See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/279869568

Acid rain

Article in Economic Policy · October 1990

DOI: 10.2307/1344480

CITATIONS READS

62 29,156

3 authors, including:

David M. Newbery
University of Cambridge
339 PUBLICATIONS 17,460 CITATIONS

SEE PROFILE

All content following this page was uploaded by David M. Newbery on 24 October 2017.

The user has requested enhancement of the downloaded file.

ACID RAIN

David M Newbery1
Department of Applied Economics Cambridge and CEPR

(Revised 6 August 1990)

1. Introduction

Air pollution is not a new phenomenon - Londoners in the twelfth century complained about the
noxious fumes from burning sea coal, and the corrosive effects of sulphur dioxide (SO2)
dissolved in rain has been well understood for at least a century. But the focus of concern
constantly shifts. In Britain the process started with the Smoke Abatement Acts of 1853-56, and
via various other measures to the landmark Clean Air Act of 1956. The cause for concern were
the very obvious health hazards associated with the unregulated burning of coal, and in particular
the large number of people, estimated at 4,000, who died in the great London smog of December
1952. America has also been concerned with reducing coal pollution, but was also active in
reducing automobile pollution from quite early on. Here the impetus was the deteriorating air
quality in urban areas like Los Angeles and Washington DC, where photochemical smog led to
high levels of ozone. This was traced to exhaust emissions of hydrocarbons and nitrogen oxides
(NOx), and California lead the way in introducing successively tighter emissions controls. In
Europe, the impetus for environmental policy developed because acid rain - from the sulphur
dioxide and NOx emissions from the burning of coal and oil - is no respecter of national
boundaries. Each locality and country discovered that some of the immediately harmful effects
of burning coal could be avoided by building tall chimneys, but this merely dispersed the
pollutants elsewhere, often great distances to other countries. These countries could not directly
control the deposition of acid rain, and could instead only complain and negotiate for some
coordinated solution to the perceived problem. Different countries responded to different facets
of the pollution problem. The Scandinavian countries were troubled by the death and
disappearance of fish from lakes and rivers. Germans worried about forest die-back. Glasnost
revealed the full extent of the environmental disasters in East Europe, and provided the focus for
local hostility to the environmental insensitivity of central planning.
Environmental awareness has grown rapidly over the past decade, and with it the growing
realisation that we live on a rather small and fragile planet. The green movement has had to
work hard to capture the attention and imagination of the public and politicians, and has had to

1
Research support from the British Economic and Social Research Council’s grant Privatisation and
Reregulation of the Network Industries is gratefully acknowledged. I am indebted to Margaret Clark for assistance
with the literature search, to Michael Hanneman for bringing to my notice the paper by Mäler (1989), to R G
Derwent of Harwell for his extremely thorough and helpful scientific comments and advice, and to David Pearce,
John Vickers and David Begg for careful comments. None of these is to be held responsible for the interpretations
I have chosen.
resort to emotional arguments to get over its message. There is no doubt that the hidden
environmental costs of current technology may be high, even life-threatening, but it is also clear
from recent experience that the costs of carelessly designed environmental regulation can also
be high. As economists, we have a duty to argue for cost-effective environmental policies.
Inefficient policies not only achieve less than they should, but they also run the risk of alienating
taxpayers and consumers who ultimately pay for the regulation and may undermine the aims of
the environmental movement. Economists, with honourable exceptions, have tended to ignore
environmental economics because it seems to raise few new ideas. Most of the useful techniques
have been the stuff of undergraduate welfare economics since Pigou’s day, and though each
generation adds to the stock of knowledge and technique, the subject has not been at the
theoretical cutting edge for some time. This might not have mattered if economists had been
supplied with accessible facts with which to clothe the theory and to bring the policy issues into
sharp perspective. But these facts are largely produced by scientists unfamiliar with the
economic style of argument, and often unconcerned with economic costs and benefits. There
has been too little communication between the disciplines.
It is interesting to compare the situation in the US. The style of regulation exemplified
by the Environmental Protection Agency, and the separation of legislative and executive power
means that environmental legislation has to be argued in a quasi-judicial way before being
enacted, and economists have been centrally involved in the ensuing debates—not necessarily
successfully. As a result of having to make a quantified case in public, economists have
investigated the scientific evidence, have conducted empirical enquiries, and have identified the
gaps in our knowledge. Environmental economics has received a considerable impetus and a
solid body of theory and evidence on which to build. We in Europe lag behind, though there are
signs that the times are changing.
This contribution deals with a small part of the environmental debate, that concerned with
acid rain. It is an important topic—not as important as the greenhouse effect, which is global
in scale, probably not as important as traffic congestion, which is a domestic matter for each
country. Nevertheless, substantial sums of money have been spent and are now being committed
in an attempt to alleviate the problem of acid rain. The thrust of this paper is that this
programme as currently interpreted is flawed, unnecessarily expensive, and if it succeeds, it runs
the risk of high political cost. Relatively simple economic principles applied to the appropriate
facts ought to be able to achieve the same environmental benefits at substantially lower cost, and
in a more decentralised and less politically problematic way.
I make no apologies for the high ratio of facts to theory in what follows. The
environmental debate has been long on emotional argument and short on substance for too long.
I am not an expert in this field, and have had to rely on secondary sources for the data. On the
crucial issue of quantifying the benefits of reducing acid rain I have not been able to find
adequate evidence and so cannot finally quantify the efficient policy. But I have found enough
evidence to cast considerable doubt about the priorities for abatement, and to suggest where
research effort should be concentrated. Several findings surprise me. Fish death from acid rain
is sad, but economically unimportant. Tree death may be far more important, though there are
worrying uncertainties about the cause and cure of this problem. Health problems associated
with coal emissions, particularly the combination of SO2 and particulates (smoke particles) are
potentially of the first importance, those associated with NOx and ozone seem trivial.
 3

2. Acid rain and its effects

In order to understand the acid rain problem it is necessary first to describe the causes and
consequences of acid rain. Considerable scientific research over the past decade has illuminated
the phenomenon, though uncertainties remain. The next step is to identify the sources and
measure the amounts of pollutant released, and their destination. What is it that causes the
damage, where does the main damage occur, and what are economically the most expensive
consequences of acid rain? Finally, one needs to determine the techniques available for reducing
emissions, and the costs of abatement, in order to identify cost-effective abatement policies.
This last step is usually ignored by ecologists and politicians, who are content once they have
found ways of reducing acid rain to press for the maximum (politically?) feasible degree of
abatement. This section addresses each issue in turn.

2.1 Defining acid rain

Acid rain is normally understood to include the deposition of the acidic combustion products
sulphur dioxide, SO2, various nitrogen oxides, NOx, and chloride, Cl-, either as dry gases or
particles, or as wet deposits in rain, snow, sleet, hail, mist or fog. These pollutants usually
undergo a series of chemical transformations into sulphuric acid, H2SO4, nitric acid, HNO3, and
hydrochloric acid, HCl. These acids in turn have a variety of effects on the environment, both
directly and indirectly in causing the release of further harmful chemicals such as aluminium.
Acid rain can be measured in a variety of ways—in terms of tonnes of the original gases
released, or tonnes of elemental sulphur, or in terms of the acidity of the rainfall, run-off, streams
or lakes. Acidity is measured in pH units on a logarthmic scale.1 As the scale is logarithmic,
rainfall with a pH of 5 is 10 times as acid as that with a pH of 6. Unpolluted rain is slightly
acidic from dissolved carbon dioxide and has a pH of about 5.6. Sea water is naturally alkaline,
having a pH of 8.3.
Most SO2 comes from large combustion plants—thus in 1987, 85 percent came from
large combustion plants, and 73 percent from power stations. Of UK emissions from fossil fuel
combustion, 79 percent came from coal combustion and 12 percent from fuel oil.2 Sulphur
dioxide pollutes the environment through two different routes. Much of the SO2 gas falls to
earth within 300 km of the source in its dry form, and this process is described as dry deposition.
Long range transport occurs because SO2 is oxidised to sulphate particles, which are not readily
deposited in dry form. Their main removal is by scavenging in rain-making processes as wet
deposition, which may occur 1000-2000 km from the source. Long term monitoring of rain
chemistry has been carried out since 1978 only at between 6-12 sites throughout the UK. During
the period 1986-89 a rather dense network of up to 60 sites was operated to establish the spatial
pattern of rain chemistry across the UK. The basis spatial pattern has changed little from year

1 +
pH is defined as the negative logarithm of the hydrogen ion (H ) concentration, having the perverse effect that
lower numbers correspond to higher acidity. Thus 0 is the most acid, 7 is neutral and 14 is most alkaline. Lemon
juice has a pH of 2, milk of magnesia 10.5.

2
UK data are taken from the Digest of Environmental Protection and Water Statistics, 1988, (Department of
the Environment, 1989). Figures after 1970 are based on revised emission factors described therein.
 4

to year during this period and so 30 sites have been retained for longer term monitoring.3 Wet
deposition can be reported in two ways—by its intensity and cumulative deposition. Intensity
is shown by the maps of the average acidity of precipitation (in pH), and cumulative deposition
by wet deposited acidity in grams of hydrogen ions per square metre per year. Deposited acidity
is the product of the acidity of the rainfall and the amount of rain—wetter areas in the west may
have more acid deposited even though the precipitation is less acidic.

2.2 Measuring acid rain

The European Monitoring and Evaluation Programme (EMEP) was set up in 1978 to monitor
the movement of pollutants, and to determine where the deposition of pollutants released from
each source occur. Until recently the only pollutant tracked was SO2, though now NOx is also
monitored. The surface of Europe is divided into squares with grid lines 150 km apart. There
are about 720 grid line intersections on land and about 100 monitoring sites which are used in
the EMEP model and these are termed arrival points. Using detailed meteorological information,
the track of air which arrived at each of the 820 or so points is followed backwards in time for
96 hours. An air parcel is then studied forwards in time as it follows each back-track precisely,
picking up pollution and depositing pollution until it arrives back at its arrival point. This whole
procedure for each of the 820 points is repeated at six-hourly intervals 365 days of the year. The
model also keeps a record of the pollution produced by each country. Not all the deposition can
be traced back to an identified source, as meteorological data is only accurate enough to track
back for 96 hours. Table 1 gives a subset of the basic data from this exercise for 1987, and is
to be read as follows.4 Britain (GB) appears as a receiver in the left hand column and as an
emitter in the top row. Looking along the row against GB the table shows that Britain received
14,000 tonnes of sulphur5 (ie about 27,000 tons SO2) from France, 11,000 tonnes from West
Germany (DE), and 571,000 tonnes from domestic sources. Looking down the column headed
GB the Table shows that Britain emitted 1,271,000 tonnes sulphur whose final destination could
be established, and of this 43,000 tonnes fell on France, 45,000 tonnes on Germany, and 437,000
tonnes on North Africa within the monitoring area (demonstrating just how far the plume can
travel).
The large numbers on the diagonal of Table 1 shows how important domestic sources of
pollution are. The large off-diagonal numbers indicate where the major impacts of one country
on another occur, and it is striking that they primarily occur in East Europe, confirming the view
that central planning has been an environmental disaster for its participants.
Table 2 presents the information from the same programme in a different way. The first
two columns give total emissions (not just those whose final destination can be identified) for

3
Personal communication from R Derwent.
4
The fuller table is given in the Appendix, and is the source of the calculations reported below. Using the
abbreviated table leads to considerable biases in estimating total damages and the complete table should be used for
all calculations. An earlier table for 1980 is given in Environmental Resources, Limited, (1983), p40.
5
To convert from sulphur to sulphur dioxide, multiply by 1.9, or roughly double the numbers.
 5

the base year 1980 and the most recent year available, in order of magnitude. This allows an
estimate of the extent to which countries have succeeded in
Table 1 Origins of Sulphur Deposition in Europe

Thousand tonnes S/year

Emitters
Receivers CS FR DD DE BL HU IT PL ES SC SU GB OE UI Sum

Czechoslovakia CS 385 11 128 28 5 45 10 95 1 0 2 7 13 28 765

France FR 19 332 41 40 28 5 21 15 65 0 0 43 8 140 760
GDR DD 84 14 725 61 11 2 2 32 1 0 1 15 3 24 979
West Germany DE 47 69 163 330 44 3 13 23 6 0 1 45 6 64 821
Benelux BL 4 32 15 51 102 0 0 4 2 0 0 31 0 19 267
Hungary HU 31 3 16 6 1 190 12 25 0 0 1 1 28 18 337
Italy IT 13 21 15 8 2 11 353 14 10 0 1 4 19 86 562
Poland PL 145 15 310 47 10 40 10 790 1 1 18 15 21 64 1492
Spain ES 2 11 5 3 1 2 2 3 523 0 0 6 16 98 674
Scandinavia SC 17 5 48 18 6 4 2 44 0 59 33 32 30 194 501
Soviet Union* SU 107 10 167 36 8 84 13 337 1 8 2204 16 97 491 3584
Britain GB 5 14 15 11 8 0 1 3 2 0 1 571 6 60 702
Other Europe OE 95 40 97 49 8 141 136 101 29 3 95 70 825 435 2163
N Africa NA 105 136 253 131 71 64 182 194 210 28 196 437 254 821 3087
Sum Σ 1064 721 2005 823 322 594 759 1685 856 107 2558 1271 1377 2553 16695
Error 5 8 7 4 17 3 2 5 5 8 5 -22 51 11 1

Source: Acid Magazine Sep 1989, from EMEP data.

