreliance on CAC approaches, and a practical
aid to adoption of more efficient market-based regulatory
instruments.
The econometric research reported in this paper was
undertaken in.collaboration with the Center for Economic Studies,
U.S. Bureau of the Census. At the Center, we had access to the
U.S. Department of Commerce's annual 20,000-plant random survey
of pollution abatement costs and expenditures (PACE).  All told,
the estimates reported in this paper reflect the experience of
approximately 100,000 U.S. manufacturing facilities.  In depth
and coverage, they are by a considerable margin the most complete
estimates available
The paper offers U.S. estimates as a reference guide for
benefit-cost analysis in developing countries.  Their credibility
depends on the answers to three basic questions:
(1)  If the primary interest of our work lies in assistance to
* new regulatory institutions in developing coutries, why use
a U.S. database at all?
Until recently, most developing countries have had little
formal regulation of air pollution.  Data on pollution abatement
costs is only generated by abatement, so we cannot reasonably
expect to assemble a large, comprehensive and reliable database
1 This approach to pollution control relies on enforcement of
fixed effluent standards or mandated installation of abatement
technology, with little or no regard for intersectoral differences
in abatement costs.
E-1



on abatement costs by pollutant from developing countries.
Second, the U.S. has the largest and most diverse manufacturing
sector of any OECD economy. For a variety of historical,
geographical and political reasons, its manufacturing sector is
characterized by enormous variation in equipment vintages,
processes, operational efficiency, and degrees of pollution
control. U.S.-based estimates are therefore broadly
representative at the sectoral level. Third, the PACE survey
seems to be the largest and most complete of its kind. Fourth,
the U.S. estimates are built up from a complete accounting of
costs-, including capital, labor, energy, materials and services.
The are not, therefore, idealized engineering estimates but
numbers for thousands of plants under actual operating
conditions. 
(2)  Can we reasonably apply U.B.-based estimates to developing
countries without modification?
Our estimates are based on high mandated levels of pollution
control in the U.S. -They also reflect U.S. input costs. While
some inputs to abatement are traded at roughly constant prices in
international markets, the non-traded inputs will generally be
more costly in higher-income economies.2  U.S. abatement costs
should therefore be higher than those in developing countries
unless protection or scarcity of-engineering skills have very
strong countervailing effects. Taking the regulatory and cost
factors into account, we think that the U.S.-based numbers
provide conservative upper-bound estimates of pollution control
costs in developing countries. Regulatory options which are
estimated to have high net benefits using these numbers would
probably look even better if local abatement cost data were
available.
(3)  Why, if direct survey evidence was available, did we use
econometric analysis?
Air pollution abatement is an activity of the firm,
characterized by multiple inputs and multiple outputs. The
latter are abatement volumes for all pollutants controlled by the
firm. Separate abatement activities have joint and common costs,
like other productive activities. Although firms have their
accounting conventions, the truth is that imputation of these
costs to separate activities is an exercise in inference from
observed experience.  In such a case, there is no substitute for
econometric cost function estimation from large samples.
Table 1 reports average abatement costs in SUS (1993) per
ton for 37 sectors defined by the International Standard
Industrial Classification (ISIC). All manufacturing activities
are subsumed in these 37 estimates.  Seven air pollutant
2 For a useful discussion and statistical analysis of this
relationship, see Dollar (1992).



categories are included: Suspended particulate matter; sulfur
oxides; nitrogen oxides and carbon monoxide; hydrocarbons; lead;
hazardous (toxic) emissions; and other emissions. The abatement
costs in Table 1 can be used in two ways:
1) Efficient Command-and-Control Regulation: Application
of direct CAC regulation will be much more efficient if it is
informed by information on relative abatement costs. For reasons
explained in the paper, the estimates in Table 1 should be read
as the costs associated with attainment of a relatively uniform
(and strict) concentracion standard for air emissions across
sectors.  Intersectoral variation in plant-level costs reflects
significant differences in the average scale of abatement, number
of emissions sources, pre-treatment concentration levels in the
waste stream, and myriad technical factors.  To attain near-
-niform emissions standards, plants in different sectors have to
incu.s very different average costs of abatement.
With scarce resources for monitoring and enforcement, many
new regulatory institutions will want to focus on industry
sectors which are the largest emitters of locally-dangerous
pollutants.  Once the relevant sectors are identified, targeting
should be informed by the relative cost of abatement. Consider,
for example, the case of particulate emissions from-pulping and
steelmaking facilities. If both are heavy local polluters, then
the cost estimates in Table 1 would imply focusing on pulping
facilities because their abatement costs are only 25% of those in
steelmaking.
2) Cost-Effective Regulatory Strategies: If complete
sectoral data on emissions are available, then a more
sophisticated approach is warranted. We illustrate with a simple
example. Suppose the area in question has five sectors which are
significant emitters of suspended particulates. We order the
sectoral emissions and cost data by ascending average abatement
cost, as follows:
Sectoral
Average
Particulate Cumulative   Abatement
Emissions  Emissions   cost
isic   Sector                (tons)     (%)    ($US/ton)
3690 Non-Metal Products    400       40        20
3411 Pulp, Paper           100       50        43
3230 Agricultural Chemicals   250    75       127
3710 Iron, Steel           200       95       182
3520 Other Chemicals        so      100       212
Total   1000
These numbers provide a guide for economically sensible
strategy, whether it is command-and-control or market-based
E-3



in nature. When sorted by ascending unit cost, the numbers
trace out a rough, aggregate social marginal cost curve:
cumulative abatement can be approximately related to the-
unit cost of abatement. Suppose, for example, that we want
to set quantitative targets for emissions reduction.  The
cheapest abatement will be in Non-Metal products
(principally cement), where 400 tons of particulate can be
abated for about $20/ton. Next cheapest-is Pulp and Paper,
where 100 more tons can be eliminated at $43/ton. Abatement
from these two sectors will eliminate 50% of total
emissions.  We can therefore state, with rough accuracy,
that the marginal social cost at 50% abatement of
particulates is $43/ton.  At 75%'atktement, this rises
sharply to $127/ton;- at 95% abatement to $182/ton. Thus, a
move from 40% to 95% abatement entails a ninefold increase
in the marginal cost of abatement. Such large step
increases should be enough to make policy makers think hard
about the optimal degree of abatement in particular
circumstances.
We can also reverse the logic of the example and use it
as a guide for setting emissions charges.. Our-numbers tell
us that an emissions charge of $30/ton should give us
approximately 40% abatement, since it will be generally
cost-effective for polluters in Non-Metal Products to reduce
emissions substantially (at $20ton) rather than pay the
charge. Doubling the charge (to $60) should buy another 10%
abatement. To effect 95% abatement the charge would have to
be tripled again, to something over $182/ton.
We hope that these simple examples will suffice to
illustrate the potential utility of our cost estimates.
Complete information is included below, in Table 1.



TABLE 1
A VERA GE AA TEMENlT COT BY SECTOR, 1979-1985 1$ 1993/toni
______________________ Pa,tlclat.ea Sulfu Ox0M  N02, C02 Hydrnarbhau   Lad Hay Emliafns Other
3reo Food  -                      86      521    229        162 46612         55   55
3130 BevWaas   -                 156      271  11918    11918    76           76   76
a140 Tobacco                     268       167    167    12030   128         128  167
32wTeztf1es                      396       396   1379      1379  1191       1191 1191
321i Spinning                    272       535   1431       188   781        188  188
3aw Apparel                      445        61     61        61    61         61  61
La Leather                       132      377   8430        633   132        324 427
3240 Footwear                    540      207.   993       1558 22141       1197  901
33 1 Wood                         47        38     38        38    38         38   38
3e Funiture                       43        25     25        25    241    13335   25
a to. Pap.r Products              87  _  364    472         472    871        87   87
.   341, Pulp, Papeu              43       155     20        20    35         35   35
iaao Printng    <4241                      117    309      -307   117       5604  117
3s11 Industrial Chemicals         46        75    304       213  1300        311  206
as,2 Agncultural Chemicals       127       519    889       341   532       1069  320
_________________=____.       . .    82  562   207      123   435          63    264
.7io Chemical Products           212      681      48       157    29         29  204
__ _        _Drugs_-      _ |   269        1045    451        173   300        300  300
im3 Refineries                   328       165     59       121  2750       2750 2750
. 4o PetoreUM, COal               59      1942     77        77    42       - 42  42
asRubb_r               -      219     1107    343        343  1010-   1010 1010
sw Plast_cs_-- 219                       2415    235        235  1132       1132 1132
36re    ______ _                 185       106   3792      3792 4709        4709 4709
-_a_ Glass .                     186       550    339       339   186        186  186
3650 Non-Metal Products, n..c.    20      213   1647       1658    11         11  11
3710 ln, Stee                    182       528    115      1203   779        253   18
me Non-Ferrous MeOas             340       152     49       622  1284       1153  177
iao Metal Products               343      1563    461       399   161        161  427
meo Other Machinery              254       855    515       515   138        138  138
35s Offce, COMpvEng A hrey       245       245    864       937   248      17458 245
38o Other Electrcal Macbinery    373      483   1559        215   365        166  166
fz Radbi TV                      394      1854    904      1096  738        1491 1138
3840 Transport Equipment         635      1266    468      1008  468        2331  468
3W4 Shipbualding                 125      832   2229       2229    84         84   84
3 wMotor Vehides                 350    .1523   1155       2441 21483        159  159
iasa Professional Goods         1208     3046    872       1376   995       1634 995
3900 Other Industres              381       26    110       110    26         26   26
The cost estimates in the Appenrix Table 4.2 are in $1979. Those have been inated to
S1993 for this executie surmmary, using an annual compound rate of inflation of 4.6%. 7he ltter
was computed using fixed-weight price indices for pollto abatement and control for 1979 ane'
1989 and was obtained from "Pollution Abatement and Control Expenditures, 1972-90", Survey
of Curwnt Business, June 7992.
E-5.



