Air pollution

EXECUTIVE SUMMARY

Air pollution is an inevitable consequence of living.  It can be eradicated only at costs which would be totally unacceptable.  Air pollution comprises many different elements.  As a phenomenum, it is associated with large population concentrations and industrial emissions.  It has, however, much diminished as a cause of harm and irritations to those potentially suffering its consequences most severely -- people located within conurbations.  This is notwithstanding continued population growth.

More benign levels of air pollution over recent year scan be traced to a number of features:

• industrial change, whereby a lower proportion of needs in progressively richer societies come from the outputs of "smokestack" industries;
• the dispersion of "smokestack" industrial facilities away from central city areas to the peripheries of conurbations',
• regulatory controls on emissions from individuals (cars and household heat generation) and industrial facilities.

Countering air pollution presents an unusual challenge to those designing taxes, charges and remedial expenditures.  Those who are affected by pollution cannot easily come together so that they may bargain with polluters over a mutually agreeable level of emissions.  Moreover, the great bulk of pollution is emitted by numerous sources -- and further, the "polluter", and its individual "victims", are by and large the same.

Markets, incorporating vested ownership and tradeable rights, are generally the most efficient means of bringing about optimal abatement expenditures.  But, given the major problems of monitoring vast numbers of minor sources of air pollution, with present technology, there are at present some apparently insuperable problems involved in developing and monitoring the contractual approaches which can allow market solutions to operate.  "Command and control" solutions remain the best option for these sources.

In the case of major sources of air pollution, the prospects of enlisting market based mechanisms offer much more promise.  In particular, tradeable rights in pollution have been demonstrated to save considerable costs where some sources can achieve abatement levels more cheaply than others.  For this reason, tradeable rights offer greater flexibility and cost savings than the alternative market based instruments, taxes or charges.

INTRODUCTION

Air pollution agents are manifold.  Those specifically targeted for control normally include particulate matter (smoke), ozone, sulphur dioxide, carbon and lead.

Over the years the problem of air pollution in western countries has been successfully addressed.  In London, in 1952, some 4000 deaths resulted from an extended period of air pollution.  Today the air is much cleaner, notwithstanding much increased traffic and higher energy generation.  Diseases associated with pollution, like influenza, pneumonia and tuberculosis, were responsible for about one quarter of deaths at the turn of the century and now account for less than 5% in a population where life expectancy has increased by over a half.  It might be said that the market for death is declining and the market share of pollution related causes falling! To be sure, much of the improvement stems from factors like improved medical treatment, but to a major extent it is due to a cleaner urban environment.

Curiously, one of the patron saints of environmentalism, Paul Ehrlich, also takes the view that air pollution is a readily resolvable problem.  Ehrlich's view might be conditioned by his ideological battles within the environmentalist movement in the course of which he has sought to propel population growth to an ascendancy which others have rejected.  In his interview in The Ecologist (1973) he said:

... from the point of view of an ecologist ... (air pollution is) one of the relatively trivial problems.  It is amenable to rather rapid technological cure and is just a symptom of some of the things we're doing, rather than something ecologically serious.

Of course, it could be argued that Ehrlich, who was forecasting widespread famine by the early 1980s as a result of Malthusian analyses of population growth, has been discredited (apart from within a particularly bizarre wing of the environmental movement).

AIR POLLUTION IN AUSTRALIA

Air pollution levels in major Australian cities have generally shown an improvement over recent years.  In Melbourne, sulphur dioxide levels have trended downwards and in 1988 were less than one third of the maximum acceptable peak levels.  Chart 9.1 illustrates peak one hour and 24 hour sulphur dioxide (SO₂) and airborne particle trends for the industrial suburb of Footscray.  SO₂ levels in Australia are low by world standards because of the low sulphur fuel used.

Chart 9.1:  Peak one hour and 24 hour sulphur dioxide (SO₂)
and airborne particle trends -- Melbourne (Footscray)
Source:  Vic. EPA

Carbon monoxide (CO) levels have also trended down to magnitudes well within the maximum acceptable, although nitrogen dioxide (NO₂) levels have remained relatively close to their "maximum acceptable level" and ozone (Os) levels are above those defined as acceptable.

Broadly comparable findings to these were reported for Brisbane by Verrall and Simpson (1988).  At the time of Queensland's Clean Air Act of 1963 and the coming into force of regulations giving effect to it (in 1968) major pollutive industries in Brisbane included four coal fired electricity generating stations, brickworks and a host of other coal and wood burning facilities.  Since then, although there has been a six-fold increase in the number of industrial premises, other factors have acted to diminish pollution levels including:

• coal powered electricity generation has ceased within the metropolitan area;
• railways have been electrified;
• domestic burning has been banned.

From the late 1970s, ozone levels have remained similar to those observed earlier;  NO₂, lead and SO₂ show slight declines;  smoke has declined markedly (and visibility has improved);  and CO has shown a major decline (see Charts 9.2 to 9.4).