Note: Sulphur dioxide figures will be about twice as large. Total wet plus dry deposition of sulphur for period
11.12.87 to 6.11.88, sums may not add due to rounding errors (shown).
* Norway plus Sweden
** European part of USSR within EMEP area of calculation.
*** Deposition in North Africa within areas of calculation
UI = unattributable to any country, plus small amount from NA

moving towards the target 30% reduction now widely accepted. The next two columns give
depositions within the country, and an estimate of the fraction of depositions which can be traced
back to domestic sources (using Table 1 data). This confirms the importance of the diagonal
element in Table 1, and the importance of domestic sources of pollution. The final column gives
the ratio of exports of SO2 to imports. As total depositions in 1987 are only 78% of total
emissions, this ratio can be expected to be significantly above unity on average. The smaller it
is, the more the country is sinned against, than sinner. Britain stands out as the greatest sinner
on this criterion, and the Scandinavian countries as those most sinned against.
 6

Table 2 Sulphur Emissions

Thousand tonnes or per cent

Emissions Change Depositions Own contrib Exp/Imp

1980 1987 % 1980 1987 tot. depos

Soviet Union* 6400 5100 -20 5101 3584 0.61 2.1

GDR 2500 2500 0 963 979 0.74 7.0
Poland 2050 2270 +10 1443 1492 0.53 2.1
United Kingdom 2335 1840 -21 803 702 0.81 9.7
Spain 1625 1581 -3 670 674 0.74 7.0
Czechoslovakia 1550 1450 -6 818 765 0.50 2.8
Italy 1900 1252 -34 916 562 0.63 4.3
German Fed Rep 1600 1022 -36 1083 821 0.40 1.4
France 1779 923 -48 1160 760 0.44 1.4
Hungary 817 710 -13 416 337 0.56 3.5
Yugoslavia 588 588 0 662 497 0.39 5.0
Bulgaria 517 570 +9 293 235 0.65 5.0
Belgium 400 244 -39 162 121 0.41 2.7
Greece 200 180 -10 150 119 0.38 1.8
Turkey 138 177 +22 209 210 0.29 2.7
Finland 292 162 -44 273 210 0.23 0.7
Denmark 219 155 -13 110 83 0.37 2.4
Netherlands 244 141 -42 175 139 0.23 1.0
Portugal 133 116 -13 83 83 0.42 1.7
Sweden 232 116 -50 333 307 0.12 0.3
Romania 100 100 0 405 330 0.10 0.2
Irish Republic 110 84 -24 66 68 0.31 1.3
Austria 177 75 -58 282 207 0.09 0.3
Norway 70 50 -29 199 194 0.07 0.2
Switzerland 63 31 -51 121 70 0.11 0.4
Total 26078 21471 -18 20484 16695 0.52

Source: Acid Magazine Sep 1989, from EMEP data.

Note: Sulphur dioxide figures will be twice as large. Countries ordered by 1987 emissions.
* European part of USSR within EMEP area of calculation.
Figures for emissions in 1987 based on interpolation except for USSR, UK, Czecholslovakia, Finland and Netherlands.

Only part of total SO2 emissions come from man-made sources; other important sources
include volcanoes, biological decay and forest fires. These natural sources might account for
80 - 290 million tonnes per year world-wide, compared to total man-made emissions of 75 - 100
million tonnes. The levels and mechanisms responsible for natural emissions are imperfectly
understood, but they may play an important role in the European acid rain problem.
The information generated by EMEP is remarkably useful, not only in quantifying the
level of pollution, but also identifying efficient and feasible abatement policies. The information
on deposition can be used to draw maps showing the average acididy of precipitation over
Europe using contour lines of increasing levels of acidity. Such maps show that in 1987 most
of Yorkshire and the East Midlands had precipitation of average acidity below pH 4.3, (ie 5
 7

times as acid as ‘normal’ rain with a pH of 5.0) whereas Wales, South-west England and the
west coast of Scotland was above 4.6, and so less acid. Substantial areas north of a line joining
the Wash and Liverpool received more than 0.05 grams H+ per square metre by wet deposition
(ie 0.5 kg per hectare or 50 kg per square km), with Wessex, East Wales, Northern Ireland, and
North-east Scotland receiving less than 0.02 grams. Most of the sulphur deposition in the UK
occurs through dry deposition as SO2, particularly in the south and east of the country. In the
north and west where rainfall is more frequent and intense, wet deposition of SO2 and sulphate
particles becomes more significant.
In a European context, the lines of equal rainfall acidity show the highest concentrations
in Germany and Poland, with pH below 4.1, with most of Southern France, almost all Italy,
Spain, Portugal, West Yugoslavia and West Greece having a pH greater than 4.9 (ie less acid
than any part of the UK and Eire).
The EMEP tables can also be used to throw light on the political economy of pollution
control. Consider first the column in Table 2 which gives the fraction of total deposition which
can be attributed to domestic sources. The unweighted average of these figures is 41% (with a
standard deviation of 21%). The weighted average (heavily influenced by the larger countries,
especially the USSR, which, by their size, have higher domestic absorption) is 52%. The
weighted average domestic absorption as a fraction of total production is 33%, and the
unweighed average is 29% (with a standard deviation of only 5%). What this means is that the
average unilateral cost of reducing a tonne of domestic deposition is equal to the cost of reducing
domestic emissions by about 3 tonnes. The ratio of total European depositions to total European
emissions is 66%, so that if all European countries acted in concert, the cost of reducing
depositions by one tonne would be only half as great. Put another way, many countries could
reduce depositions within their borders by about 50%, but at twice the cost per tonne reduced
as if all countries acted together. There are thus considerable benefits to coordinated action, but
these should not be dramatised—SO2 pollution is far from a pure public good at the country
level, and self interest ought to go a considerable way towards alleviating the problem.
The next question one can ask of the EMEP data is whether there are significant
opportunities for bilateral bargaining between pairs of countries over pollution levels. One way
to identify such is to look for instances where the volume of bilateral pollution trade is large
relative to total depositions, and where trade is bilateral rather than unilateral. The volume can
be measured by one-half exports plus imports, and bilateralism can be measured by the
difference between exports and imports. Table A1 of the Appendix gives the net exports of each
country and can be used to identify the extent of bilateralism. The following country pairs have
a difference between these two measures of 5% of depositions or less for the smaller of the
partners: Czechoslovakia-GDR; Czechoslovakia-Hungary; Czechoslovakia-Poland; GDR-
Poland; Poland-Hungary; Soviet Union-Czechoslovakia; Soviet Union-GDR; Soviet Union-
Hungary; Soviet Union-Poland. It is notable that significant balanced trade in pollution is
confined to Eastern Europe, and does not affect any of the other countries identified in Table 2.
(It may be substantial for smaller countries as a proportion of their deposition.)
Another possible question to ask is which pairs of countries have large net trade balances
in pollution which might lead to financial negotiations over pollution levels. The following
countries have net imports from another country which are greater than 5% of total depositions:
 8

Poland from GDR (19%); Denmark from GDR (12%); Scandinavia from GDR (10%);
Scandinavia from Poland (9%); Soviet Union from Poland (9%); Czechoslovakia from GDR
(6%); Scandinavia from Soviet Union (5%). Scandinavia is thus the only West European
country which receives large net imports from single countries—otherwise it is the Eastern bloc
countries that stand out as large net importers from each other.

Nitrogen oxide emissions Nitrogen oxide, or NOx, emissions are measured in terms of tonnes
nitrogen dioxide equivalent, NO2. Table 3 gives per capita emission levels for both SO2 and
NOx for the major member countries of the UN Economic Commission for Europe Convention
on Long-range Transboundary Air Pollution. It shows that the UK is not high in comparison
with Europe and North America taken together, but is rather higher than the West European
countries in the sample. The table shows that whereas SO2 has decreased between 1980 and
1985, NOx has if anything increased.
Table 4 gives further information about NOx for 1985. The first column shows that
mobile sources contribute about one-half of all NOx emissions in Western Europe (actually
OECD Europe), though the range is from 28 per cent in Ireland to 84 per cent in Norway. Much
of the rest comes from large combustion plants—thus in Britain 35 per cent of the total came
from power stations. The next column shows total emissions of NOx from mobile and stationary
sources in relation to total energy use. The coefficient of variation (CV) is 34 per cent, showing
that emissions correlate quite closely with energy consumption, but there are important
variations in the degree to which energy use causes NOx pollution. Britain does poorly by this
score, almost as badly as Portugal and Greece. (Figures for Luxembourg seem rather low and
may be explained by some energy sales, especially of transport fuel, being consumed abroad.)
Column (3) shows mobile emissions of NOx per unit of GDP (which has a lower CV than total
emissions per unit of GDP). Column (4) gives mobile emissions in gm NO2 per km driven by
cars. White (1982, Table 2) shows that if there were no emissions regulations, then for the US
64 percent of total NOx emissions would come from cars, and the balance of 36 per cent from
trucks. Uncontrolled emissions are 5.44 gm/km for cars, and 38.6 gm/km from heavy diesel
trucks. The expected uncontrolled emissions per km driven by cars alone might therefore be 8.5
gm/km (ie 5.44/0.64) if the proportion of car km in total vehicle km were the same as the US.
The European average is 3.15 or only 37 per cent of that predicted for uncontrolled emissions.
Perhaps more impressive, if the total emissions from all mobile sources are attributed entirely
to cars, then the average achieved is about as good as those achieved in the US by cars of later
than 1976 model year (White, 1982, Table 10). It is interesting to note that Britain does quite
well by comparison with other countries—at least as far as emissions from mobile sources are
concerned. The last two columns relate mobile emissions to two transport fuels. Column (5)
gives mobile emissions in kg per tonne of gasoline, again showing Britain in a favourable light.
The final column gives total mobile emissions divided by total fuel consumed in the transport
sector, and thus accounts for diesel emissions, which are potentially quite serious.
 9

Table 3 Emissions per head

Kilograms/head/year

SO2 NOx
1980 1985 1980 1985

United Kingdom 85 65 35 33

Austria 47 18 29 28
Czechoslovakia 202 203 78 73
France 65 31 34 29
Germany Fed Rep 52 43 49 49
Greece 83 73 13 15
Irish Republic 63 39 21 19
Netherlands 28 25 35 34
Poland 115 116 5 18
Spain 87 75 21 24
Sweden 64 33 38 37
Switzerland 19 15 31 33

Canada 193 150 72 72

United States 102 90 90 83

Averages: All 86 70 39 39
Europe 76 61 32 32
West Europe 59 42 31 30
Ratios:
UK/sample 0.99 0.93 0.89 0.84
UK/Europe 1.12 1.06 1.08 1.01
UK/W Europe 1.43 1.56 1.14 1.10

Source: UNECE (1987) National strategies and policies for air pollution abatement

Note: Averages are unweighted.

Nitrogen oxides also come from natural as well as man-made sources. Again estimates
are very imprecise, but natural sources may account for 20 - 90 million tonnes compared to
estimated total man-made emissions of about 90 million tonnes. One might therefore argue that
perhaps only half the total acid rain emissions are man-made, but whereas natural emissions are
world wide, man-made sources are concentrated in the northern hemisphere, and specifically in
Europe and North America.
 10

Table 4 NOx emissions from mobile sources

Mobile Total Mobile

Country total/ kgNOx/ NOx/ gm NOx/ NOx/ NOx/
% toe gdp km car gas fuel

(1) (2) (3) (4) (5) (6)

Austria 68 10.1 1.27 3.76 61 37

Belgium 55 9.0 0.84 2.73 44 22
Denmark 34 15.9 1.04 2.48 61 35
Finland 58 14.8 1.80 7.62 86 47
France 66 13.1 1.29 2.81 62 38
Germany 59 14.6 1.54 3.73 70 46
Greece 63 18.6 3.35 4.36 75 44
Ireland 28 10.7 1.08 22 14
Italy 51 14.8 1.06 2.03 70 34
Luxembourg 64 6.9 2.32 5.61 45 28
Netherlands 60 10.5 1.53 3.27 95 56
Norway 84 12.3 2.27 9.65 113 81
Portugal 38 19.6 3.38 4.13 123 54
Spain 46 18.5 1.49 4.50 67 36
Sweden 68 9.2 1.48 3.26 52 37
Switzerland 74 11.1 0.92 3.04 49 41
UK 45 16.4 1.65 2.96 44 32
Europe 49 14.1 1.42 3.15 63 39

SD unweighted 14 4.76 0.83 2.01 25 14

CV 0.29 0.34 0.58 0.63 0.40 0.37

Sources: UNECE (1987) and OECD; emissions for 1985 or latest available year; other figures for 1985 or
1986.

Notes: Col (1): NOx Emissions from mobile sources divided by total emissions NOx; Col (2) total emissions
divided by total energy consumption in tonnes oil equivalent (toe); Cols (3-6): Emissions from mobile
sources divided by: GDP; km driven by cars; kg gasoline; kg total road transport fuel.
SD is standard deviation; CV is coefficient of variation.