1. Introduction
In this paper, we develop estimates of air pollution
abatement costs for U.S. manufacturing sectors at the four-digit
level of International Standard Industrial Classification (ISIC).
We utilize a very large plant-level database for the period 1979-
1985, a high point in industry response to the abatement
requirements of the U.S. Clean Air Act. Our estimates measure
both marginal and average total abatement costs for seven
categories of air pollutants.3  As far as we know, they provide
the only consistent summary of abatement costs across all
manufacturing sectors.  As such, they offer new information to
policy makers who are concerned with evaluating the benefits and.
costs of pollution control.  The cost estimates also provide
evidence on the relative economic efficiency of command-and-
control-regulation, as practiced in the U.S. The greater the
variation in sectoral abatement costs for a given pollutant, the
greater the inefficiency of the abatement efforts for society at
large.
In developing our estimates, we have used a model of joint
pollutant abatement cost which is discussed in Section 2.
Section 3 describes the data, which have been-created by merging
plant-level information from two sources: The U.S. Department of
Commerce's annual 20,000-plant random survey of pollution
3 These are particulates, sulfur oxides, nitrogen oxides and
carbon monoxide, hydrocarbons, lead, hazardous air pollutants and
a residual "Other" category.



abatement costs and expenditures (PACE), and the U.S. Census
Bureau's Annual Survey of Manufactures.4  The PACE data for 1979-
85 include reported pollutant abatement and cost, while the
Census provides detailed sectoral identifiers.  tn Section 4, we
present our econometric results and the implied estimates of
pollution abatenent costs.  Section 5 summarizes the paper.
2. The Abatement Cost Function
Until recently, the scarcity of appropriate plant-level data
has prevented detailed empirical studies of average and marginal
abatement costs by pollutant.5  Policy analyses have frequently
developed abatement cost estimates from engineering models.
However, failure to rely on behavioral data has led to
considerable estimation-errors in some cases.6 This paper
attempts to advance the state of the art by analyzing plant-level
data on end-of-pipe abatement and related costs.
4 While random sampling reduces the scope for assembling a
panel of firms from PACE, it does assure very broad representation
over time. The estimates reported in this paper incorporate some
information from approximately 100,000 facilities.
For a recent exception see Mundle, et. al. (1994).
6 A-useful illustration is provided by the trading price for
So2 emissions permits under the U.S. Clean Air Act (Hamilton,
1994).  Using engineering models, the U.S. Government forecast a
price around $600/ton before- the trading system was instituted. In
fact, permits are currently trading around $150/ton.
2



2.1 Abatement Technology and Economics
End-of-pipe abatement volume is the difference between
influent and effluent, where influent is the waste water or gas
from production before treatment and effluent is the residual
emitted after treatment.  The presence of some pollutant in the
waste stream registers as a concentration, typically measured in
&g/l3 for air eimissions.  From an engineering perspective,
'abatement' means installing and operating processes which reduce
influent concentrations from one or more sources to target
effluent concentrations. Box 1 (p. 5) includes illustrative-
descriptions for three pollution-intensive industry sectors,
while Table 2.1 (p. 6) provides complementary average-abatement
volume and cost estimates from the Appendix tables.
The engineering literature identifies four major factors
which determine the cost of abatement: Pollutant type, diversity
of emissions sources, scale of abatement, and pollutant
concentration in the waste stream.7 When air emissions from
industrial processes are concentrated at a few points, most
abatement can be accomplished using three basic systems:
Dry systems use gravity, centrifugal force or fabric
filters to trap pollutants. Gravity/centrifugal
systems are low-cost, but appropriate only for removal
of relatively heavy particulates. Filter systems trap
fine particulates;-costs vary widely according to waste
stream characteristics and desired removal rates.
7 The discussion in this section (including Box 1) relies
heavily on Sell (1992).
3



.wet scrubbers use streams of water or other liquids to
increase collection efficiency. Wet collectors can
cool and wash waste streams and remove-gases.  High-
efficiency systems have high operating costs. Scrubbers
come in a great variety of sizes, configurations and
efficiency levels, depending on the problem. Their
slurry can be a serious source of water-pollution if it
is discharged because hazardous pollutants may be
highly concentrated.  Thus, additional treatment may be
needed to complement the cleaning of air pollutants.
Electrostatic precipitators use the principle of
electrostatic attraction to trap pollutants by
channeling the waste stream between two electrodes - a
high-voltage discharge electrode and a. grounded
collecting electrode. Plate precipitators allow for
the dry collection of very fine particles and are
highly efficient. Pipe precipitators work better for
liquid aerosols or fumes. In general precipitator
systems have low maintenance costs, although dry
systems ar.- subject to explosion dangers if they are
operated beyond rated capacity.
2.1.1 Pollutant-specific factors
Simpler, less energy-intensive collection systems are less
efficient and generally suitable only for large, dense
particulates.  Fine particulates and other air pollutants require
more sophisticated, costly collection devices.  This difference
can be seen, for example, in Table 2.1, where average abatement
costs for sulphur oxide removal are.at least three times higher
than those for-particulates.B
8 In some sectors, removal of fine particulates from diverse
sources may increase abatement costs substantially. As Table 1
indicates.,.sulphur oxide abatement costs by sector are generally,
but not always, higher than particulate abatement costs.
4



BOz  IP iotwCza  PRocSuSuz  aIND uuszouvu CwraaCL
IN TRSs POLDTZIaS-3TBrSZVZ 8ECTMS
In this box we describe the basic processes and smissions control problems in
three sectorss Iron and Steel, Pulp and Paper and Cement. All thre sectors are
extremely energy-intensive, and large installations frequently maintain their own
power-generation facilities. When the latter burn fossil fuels, they have comon
control problem. which focus an particulate and sulphur oxide removal. Sectoral
abatement costs therefore reflect a mix of common costs from power plant emissions
control and costs which reflect differences in mctoral production procsses.
nRON aJD  ST'-
Pig Iron Pr ducti     . 
The production of pig iron involves coking and blast furnace operation. During
coking, there are many ponsible emissions sources.  s02 and particulates are
emitted into the atmosphere during charging (filling the oven), from door and lid
leaks, and from quenching the coke with wastewater. Coke oven gas is often burned
for fuel elsewhere in the plant and can also produce these emissions.
Diverse emissions sources require divrse  solutionas   Oven lid and door
maintenance; putting baffles in the quenching towers; using clean, not waste,
water for the quenching process; stack gas cleaning of the 502; using negative
oven pressures during the charging; and hooding many of the steps to capture
releases.  In general, coke oven emissions are very difficult to control.
Blast furnace operation produces primarily S02 and Co.  Sulphur in the coke
oxidizes to form S02, which eocapes from the top of tho furnace.
8t selmkisg.              -.. 
Air pollution problems in this process come mostly from fume generation by the
furnaces and from pouring operations.  Open hearth and basic .oxygen units u-
electrostatic precipitators or venturi scrubbers -to remove particulates  newer
electric furnaces use dry collection in bag houses (fabric filtering
installations).
PUL    D PAP--
Specific environmental effects depend on the process used. Xraft (or sulfate)
pulping processes are dominant in the industry and produce most of the pollution.
Kraft pulping uses chemicals, heat and pressure to dissolve the wood material.
Sulphur-containing gases are released during this process;  after concentration
of waste 'liquor' into solid residuals, burning of the latter for organic residue
removal also rolease sulphurous waste gases. Most plants use wet scrubbers to
remove sulphur gases and particulates from the-waste stream.
The major environmental problm for this sector is handling the waste kiln dust,
particularly from plants employing wet processe. Dust removal is generally. done
-with electrostatic precipitators.
.               ,           ,--    .                                .~~~