2.3 Assessing the damage caused by acid rain

Acid rain has ecological consequences in that it affects the soil, vegetation, especially forests,
lakes (and hence fish). It causes economic damage to man-made structures (buildings, fabrics,
metals), and it can affect human health. The ecological consequences are complex and still
subject to scientific uncertainty and hence dispute. Soils vary widely in their ability to buffer
(ie neutralise) acid rain, and natural processes add to the man-made sources of acid rain. Recent
work undertaken for the UNECE Convention on Long-range Transboundary Air Pollution6
attempts to establish critical loads for various kinds of soils, which, if exceeded, would mean that
the soil could no longer neutralize additional acid rain depositions.7 Many sensitive areas

6
and reported in Acid News No 3, October 1988
7 + 2
Two measures are given for these critical loads - acid input in keq H ions/km /year and sulphur deposition
+
in kg S/ha/year. (It appears that 1 keq H is 1 kg.) For very slow weathering rocks (Class 1, which have poor
+
buffering ability and hence low critical loads, the figures are < 20 keq H or < 3 kg S/ha; for moderate weathering
 11

especially in Scandinavia experience both high rates of deposition and soils for which critical
loads are low.
The effects of acid runoff on lakes have been intensively studied in Scandinavia and the
UK (and doubtless elsewhere). One of the main mechanisms leading to the decline in fish stocks
is the release of aluminium caused by acidification, rather than the direct effects of acid (see eg
Environmental Resources, Limited, 1983). Palaeoecological studies of core samples can trace
back acidity levels into the distant past and show substantial falls in pH in many lakes after the
industrial revolution. Thus Battarbee et al (1988) analyzed lake acidification in sensitive areas
in the UK and found that before 1850 most lakes studied had pH levels of about 6.0, but that
since then pH values had declined by 0.5 - 1.5, (ie acidity had increased by between 3 and 30
times) depending on deposition rates and buffering capacities. Other studies show similar trends,
and also show that lake acidification may be reversed (even if temporarily) by the addition of
3
lime either to the lake or the rivers in the cachement area. This is expensive, as about 5g/m of
bicarbonate (usually in the form of limestone) are required to raise the pH from 4.5 to 6.5, and
it does not by itself restore the lake to its original condition—restocking may also be required.
(See eg Dudley, et al 1985; Britt, 1986.)
One implication which does not seem to have been adequately emphasised is that acid
rain is a stock pollutant as well as a flow pollutant. That is, part of the final damage caused will
depend on stocks of acid in the environment, not just the rate at which acid rain is deposited.
Even if the environment is capable of neutralising or disposing of some of the acid each year,
if inflows exceed this rate of disposal, then the stock of acid will increase. In the UK it is
believed that current levels of soil acidification are a legacy of the Industrial Revolution, and that
water quality will not be restored until the soil recovers. This recovery is a slow process and
relatively insensitive to near-term rates of emission reduction, requiring liming for rapid
recovery.8 This may go some way to explaining the paradoxical relationship between
decreasing levels of SO2 emissions on the one hand, and apparently deteriorating ecological
conditions on the other. On the other hand, Battarbee et al (1988) note that acid deposition has
been declining in Scotland over the past 15 years, and that the uppermost sediments are already
recording an improvement, which suggests a possibly swift improvement if deposition levels
could be further reduced. The speed of response will presumably depend on the ecological
circumstances, but may imply high rates of ‘depreciation’ of acid stock levels.
Lake and fish damage appears quite well understood compared to the damage suffered
by trees. The problem was highlighted in Germany in the early 1980s, and shown to occur
elsewhere. Forest damage has been attributed to acid rain, weather changes and droughts, the
age of trees, fragility of soils at high altitudes, and inappropriate forest management. Ozone
attack appears to be important, and may have synergistic interactions with acid rain
(Environmental Resources, Limited, 1983). Even if the exact proportion of damage attributable

capacity soil (Class 3) the figures are 50-100 keq and 8 - 16 kg respectively, and for very rapid weathering soils in
Class 5 > 200 keq and > 32 kg respectively. Currently deposition rates in parts of West Germany are in the range
50 - 80 kg S/ha/yr and in the worst affected parts of Europe the figures exceed 100 kg. 1 kg of sulphur is about 2
kg of SO2.
8
Personal communication from David Pearce
 12

to acid rain is not known, there seems widespread agreement that reductions in acid rain would
be beneficial to forests. Similar uncertainty pervades the study of crop damage, though again
ozone appears to be more directly harmful than acid rain. To the extent that ozone plays a major
role in crop and forest damage, NOx, which is a major contributor to ozone production, is more
damaging than SO2.
Damage to buildings and materials occurs primarily in urban areas as a consequence of
relatively high concentrations of SO2, with little effect detected from exposure to NOx. The
effects have been observed and correctly attributed for centuries, and the estimated damage costs
are thought to be high.9
Health effects of intense pollution can be dramatic—it is estimated that 4,000 people died
in the great London smog of December 1952. Similar levels of SO2 concentration were attained
in a subsequent episode in London from 3 - 7 December 1962, after the Clean Air Act of 1956
had lead to a dramatic fall in smoke concentrations. This time an estimated 340 died, suggesting
that the earlier episode was so deadly because of synergistic interactions between smoke
particles and SO2 (Park, 1987, p127). It appears that it is the gas SO2 that is harmful, rather than
the wet form of acid rain. Acid rain in its wet form can have indirect effects by releasing toxic
heavy metals into water supplies. Individuals vary considerably in their tolerance to these gases,
but there is some evidence from epidemiological studies that long-term exposure at lower levels
than these dramatic episodes can be harmful to health (Park, 1987, p127. See also Pearce and
Markandya, 1989 for a summary of the extensive economics literature on the health impacts of
SO2.)
If SO2 and particulates are lethal, the health case against NOx is at best unproven, for it
appears that NOx is much less active biologically. There is a certain irony in the fact that the
impetus to reducing automobile emissions initially came from California, where it was
suspected, and later established that photochemical smog was caused by vehicle exhaust. The
case mounted by the US Environmental Protection Agency for reducing emissions was based
on the supposed adverse health effects of high concentrations of ozone, though subsequent
studies (Lave, 1982; White, 1981; 1982) cast considerable doubt on the evidence. To quote
White (1981, p59-60): ‘. . the ozone-related health effects under discussion were short term and
reversible. .. Thus far, ozone exposure has not been demonstrated to have long-term debilitating
consequences in humans. ... the contrast with other studies of pollutants, such as particulates and
sulphates, was striking. .. Particulates and sulphates probably killed; ozone appeared to do little
more than cause coughing!’

2.4 Measuring the costs of acid rain damage

Estimates for the costs of different types of damage are scattered in the literature, and vary
greatly in their reliability. Pearce and Markandya (1989) provide a useful methodological
discussion of cost-benefit analysis applied to environmental pollution, summarise a variety of
these estimates, and note the criticisms to which they are vulnerable. One approach familiar to
economists is to ask what people would be willing to pay for property located in less rather than

9
Exactly how high the damage costs might be is currently the subject of active research, especially in Germany,
but at the time of writing the buildings component of damage cost estimation has not reported.
 13

more polluted air, as reflected in the response of property values to pollution levels. Most of the
estimates here come from the 1960s and 1970s, and suggest that for each 1% increase in SO2
concentration, property values fall by between 0.06 - 0.15 of 1% of their value for a house of
average value.
An alternative approach is to look at the direct economic costs caused by the acid rain,
and this has been done by for the Netherlands for 1986. It was estimated that current costs were
about $53 - 175 million per year,10 but if the costs of dealing with future damage were taken
into account (loss of timber etc) this might rise to $120 - 380 million. Looking at the current
costs, the large proportion of the total comes from agricultural damage, thus extra liming of the
soil to counteract acidification might cost $18-60 m, and falls in crop yield might be $36 - 360
m. One should of course be most wary about estimating the value of lost agricultural output
given the distortions of the CAP. Indeed, as an aside, agriculture is responsible for considerable
ground water pollution (notably nitrates, and possibly algal blooms). Much of this is in turn the
consequence of intensive agricultural practices induced by the high agricultural prices enjoyed
under the CAP, notably high fertiliser levels. If for various reasons it is difficult to reform
agricultural output price levels, then there is a strong case for raising agricultural input price
levels to the same ratio to world prices as output prices enjoy. This would improve the
efficiency of resource use and reduce the deadweight losses associated with the CAP. Thus if
output prices are twice import parity levels, then fertiliser prices should be taxed to raise their
price to twice world market levels. This would go some way to reducing another form of
environmental pollution. (See Newbery, 1989, for the details of the arguments on efficient input
taxation.)
Other ecological damage estimates are given in Environmental Resources, Limited,
(1983). German forest damage was put at $0.25 billion p.a., and rough estimates of potential
EEC wide damage can be deduced from the annual value of spruce and fir forestry production
of $6.6 billion p.a. Thus if 20% of forests are adversely affected so that their production drops
by 10%, the loss is $0.13 billion p.a.11 OECD (1981) estimated that the value of fish loss in
Scandinavia was $38 million p.a., and in Scotland might be $0.7 m p.a.
Acid rain causes material damage, whose costs, excluding the costs of restoring historic
buildings, have been estimated by UNECE (1982) at between $4 - 17 per head (ie $1.0 - $4.7
billion for the 1983 European Community as a whole). The figures in the Netherlands are
somewhat higher ($10-19/head) and in Germany are estimated at $19/head. Environmental
Resources Limited, (1983) gives figures from OECD (1981) for the estimated total corrosion
damage for 12 OECD European countries to galvanised steel and its paint coatings in 1974. For
the UK the figures were $5.9 billion p.a., for W Germany $9.5 billion p.a., and for Belgium,
Luxembourg, Denmark, France and the Netherlands together $3.3 billion, or in total $18.7
billion. How much of this can be attributed to acid rain is still under study.

10
Unless specifically stated to the contrary, all cost estimates have been expressed in mid-1989 purchasing
power, updating from the US CPI.
11
But note that there is still, considerable disagreement as to how much forest damage has occurred, and how
much is caused by acid rain.
 14

Reducing car emissions would also reduce photochemical smog in some areas—-
particularly those which experience temperature inversions combined with strong sun. Los
Angeles is the leading example, but clearly Athens suffers similarly. There is no doubt that those
living in such areas would be willing to pay for reductions in smog levels, and Schechter et al.
(1989) estimate that households in metropolitan Haifa, Israel, would be willing to pay £12 (1987
£) per household per year to reduce pollution levels by 50 per cent. It is difficult to imagine that
this would amount to a large total sum for Europe as a whole compared with the other damage
costs, given the relative infrequency of photochemical smog in more Northerly climes.

2.5 The costs of abatement

The Department of Environment estimated the costs to the UK of retrofitting 6GW of coal fired
plant with Flue Gas Desulphurisation (FGD) and all 12 major coal fired power stations (23GW)
with low NOx burners at over £1 billion. It now seems doubtful that more than a small part of
this programme will go ahead, as the liability to install FGD would make the privatisation sale
of the CEGB unattractive. Instead it appears that the successor companies to the CEGB,
Powergen and National Power, will meet the emissions standards by a combination of installing
high efficiency gas turbines and importing low sulphur coal. The impact of this on British Coal
will be substantial, and it is in interesting example of how the (private) cost-minimising solution
to the emissions standards may differ from centrally imposed solutions. The Government also
intends to apply the EC large car emission standards ‘as soon as practical, probably in the early
1990s,’ at an estimated annual cost of £550 million. The second stage, applying to small cars,
was estimated to add an additional £250 million per year, a total of about 4 per cent of UK
motoring costs. (Department of Environment, 1988, 7.14-15.) This section examines various
estimates of the costs of abatement in somewhat more detail to see if mandating particular
solutions is likely to be cost effective, and to check on the consistency and plausibility of various
estimates. The results are summarised in Table 5 and then briefly explained.
The sources of these estimates are as follows. Environmental Resources Limited (1983)
gives estimates of the capital cost of FGD at $175-200/kW or about 15-20% of the capital cost
of the plant. Retrofitting, where practical, may increase this cost by a further 30-50%. FGD
reduces thermal efficiency by about 2% (eg from 36% to 34.1%) and so can increase the
operating costs by 10-20%. The first station which the CEGB plans to retrofit is Drax A+B,
which has a total capacity of 4000 MW, and which burns 11 million tonnes of coal per year with
a sulphur content of 1.7%. If 95% of this were previously released as SO2, the annual emissions
would be 178,000 tonnes S, or 338,000 tonnes SO2. After fitting FGD, 90% of the SO2 will be
removed, and the reduction in SO2 would be 287,000 tonnes. The Layfield report gives the costs
of FGD as £17/kw/year, broken down into £5 capital, £2.5 operating and £9.5 for loss of thermal
efficiency. This last figure depends on the cost of coal which has since fallen and Jeffrey (1988)
estimates the efficiency cost as £6.5. Using these figures the annualized cost of the Drax
programme would be £56m, or about £200/tonne SO2 reduction. Longhurst et al (1987) gives
the CEGB’s estimates of the cost of retrofitting FGD for a 2000MW plant as £160 million plus
£35 million is lost output, or £440-£740/tonne of sulphur removed from the gas stream (ie £230-
390/tonne SO2 removed). Dudley et al (1985, p121) try to cost a programme to achieve a 60%
reduction in SO2 in UK emissions from its 40 GW coal-fired capacity using CEGB data and data
 15

presented at the Layfield enquiry. The levelized lifetime cost of sulphur abated ranges from
£442/tonne to £738 with an average value of £550, all in £1983. (The average figure is thus
£745 at £1989, or $1200/tonne S. The costs per tonne SO2 removed would be about half this.)
Brackley (1987) estimates the cost of SO2 removal using FGD as $1,150 /tonne S removed with
90 per cent removal, rising to $3,000 /tonne S for the next 5 per cent removed in going from 90
to 95 percent removal, in both cases using coal with 1% sulphur content. This figure is very
close to Dudley’s estimate.
The effect on the cost of electricity generation would be about 10 - 15 per cent of the cost
of generation from coal-fired power stations, or possibly 6 percent of the price paid by
customers.12 This can usefully be compared with the predicted size of the ‘nuclear levy’ of 11
per cent of the sales price, which will be paid by consumers of fossil-fuel generated electricity
after the privatisation of the CEGB to cover the cost of supplying 20 per cent of total electricity
by non-conventional (mainly nuclear) means.13
Dowlatabadi and Harrington (1989), in a rather critical account of US estimates of the
costs of large programmes to reduce total emissions by 8 million tonnes per year from the
current levels of 25 million tonnes, cites various estimates of the average cost per ton SO2
reduction. Thus the Congressional Budget Office (1986) gives the least cost method of making
this reduction (by allowing the utilities to choose how best to meet the standards) as $360/ton,
and $400/ton if they must continue to use the same coal as originally instead of substituting to
lower sulphur content coal. These are average costs, and there is considerable agreement that
the marginal costs of high levels of sulphur removal are substantially higher than small
reductions. If the marginal cost were twice the average cost at this programme level then the US
figures would be close to the CEGB figures. Estimates based on German data applied to East
Germany suggest that the capital cost per kw capacity are DM600 using the Wellman-Lord
method.14 At present the capacity of the 15 larger plants is 13.3 GW which is responsible for
1.98 million tons of SO2 per annum. The capital cost would be DM 8 billion, and the estimated
annual cost of operating less the value of the sulphur sold on world markets would be DM 1.6b,
or DM 600 per ton SO2 removed, or $300 per ton SO2, which appears rather low.

12
Based on estimated capital costs of £100-£200/Kwe capacity, 5 - 7% discount rate, 30-40 years lifetime, 70%
load factor, and 1.7% sulphur content coal (ie for Drax). The calculated figure is reassuringly close to that given
in 1984 by the CEGB in HMSO (1984, para 5.93).
13
This suggests that the cost of non-conventional alternative is 44 per cent of the final price (ie 11% borne by
80% of the total allocated to the 20% non-conventional), which seems unreasonably high. But detailed estimates
of the cost of nuclear power are not yet available.

14
See Acid News, No 3, July 1989, p9.
 16

Table 5 Estimates of costs of reducing SO2 by various means

Source Action Cost 1989$/tonne

SO2 removed

A moving to low sulphur gas oil $2,560

A moving to low sulphur fuel oil 640
A Fluidised bed combustors (FBC):
new boilers 96
existing 2,240
A Flue gas desulphurisation (FGD):
new plant 256
existing plant 640

B Drax FGD new plant 350

C FGD retrofit 2000 MW plant 400-750
D 60% reduction from 40 GW CEGB coal capacity 600

E FGD 90% removal 600

E FGD marginal cost of next 5% removal 1,600

F US coal generators, coal switching, Av cost 400

F US coal generators, no coal switching, Av cost 460

G East Germany, Wellman-Lord FGD, net of S sales 300

H move from 2.15% to 1% sulphur heavy fuel oil 380

H move from 1% to 0.7% sulphur heavy fuel oil 825

Sources: A Environmental Resources Limited (1983, p137, uprated by 1.28 to $1989).

B Based on Layfield (1987) and Jeffrey (1988).
C Longhurst et al (1987).
D Dudley et al (1985, p121).
E Brackley (1987).
F Congressional Budget Office (1986) in Dowlatabadi and Harrington (1989). These are
average costs. Marginal cost might be twice average cost.
G Acid News, No 3, July 1989, p9.
H Alfsen et al (1986).

An alternative option for reducing SO2 emissions is to switch to lower sulphur content
fuels. Thus Alfsen et al (1986) calculate the cost of switching from high to low sulphur heavy
fuel oil as 2,300 NOK/ton SO2 removed (ie $380) when moving from 2.15% to 1% HFO, and
5000 NOK/ton removed (= $823) when moving from 1% to 0.7% HFO. Table 5 shows similar
calculations of switching from high to low sulphur fuels for the EEC given in Environmental
Resources Limited (1983, p137).
What stands out from Table 5 is the wide variation in the costs of the most common
proposed method of dealing with large power stations—flue gas desulphurisation, or FGD. In
part this variation may be explained by differing degrees of sulphur removal—estimate E shows
 17

that the marginal cost is sharply increasing. The East German estimates may be based on lower
construction costs or a more optimistic view of the value of recovered sulphur. The figures for
FGD from source A seem rather low when compared to other estimates. Unfortunately there are
no European estimates for the important option of shifting to low sulphur coal. In part this is
because, unlike oil, there is no clearly defined world market price for the two grades of coal that
would allow a robust estimate to be made of the differential cost of shifting from high to low
sulphur coal.
The study by Environmental Resources Limited (1983) concludes that the estimated total
damage caused by acid rain might be in the range $0.6-4.5 billion per year, of which the larger
part is damage to buildings, then to forests, then to crops, with fisheries negligible. The costs
of reducing SO2 emissions in the year 2000 by 10-27% by adopting FGD and FBC in new
boilers might be $1.9-6.5 billion per year, depending on the amount of new capacity installed.
The cost of retrofitting might be $9 billion to achieve a further 38-42% reduction. The cost of
new FGD might be to raise the cost of generating electricity from coal by 10-15 per cent, and
the effect of retrofitting existing stations might be to raise generating costs by nearly twice that.
Since the price of electricity would be determined by the marginal cost of new plant, the effect
of forcing older plant to adopt FGD would be to decrease their profitability and equivalently, to
write down their capital value sharply. Once installed, though, their operational life should be
the same as new plant. New gas-fired generating capacity will become relatively more attractive
than coal-fired capacity, as low sulphur gas is more readily available. Recent developments
which have greatly raised the thermal efficiency of gas turbines makes these cost-effective for
base load at current European gas and coal prices, and may offer a lower cost alternative way of
reducing acid rain emissions. The combined impact of more stringent emissions standards,
movements towards a single market in energy with the consequential reductions in Government
subsidies and protectionist policies, and the new gas turbine technology may sound the death
knell for the European coal industry.

2.5.1. Costs of reducing NOx emissions Estimates of the cost of reducing NOx emissions are
somewhat harder to find, partly because most studies have concentrated on emissions from cars,
where other pollutants were also being reduced, and partly because the technology for removing
NOx from stationary sources is not yet commercially proven. There are extensive studies of the
cost of reducing NOx emissions from vehicles by the use of catalytic converters and engine
modifications, mostly for the US. It is possible to use the most recent to calculate costs per
tonne NOx removed. Using the information supplied by Crandall et al (1986) it appears that the
costs of introducing the 1975 standards compared to no regulation might be $5,500/tonne
removed, and the additional costs of introducing the 1981 standards starting from the 1975
standard would be $9,000/tonne removed—illustrating the rising marginal cost of abatement as
emissions standards are tightened.15 Crandall et al (1986, p114-5) estimate that the programme
costs for the US of the more stringent 1984 standards might be about $20 billion per year with
a replacement rate of 10.5 million cars, which is several times rather optimistic estimates of the

15
These are based on a vehicle lifetime of 12 years, annual distance driven of 8,000 miles, and a real discount
rate of 3%. Higher discount rates increase the cost.
 18

potential benefits of reducing pollution. (Safety regulations in contrast, though expensive, seem
to have been justified on cost-benefit criteria.) Another way to put the costs of emissions into
perspective is to note that the capital cost of emissions equipment in 1981 was 6.7 per cent of
the new car price, and the reduced fuel efficiency increases fuel consumption by 5 - 15 percent.
The British estimate that emissions controls might raise driving costs by about 4 per cent is
consistent with these figures.
Recent figures from Europe given in IME (1987) suggest that the cost of meeting
proposed emissions standards are $5,800 - $10,600/tonne NO2 removed, with lean-burn engines
somewhat cheaper than three-way catalytic converters. This compares with costs of between
$200 - $4,000/tonne removed from stationary sources (using modifications to burner design) and
between $4,000 - $10,000/tonne removed using (relatively untried) catalytic converters. The
best method of dealing with emissions from large combustion plants is disputed. Selective
catalytic reduction is costly (perhaps $150-200 per kw capacity) and possibly unreliable. Low-
nox burners are currently under development and are increasingly adopted in new plants.
According to The Economist (Dec 23, 1989) there is a new technology, tested on a 150 MW
power plant in West Germany, and in about 20 industrial applications, called selective non-
catalytic reduction. The technique involves injecting a urea-based liquid (‘Noxout’) into the
furnace to reduce the NOx to nitrogen. The capital cost of retro-fitting is low ($15-20/Kw
capacity) and the cost of NO2 removal is about $1,500/tonne, or less than one quarter that of
selective catalytic conversion.

3. Acid rain and environmental policy

Acid rain is no respecter of national boundaries and has thus raised international concerns.
Debates about international rights and obligations appear to be based on legalistic rather than
economic principles—of equity, uniformity and the appeal to simple principles embodying such
notions, rather than to the more finely adjusted and individually variable notions of costs and
benefits. We shall see first the outcome of this international debate, and then contrast its
proposed solution with an economically rational solution.

3.1 Policy to date

The debate on acid rain and on appropriate responses has been conducted in two different
forums. The initial pressures came from the UNECE Convention on Long-Range Transboundary
Air Pollution (LRTAP) which has 34 members from Europe and North America. Much of the
pressure here was exerted by the Scandinavian countries and Canada, who are both large net
importers of acid rain because of their unfortunate downwind location relative to their polluting
neighbours in Europe and the United States. In 1982, Norway and Sweden pressed for the
signatories to reduce SO2 emissions to 30% below 1980 levels by 1993. This led to an informal
‘30% Club’ founded in Ottawa in March 1984, and, in July 1985 21 countries, but not including
the US and the UK, signed a protocol at the third meeting of the UNECE LRTAP Convention
in Helsinki.
Whereas the Scandinavians were initially primarily concerned with the acidification of
lakes and streams and consequent loss of fish, West Germans were worried about the impact
 19

their own industry was having on the environment, concerns which were reflected in the growing
political power of Green parties in the early 1980s. An emotive campaign in 1982 drew attention
to the problem of Waldsterben or forest death, in which official estimates showed that over half
the forest area had suffered damage, attributed to acid rain. For a variety of political reasons
described in more detail in Berkhout et al (1989), a Large Combustion Plant Ordinance
(Grossfeuerungsanlagen-Verordnung or GFAVo) was enacted in June 1983, under which flue
gas desulphurization equipment would be fitted to 37 GW of coal fired power stations and to the
early closure of 12 GW. Not surprisingly, industry protested that the costs of this programme,
which were to be borne by electricity consumers, would harm West Germany’s competitive
position in international markets, and this led the government to press for similar standards being
adopted for the whole of Europe. The European Commission proposed a Large Combustion
Plant (LCP) Directive based on the GAFVo in December 1983, calling for a cut in SO2
emissions by 1995 by 60% to 40% of their 1980 level. After much debate, described in Skea
(1988), the UK finally agreed to reduce SO2 emissions from existing large plant to 20% below
its 1980 level by 1993, by 40% below by 1998 and to 60% below by 2003, and nitrogen oxides
(on the same basis) by 15% to 1993 and 30% by 1998. The Directive also provides for stringent
emissions standards for new large combustion plants which the UK accepted.16 In November
1988, the UK Environment Minister, Lord Caithness, signed a UN protocol in Sofia committing
the UK and most leading industrial countries to freeze the level of nitrogen oxides at 1987 levels
until 1994 and by 1996 to agree to further reductions based on critical levels.
The proposed agreements can be described as aiming at uniform reductions in emissions
from each country relative to an arbitary starting date, and without regard to the costs and
benefits at the country level.

3.2 The economics of designing a policy for acid rain

Welfare economics has developed a range of theories to deal with externalities and public good
(or bad) problems such as acid rain, which hardly need rehearsing here. Acid rain differs from
‘greenhouse gases’ like CFCs and carbon dioxide in that it is not a pure public good (or bad) in
the Samuelson sense. Acid rain causes damage where it is deposited, and there appears to be a
reasonably linear relationship between emissions and depositions, as Table 2 suggests for the
aggregate. Moreover, the EMEP is predicated on there being a predictable and stable relation-
ship between the location of the source and the deposition, at least averaged over a year.17
Detailed work, reported in Derwent (1988, 1990), suggests that both SO2 and NOx depositions
can be statistically predicted, given the location of the source, and that depositions at specified
sites can be traced back to their originating source. The main differences between the two
pollutants is that a higher fraction of SO2 is deposited within 100 km of the source, and that a
larger fraction of SO2 can be accounted for by individually identified large stationary sources,

16
The details are reported in Our Common Future: A perspective by the UK on the Report of the World
Commission on Environment and Development, Department of the Environment, July 1988.
17
In the short run, the location of the deposition will depend on the current weather conditions and wind
direction. Over a year or possibly longer these will follow predictable patterns in aggregate.
 20

predominantly power stations. Nitrogen oxides are produced both by large stationary sources
and by numerous small mobile sources—road vehicles.
Technically, acid rain appears to be a depletable or rival good in consumption, in that if
one tonne of sulphur falls on a given local area, then that tonne cannot fall elsewhere, and
reduces the amount which will harm others by that amount. It might appear to be a simple
bilateral externality of the kind considered by Coase (1960). But there are two important
differences from the simple case of bilateral externality in which a well-defined polluter deposits
pollution on a single well-defined recipient. On a given day, the plume of pollution from a
source can be tracked, and the amounts of acid rain deposited along its flight path estimated.
The plume may pass over many different cells, and thus deposit pollutants on a number of
different countries. Even over a short period of time, then, more than one recipient will be
affected by the single pollutor. Over longer time periods the number of recipients will be larger
as the wind direction and strength vary, even if the average deposit received by a given country
from a given source is well-defined over some time period. If the recipients are to bargain over
reductions in acid rain, they will have to agree among themselves how to coordinate their
bargaining, and how to share any costs involved.
The second complicating fact is that reducing emissions requires a fixed durable
investment in FGD equipment, as well as variable costs (largely additional energy costs to
operate the FGD). Even if there were only one recipient at any date, there will be a number of
recipients over the life of the FGD equipment, all of whom benefit from its installation. Again,
there are joint beneficiaries, and the FGD plant itself is like a public good, available to all
potential recipients once installed. Third, there are several alternative ways of reducing the
damage caused by the main source of SO2, power stations. One is to reduce emissions by
investment in FGD, another is to reduce emissions by reducing the amount of electricity
generated by a given station. In an integrated electricity grid, power stations are ranked by merit
order, with the cheapest variable cost plants given the highest rank, and the most expensive the
lowest. The despatcher calls for power from the highest ranked stations and moves down the
merit order until total demand is satisfied. On a given day the wind direction and strength may
cause the plume from one power station to deposit acid rain where it does high damage, while
that of another power station may fall largely on the sea or on areas where damage costs are low.
If the power despatcher includes in the costs of generating electricity the social cost of the
damage done by emissions, and if these costs are varied with changing weather patterns, then
the position of the power station in the merit order will change with the weather, and the damage
done by the pollution will be reduced, even if the total amount of pollution remains roughly the
same.18 Much the same result can be obtained by holding stocks of low-sulphur coal, which
are burned when the wind direction is adverse in preference to the use of normal (higher sulphur)
coal.
Finally, if the public good and coordination problems facing the recipients of pollution
from a given country can be overcome, then it should be possible for these recipients to bargain