Table 2.1
Average Abatement Volume and Cost
for PACE Facilities
Average Abatement Valume
(Tons)
Total
Suspended     Sulphur
Sector            ISIC   Particulates   Oxides
Cement            3690     34,773          613
Pulp and Paper    3411     28,029        1,672
Iron and Steel    3710     12,336        1,108
Average Cost-of.Abatement-
($US 1993/Ton)
Total
Suspended     Sulphur
Sector            ISIC   Particulates   Oxides
Cement            3690    . .20          213
Pulp and Paper    3411       43          155
Iron and Steel    3710      182          528
Source: Tables 1, 3.2 (Abatement volumes in Table 2.1 are means
from non-missing cells in Appendix Table 3.2)
2.1.2 Divarsity of emissions sources
A major problem for control costs arises when significant
emissions sources are dispersed and highly varied.  For example,
Box 1 describes the diversity of iron and steel emissions as
compared with those of pulp, paper and cement. The consequences
are suggested in Table 2.1: Average abatement costs for iron and
. , .6



steel are much higher than those for the other sectors.
2.1.3 Scale of abatement
Scale economies may apply to some abatement processes,
yielding declines in marginal and average cost as treatment
volume increases. The impact of this factor seems clearly
apparent for Total.Suspended Particulates in Table 2.1: Average
abatement cost drops sharply as abatement volume increases.
2.2 Joint Production of Abatement
Abatement is frequently a joint process, since both
scrubbing and electrostatic precipitation systems can be used to
reduce influent concentrations for several pollutants at the same
time. For example, as noted in Box 1, stack.scrubbers can be
used to remove both TSP and SO2 from air emissions. Even where
separate equipment is called for, there may be common use of
skilled and unskilled labor, materials.and energy.
Generally then, abatement can be characterized as a
multiple-output process whose costs are affected by diminishing
returns, pollutant-specific technology requirements, scale
economies, and diseconomies associated with diversity of
emissions sources.
2.3 Model Specification
Our regression model is specified in equation (1) below.
7



We assume that the abatement cost function is separable from the
firm's production cost function, reflecting purely end-of-pipe
activity.9 In light of the preceding discussion, we estimate
separate regressions by sector, using sectoral identification as
our control for the influence of many unobservable factors:
Influent concentration, differential reduction of influent,
abatement scale, source diversity,-and more detailed engineering
considerations. The quadratic specification of the cost function
allows for testing possible pollutant-specific scale economies.
(1) Cij  Po=    + EiPjkAijk +    3PjkA3jk + ij
where  C4. = Total cost of abatement for end-of-pipe air
pollution control by plant.i in sector j
A;jk = Quantity of air pollutant k abated by plant i
sin = A random disturbance term
There are various techniques for allocating the fixed cost
component (p0) in.(1) across pollutants to yield average cost
estimates. We opt for the simplest in this paper, averaging the
joint and common cost across the sum (in tons) of all pollutants
9 The PACE survey attempts to maintain this distinction, and
we think that our assumption is a reasonable approximation.
However, it will not hold true in all cases. For example, the
description of iron and steel. production in Box 1 notes the
importance of costly adjustments such as.negative oven pressure for
the reduction of emissions. In such cases, there is clearly some
interaction between processing and abatement costs.
8B 



abated.10  Sectoral marginal and average costs are therefore
computed as follows:
(2a)   MCjk = Pik + 2PjjkAijk
(2b) AFCjk = POj/ZkAijk (Aijk= plant-level mean abatement of k)
(2c)   ACik   MCIk + AFCjk = Pjk + 2PjjkAijk + POj/EOkAijk
We estimate Equation (1) for two sets of pollutant
categories, depending on data availability.  During the period
1979-1982, PACE aggregated air pollutants into four categories:
1) particulates; 2) sulfur oxides; 3) nitrogen oxides, carbon
monoxide and hydrocarbons; and 4) other pollutants, including
lead and toxic emissions.  During 1983-1985, pollutants were
surveyed in seven categories: 1) particulates; 2) sulfur oxides;
3) nitrogen oxides and carbon monoxide; 4) hydrocarbons; 5) lead;
6) hazardous emissions;  7) other.
We also test whether we can pool the data over 1979-1985.
This pooling involves aggregating the seven categories to the
four categories for the 1983-1985 period. We test whether we can
pool the pollutants in both linear and quadratic forms of
equation (1). For sectors in which we cannot reject pooling by
pollutant category over 1983-1985, we gain efficiency by
estimating the equation for the four categories over the whole
10 An alternative method would be to allocate joint and common
costs among pollutants abated in proportion to their variable
costs. This approach is called "axiomatic cost analysis." See for
example Mirman, Samet and Tauman (1983).
9



period 1979-1985.
Because our focus is on internationally-comparable
estimates, we have recoded all U.S. plants to ISIC (International
Standard Industrial Classification) categories using a
concordance made available by the U.S. Census Bureau. Our data
set provides observations on Ci and Aijk for a large sample of
plants within each ISIC. We express all costs in U.S. 1979
dollars; volumes are in tons.13 We estimate the regression
equations using OLS, because command-and-control regulation has
forced U.S. plants toward similar effluent concentrations .it.a
little regard for relative abatement costs. Several WLS
corrections have been tested and rejected.
3.  Econometric Issues -- Data and Estimation
Since the 1970's, the United States Department of Commerce
has implemented an annual survey of pollution abatement costs and
expenditures (PACE).3- The survey summarizes annual capital and
operating costs of abatement for a random sample of approximately
1 Note that the regressions reported in Tables 4.1 and 4..2
were run on data adjusted to constant $US 1979. These results are
inflated to $US 1993 for presentation in the Executive Summary
(Table 1).
12 The annual survey continues, but its scope and relevance
were seriously compromised in 1986 by the elimination of questions
on abatement volumes. We are therefore unable to exploit more
recent PACE data for the work reported in this paper.
10



20,000 establishments identified at the four-digit SIC level.13
The PACE data also record annual abatement volumes for many air,
water and solid waste pollutants. In this paper, we focus-on the
air pollutants."
We have used common identifier numbers to merge the PACE
data with the Longitudinal Research Database (LRD) of the US
Bureau of the Census.15 The LRD pools annual data at the
establishment level from the U.S. Census of Manufactures and the
Annual Survey of Manufactures. It includes detailed information
on location; ownership; inputs (labor, energy, materials, plant
and equipment); and production of goods and services at a highly
disaggregated level.  For this study, we have used the LRD only
for sectoral identification of the PACE facilities.
We have translated the U.S. coding of our sample
establishments (4-digit SIC) into 37 four-digit ISIC categories.
The latter are listed in Appendix Table 3.1.  For each sector and
pollutant category, Appendix Table 3.2 displays mean tons abated
during the periods 1979-1982 and 1983-1985. These estimates
13 Specifically, the annual operating expenses include the
cost of capital services in the form of depreciation; labor costs;
materials, supplies, fuel and electricity; and contracted services,
equipment, leasing and other. The survey also quantifies annual
capital expenditures for abatement equipment.
14 Regression-based estimates for the major regulated water
pollutants are also available from the authors on request.
See McGuckin and Pascoe (i988).
11



reflect abatement by the representative establishment in each
ISIC. Volumes for the chemical and resource processing
industries (ISIC 3411 - 3690) are particularly noteworthy for
most air pollutants. The more disaggregated information for
1983-1985 makes it clear that lead and some hazardous air
pollutants have small average volumes at the plant level.
4. Results
Because the complete results for all 37 ISIC industries are
quite lengthy, we do n;ut discuss them in detail here.  Full
results are available from the authors on request.  In this
section, we present estimates of marginal and average total
abatement costs for cases where statistically significant
regression coefficients have been obtained.  These cost estimates
are provided in Appendix Tables 4.1 and 4.2.  Separate cost
estimates are presented for 1979-1982, 1983-1985 and (where
pooling cannot be rejected) 1979-1985.  Cost estimates are always
presented for all seven pollutants (denoted P1 - P7) . In the
four-pollutant cases, cost estimates for the aggregate categories
are assumed constant across the constituent pollutants.16  Tables
4.1 and 4.2 also identify whether the linear or non-linear
1 6
16 For example, the estimated costs for the third category of
pollutant will apply. to both P3 and P4. while the estimated costs
for the fourth category will apply to P5, P6 and P7.
12