18
Estimates presented by Sir Sam Edwards of the Cavendish Laboratory, Cambridge, suggest that this may be
a very cost-effective way of reducing pollution costs for UK generators.
 21

towards an efficient level of pollution, as argued by Coase (1960). How this might occur is
discussed below.
A number of very important economic consequences flow from these facts. First, the
damage done by acid rain depends critically on where the deposition occurs. If it occurs over
the sea, it is unlikely to have much harmful effect, as the sea is naturally alkaline. If it occurs
over sparsely populated areas with class 5 (acid tolerant) soils, then again the damage may be
low. If it falls on rivers and lakes, it may be very damaging ecologically, but financially the
costs look trivial compared to property damage. If it falls on densely populated areas, the
property damage will be substantial. As a consequence, the benefits of reducing emissions will
vary significantly from source to source, since different sources will have very different
deposition patterns. Second, the costs of reducing SO2 emissions will vary significantly with
the source. The cost per tonne SO2 reduction will depend on whether an integrated FGD system
can be installed at the time of construction of a new station, or whether it must be retrofitted.
If it is retrofitted, the cost will depend on the number of Gwhours the plant will produce over the
remainder of its life. For a base load plant with a long expected life, the cost will be low, but for
a plant near the end of its life, or one which is primarily used for peaking, the cost will be high.
If the object is to maximise the Europe-wide net benefit of pollution reduction, then three
types of relationships are needed. EMEP divides Europe up into 150km square cells and
measures depositions in each cell. This is probably be the finest degree of disaggregation
practical, at least in the medium run. For each of these cells one needs a damage function,
relating total damage to depositions. If qi is total quantity of pollutant deposition in EMEP cell
i, then the total damage is Di = Di(qi). For each source, j, one needs a cost function describing
the cost of emission abatement. If ej* is the uncontrolled or initial level of emission, and if ej
is the actual level of emissions then the amount of abatement is aj = ej* - ej. The cost borne by
j is Cj = Cj(ej*-ej), with Cj(0) = 0, and the marginal cost of abatement, dCj/daj > 0.
The final component is a transport matrix, relating emissions at various sources to
depositions at various destinations. If tij is the amount of deposition at location i per tonne of
pollutant emitted at source j, then the vector of depositions q = Te, where T is the matrix whose
ij-th element is tij . Total social costs to be minimised by the choice of e are

S Cj (ej ej ) Di ( tij ej ) .
j i j

The first-order conditions are

Di .tij Cj , j 1, 2,..., n .
i

Detailed figures on the damage functions are not readily available, but it is interesting
to consider the implications of three different measures of damage. First, we need to measure
the intensity of pollution produced by a given source, and the natural measure is SO2 per square
mile of the recipient country per tonne emitted. The total damage done will then depend on what
is damaged. Given the dominance of property damage one possibility is to make the marginal
damage proportional to capital stock times pollution intensity. Property (and possibly other
 22

relevant components of damage) are likely to be proportional to GDP, and so if we multiply

tonnes deposited per tonne emitted by GDP per square mile, then we have a rough measure of
property damage. Given the unreliability of estimates of GDP from socialist countries, the
measure of GDP finally adopted are the purchasing power parity measures for 1985 given in
Summers and Heston (1988). The T matrix can be derived from the fuller version of Table 1,
given as table A2 of the Appendix, and is presented as Table A3. The resulting marginal damage
costs are shown in Appendix Table A4. It is more useful to compute an index of these costs and
Table 6 below gives these figures for various source countries, ranked by the damage they
export, relative to Britain. The index of property damage in the column headed ‘prop’.
An alternative measure might be the number of population in the country affected, again
multiplied by the pollution intensity. This index is given in the column headed ‘Pop’. Finally,
as a measure of willingness to pay for reductions in pollution, the third measure multiplies
pollution intensity by GDP per head, and is given in the column headed ‘GDP’.
Four countries appear to be as or more damaging than the UK by the property damage
and by the population criterion, and eleven by the GDP measure, though one should remember
that the estimated damages have been very crudely estimated. As SO2 does not travel very far,
and as the damage is weighted by the GDP density, countries like Belgium which are situated
near wealthy neighbours yield high damage costs per tonne SO2 released. At the other extreme,
the Soviet Union, which produces a vast amount of SO2, deposits most of it within its own
borders, and even here the damage cost is low because of the low population and GDP density.
It is interesting, largely as an illustrative exercise, to see what kinds of SO2 reduction
would be justified given the marginal costs of abatement estimated by Amann and Kornai
(1987). That source gives plots of total and marginal abatement costs by tonnage abated by
country for the year 2000 (ie allowing adequate time to undertake adjustments, and predicting
future levels of emissions without abatement). These plots have been used to determine
marginal costs of abatement and the levels of reduction which would be justified—the level at
which the marginal cost of abatement is equal to the marginal damage which would be
appropriate for a 30% overall reduction in emissions.19 If we had absolute measures of the
damage then it would be possible to determine the appropriate total and individual reduction
which is cost-justified, but as we only have an index of relative damage, this alternative finds
the least-cost method of making the 30% reduction.
Table 6 gives the implied reductions which minimise costs. In some cases the estimate
is reasonably robust, assuming the correctness of the underlying abatement cost figures. In other
cases, indicated with a single star, the marginal cost of abatement was flat at a level close to the
estimated marginal damage, and so the appropriate level of reduction is poorly identified. Figure
1 illustrates the problem for the UK. When the damage is measured by the third measure based
on ability to pay (GDP/head) the marginal damage and marginal costs almost coincide over a

19
The calculations were done iteratively, first linearising the estimated marginal abatement costs, then finding
the level of marginal abatement cost corresponding to the index number of 100 in the damage index at which the
overall reduction of 30% was achieved. This cost was then multiplied by the damage index for each country to give
the country level abatement cost, and this figure was used to read off the appropriate level of abatement from the
detailed abatement cost schedules. The overall level of reduction was then calculated, and is as shown in Table
6—close to 30% except for the last column.
 23

range, whereas for the other cases the marginal damage intersects the MC schedule at a vertical
step.
The reductions proposed vary significantly by country and also according to the measure
of damage. They are not simply related to the damage index, as abatement costs vary
considerably by country. Thus although Luxembourg has a rather high damage index, the
marginal abatement costs are so high that very low reductions are justified. In some cases the
abatement costs of further reductions are high relative to the benefits of the reduction and for
these countries zero reduction is proposed.
One obvious conclusion is that a uniform reduction in pollution from an arbitrary date
like 1980 is most unlikely to maximize social net benefits. Instead, abatement should take place
in an order determined by the size of the excess of social benefits over abatement costs, con-
tinuing until the net benefit falls to zero. In the illustrative case shown in the table, Britain
appears to warrant a more radical abatement programme than most of her EC partners, primarily
because abatement costs are quite low and damage costs appear to be high. It should be stressed
that this conclusion is only as good as the data on which it rests and the methodology for
estimating damage costs.
The exercise reported in Table 6
makes strong assumptions about costs and Figure 1
benefits, and it is therefore important to look Abatement cost function for UK
at other similar exercises based on an Marginal costs (DM/t) (Thousands)
35
economic approach to costs and benefits.
Mäler (1989) has estimated cost and damage 30
functions for the same set of European
25
countries and used the 1984 EMEP
transmission matrix to compute the efficient 20
level of abatement. He assumed the costs of
15
abatement were quadratic (ie the marginal
costs were linear, which is a poor assumption 10
for many of the countries)20. As in Table 6
5
the cost schedules were calibrated to cost
estimates made at IIASA for the year 2000 0
0 1 2 3 4 5
using the schedules in Amann and Kornai
Abatement (kt SO2) (Thousands)
(1987). He also assumed that damage costs
were constant per tonne of sulphur deposited MC abatement Prop damage
Pop damage GDP/hd damage
in each country (as was done in Table 6), but
that the damage ‘represents the evaluation of Amann and Kornai (1987)
the damage that respective governments
make today.’ These estimates are made as follows. If each country takes the level of emissions
of all other countries as given (the non-cooperative solution), and if each country then balances

20
In an earlier draft this assumption was employed as a first approximation and it resulted in substantially lower
reductions for the UK in particular.
 24

the marginal cost of domestic abatement with the marginal benefits of reduced domestic
deposition, then it will solve the equation

Di .tii Ci .

Table 6 Index of damage per tonne SO2 emitted

Index of damage Reduction %

Source prop pop GDP prop pop GDP

Belgium 175 160 149 39 30 26

Netherlands 164 154 121 65* 40 18*
GDR 156 139 155 83 80 75
France 100 92 146 29 28 28
UK 100 100 100 60 60 40*
FRG 95 100 136 36 28 23*
Luxebourg 84 69 88 6 6 6
Switzerland 78 68 97 25 25 25
Czecholsovakia 68 80 116 55 54 52
Italy 65 76 82 39 29 22
Denmark 46 41 111 31 21 35
Hungary 46 69 104 60* 65 70
Austria 45 51 92 30 30 30
Poland 43 65 95 26 33 33
Yugoslavia 35 55 77 63 75 75
Spain 30 37 69 10 11 36*
Ireland 27 33 51 11* 11* 12*
Portugal 25 49 53 10* 40 30*
Bulgaria 18 33 50 17 26 27
Greece 13 27 38 10* 60* 60*
Sweden 10 9 109 0 0 3
Albania 9 26 17 0 10 0
Romania 8 20 45 0 7* 15*
Turkey 8 25 30 0 0 0
Norway 7 5 104 0 0 6
Finland 7 7 105 0 0 18
USSR 5 6 73 2 3 20
Average 54** 59** 89** 30*** 30*** 33***

Notes: Damage is deposition per tonne emitted times damage measure, which is
GDP/sq. mile (prop), population, or GDP/head, taking damage by UK as 100.
Countries are ranked by total damage using the first measure of damage.
Reductions are those required to achieve an overall reduction of 30% at
minimum cost according to the three measures of cost. Damage index for USSR
are related to population densities in the EMEP area alone.
* estimate unreliable as MC of abatement flat
** unweighted average
*** weighted average

If C’ i and tii are known, then D’ i can be deduced. On the strong assumption that the
Governments of European countries were rational, selfish and non-cooperative, the revealed
marginal willingness to pay for abatement can thus be deduced. (Mäler, 1989; 1990, p90).
 25

His estimates of the efficient

level of abatement involves an Figure 2
aggregate reduction of 8.9 million Relation between proposed reductions
tonnes sulphur, or a reduction of 39%,
Other damage reduction %
yielding a net benefit on DM6.3 billion. 100
The main conclusion to be drawn from
this estimate is that the efficient
80
solution requires very different re-
ductions in each country, and that a uni-
form percentage reduction would yield 60
much lower social benefits. His basic
table gives the percentage reductions
and net benefits for selected countries 40

shown in Table 7. For comparison the

percentage reductions for the three 20
different measures of damage from
Table 6 are presented next to Mäler’s
estimates. The relation between the 0
0 20 40 60 80 100
different proposed reductions is also
Property damage reduction %
graphed in the scatter diagram in Figure
2. In interpreting the figure it should be Population GDP/head Maler

remembered that Mäler’s estimated

overall level of reduction is 39 percent,
rather than the 30 percent used in the calculations of Table 6, so that his figures will be on
average higher. It is interesting (and rather surprising, given the very different approach
adopted) that the order and the degree of pollution reduction in Mäler’s calculation is not so
different from those appropriate to some measures of Table 6. Table 8 gives the correlation
matrix between the various proposed reductions. As expected, Mäler’s measure has a lower
correlation with those proposed here than each of the three measures have with each other.
Table 7 shows, on Mäler’s figures, that not all countries are net beneficiaries of the
efficient plan, and the losers might not agree to join the agreement. The fact that most countries
gain from a move to the efficient plan is also rather misleading, for if countries have to agree to
the percentage of the efficient level of pollution reduction to jointly undertake, then many
countries will lose from the final increments—say, in going from 90% to 95% of the plan. Mäler
argues that in a repeated game with appropriate side payments, supporting the efficient solution
could be made individually rational—deviating countries would trigger punishments in the form
of increased depositions, and would lose any side-payments.
 26

Table 7 Net Benefits from efficient solution

Country Percentage reductions Net Benefit

Prop Pop GDP/hd Mäler DM m.

GDR 83 80 75 80 11
Netherlands 65 40 18 62 565
Yugoslavia 63 75 75 79 344
Hungary 60 65 70 77 5
UK 60 60 40 81 -365
Czecholsovakia 55 54 52 75 152
Belgium 39 30 26 36 191
Italy 39 29 22 33 -81
FRG 36 28 23 86 328
Denmark 31 21 35 86 119
Austria 30 30 30 21 324
France 29 28 28 10 879
Poland 26 33 33 27 599
Switzerland 25 25 25 23 192
Bulgaria 17 26 27 43 -7
Ireland 11 11 12 38 71
Spain 10 11 36 14 -29
Portugal 10 40 30 19 10
Greece 10 60 60 86 53
Luxembourg 6 6 6
USSR 2 3 20 2 1505
Albania 0 10 4 45 24
Romania 0 7 15 83 422
Finland 0 0 18 14 -2
Norway 0 0 6 6 272
Turkey 0 0 0 9 68
Sweden 0 0 3 4 606
Total 30 30 33 39 6290

Sources: Mäler (1989); Table 6 above

Note: Net benefits as estimated by Mäler.

Table 8 Correlations between different proposed reductions

R2 percentages

Prop Pop GDP/hd

Prop
Pop 72
GDP/hd 52 82
Mäler 38 46 39
 27

3.2.1 Bargaining over abatement Let us return to consider the efficacy of bilateral or multilateral
bargaining over pollution abatement, in the spirit of Coase. One can imagine two possible
allocations of property rights. The status quo is one in which each country is free to pollute its
neighbours. The natural alternative is one in which each country agrees to a certain annual level
of emissions—eg 70% of the 1980 measured level. Emissions above that level are only
acceptable with the agreement of the recipient countries. In the second case, the polluter would
have to pay for increased emissions, and a natural offer would be an amount between the
marginal cost of abatement and the marginal damage done to the recipient. This principle is
termed the Polluter Pays Principle or PPP. The important question of how this should be
decentralized within the polluting country will be considered below, but in principle any new
pollution source would have to buy the right to release a given number of tonnes of pollution
from an existing source, who would then reduce its emissions by that amount.
In the status quo situation, new polluting sources could set up and pollute at no apparent
cost. The costs of damage would be borne by the recipients (the Victim Pays Principle, or VPP).
Does this mean that the status quo would favour entry by more polluters? Ignore for the moment
the problem of coordination among the recipients, and suppose that there is one representative
recipient who is willing to bargain with the polluting country, and one polluter within the
country, such as the CEGB in Britain. Suppose also that there are no market failures in selling
electricity, of the kind considered in the next section. Then it is a standard Coase result that the
allocation of property rights does not affect the efficiency of the final outcome. The first task
would be to agree a reference path of pollution in the absence of any cleanup—presumably based
on a forecast of energy demand and emissions per GWhr. The recipient representative would
then offer to pay the CEGB for reductions below this level. In both cases the social cost of
producing electricity would be the same and would include the social cost of releasing pollutants.
In the first case (PPP) the CEGB would pay the recipient for extra emissions, and in the second
case (VPP) the CEGB’s total profits would decrease for each extra tonne released—the cost of
emissions would be the reduction in profit.
There is a further similarity in the operation of the two forms of property right, in that
recipient countries may find the benchmark figure of 70% of 1980 levels too high, in which case
that becomes the new status quo point. Economic growth which raises the wealth density is
likely to make further absolute reductions cost-effective, especially as technical progress should
continue to lower abatement costs. The recipient would then wish to pay for reductions below
the status quo, just as before. The only difference between the two cases would be in the size
of the lump-sum transfer from recipient to polluter. In the second case the recipient would pay
the polluting country the damage cost of the difference in 70% of 1980 levels and the reference
emission path.

3.2.2 Complications caused by regulation and privatisation Suppose the electricity supply
industry (ESI) is privately owned but regulated, as in the US. If regulation is based on actual
costs (again as in the US), with the allowable price related to the average cost, then the average
cost of electricity is lower under VPP than under PPP, and hence prices will be lower, demand
higher, and the level of pollution will be inefficiently high. The appropriate method of price
regulation would be to determine benchmark costs based on a pre-determined level of emissions
 28

control, and allow the generators either to keep payments for reductions, or force them to pay
for emissions above the status quo level. Naturally this particular problem does not arise with
price-based regulation of the RPI-X typified by the regulation of British Gas or British Telecom.
The present proposals for privatising the electricity supply industry in the UK involve
splitting the CEGB into three components. All nuclear power stations will remain in public
ownership (and produce no acid rain pollutants anyway). About two-thirds of the remaining
capacity will be allocated to National Power, the rest going to Powergen. The grid will be owned
by the twelve distribution companies (Discos). New entrants and existing private power
suppliers will be free to sell to the grid or Discos. The price of electricity sold by the Discos will
be regulated, but not that sold by the generators, who will be subject to normal market
competition.
Suppose that the British government accepts a target level of abatement (eg 70% of 1980
levels), and must decide how to implement this target. The two standard instruments available
are emissions taxes (per tonne of sulphur or NOx emitted), or standards (eg tonnes SO2 emission
per GWhr). Standards applied uniformly are typically inefficient, and economists usually argue
that their inefficiency can be eliminated by auctioning a fixed number of licenses to pollute 1
tonne SO2 in a given year, equal to the total allowable emissions. If polluting firms are
competitive, and the tax is appropriately set (at the auction price) then standard arguments show
these two methods to be the same under certainty.21 But the British ESI will not be
competitive, at least in the near future until considerable entry takes place, and so this argument
does not apply. In the Appendix a simple Cournot duopoly model is constructed to see how an
auction market in license to pollute might work.
The model shows that if firms compete in the input market (for licenses to pollute) as
well as in the output market, then it may pay them to raise the cost of inputs to each other, as a
way of increasing the final product price. The idea that raising rival’s costs may be a preferable
competitive strategy to predatory pricing is not new, though the usual assumption is that firms
are differentially affected by input costs (see eg Salop and Scheffman, 1983). The normal
argument is informal, if intuitive, and ignores some of the subtleties in modelling competition
in both input and output markets. The model in the Appendix considers two identical firms
which can reduce emissions at a cost, and which are allocated or bid for licenses to pollute. If
they behave as Cournot duopolists in the output market, then their joint profits will be higher
under an asymmetric allocation of licenses than with the efficient symmetric allocation. The
reason is that an asymmetric allocation raises marginal production costs of the smaller firm and
allows the two firms to compete less intensely, to their joint benefit. What this implies is that
there is a mutually beneficial transfer price for licenses between the two firms which
compensates the firm with the smaller number of license more than its fall in profits.
Unfortunately the increase in joint profits is smaller than the increased deadweight loss of the
inefficient allocation of licenses, so net social surplus decreases.

21
Just as tariffs and quotas are equivalent under certainty, though not, as Newbery and Stiglitz (1981) show,
under uncertainty.
 29

Consequently, taxes and quotas or auctioned licenses are not equivalent in the presence
of imperfect competition—for taxes can in principle mimic the competitive market provided the
government knows that abatement cost function and can set the tax level appropriately.
There are additional arguments differentiating licenses and taxes. If there is uncertainty
about the location of the damage and abatement cost functions, then it may be preferable (in the
sense of lowering the expected social cost) to choose one instrument rather than another. In
Appendix B taxes are shown to dominate quotas for a competitive industry in which here is
uncertainty about the slope and level of the marginal cost of abatement schedule and similarly
about the slope and level of the marginal damage cost schedule, provided the expected slope of
the marginal cost of abatement exceeds the expected (absolute value of the) slope of the marginal
damage schedule. This seems plausible, for the damage schedule is likely to be rather flat if
damage is proportional to depositions, while the age distribution of large power plants will
introduce an upward slope into the abatement schedule. The proportional relative advantage of
2
taxes over quotas will be (b - d)σ /(b + d), where b is the slope of the abatement schedule, -d is
the slope of the damage schedule, and σ is the coefficient of variation of the slope of the abate-
ment schedule, a measure of the uncertainty about the marginal costs of abatement.

3.2.3 Choosing policy instruments Licenses to pollute which are tradeable and are auctioned off
have several attractions. The first is that they are the natural instrument to meet international
agreements couched in terms of total emission levels, typical in recent negotiations. Second,
they can overcome organized resistance from the industry affected, since firms can be allocated
licenses proportional to current emission levels. The costs of abatement fall on consumers (who
forgo the revenue from auctioning off the licenses), and on new entrants, who have to buy
licenses from incumbents. This anti-competitive feature has already been discussed, and is a
reason for industries favouring stringent standards with appropriate grandfather clauses. Third,
other countries can negotiate further reductions by buying up licenses, providing the issuing
authority recognizes their property rights in such licenses.22
The limitations of licenses are that they allow anti-competitive behaviour, they are
inferior to taxes in maximizing the expected net benefit of control, and if they are allocated
rather than auctioned, they involve a lump-sum transfer to industry, of a kind comparable to
giving away rather than selling state-owned assets upon privatisation. If instead, taxes are used
(or licenses are sold) then the government can reduce other taxes by the amount of extra revenue,
and so reduce deadweight losses elsewhere in the economy. Put another way, most taxes raised
revenue at the cost of some inefficiency, but corrective taxes raise revenue while improving
efficiency. The latter are highly attractive sources of revenue, not to be lightly foregone. There
is a further limitation which has not yet been mentioned, arising because of the nature of the
abatement technology. Flue gas desulphurisation or denitrification requires heavy fixed
investment with a lifetime equal to that of the plant—perhaps up to 40 years. Its cost-effective-
ness will depend on the costs of emission in the future, and these may be harder to predict if they

22
If the issuing authority interprets the international agreement as only limiting the total amount of emission,
and not the total amount licensed, then foreign purchase with the intention of reducing the total emission would be
thwarted by an increase in the number of licenses.
 30

are determined by a spot market in licenses at each future date, than if they are determined by
a tax which is announced and expected to remain relatively constant.23 This objection can be
overcome by issuing plants with an estimated (but non-negotiable) lifetime supply of licenses,
so that the plant owner would expect neither to be a net seller nor net buyer of licenses for the
expected lifetime of his plant. The difficulty with this solution is that licenses for new plant
would have to come from the liquidation of allocations from existing plants, though one should
not exaggerate this difficulty. Old plants will have to be retired, and would at the predetermined
appropriate date release licenses for reallocation.
Licenses work well for large stationary sources which are the main source of SO2, but
are not immediately applicable to small mobile sources, like vehicles which are increasingly
responsible for NOx. Licenses could be issued or auctioned to car manufacturers who would
fulfil their quota, T tonnes NO2 equivalent per year, by selling n vehicles each of which emits
no more than t tonnes per year of normal driving, where nt ≤ T. Subsequent vehicle testing
would ensure that emissions were kept down to t.
Given the advantages and disadvantages of taxes and licenses, is there some compromise
system which might do better than either pure instrument? The anti-competitive problem can
be met by the government or its agency being willing to buy or sell licenses at a predetermined
price. This would require an international agreement about the form in which abatement would
be encouraged. If such agreements insist on quantity controls, then the agency would find it
difficult to hold the price of licenses down to a satisfactory level needed to prevent the incentive
to raise rivals’ costs. If the international agreements merely defined the status quo for
bargaining, then recipient countries could establish a market price for licenses, and might, if they
overcame their coordination problems, be able to act as a quasi competitive force.24 This, quite
part from the increased efficiency of such a system, provides another argument for attempting
to negotiate such treaties.
The main arguments against using pollution taxes are that recipient countries may fear
that it has little effect on pollution levels, while the polluting companies see the tax as harming
their profits. The first worry can be greatly alleviated by combining the tax with emission
standards on new plant, designed to be cost-justified at the tax rate. (It may be enough to set
limits which firms would choose to better given the tax, but which are sufficient to allay doubts
on the part of recipients.) The second problem is primarily one of dealing with firms with
existing inappropriate plant. It may be sufficient to draw up a reference emission plan for each
large plant, specifying the maximum allowable emissions in each of the next 20 years (perhaps
falling by 5% of the initial value each year, down to zero). The plant would be entitled either
to licenses or a tax rebate to this amount, but would have to pay for emissions above this level.

23
It may be sensible to index the tax to per capita income in the recipient countries to allow for increasing
damage costs, and this will be no less predictable than most taxes.
24
If damage per tonne is relatively independent of pollution levels, and relatively independent of the year in
which it occurs, then recipient countries can arbitrage the price from year to year, and the price should be
maintainable within moderate bands, reducing the ability of dominant polluters to manipulate the price of licenses.
 31

3.2.4. Licenses and charges Hahn (1989) provides a useful survey of the experience in dealing
with environmental problems in four countries. The instrument most economists find attractive
is the marketable permit, and he describes two important applications in the US. In 1981
Wisconsin allowed firms to trade the rights to pollute the Fox River, but resulting cost-savings
were minimal. Hahn finds this unsurprising, as the main sources were regulated municipal
utilities and oligopolistic paper and pulp manufacturers. The arguments above suggest that in
both cases the incentives to trade to achieve efficiency will be muted, and it is interesting to have
this empirical confirmation.
A much more important experiment was the US Environmental Protection Agency’s
emissions trading policy, with its various components allowing firms to meet firm level or area
level standards by either internal reorganization of their pollution activities or by external trading
or bargaining. The resulting cost-savings have been substantial, and Hahn cites estimates of cost
savings of between $525 million to $12 billion, the wide range reflecting very incomplete
monitoring of the cost-savings achieved within firms. Most of the gains came from internal
trading, and pressures from environmental groups appears to have lead regulators to downplay
the property rights nature of permits, thus undermining their marketability. Hahn concludes that
‘emissions trading is best viewed as an incremental departure from the existing approach.
Property rights were grandfathered.’ (Hahn, 1989, p101.)
The most successful example of permit trading was lead trading in gasoline, where
refineries could buy or sell the right to lead levels in gasoline, based on an existing standard per
gallon. About 15% of total lead rights used were traded, with an active spot market and an
intertemporal market via ‘banking’. The systems appears to have worked well because it is easy
to monitor and there was widespread agreement about the objectives of the programme, which
were to gradually phase out lead in gasoline. Perhaps the most difficult aspect of environmental
regulation is this problem of reaching a well-defined consensus on what to do—much of the
environmental opposition arises because of distrust as to the objectives of the various parties.
If marketable permits have been rather disappointing, there appears to have been
reasonably widespread acceptance in Europe at least for charges for pollution (primarily for
water borne and solid pollutants). The revenue from these charges is normally earmarked for
environmental cleanup, so the link between the charge and the resulting clean-up is seen to be
fair, given that it would have presumably cost the polluter more to clean up the pollution itself.