version of the cost function is reported.17
Several conclusions can be drawn from Table 4.1.  First,
many sectors have nonlinear abatement cost functions.  Some
marginal costs rise with abatement (due to the technological
difficulty of increasing abatement rates), while others fall
(due, presumably, to the scale effect once some threshold level
of abatement is achieved).
Second, we accept the pooling hypothesis (and associated
pooled cost estimates) for approximately 30% of the industries
(3130, 3210, 3310, 3410, 3411, 3540, 3610, 3620, 3820, 3841 and
3900 -- see Table 3.1).18
Third, marginal abatement costs vary considerably by
pollutant and sector. Notice, for example, that marginal costs
for ISIC 3610 range from $69.04 for particulates to $2484.90 for
lead, hazardous air pollutants and other pollutants. These costs
are substantially less for ISICs 3410 and 3411.
Fourth, the separate estimates for 1979-1982 and 1983-1985
reveal a very broad pattern of variation. For some sectors and
17 The choice of linear or non-linear version is determined by
the relative statistical performance of each.  In a few cases, the
version that was selected on overall statistical merit (that is,
across all pollutants simultaneously) did not produce a usable cost
estimate for a given pollutant.   In such cases, we report the
pollutant-specific cost estimate from the relevant "next-best"m
regression.
18 Where pooling cannot be rejected, these estimates are the
most efficient.
13



some pollutants, marginal abatement costs decline over time
(e.g., particulates in 3211 and 3520 and sulphur oxides in 3512).
For other sectors and pollutants, marginal costs rise with time,
perhaps reflecting scale effects or (more likely) changes in
subsectoral composition of plants surveyed (e.g., 3511: nitrogen
oxides plus carbon monoxide; hydrocarbons).  In both cases, the
changes are probably due to some combination of technical change,
scale effects and alteration in subsectoral composition.
Many estimates are not statistically distinct 19 In some
cases, estimates seem affected by small sample sizes or joint
abatement costs which are far larger than abatement quantities
for particular pollutants.  For example, the marginal cost
estimates for lead and hazardous air pollutants are quite large
for some sectors (e.g., 3110, 3320, 3843, and 3900).
Fifth, we should note that separate regressions for
individual years (1979 through 1985) did not yield generally
robust results and are not reported.
Finally, we report the results of a fixed effects regression
of marginal abatement costs for all establishments whose sole
pollutant was particulates.  Pooling of these establishments
increases degrees of freedom and estimation efficiency
considerably. Our fixed effects model allows for slope and
19 We do not test statistically whether the cost estimates
differ across the two periods.
14



intercept dummies for establishments in each ISIC. For this
simpler abatement function, our results suggest that the marginal
cost of abatement is fairly constant at approximately $5.54/ton
for many sectors. Some sectors do reveal distinct abatement
problems with much higher marginal costs (e.g., ISIC 3140 =
$61.17; ISIC 3530 = $98.75; and ISIC 3843 = $81.24). The
estimates of average total abatement cost in Table 4.2 extend
these interpretations, but remain consistent with them.
5.  Sunary and Conclusions
in this paper, we have estimated the costs of abating major
air pollutants. Our results reveal very high variances in
abatement costs for individual pollutants by sector: maximum/
minimum ratios are frequently near ten, and occasionally near one
hundred. They also show large differences in the means and
variances of costs across pollutants..
For environmental policymakers, these results suggest an
important lesson:  Command-and-control (CAC) regulation in the
U.S. seems to have reduced pollution at a very high cost.
Optimal regulation would attain the desired reduction in
pollution while equalizing the-marginal cost of abatement across
sectors. Where they are feasible, market-based instruments such
as emissions charges and tradable permits are optimal in this
sense. In principle, regulators who were properly informed about
15



abatement costs could approach this optimum using CAC-type
methods. In practice, we can see that nothing like this has
occurred. For the same pollutant, intersectoral abatement costs
can differ by more than a factor of 100.
While command-and-control regulation has certainly not been
optimal for the U.S., it has undeniably been useful for this
exercise. By forcing roughly uniform emissions standards on
sectors with very different pollution control problems, it has
permitted us to estimate intersectoral differences in costs.
These are so large that environmental policy-makers would be
unwise to ignore them. Table 1 provides our current best
estimates of average abatement costs by sector and pollutant.20
When combined with information on the main sectoral sources of
air pollution in a particular region, Table 1 can be used as a
guide to setting priorities for pollution control. Since
regulatory resources are scarce, it undoubtedly makes sense to
focus initial efforts on those sectors characterized by
relatively large contributions to total emissions and relatively
low average costs of abatement.
20 These estimates come from Table 4.2 and are inflated to
$1993, a.s discussed in the Executive Summary.
- 16 --16



Table 3.1
Industry Sectors
(International Standard Industrial Classification)
ISIC    SECTOR
3110    Food Products
3130    Beverages
3140    Tobacco
3210    Other Textile Products
3211    Spinning, Weaving
3220    Wearing Apparel
3230    Leather & Products
3240    Footwear
3310    Wood Products
3320    Furniture, Fixtures
3410    Other Paper Products
3411    Pulp, Paper
3420    Printing, Publishing
3510    Other Industrial Chemicals
3511    Basic Industrial Chemicals
3512    Agricultural Chemicals
3513    Synthetic Resins
3520    Other Chemical Products
3522    Drugs and Medicines
3530    Petroleum Refineries
3540    Petroleum & Coal Products
3550    Rubber Products
3560    Plastic Products,
3610    Pottery, China, etc.
3620    Glass & Products
3690    Non-Metal Products n.e.c.
3710    Iron and Steel
3720    Non-Ferrous Metals
3810    Metal Products
3820    Other Machinery n.e.c.
3825    Office & Computing Machinery
3830    Other Electrical Machinery
3832    Radio, Television, etc.
3840    Transport Equipment
3841    Shipbuilding, Repair
3843    Motor Vehicles
3850    Professional goods
3900    Other Industries
17



REFERENCES
Dollar, D., 1992, "Outward-Oriented Developing Economies Really
Do Grow More Rapidly: Evidence from 95 LDCs, 1976-1985,"
Economic Development and Cultural Change, pp. 523-544.
Hamilton, M., 1994, "Selling Pollution Rights Cuts the Cost of
Cleaner Air," Washington Post, August 24,- p. FP.
McGuckin, R.H. and G. A. Pascoe, 1988, "The Longitudinal Research
Database:  Status and Research Possibilities," in Bureau of
Economic Analysis, US Department of Commerce, Survev of Current
Business, November.
Mirman, L.J, D. Samet and Y. Tauman, 1983, "An Axiomatic Approach
to the Allocation of a Fixed Cost Through Prices," Bell Journal
of Economics, 14(1).
Mundle, S., S. Mehta and U. Shankar, 1994, "Incentives and
Regulation for Pollution Abatement, With an Application to Waste
Water Treatment," paper presented to the 50th Congress of the
International Institute of Public Finance, August 25-29, Harvard
University.
Sell, N., 1992, Industrial Pollution Control: Issues and
Techniques (New York: Van Nostrand)
18



TABLE 3.2
MEAN TONS ABATED
ISIC                   3110        3130        3140       3210       3211       3220       3230       3240       3310
_____________________________      _ __ _,,Food  Beverages     Tohacco    Taxtile    spinning    Apparel    Leather    Footwear    Wood
P1 - Particulates                              _
Four Pollutants 1979-1982               1474.9        2770.8      1348.9      93.3       574.7                134.7        18.1  2345.5
Four Pollutants 1983-1985              ___            1170,0                 183.3                       ___                      2333.5
Seven Pollutants 1983-1985              1429.3        1170.0      1968.0    163.3       658.6                 252.4        13.4  2333.5
P2- Sulfur Oxides                                                   .  ___                               _       ....     _
Four Pollutants 1979-1982                 5B.2         144.7        41.2       3.3         9.1 _               14.3         0.1      6.5
Four Pollutants 1983-1985                              107.0                   1.8                                                   2.5
Seven Pollutants 1983-1986                40.6         107.0        34.8       1.8       28.2                   1.4                  2.5
P3 - N02 and CO                                                                         _
Four Pollutants 1979-1982                165.7           1.8        15.6    112.9         37.7                 1 11.8       0.7     23.2
Four Pollutants1983-1986            ___                  9.4 .               142.1  . _                                             58.0
Seven Pollutants 1983-1985                 6.5           7.4         4.5       4.1         6.3                  2.4         2.2     t3.3
P4  H &drocapbons   - 
Four Pollutants 1979-1982                 13.7          26.0        15.0      20.9        13.3 ____0.4                      0.2     14.1
Four Poltltants 1983-1986                                0.3                  18.1 ._.___71.6
Seven Pollutants 1983-1985                12.            2.0         7.0    138.0        24.1                  47.6         2.3    42.8
P6 - Load                                                                                             _
Four Pollutants 1979-1982                                                                                                          -
Four Pollutants 1983-1986
Seven Pollutants 1983.-1985                                                 .              3.7 _0.01                      0.008    0.002
P6 - Hazardous Emissions                        _____   ____   ___    ____
Four Pollutants 1979-1982                                :-     .                       .
Four Pollutants 1983-1985        .   .                       _ - . ..
Seven Pollutants 1983-1985                 0.4                         -      15.4        2.5                    5          0.2      0.8
P7 - Other                             _        _      .               .           _           _
Four Pollutants 1979-1982                                  .                          .
Four Pollutants 1983-1985                   .
Sevwn Pollutants 1983-1986           1    13.71          0.3         0.4       0.7       13.7                  5.01         2.1     70.8
19