3.3 A possible strategy

To date, inter-government negotiations have been guided by the principles of uniform reduction
in emissions from a benchmark level (usually 1980 actual emissions) for large combustion
plants, or towards uniform standards for mobile sources (i.e. cars). The attraction of this is that
it appears to impose equal costs on the participants, so that it gives no commercial advantage or
disadvantage to producers in each country. In fact, of course, countries which have become
heavily dependent on nuclear power (France), hydroelectricity (Scandinavia), or gas
(Netherlands), can achieve these emission reductions at lower costs than those like the UK
and/or Germany still heavily dependent on coal. There are two serious problems with this
approach. The first is political, and may be sufficient to derail the negotiations—some countries
are net losers from such negotiations. The second objection is economic—the reduction in
 32

aggregate pollution damage is done at higher than least cost. Both probelms can be solved by
the natural solution which allows beneficial and efficient bargaining. This would involve first
agreeing a benchmark trajectory of allowable emissions (not necessarily equal for all countries),
and then facilitating bargaining over deviations from this level. This might best be done by first
estimating the marginal damage costs (or willingness to pay for abatement, if higher) by EMEP
cell, then calculating the appropriate cost sharing formula for the group of countries affected by
polluter j. Thus for country i its share of the payments to polluter j would be

Di tij
αi .
Dk tkj
k

The recipients would then appoint a negotiator with powers to levy charges on recipients
proportional to these cost fractions, up to the total damage level. The negotiator could then bribe
the polluting country to make additional reductions.
The main problem lies in decentralising this within the polluting country. One solution
would be to compute cost-effective emission levels for new plants (i.e. the level at which the
marginal cost of further abatement is equal to the marginal damage) and set this (or a somewhat
more lenient standard) as the benchmark for new plant. Deviations from this could then be taxed
(if excessive) or rebated (if lower). Old plant could be similarly dealt with, at least for the large
combustion plants, which are not too numerous in any country. Again, the first step is to
determine the benchmark, presumably starting near current emission levels, but falling to zero
at the estimated economic date of plant retirement. Where retro-fitting is evidently cost-
effective, then the benchmark might be the retro-fitted emission level. The emission tax would
be uniform over all plants, new and old, but would vary with location.
The current tax rate would be the amount offered by the recipient negotiator. The main
problem is that FGD requires a large fixed investment to amortise over up to 40 years, and thus
requires the investor to calculate the future benefits (i.e. taxes avoided) of installation. If these
future benefits are uncertain, then abatement may be deterred. It would pay the recipient
negotiator to offer formal contracts, possibly indexed to recipient GNP, in order to reduce
uncertainty and thus induce greater abatement for the same cost. This would not completely
solve the problem, at least in the UK, for the following reason. The argument advanced above
suggested that it might be in the interests of the duopoly electricity generators to face
competition from a high cost fringe of producers. It might therefore pay National Power and
Powergen to sell off their least efficient, most polluting plant to independents. The buyers
would be uncertain of the likely future demand for their power (assuming they are unlikely to
secure long-term supply contracts with the Discos for more than a small fraction of their output).
Even if they knew the future payment per tonne SO2 reduced, they would be uncertain of their
likely future electricity (and hence SO2) production. They are unlikely to install FGD, and their
operating costs will thus be raised (by the emissions tax). This in turn will raise the equilibrium
price of electricity, and increase the profits of National Power and Powergen. In short, the
asymmetry in abatement costs for new base load plant and old coal-fired plant makes the case
for raising rival’s costs stronger than that given in the Appendix.
 33

One should not exaggerate this risk. Encouraging divestiture by the duopoly generators
will at least increase the number of competitors, and should make entry-deterrent collusion more
difficult, for the future benefit of consumers. It may also be that retro-fitting is not cost-effective
for the kind of plant likely to be sold, in which case the problem does not arise. Nor does it seem
likely that emissions standards and taxes would discriminate against new entrants, because the
best practice generating technique for new entrants appears to be high efficiency gas turbine or
gas-based combined heat and power systems, of modest capacity (130-300 MW). Emissions
from such systems are naturally low, giving them an additional advantage over the existing large
coal-fired conventional power stations.
If this seems utopian, it nevertheless appears likely that some international agreement on
emissions will be reached, and it is in any case in each country’s own interest to reduce
emissions, as such a large fraction is deposited within its own borders. This being the case,
countries like the UK which have, or will have a privately owned electricity supply industry will
still need a decentralised system of abatement. Standards, appropriately chosen to be cost-
effective, may be satisfactory for new sources (and especially for small sources, like vehicles
where the cost of individually monitored charges is likely to make them uneconomic). They are
not so suitable for existing sources where some abatement is cost-effective. Here taxes (or
charges) have obvious advantages, though bargaining between the relevant government agency
(Her Majesty’s Inspectorate of Pollution) and the individual firm may be a satisfactory
alternative. The attraction of an explicit tax or charge is two-fold—it signals to recipient
countries the advantage of further negotiations (with side-payments) and it signals the advantage
of developing more cost-effective abatement strategies. In their absence, the generators
themselves have no such inducement, as they risk having the new technology mandated in more
stringent standards (now cost-effective).

4. Summary and conclusions

In many ways, the debate on acid rain is quite encouraging from an economists’ perspective.
EMEP has quantified pollution flows and provided the raw data for bilateral bargaining. The
least satisfactory element so far is that of quantifying the damage costs (rather than the levels of
deposition). Much commendable effort has been directed to identifying the ecological path-ways
by which acid rain harms the ecological environment, but insignificant effort to quantifying the
cost of this harm. It appears that property damage is more important than ecological damage,
and that forest and crop damage is far greater than lake and river acidification, yet very little
recent work is directed to quantifying property damage. No doubt this reflects the emotional
level at which most environmental issues are discussed, but one might hope that the best way to
lower the emotional temperature is to increase the factual and economic content of such debates.
The main findings reported here are that it is likely to be excessively costly to aim at a
uniform reduction in emissions in all countries. Since some countries will gain while others lose
from such proposals, it is likely to be politically difficult or costly to reach agreement. Instead,
it seems more sensible to aim at reducing emissions where the benefits of reduced damage
exceed the marginal costs of abatement, and to aim at methods of coordinating payments for
abatement to overcome the ‘free-rider’ problem.
 34

The other finding, less secure given the present state of scientific knowledge, is that
reducing NOx appears less urgent and less attractive than reducing SO2. Certainly, the cost of
reducing automobile emissions is high, the health gains small, and the ecological benefits
uncertain. On the other hand, it may be cheap to reduce NOx emissions from stationary sources,
using better burner designs and chemical additives. If one priority is to be singled out, it is that
of reducing the appalling pollution in the socialist countries in order to secure primarily health
benefits. Given the inefficiency of energy use in these countries, it is likely that substantial
reductions can be achieved at relatively low cost and high benefit, in many cases just by closing
down unprofitable enterprises, and perhaps moving towards a more integrated energy market,
relying more on trade and less on self-sufficiency.25

25
See Hughes (1990) for a discussion of energy and environmental issues in Eastern Europe.
 35

Appendix A
Table A1 Net exports of Sulphur in Europe
Thousands of Tonnes Sulphur per year

Exports to
BG CS DK FR DD DE GR HU IT NL NO PL RO ES SE TR SU GB YU
Net exports from
Bulgaria BG
Czechoslovakia CS -4
Denmark DK 0 2
France FR 0 8 0
GDR DD -3 -44 -10 -27
West Germany DE -1 19 -4 29 102
Greec GR 14 2 0 0 2 0
Hungary HU -9 -14 0 -2 14 3 -3
Italy IT -5 3 0 0 13 -5 -5 -1
Netherlands NL 0 1 -1 6 4 21 0 0 0
Norway NO 0 5 6 2 15 6 0 1 1 1
Poland PL -4 50 1 0 278 24 -2 15 -4 2 -11
Romania RO 18 28 0 3 18 5 2 60 13 0 0 32
Spain ES 0 1 0 -54 4 -3 0 2 -8 -1 0 2 -1
Sweden SE 0 12 13 3 33 12 0 3 1 2 -1 32 0 0
Turkey TR 12 3 0 1 4 1 5 5 4 0 0 3 0 0 0
Soviet Union* SU 12 105 8 10 166 35 1 83 12 3 -9 319 -5 1 -16 -5
Britain GB 0 -2 -7 -29 0 -34 0 -1 -3 -17 -19 -12 -1 -4 -13 0 -15
Yugoslavia YU 4 13 0 5 17 7 -2 24 45 0 0 13 -38 3 -1 -5 -23 32
N Africa** NA 37 105 43 135 253 131 26 64 179 32 7 194 6 209 21 22 196 437 61

Source: Acid Magazine Sep 1989, from EMEP data, and Table A2.

Note: Sulphur dioxide figures will be about twice as large. Total wet plus dry deposition of sulphur for period 11.12.87 to 6.11.88
* European part of USSR within EMEP area of calculation.
** Deposition in North Africa within areas of calculation
 36

Table A2 Origins of Sulphur Deposition in Europe

Thousands of Tonnes Sulphur per year

Emitters
AL AT BE BG CS DK FI FR DD DE GR HU IS IE IT LU NL NO PL PT RO ES SE CH TR SU GB YU NA UI Σ Receivers
AL 5 0 0 1 1 0 0 0 1 1 1 1 0 0 4 0 0 0 1 0 0 0 0 0 0 0 0 0 0 8 30 Albania
AT 0 18 2 0 27 0 0 12 24 18 0 12 0 0 30 0 1 0 15 0 0 2 0 0 0 0 4 8 0 27 207 Austria
BE 0 0 50 0 2 0 0 19 7 15 0 0 0 0 0 0 4 0 2 0 0 1 0 0 0 0 11 0 0 8 121 Belgium
BG 0 0 0 152 4 0 0 0 3 1 3 10 0 0 5 0 0 0 5 0 4 0 0 0 1 5 0 14 0 22 235 Bulgaria
CS 0 4 4 0 385 0 0 11 128 28 0 45 0 0 10 0 1 0 95 0 0 1 0 0 0 2 7 9 0 28 765 Czechoslovia
DK 0 0 1 0 2 31 0 1 12 7 0 0 0 0 0 0 1 0 4 0 0 0 0 0 0 1 7 0 0 9 83 Denmark
FI 0 0 0 0 4 1 48 0 8 3 0 2 0 0 0 0 0 0 12 0 0 0 3 0 0 57 3 1 0 62 210 Finland
FR 0 0 23 0 19 1 0 332 41 40 0 5 0 1 21 0 5 0 15 2 0 65 0 1 0 0 43 3 1 139 760 France
DD 0 0 7 0 84 2 0 14 725 61 0 2 0 0 2 0 4 0 32 0 0 1 0 0 0 1 15 1 0 24 979 GDR
DE 0 1 29 0 47 3 0 69 163 330 0 3 0 0 13 1 14 0 23 0 0 6 0 1 0 1 45 1 0 64 821 FRG
GR 0 0 0 17 2 0 0 0 2 0 45 3 0 0 5 0 0 0 2 0 0 0 0 0 1 2 0 6 0 28 119 Greece
HU 0 3 1 1 31 0 0 3 16 6 0 190 0 0 12 0 0 0 25 0 1 0 0 0 0 1 1 23 0 18 337 Hungary
IS 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 15 21 Iceland
IE 0 0 1 0 0 0 0 2 2 1 0 0 0 21 0 0 0 0 1 0 0 0 0 0 0 0 17 0 0 19 68 Ireland
IT 0 1 2 0 13 0 0 21 15 8 0 11 0 0 353 0 0 0 14 0 0 10 0 2 0 1 4 16 3 83 562 Italy
LU 0 0 0 0 0 0 0 2 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 7 Luxembourg
NL 0 0 15 0 2 0 0 11 8 35 0 0 0 0 0 0 32 0 2 0 0 1 0 0 0 0 20 0 0 10 139 Netherland
NO 0 0 1 0 5 6 2 2 15 6 0 1 0 0 1 0 1 13 11 0 0 0 5 0 0 10 19 0 0 91 194 Norway
PL 0 3 6 1 145 5 0 15 310 47 0 40 0 0 10 0 4 0 790 0 1 1 1 0 0 18 15 11 0 64 1492 Poland
PT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 35 0 19 0 0 0 0 0 0 0 26 83 Portugal
RO 0 1 0 22 28 0 0 3 18 5 2 61 0 0 13 0 0 0 33 0 34 1 0 0 1 17 1 39 0 43 330 Romania
ES 0 0 1 0 2 0 0 11 5 3 0 2 0 0 2 0 0 0 3 15 0 523 0 0 0 0 6 1 1 97 674 Spain
SE 0 0 2 0 12 13 8 3 33 12 0 3 0 0 1 0 2 4 33 0 0 0 37 0 0 23 13 1 0 103 307 Sweden
CH 0 0 1 0 2 0 0 13 4 4 0 0 0 0 14 0 0 0 1 0 0 3 0 8 0 0 3 0 0 14 70 Switzerland
TR 0 0 0 13 3 0 0 1 4 1 6 5 0 0 4 0 0 0 3 0 1 0 0 0 61 12 0 5 1 86 210 Turkey
SU 0 3 5 17 107 9 22 10 167 36 3 84 0 0 13 0 3 1 337 0 12 1 7 0 7 2204 16 24 0 491 3584 USSR*
GB 0 0 5 0 5 0 0 14 15 11 0 0 0 6 1 0 3 0 3 0 0 2 0 0 0 1 571 0 0 60 702 Britain
YU 1 3 1 18 22 0 0 8 18 8 4 47 0 0 61 0 0 0 24 0 1 4 0 0 0 1 32 192 1 74 497 Yugoslavia
NA 3 5 39 37 105 43 18 136 253 131 26 64 0 17 182 0 32 7 194 13 6 210 21 1 23 196 437 62 11 810 3087 N Africa**
3 5 6 4 5 7 3 8 7 4 4 3 0 2 3 8 3 5 2 5 5 5 4 1 5 -22 7 3 180 31 Rounding Errors
Σ 12 47 202 283 1064 121 101 721 2005 823 94 594 0 51 759 5 115 28 1685 67 65 856 79 17 95 2558 1271 424 21 2532 16695 Total
25 75 244 570 1450 155 162 923 2500 1022 180 710 3 84 1252 6 141 50 2270 116 100 1581 116 31 177 5100 1840 588 0 21471 Production 1987