TABLE 3.2
MEAI TONS ABATED
ISIC                3320       3410      3411   3420          3611          3612        3513         3520        3522      3530
Furnture    Paper     Pulp     Pdn.      Ind. Chem.    Ag. Chem.    Reins     Chem. Prad.    DaOga    Peo.
P1 - Particulates                                 _        _                  .                  .                             -
Four Pollutants 7979-1982          2758.5   1382.8   26585.4      39.0        7353.6       1296.8   4820.0            463.2  977.6      1905.1
Four Pollutants 1983-J985                     37.3   25760.5    197.7 _                                                __      447.1 
Seven Pollutants 1983-1985         2047.8     37.3   28750.5    197.7       242754.4       1350.5   4783.7            738.2   447.1    1851.4
P2 - Sulfur Oxides            _      _ .           .                   _                                    _
Four Pollutants 1979-1982             3.3     75.6    1837.3       0.2        1879.8        31 1.8    1158.8           40.0   265.8   26687.8
Four Pollutants 1983-1986                     24.9    1589.6       0.2                                                          81.8
Seven Pollutants 1983- 1986           2.5     24.9    1559.6       0.2       69102.0        654.4      82.0            24.6     81.8   33267.2
P3- ,N02 and CO                           _       .. ._._.
Four Pollutants 1979-1982            31.6    26B.4     305.1    696.4         3205.8       1237.0   2233.9           1792.E   318.2   49095.3
Four Pollutants 1983-1985                      8.6     379.8   1518.0                                                          S .  526.6  _
Seven Pollutants 1983-1985           77.8      0.1     315.7    397.2         1683.2        270.0    217.7          11257.5     12.4   35138.4
P4 - Hydrocarbons                       .                                              . __. _                           ._.
Four Pollutants 1979-1982            22.3     17.0    1002.5       2.4         876.6        383.2    165.1             46.3    68.9    1182.8
Four Pollutants 1983-1985                      3.7    1044.7      19.1                            _                            154.6
Seven Pollutants 1983-1985  _         5.2      6.4      60.8   1120.8         1656.0        824.5   1911.3           1003.4   514.2   10044.6
P5- Lead
Four Pollutants 1979-1982           .                                        _                                                                     :
Four Pollutants 1983-1985       .      .                        .                               .                                     .
Seven Pollutants 1983-19865                              0.1 ___                17.7                    e.s             0.8      0.3      73.6
P6 - Hazardous Emissions                                          _                  _             _       ._-
Four Pollutants 1979-1982....
Four Pollutants 1983-1985                 _                                     
Seven Pollutants 1983-1985            0.2     0.03      27.3       0.2         216.1         22.8    605.3              9.7      1.8     142.6
P7 - Other                                              _.                                                          ..
Four Pollutants 1979-1982                  .                                                                                     .
Four Pollutants 1983-1985                                      .                                                     .
* Seven Pollutants 1983- T985        11.5      3.7    1017.3      18.9        1011.1      1285.3       63.4          1142.9    182.5      79.6
20



TABLE 3.2
MEAN TONS ABATED
/sic                   3540      3560      3560      3610      3620      3690      3710           3720           3810
.___________:__________. ___________  Coa/    R"bber    PlAito    Poiery   G0ass    N*Metal    Iron   N_Ferrous Metal            Metal
P1 - Particulates               .                                                .
Four Pollutants 1979-1982               3777.0     525.2    290.0       408.0   505.2   34936.0   15205.6                4530.3   225.0
Four Pollutants 1983-1986               2623.8     581.5                508.4   685.9   34690.6
Seven Pollutants 1983-1985              2623.8     581_5 _ __           508.4   685.9   34690.6    9466.1                2895.0   201.5
P2 - Sulfur Oxides                             _                    _                                       _____________
Four Pollutants 1979-1982                 42.7      63.5       2.8        0.1    29.2      257.0      796.2             6527.9      7.0
Four Pollutants 1983-1985                 33.61     37.'                 0.03    15.4      791.6
Seven Pollutants 1983-1985          __-33.          37.1 _               0.03    15.4      791.6    1418.7              4468.0    10.5
P3 - N02 and CO 
Four Pollutants 1979-1982               1919.9      89.0    132.2         6.3    27.8       81.5    3299.3                154.5   123.0
Four Pollutants 1983- 1986              1527.4      70.2                 12.4    61.7      223.2
Seven Pollutants 1983-1985               S40.1      21.E                  2.1    59.1      161.7    2040.5                 18.4    19.4
P4 - Hydrocarbons
Four Pollutants 1979-1982                  3.3      40.7      35.3        0.2     7.3      353.8      800B6               724.5    21.5
Four Pollutants 1983-1985                  0.3       3._  _               4.8      10      102.5 _
Seven Pollutants 1983-1985               987.3      48.7                 10.3     2.6       61.5       64.5               55.2   122.5
P6 - Lead                            _____
Four Pollutants 1979-1982                                        .                  .
Four Pollutants 1983-1985                                      ..             _    .           .                             _
Seven Pollutants 1983-1986                0.01                           0.02     0.6        0.1       13.3               93.1      1.1
P6 - Hazardous Emissions               =__  _
Four Pollutants 1979-1982
Four Pollutants 1983-1985                                                  .     .                 ..  _
Seven Pollutants 1983-1985                 0.2      0.02 _-                       0.3       27.7      1                  124.4      3.4
P7- Other               .                                                                         _
Four Pollutants 1979-1982
Four Pollutants 1983-1985                                                                     _                                  .
Seven Pollutants 1983-1985                 0.21      3.ea                 4.8     9.2       74.8      184.2              776.9      6.5
21



TABLE 3.2
MEAN TONS ABATED
ISIC                     3820         3825         3830       3832       3840      3841      38          3850      3900
. __________._________.____   Machine,y    Computing    BfotifIvl    Radlo    Tiampon    Shi               Vehice     P. Gooad     Wthar
P1 - Particula tes. 
.four Pollutants 1979-1982                  404.0         109.9        583.2   186.2        159.9   420.9      1022.1        23.8  245.4
Four Pollutants 1983-1985                   203.5                                             _     204.0        _                 1080.6
Seven Pollutants 1983-1985                  203.5          72.3       399.6   101.5         178.8   204.0       547.4        28.7  1 080.5
P2 - Sulfur Oxides                                            _  . _
Four Pollutants 1979-1982                    62.6           5.2        31.0      8.0         S5.0   185.4       120.2         1.2    0.2
Four Pollutants 1983-1985                    48.7                      _                            141.7        _                    0.3
Seven Pollutants 1983-1985                   48.7           4.5        46.0      5.0         20.1   141.7       142.0         0.8    0.3
P3 - N02 and CO                                                                                _
Four Pollutants 1979-1982                    37.0         211.3       168.8   148.7          27.8    28.1       139.9        17.4    7.8
Four Pollutants 1983-1985                    26.0                                                     28.8                          444.7
Seven Pollutants 1983-1985                    7.8           0.6         8.5    51.8          18.6    22.8        11.3         7.4    1.2
P4 - Hydrocarbons                                        .______._.
Four Pollutants 1979-1982                    30.1           4.7         84.8     1 8.0        7.3    13.4       407.2         2.8   10.1
Four Pollutants 1983-1985                     7.1                                                      1.9                         1376.4
Seven Pollutants 1983-1986                   18.2         289.7        62.4   114.3          17.3     B.0       110.2        11.3  443.5
P5 - Lead
Four Pollutants 1979-1982                             . .__                                                        .
Four Pollutants 1983-7985                                                                                    .              .
Seven Pollutants 1983-1985                    0.5                       87.2     0.2        0.004   0.008         1.3         0.1  0.119
P6 - Hazardlous EmlssIons_____
Four Pollutants 7979-1982
Four Pollutants 7983-1985            _                          __._.. _._.
Seven Pollutants 1983-1985                    0.8           2.1         0.9    14.6           1.4     1.2         8.3         4.4   89.9
P7- Other       '_._.__
Fo-ur Pollutants  1979-1982          _ _ _ _ _     _ _ _ _ _     _ _ _ _       _ _     _  _  _ __             _  _  _ _ _ _ _
Four Pollutants 193-1985                                                           _
Sevevn Pollutants 1983-1985                   5.81          1.8         12.51   13.81         7.9      0.8       11.4         8.9  1286.
22



TABLE 4.1
MARGINAL ABA TEMENT COSTS ($1979)
/SIC    _   3110                                   3130         3140       3210      3211        3220      3230       .3240      3310
.___________________________  Food  Beverages.    Tobacco    Textile    Spinninng    Appa,el   Leathwr| Footwear    Wood
Marg_nal Cost (79-82)                       NL _   _                           NL          NI          L         L            L       L
Pl - Pariculates                         14.77                    107.90   405.62       49.38     204.95    261.67        72.11     5.47
P2 - Sulfur Oxidos                      381.281                           4777.57      370.80      _        311.11
P3 - NO2 and CO                           2.99                             187.24       92.52                311.11      555.38
P4 - Hydrocarbons                         2.99                             187.24       92.52 _                          555.38
P6 - lead -____                                                            490.00                  20.28 .
P6 - Hazardous Emisslons            .___._._490.00                                                 20.28                           _
P7 - Other                           .                         .           490.00                  20.28
Margf al Cost (83-85)                        _L                                NLI          L      _            NL            L      NL
P1 - Particulates                        17.82                                          39.69                                      14.38
P2 - Sulfur Oxides                      118.54
P3 - N02 and CO                         182.7                              698.99     1327.18              8550.88 
P4 - Hydrocarbons                       111.01       _          12869.60    712.09     453.33               223.47       603.73
P6 - Lead                            49,723.02 _             _           _      _      633.16             :           23,142.34
P6- Hazardous Emisslons                 747.32                             _   _                         _  191.06       773.40
P7 - Other_'                                                                        .                       300.10      457.51
WMarginal Cost (79-85J                                   NL                    NL                                                     L
P1 - Particulates_                                     42.42                        _                                              4.78
P2 - Sulfur Oxides                                    103.91
P3 - N02 and CO                                     6323.80  _             624.76                _  _
P4 - Hydrocarbons                                '  6323.80                524.78
P6 -Lead          -           -                                            424.44
P6 - Hazardous EmissIons                                                   424.44
P7 - Other                                        .                      ' 424.44         _                     _              __
.Margnal Cost (79-86)                     .                       .. _                                                               _
PI - Particuate:                          5.54          5.54       61.17     5.54       35.62     223.58      5.54_        5.54    S.45
*(Particula tes only)
*   '* L - Linear, NL - Non-Llnear
23



TABLE 4.1
MARGINAL ABA TEMENT COSTS (s 979)
/SIC                            3320       3410    3417    3420         35t1         3J12        3513  |      3520       3522    3530_
| Furnitura    Paper   Pulp    Printing    Ind. Chem.    Ag. Chem.    Resins    Chem. Prod.    Dwgs    Petrm.
Marginal Cost (79-82V                 NL       NL      NL         L            NL           NL        NL             NL       NL       NL
P1 - Particulates                   6.30    28.58   15.74    327.23          7.83        45.24     17.34         188.69   115.20   336.43
P2 - Sulfur Oxides                         178.481 67.79                    42.35       295.28    232.48                  386.29    64.72
P3 - N02 and CO                   _        214.70             113.21        7222         158.56    40.25         1¶9.9    72.9B        6.99
P4 - tHydrocarbons                _        214.70   _        113.21         72.22       168566    40.26           19.99    72.a5    16.99
P5 - Lead          .                        43.36   31.59                  170.83       212.12                            217.78  2923.71
P6 - Hazardous Emissions     _              43.3B   31.69                  170.53       212.12                            217.78  2923.71
P7- Other                                   43.36   31.69      .           170.53       212.12                            217.78  2923.71
Marginal Cost (83-86)                 NL        L     NL          L            NL           NL        NL             NL        L        L
Pi. - Partlculates                  12.89           13.41                   13.42                  19.77          29.09    89.83
P2 - Sulfur Oxides           .  1620.60  178.14   76.29       92.11         10.12       168.22   317.20          696.35  827.51    97.80
P3 - N02 and CO                            417.88             59.52        225.30       700.05    130.18                  30B.33    32.91
P4 - Hydrocarbons                 _        417.86                          127.40       114.79    40.19          117.34     5.95    98.07
P6 - Lead                                            8.57   5859.b5       1189.70                 413.83 
P6 - Hazardous Emissions        14,215.16    -       8.57                  133.45       839.09      7.52                       _
P7 - Other                                           8.57                   20.58        38.44   231.23          187.04
Marginal Cost (79-85)                          NL     NL                                                      _
P1 - Particulates                           28.85   12.78 _                                     _
P2. - Sulfur Oxides .                      178.83   72.12
P3 - N02 and CO                            234.39                                                                              _
P4 - Hydrocarbons                          234.39                                                      _
P5 - Lead                                   28.79    8.08       _                                  _-
P6-- Hazardous Emissions    .               28.79_  8.08
P7 - Other                                  28.79    8.08                        _____________
Marginal Cost (79-8:)                                                                                                     -         -
PI - Particulates*                  5.45     5.54    5.54      6.54          1.65         5.54      5.54           5.54     5.54    98.75
*/Particulates only)
** L - Linear, NL - Non-Linear
24



TABLE 4.1
MARGINAL ABATEMENT COSTS ($1979J
Is/c                                      .3540     3550      3660      3610      3620       3680      3710          3720          3810
Coal    Rubber    Plestic   Ponery     llass    NMetal    Iron       N-Ferrous Metal    Metal
Marginal Cost (79-82)                          NL        NL        NL        NL         L         NL       NL                 Nl       NL
PI - Particulates                           10.89     52.44     84.01    186.01                 4.51    77.10              173.27  101.46
P2 - Sulfur Oxides                        1914.63    312.98   1256.79              205.14     138.36   639.79               64.55  949.68
P3 - N02 and CO                              8.79    169.15     92,44   2134.64    128.48     911.98    69.34                      135.69
P4 - Hydrocarbons                            8.79    169.15     92.44   2134.64    128.48     911.98    69.34                      135.69
P5 - Lead_                                           343.54   571.4 3    -                                                 136.91
P6 - Hazardous Emissions                     _       343.54    571.43  _                                   _               136.91
P7 - Other                                           343.54    571.43                                                      136.91
Marginal Cost (83-85)                          NL        NL                  NL        NL         NL       NL                 NL       NL
PI - Paricula tes                            7.21_  33.72                                       4.60    97.60_             137.55   93.48
P2 - Sulfur Oxides                         227.32    308.30                                    77.06     4.32              45.14  547.53
P3 - N02 and CO                             44.03                       5907.00    220.22     835.38    33.34                      184.63
Hydrocarbons                                18.91                       2662.79    196.34     847.33  1195.59             612.08  119.11
P5 - Lead                                _                                       22,536.78             811.80             1182.05
P6 - Hazardous Emissions                 _                          _                                  250.58            1029.47
P7 - Other  .                                       1127.01             3687.96                                                    284.14
Marglnal Cost (79-85)                          Nl        NL                  Nl         L
P1 - Partculates                             9,25'    25.1 8              5_9.04                                         _
P2 - Sulfur Oxfdes                        1014.70    498.96                       _194.69                      ___                _
P3 - N02 and CO                             1871      93.99             1 gg4.90     Ml 6                              _
P4 - Hydrooarbons                           1 e.78    93.99             1954.80     81.61
P5 - Lead                                   _        447.48 _           2484.90                    _ _
P6 - Hazardous Emissions                             447.48             2484.90 _
P7 - Other                                           447.48             2484.90 _       _           _                               3
Mar mal Cost (79-85)                     _                                                                   1
P1 - Partculates*                            5.54      5.54      5.54-    5.54       5.654      3.25    84.541              5.54   53.32
* (Partdculates only)
* L - Linear, NL - Non-Linear
25