Source: Acid Magazine No. 8, Sep 1989, p8 Notes: As for Table A1

 37

Table A3 Depositions per tonne emission

percent emissions deposited

Emitters
Receivers AL AT BE BG CS DK FI FR DD DE GR HU IS IE IT LU NL NO PL PT RO ES SE CH TR SU GB YU
Albania 20 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Austria 0 24 1 0 2 0 0 1 1 2 0 2 0 0 2 0 1 0 1 0 0 0 0 0 0 0 0 1
Belgium 0 0 20 0 0 0 0 2 0 1 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 1 0
Bulgaria 0 0 0 27 0 0 0 0 0 0 2 1 0 0 0 0 0 0 0 0 4 0 0 0 1 0 0 2
Czechoslovakia 0 5 2 0 27 0 0 1 5 3 0 6 0 0 1 0 1 0 4 0 0 0 0 0 0 0 0 2
Denmark 0 0 0 0 0 20 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
Finland 0 0 0 0 0 1 30 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 3 0 0 1 0 0
France 0 0 9 0 1 1 0 36 2 4 0 1 0 1 2 0 4 0 1 2 0 4 0 3 0 0 2 1
GDR 0 0 3 0 6 1 0 2 29 6 0 0 0 0 0 0 3 0 1 0 0 0 0 0 0 0 1 0
West Germany 0 1 12 0 3 2 0 7 7 32 0 0 0 0 1 17 10 0 1 0 0 0 0 3 0 0 2 0
Greec 0 0 0 3 0 0 0 0 0 0 25 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1
Hungary 0 4 0 0 2 0 0 0 1 1 0 27 0 0 1 0 0 0 1 0 1 0 0 0 0 0 0 4
Iceland 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Ireland 0 0 0 0 0 0 0 0 0 0 0 0 0 25 0 0 0 0 0 0 0 0 0 0 0 0 1 0
Italy 0 1 1 0 1 0 0 2 1 1 0 2 0 0 28 0 0 0 1 0 0 1 0 6 0 0 0 3
Luxembourg 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 0 0 0 0
Netherlands 0 0 6 0 0 0 0 1 0 3 0 0 0 0 0 0 23 0 0 0 0 0 0 0 0 0 1 0
Norway 0 0 0 0 0 4 1 0 1 1 0 0 0 0 0 0 1 26 0 0 0 0 4 0 0 0 1 0
Poland 0 4 2 0 10 3 0 2 12 5 0 6 0 0 1 0 3 0 35 0 1 0 1 0 0 0 1 2
Portugal 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 30 0 1 0 0 0 0 0 0
Romania 0 1 0 4 2 0 0 0 1 0 1 9 0 0 1 0 0 0 1 0 34 0 0 0 1 0 0 7
Spain 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 13 0 33 0 0 0 0 0 0
Sweden 0 0 1 0 1 8 5 0 1 1 0 0 0 0 0 0 1 8 1 0 0 0 32 0 0 0 1 0
Switzerland 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 26 0 0 0 0
Turkey 0 0 0 2 0 0 0 0 0 0 3 1 0 0 0 0 0 0 0 0 1 0 0 0 34 0 0 1
Soviet Union* 0 4 2 3 7 6 14 1 7 4 2 12 0 0 1 0 2 2 15 0 12 0 6 0 4 43 1 4
Britain 0 0 2 0 0 0 0 2 1 1 0 0 0 7 0 0 2 0 0 0 0 0 0 0 0 0 31 0
Yugoslavia 4 4 0 3 2 0 0 1 1 1 2 7 0 0 5 0 0 0 1 0 1 0 0 0 0 0 2 33
N Africa 12 7 16 6 7 28 11 15 10 13 14 9 0 20 15 0 23 14 9 11 6 13 18 3 13 4 24 11
Errors 12 7 2 1 0 5 2 1 0 0 2 0 0 0 0 50 6 6 0 2 5 0 4 13 1 0 -1 1
Sum (%ident) 48 63 83 50 73 78 62 78 80 81 52 84 0 61 61 83 82 56 74 58 65 54 68 55 54 50 69 72
Product 1987 100 100 100 100 100 100 100 100 100 100 100 100 100 100100 100 100 100 100 100 100 100 100 100 100 100 100

Source: Table A2
 39

Appendix B

The choice between taxes and quotas under uncertainty

Suppose that the abatement schedule gives the marginal cost of abatement p in terms of the level
of abatement, x, as

p = a + bθx + ε , Eθ = 1; Eε = 0; Eθ ε = 0,

where θ represents the uncertainty about the slope parameter b and ε represents uncertainty about
the level of initial cleanup costs, assumed independent. If w is the willingness to pay for
cleanup, and X is the current level of emissions, then suppose that

w = c + η + dφ(X - x), Eφ = 1; Eη = 0; Eφη = 0.

Again, φ represents uncertainty about the slope, η about the level when there is no pollution (ie
x = X). Assume that all four random variables are independently distributed.
The net social surplus (consumer surplus plus profit) when the level of abatement is x is
given by

x
⌠
S =  (w - p)dx = S(x).
⌡
0
If the government must choose the level of quota, x, before knowing the parameters of the
system, then x is chosen to max ES(x). The solution is

2
c+dX-a (c+dX-a)
x = ——————; ES = ————————.
b+d 2(b+d)

If the government must choose a tax rate t, then x = x(t), given by

t - a - ε dx 1
x = —————————; —— = ——.
bθ dt bθ

The value of t is now chosen to max ES[x(t)], and the result is

2 2 2 2
(c+dX-a) (E1/θ) (c+dX-a) (1+2σ )
ES = ————————————————— ≈ ————————————————
-2 2 2
2[bE(1/θ)+dE(θ )] 2[b(1+σ )+d(1+3σ )]
 40
2 2
where σ = E(θ-1) is the coefficient of variation squared of the abatement slope. Taxes will be
superior to quotas if ES(t) > ES, i.e if

2 2 2
b[(Eψ) - Eψ] > d[Eψ - (Eψ) ], ψ ≡ 1/θ.

A sufficient condition for this to hold is that σ is small and b > d, ie if the slope of the abatement
cost schedule is steeper than the slope of the damage schedule.

Duopoly model
Consider the following very simplified model of a duopoly electricity supply industry. Demand
is linear, and the market clearing price is p = a - Q when aggregate supply is Q. The
technology is one of constant unit operating costs (set equal to zero without loss of generality)
ignoring pollution abatement. The amount of pollution released when output is q and abatement
equipment h is installed is q/h. The unit cost of abatement equipment is r, and the generator is
allowed to release x units of pollution per year. Its profit is then pq - rh = (p - r/x)q.
Each generator behaves as a Cournot duopolist in the output market, given its entitlement to
release pollutant, and chooses q to maximize profit, assuming the other generator’s output is
given. If the other generator can release y units of pollution, then

1 2 1
q = —(a - {— - —}r),
3 x y

and the market clearing price is

1 1 1 
p = — a + r — + —  ,
3 x y 

providing in both cases that the other generator produces a positive output. If not, then the
generator is assumed to behave as an unconstrained monopolist. If the total pollution level is
fixed at 1 unit, so that y = 1 - x, then individual profit, aggregate profit, consumer surplus, and
net social surplus (the sum of consumer surplus and aggregate profit) are all functions of x. At
the symmetric duopoly equilibrium x = 0.5, locally net social surplus is maximized, while net
aggregate profits are minimized given the chosen parameter values (a = 10, r = 1.5). Fig.1
shows the resulting plot. There is a discontinuity as the duopoly collapses into a monopoly, and
it is assumed that the two firms are not allowed effectively to combine to form a monopoly, and
the graphs is terminated at the point at which monopoly would occur. It is an interesting
observation that if the total licenses were to be allocated to the surviving firm, then the net social
surplus may increase as the combined inefficiency of the duopoly and the misallocation of
pollution licenses is replaced by the single inefficiency of a monopoly.
 41

The fact that aggregate

profits increases as the duopoly Fig. 1
moves away from the social opti- Emissions license reeallocations
mum towards an asymmetric equi-
librium means that there are profit- Profits and Surplus
25
able exchanges of pollution licenses
for cash between the two duopolists
20
which leave each better off than in
the symmetric equilibrium. It will
15
typically pay one firm to buy all the
licenses from the other firm in order
to become an unconstrained mono- 10

polist. If this is not allowed by

competition policy, then the firms 5
will aim at the maximum allowable
degree of asymmetry. 0
0.5 0.45 0.4 0.35 0.3

Fraction of licenses in firm 1

Profits 1 Profits 2 Total Profits

Cons surplus Social surplus

References

Amann, M. and G. Kornai, (1987), ‘Cost Functions for controlling SO2 emissions in Europe’, Working paper WP-
87-065, International Institute for Applied Systems Analysis, Laxenburg, Austria.
Alfsen, K.H., D.A. Hanson and S. Glomsrød, (1986) ‘Direct and indirect effects of reducing SO2 emissions:
experimental calculations on the MSG-4E model’, Central Bureau of Statistics Discussion Paper 20, Oslo.
Battarbee, R.W. and 15 colleagues (1988), ‘Lake acidification in the United Kingdom 1800-1986’, Report of the
Palaeoecological Research Unit, University College London, published by HMSO.
Berkhout, F., S. Boehmer-Christiansen and J. Skea (1989), ‘Deposits and repositories: electricity wastes in the UK
and West Germany’, Energy Policy, April, 109-15.
Brackley, P (1987) Acid deposition and vehicle emissions : European environmental pressures on Britain,
Aldershot: Gower
Britt, D. (1986), ‘The renovation of acid lakes and streams’, chapter 2 of Acid rain Control II: The promise of new
technology, D.S. Gilleland and J.H. Swisher (eds.), Southern Univ. Press, Carbondale.
Coase, R. (1960), ‘The problem of social cost’. Journal of Law and Economics, 3, 1-44.
Congressional Budget Office (1986), Curbing acid rain: cost, budget and coal-market effects, US Government
Printing Office, Washington DC.
Crandall, R.W., H.K. Gruenspecht, T.E. Keeler and L.B. Lave (1986) Regulating the Automobile, Brookings,
Washington DC.
Department of the Environment, (1989) Digest of Environmental Protection and Water Statistics, 1988.
 42
Department of the Environment, (1988), Our Common Future: A perspective by the UK on the Report of the World
Commission on Environment and Development.
Derwent, R.G., (1988), ‘A better way to control pollution’, Nature, Vol 331, No. 6157, 18 Feb, pp 575-8.
Derwent, R.G., (1990), Optimal strategies for sulphur emission control in Europe, Harwell Laboratory
Dowlatabadi, H. and W. Harrington (1989), ‘Policies for the mitigation of acid rain’, Energy Policy, April, 116-22.
Dudley, N., M. Barrett and D. Baldock, 1985), The Acid rain Controversy, Earth Resources Research, London.
Environmental Resources Limited, (1983), Acid rain: a review of the phenomenon in the EEC and Europe, prepared
for the Commission of the European Communities, Graham and Trotman, London.
Gilleland, D.S. and J.H. Swisher (eds), (1986), Acid rain Control II The promise of new technology, Southern
Illinois Press, Carbondale.
HMSO (1984) Royal Commission on Environmental Pollution - Tenth Report, Cmnd 9149, London, HM Stationary
Office.
Hahn, Robert W (1989) ‘Economic Prescriptions for Environmental Problems: How the Patient Followed the
Doctor’s Orders’, Journal of Economic Perspectives. 3:2, 95-114.
Hughes, G.A. (1990) ‘Energy policy and the environment in Poland’, European Economy.
IME (1987) ‘Vehicle emissions and their impact of European Air Quality’, Institution of Mechanical Engineers
Conference Publications 1987-8.
Jeffrey (1988) Energy Policy
Layfield, Sir Frank (1987) Sizewell B Public Enquiry, Report by Sir Frank Layfield, 8 vols, Department of Energy,
HMSO, London.
Longhurst, J.W.S., S.E. Green and D.S. Lee (1987), Acid deposition in the Northern Hemisphere, Acid rain
Information Centre, Manchester Polytechnic.
Mäler, K-G., (1989), ‘The acid rain game, 2’, paper presented at Conference in Wageningen, mimeo, Stockholm
School of Economics.
Mäler, K-G., (1990) ‘International Environmental Problems’, Oxford Review of Economic Policy, 6(1), 80-108.
Mandelbaum, P. (ed) (1984) Acid rain, Economic Assessment, Plenum Press New York. (334.2b95 217)
Newbery, D.M. and J.E. Stiglitz, (1981), The Theory of Commodity Price Stabilization, Oxford; Clarendon Press.
Newbery, D.M., (1990), ‘Optimal Trade Taxes on Agriculture in Developing Countries’, Economic Journal
Supplement, 180-93.
OECD, (1981), The costs and benefits of sulphur dioxide control - a methodological study, OECD, Paris.
OECD Environmental Data Compendium, OECD, Paris.
Park, C.P. (1987), Acid Rain; Rhetoric and Reality, Methuen, London. 9300 c 5317
Pearce, D and A. Markandya, Environmental Policy benefits: Monetary Valuation, OECD, Paris
Salop, S.C. and D.T. Scheffman (1983), ‘Raising Rivals’ Costs’, Amer. J. Econ. Papers and Proceedings, May,
267-71.
Schechter, M., Kim, M. and Golan, L. (1989) ‘Valuing a public good: direct and indirect valuation approaches to
the measurement of the benefits of pollution abatement’ in Fomer, H. and van Ierland. E (eds) Valuation
Methods and Policy Making in Environmental Economics, Amsterdam, Elsevier.
Skea, J. (1988), ‘UK policy on acid rain: European pressures and emission prospects’, Energy Policy, June, 252-69.
UNECE (1982) ‘Effects of sulphur compounds on materials, including historic and cultural monuments’ Draft
report ENV/IEB/WG1; Geneva.
UNECE (1987) National strategies and policies for air pollution abatement
White, L. J. (1981) Reforming Regulation: Processes and Problems, Prentice-Hall, Englewood Cliffs, NJ.
White, L. J. (1982) The regulation of air pollutant emissions from motor vehicles, American Enterprise Institute for
Public Policy Research, Washington and London.

View publication stats