TABLE 4.1
MARGINAL ABATEMENT COSTS ($7979)
3820         3825         .9830     3032       3840       .9841     .1843       3850      3900
Machinery    Computing    Elecdoal__ Radio    Tranaport  Ships    Vehcle&s   P. Goods    oth.r
Marginal Cost (79-82)                         NL            NL          NL       NL            L       NL         NL          NL       NL
PI - Particulates                           52.15                    110.01    39.04       50.96   126.89                  284.90    5.77
P2 - Sulfur Oxides                         425.83                    408,48  1718.28      427.70   374.17    1354.92      2632.73
P3 - N02 and CO                            294.98       661.33        81.64   513.09              1618.58    1080.98        65.57  134.69
P4 - Hydrocarbons                          294.98       661.33        81.64   513.09              1618.58    1080.98        6S.57  134.69
P5 - Load                                      ._                     69.93   526.38                            17.41      442.22
P6 - Hazardous Emissions                                    _9.93   52.6.35                                     17.41 _    442.22
P7 - Other                                                            5 .9.93   526.38        _                 17.41      442.22
Marginal Cost (83-85.                         NL  __L                   NL       NL           NL       NL         NL          NL        L
PI - Partculates                            89.93                    181.83   117.81      127.41   334.01      221.47      359.83   13.32
P2 - Sulfur Oxides                         288.53                  _                      424.62   632.60      118.80
P3 - N02 and CO                            122.32                   1476.22   190.17               520.62                 .193.09
P4 - Hydrocarbons                          198.34        78.24       40.49   395.21       574.13              1393.42      783.15   33.66
P5 - Lead                                                           213.00                                  2273.60               3855.22
P6 - Hazardous Emissions        .                    18384.10                803.70     1989.31                          862.11   119.65
P7- Other _                  .                   _                        ._.
Marginal Cost (79-85)                         NL                                                L__                                     L
P1 - Particulates                          62.,29                                                   21.93                            6.22
P2 - Sulfur Oxides                         303.00          __            _ ____399.53
P3 - N02 and CO                            201.42              _                            _     1146.40                           44.67
P4 - Hydrocarbons                          201.42 _                                               1 1145.40
__ _ _ _  __ _ _ _ _ __ _ _ _  _ ____-_ _                  _                _  _ _ __              _ _ _44 .6 7
P5 - Lead    .                                                        .               .                                              ._.
P6- Hazardous Emissions                    .
P7- Other                                                        ._.
.Marginal Cost (79-P6) 
P1 - Partioulaes .36.60                                   6.64       76,00       6.54      49.25     5.54      81.24         5.54    5.54
*(Parttculates only)
** L - Linear, NL - Non-Linear
26



TABLE 4.2
TOTAL A VERAGE ABATEMENT COSTS ($1979)
/SiC                                     3110         3130         3140       3210       3211        3220       3230        3240       3310
Food       Tverges      Tobacco     rextif.    Spinnfng   Apparel    Leather    Footwear    Wood
Avorage Cost (79-82;                                                           _             .                  .
___________________________________ _     7   NL                          L        NL          NL          L          l            L        L
P1 - Particulates                          47.62       _             178.41    B12.99      160.77     237.58     109.40       182.42    25.84
P2 - Sulfur Oxidos                        414.13                      68.51   4984.94      482.19      32.63     371.07       110.31    20.37
P3 - N02 ond CO                            35.84                      68.61    394.61       111.39     32.63     420.51       665.69    20.37
P4 - Hydrocarbons                          35.84                      63.51    394.81       111.39     32.53     420.51       665.69    20.37
P5 - Lead                                  32.85                      68.51    697.37      111.39      32.63      109.40      110.31    20.37
P6 - Hazardous Emissions                   32.86                      68.51    697.37      111.39      32.63     109.40       110.31    20.37
P7 - Other                                 32.85                      68.51    697.37      111.39       32.63    109.40       110.31    20.37
Average Cost (83-85)                         .          . _                            -                                            .
_______ ______ _______ ______ __ _   Nl  _             L        NL            L                   Nl            L        L
P1 - Particulates                          43.99                     110.06    177.46       129.36                31.95       394.51    23.44
P2 - Suffur Oxides                         142.65                    110,06    177.46        89,67                31.95                 20.32
P3 - N02 and CO                           208.88                     110.06    177.46     1416.86               8582.83       394.51    20.32
P4 - Hydrocarbons                         137.12                  12,779.56    889.66       89.67                255.42       998.51    20.32
P5 - Lead                              49,749.13                                           722.82                 31.95    23,536.85    20.32
P6 - Hazardous Emissions                   26.11                               177.46       89.67                237.10      1167.91    20.32
P7 - Other                                 26.11 _        _110.08    177.46                  89.67               346.14       852.02    20.32
Average Cast (79-85)                      _                                 _   __
NL                    Nl                                                       L
P1 - Particulates                       _                 83.08                211.59          -                                        24.87
P2 - Sulfur Oxides                                       144.67                211.69            _                                      20.09
P3 - N02 and CO                                         6 6364.46              736.35                                                   20.09
P4 - Hydrocarbons                                       6384.46                736.35 __                    _         _                 20.09
P5r - Lead                             _____40.65                              836.03              _________20.09
P6 - Hazardous Emissions                                  40.65                636.03                                                   20.09
P7- Other                                 -               40.66                 836.03     -          -          -                      20.09
A verage Cost (79-85                                _                                      _
PI - Partcuiates                             5i.65         5.79       61.84       7.05      35.78     228.54        7.59       20.22     5.58
*(PartIculates only)
* * L - Linear, NL - Non-Linear
27



TABLE 4.2
TOTAL AVERAGE ABATEMENT COS7S ($1979)
[SIC____________________          3320       3410    3417      3420        3517          3512        3513         3520        3622      3530
_____________    _ _______ _  Fumiture    Papor    Pulp    Printing    Ind. Chum.      Ag. Chem.    Relnhs     Chem. Prod.    DMOg     Petro.
Average Cost t79-82)                         
__                          ~~~~~~~~                ~    ~~NL  NL  NL  L  Nl         Nl       NL              NL       Nl        NL
PT - Partloulates                     19.02    44.09   25.41    411.08          34.45       117.39     47.13          192.90   149.16   347.91
P2 - Sutfur Oxides                    12.72   195.99   76.46     63.B4          68.99       367.43    262.25           26.21  420.25    76.20
P3 - N02 and CO                       12.72   232,21    8.67    197.05          98.85       230.71     70.05           46.20   106.92    28.47
P4 - H_ydrocarbons                    12.72   232.21    8.67    197.05          98.85       230.71     70.05           46.20   106.92    28.47
P5 - Leaa                 __          12.72    60.87   40.36     83.84         197.26       284.27     29.79           26.21   261.74  2935.19
P6 - Hazardous Emissions              12.72    60.87   40.36     83.84         197,26       284.27     29.79           26.21   251.74  2935.19
P7 - Other                            12.72    60.87   40.30     83.84         197.26       284.27     29.79           26.21   251.74  2935.19
A vorage! Cost (83.85)   
NL       L      NL          L       _ NL               NL       NL              NL        L         L
PI - Particulates                    26.50    73.01   22.60      41.33          14.33         18.76    40.56           33.70   138.20      2.14
P2 - Sulfur Oxides                    13.81   261.11   85.48     41.33          11.03        158.95    337.98         700.96  698.09    99.94
P3 - N02 and CO                       13.81    73.01    9.19     133,44        226.21        718.81    150.96           4.61   374.90    35.05
P4 - Hydrocarbons                     13.81   490.81    9.19    130,86         128.31        133.55    60.94          121.95    77.52   100.22
P_S - Lead                                             17.76     41.33        1190.68    _            434.__1           4.61    6867       2.14 
P6 - Hazardous Emissions          14,228.97    73.01   17.78   5901.28         134.36       857.85     37.16            4.61    68.57      2,14
P7 - Other                            13.81    73.01   17.78     41_33          22.24        57.20    262.01          191.65    68.57      2.14
Average Cost (79-85)             _______    _____
7 0 - ~~~~~~~~~NL                  NL -_____
pi - Particulates                             4 637   23.20
P2 'Sulfur Oxides                            194.35   82.64 _           _      __     _                       _         -
P3 - N02 and CO                                              ______  _______2-6i.91   10.62          _
P4- Hydrocarbons                             25 .  21.91   10.52                                                             .
P5,-L ead                                 _ 46.31   18.60               _                    ____
P6 - Hazardous Emissions                      46.31_  1B.60
P7- Other                                     48  46.311  18._0                                                               __
Average Cost 79-85)            ___                          _          _                                     _               _
P1 - Partlculates*                     5.e1     8.20    6.67      9.79           1.74          8.70     5.61            6.74     8.94    99.09
*(Partlculates only)
* L - Llnear, NL - Non-Linear
28



TABLE 4.2
TOTAL A VERAGE ABA TEMENT COSTS ($1979)
ISIC                         | 3640      3660     3660      3610      3620      3690      3710          3720
Coa/    Rubber   Nlaatlo   Pottey  |  Olhs | N-Metal    iron        N-Fearrau Metal
Average Cost (79-82)                                                .  .
NL        NL       NL        NL         L        NL       NL                 NI
P1 - Particulates                30.82    104.55    116.85    166.01    136.15     11.36    88.37            195.10
P2 - Sulfur Oxides             1934.56    365.09  1289.63              341.29    145.21   551.07              86.38
P3 - N02 and CO                  28.72    221.26    125.28   2134.64    264.63    918.83    80.62             21.83
P4 - Hydrocarbons                28.72    221.26   125.28   2134.64    264.63    918.83    80.62              21.83
P5 - Lead                        19.93    395.65   604.27              136.15       6.85    11.28            158.74
P6 - Hazardous Emissions         19.93    395.86   604.27              136.15       6.85    11.28            158.74
P7- Other                        19.93    395.85   604.271             136.15       6.85    11.28            15B.74
Average Cost (83-85)          _                        = _                                  -
NL        NL                 NL        NL        NL       NL                 NL
PI - Particulates                33.82    126.91              36.04     73.97       9.52   106.03             167.85
P2 - Sulfur Oxides              253.93    401.49              36.04     73.97      a1.98    12.75             75.44
P3 - NV02 and CO                 70.64     93.19            5943.04    294.19     840.28    41.77             30.30
P4 - Hydrocar6ons                45.52     93.19            28.84    270.32    852.25  1204.02               842.38
P5 - Lead                        26.61     93.19              36.04 22,610.75       4.92 =20.23             1212.3S
P6 - Hazardous Emissions         26.81     93.19                        73.97       4.92   259.01           1072.44
P7 - Other                       26.61   1248,42               3704     73.97       4.92    8.43              30.30
Average Cost (79-851).._.__
_________________________          NL        NL                 NL        N I  5
P1 - Pariculates                 31.76    117.12              98.88     99.21
P2 - Sulfur Oxides             1037.21    590.90              56.38    293.80
P3 - N02 and CO                  41.29    183.24            2024.74    180.82
P4 - Hydrocarbons                41.29    183.24            2024.74    180.32
P5 - Lead                        22.51    539.42 _          2514.74     99.21 _
P6 - Hazardous Emissions         22.51    539.42        _  2514.74      99.21                    _     _=
P7 - Other                       22,51    539.42 .,         2514.74     99.21                        _
Average Cost (79-865                                                     .
PI - Paricubltes          _       5.82      5.87     7.78      e.        5.83       3.26    64.57              5.85
* (Partfoulates only)
L - Linear, NL - Non-Llnear
29



TABLE 4.2
TOTAL AVERAGE ABATEMENT COSTS (*1979)
Is/c                          3810       3820         3825         3830       3832       3840       3841      3843        3850      3800
Metal    Machinery    Computing    EBectdcal   Radio    Transport    Shipa    vehrices    P. Goods    Other
Average Cost (79-82)        _        .             _ -                       _        _           _        
_______________________  NL  NL  NL    NL       NL            L      NL          NL          NL       L
P1 - Particulates             184.94        112.92       132.23       155.45   138.63      349.36   168.57       79.19      608.53   74.19
P2 - Sulfur Oxides           1033.16        486.60       132.23      453.90  1815.85       726.09   413.85    1434.11      2974.a6   68.43
P3 - NV02 and CO              219.17        355.75       793.66       127.08   61 0.68     298.39  1658.26    1160.17       407.20  203.12
P4 - Hydrocarbons             219.17        355.75       793.568      127.08   610.68      298.39  1658.26    11 60.17      407.20  203.12
P5 - Lead                      83.48         60.77       132.23       115.37   623.97      298.39    39.68      96.60       783.85   68.43
P6 - Hazardous Emissions       83.48         60.77       132.23       115.37   623.97      298.39    39.68       98.60      783.85   68.43
P7 - Other                     83.48         60.77       132.23       115.37   623.97      298.39    39.68       96.60      783.85   68.43
Average Cost (83-85J)                 _____.__ _                _                             _             _          _____
Nl.          Nl            L           NL       NL          NL       NL          NL          Nl.    L
PI - Partioulates             181 .67       183.20       129.10      243.42   284.02       329.03   412.34      294.69      883.61    21.59
P2 - Sulfur Oxirdes           635.72        381.00       129.10        61.79   164.51      626.24   674.36      192.02      278.98    8.27
P3 - N02 and CO      .272.82                218.28       129.10     1538,01   354.68       201.62   699.9r      73.22       524.06    8.27
* P4 - Hydrocarbons              207.30       291.01       207.34       102.281 559.72       775.75    78.33    1446.64      1062.13   41.93
P5 - Lead                      88.19         93.27                   274.79   164.51       201.62    78.33   22,847.82      278.98 3863.50
P6 - Hazardous Emissions       88.19         93.27    18,513.25       61.79   966.21      2190.93    78.33       73.22      961 .09  126.34
P7 - Other                    372.33         93.27       129.10       61.79   591.86       201.62    78.33       73.22      278.98    6.69
Average Cost (79-85)        _                                   _ . .                                             _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _   _ _ _ _ _ _ _   N L  _ _ _ _ _ _L                                      L__ _ _ _  
Pl  Palculates                 _            135.8;2        _,._6.80                                                                  20.16
P2 - Sulfur Oxidos           ____ -_     _ 4586.53              .                         :         444.41                           1 3.94
P3 - N02 and CO             .               274.95                                                 1190.3C0 _                        58.61
P4 - Hydrocarbons                           274.95                                                 1190.30                           58.61
P6 - Leadt                                   73.53                                              _   44.87          _                 13.94
.   P6- Hazardous Emissions   .                  73.583                                 .             __-44.87                            13.94
P7 - Otheor.                                 73.63                                                   44.87                 =         13.94
Average Cost (79-85) .                                                         _        _
P1 - PartIculatesx             53.42         36.91       234.06       76.17     8.99        49.7S     7.35      81.53        10.23    6.24
(Particulates only)
* Y L - Linear, NL - Non-Linear
30



Policy Research Working Paper Series
Contact
Thile                           Author                  Date              for piper
WPS1384 Fiscal Decentralization and       David Sewell           November1994       G. Langton
Intergovemmental Finances in the    Chrisline I. Wallich                  38392
Republic of Albania
WPS1385 Fiscal Federalism Dimensions of Tax  Robin Boadway       November 1994      C. Jones
Reform in Developing Countres   Sandra Roberts                            37754
Anwar Shah
WPS1386 EU Bananarama iII                 Brent Borrell          December 1994      G. Ilogon
33732
WPS1S87 Fiscal Decentralization and the Size  Jaber Ehdale       December 1994      C. Jones
of Government. An Extension with                                          37699
Evidence from Cross-Country Data
WPS1388 Does Voice Matter? For Public     Samuel Paul            December 1994      B. Moore
Accountability, Yes                                                       35261
WPS1389 Ruble Overhang and Ruble Shortage: Patrick Conway        December 1994      L Suki
Were They the Same Thing?                                                  33974
WPS1390 Regional Integration and the Baltics:  Piritta Sorsa     December1994       J. Ngaine
Which Way?                                                                 37947
WPS1391 Where in the World Is Population  Jeff ling              December 1994      L. Fallon
Growth Bad?                     Lant Pritchett                            38009
WPS1392 Some Economic Consequences of   Jean-Paul Azam           December 1994      C. Jones
the Transition from Civil War to Peace David Bevan                        37699
Paul Coller
Stefan Dercon
Jan Gunning
Sanjay Pradhan
WPS1393 The Interface of Trade, Investment,  J. Luis Guasch      December 1994      J. Troncoso
and Competifion Policies Issues and Sarath Rajapatirana                   37826
Challenges for Latin America
WPS1394 Macroeconomic Reform and Growth  Lawrence Bouton         December 1994      R. Marlin
in Africa: Adjustnent in Aftca  Christine Jones                           39026
Revisited                       Miguel Kiguel
WPS1395 A Typology of Foreign Exchange   Janine Aron             December 1994      R. Luz
Auction Markets in Sub-Saharan  Ibrahim Elbadawi                           39059
Africa
WPS1396 Foreign Exchange Auction Markets  Janine Aron            December1994       R. Luz
In Sub-Saharan Africa: Dynamic    Ibrahim Elbadawi                        39059
Models for Auction Exchange Rates



Policy Research Working Paper Series
Contact
flU.             -    -.       Author                  Date             for paper
WPS1307 Are Pivate Capl Fbslo"to         Uh Dadush              December 1994     J. Queen
Develophg Countrbs Suatainable?   A8hok-Dharee r                         33740
Ron Johanns
WPS 1398 The Cost of Alr Pouudon Abatement  Raynond S. Hartmnan  December 1994    E Schaper
David Wheeler                            33457
Manjula Singh