Chart 9.2:  Peak eight hour and one hour averages
for Carbon Monoxide (CO) levels -- Melbourne region
Source:  Vic. EPA

Chart 9.3:  Peak 24 hour average nitrogen dioxide (NO₂) levels -- Melbourne region
Source:  Vic. EPA

Chart 9.4:  Peak one hour average ozone (O₃) levels -- Melbourne region
Source:  Vic. EPA

The abatement of urban air pollution levels have been achieved by "command and control" regulation.  Where markets do not automatically equilibrate supply and demand because of monitoring difficulties, total permitted supply must be specified by a government authority.  Such quasi-market approaches will pay dividends when applied to some sources;  however, continuation of more directive "command and control" approaches seems to be inevitable in the case of domestic and, perhaps, automotive emissions.  In a strict sense, therefore, the achievement of efficiency largely turns on the nature of the regulation.  If market mechanisms are employed to allow polluters flexibility in meeting the levels desired, we can obtain the same outcome at a reduced cost.

ADDRESSING THE EXTERNALITY OF AIR POLLUTION

THE NOTION OF SOCIAL COSTS

Air pollution was the example used by Pigou (the economist responsible for pioneering the study of welfare economics).  In developing the notion of externalities, Pigou sought to illustrate the difference between private and social costs by posing the issue of a factory making use of inputs for which it paid, and inputs (say, the atmosphere) for which it did not pay but soiled (to the detriment of its neighbours).  If it faced diminishing returns and if each private input was valued equally, the factory's production could be represented by Table 9.1

Table 9.1:  Private marginal gain from a factory operation

Output (1)	Value of output (2)	Marginal value of output (3)	Marginal input cost (4)	Marginal economic gain to the owner (5)
	($)	($)	($)	($)
0	0	0	0	0
1	26	26	12	14
2	50	24	12	12
3	72	22	12	10
4	92	20	12	8
5	110	18	12	6
6	126	16	12	4
7	140	14	12	2
8	152	12	12	0
9	162	10	12	-2

Under these circumstances, the owner would produce up to the level (8 units) where his marginal costs equalled the marginal value of his inputs.

If uncontracted costs which cannot be charged for are added to this example, and these costs rise at a constant rate with each additional unit of output, then the marginal economic gain to society is different from that of the owner (shown in column 5, Table 9.2).

Table 9.2:  Social marginal gain from a factory operation

Output (1)	Marginal private economic gain (5)	Value of uncontracted costs (6)	Marginal social gain (7)
	($)	($)	($)
0	0	0	0
1	14	2	12
2	12	4	8
3	10	6	4
4	8	8	0
5	6	10	-4
6	4	12	-8
7	2	14	-12
8	0	16	-16
9	-2	18	-20

From his analysis, Pigou concluded that a tax equal to the uncontracted costs should be imposed -- a tax which would bring production back to four units.

Freeman, Haveman and Kneese (1973) in Chart 9.5, offer a diagrammatic version of the externalities which Pigou was describing.  They do so by examining both negative and positive cases.

Chart 9.5

Coase (1960) showed that there is no more certainty that the pollutees have a right to clean air than the factory owner has the right to use the air.  Clean air is a common, open access resource owned by neither party.  There is a mutuality of interest rather than an automatic onus upon the polluter.  Coase's analysis demonstrated that efficiency will be arrived at where exclusive rights are given either to the neighbours or to the factory owner provided that there are no costs in arriving at the transaction.  There are, of course, as many complications thrown up by this solution as there are insights provided by it.  Importantly, it is uncertain how the neighbours, in particular, could agree on an optimum level of pollution at specific compensation levels, and how they might ensure the factory's pollution is monitored effectively.

Others have pointed out that mankind begins modifying his environment as soon as his existence is significant.  Moreover, rights once seized or acquiesced assume a value.  The factory owner's costs are capitalised as rents within his production capabilities and should he on-sell the original factory, the new owner would have had the expectation that any free inputs would continue as such.

The real problem with externalities is their pervasive nature.  Most attention is focussed on the adverse externalities -- the "bads" like pollution.  However, actions of others also result in unmerited increases in wealth.  Such positive externalities occur where, for example, a neighbour maintains a highly attractive garden which raises others' enjoyment of their own property and may even increase its value.  Other forms of unmerited increases in property values occur where "gentrification" of hitherto blighted inner city properties takes place.  Similarly, certain skills like those of business economists become more highly prized when a more liberal banking regime allows increased competition in this sector.  Much the same may be true of journalists, telephone technicians and airline pilots following relaxations of regulatory arrangements in areas where they are qualified.  A road development will change the value of properties around it -- possibly reducing the value of those properties which are adversely affected by noise, and increasing the value of those which benefit from greater transportation convenience.

Wherever possible, the agents of change will attempt to garner the maximum rents from the change;  but full capture will never be practicable.  Indeed, it is the lack of full capture of rents that has been responsible for much of the trickle down' of wealth from those generating increases in wealth to the community in general.  Even where -- as in the case of intellectual properties -- new forms of rights have been developed, the full capture of the benefits of inventions is barely conceivable.  It would require the inventor to charge each user a separate price based on the user's "willingness to pay".

DETERMINING THE CORRECT OVERALL
LEVEL OF PERMITTED EMISSIONS

The various approaches to emission control which have been discussed are alternatives to "command and control" approaches.  Only a pure Coasian approach makes full use of markets.  Both taxes and tradeable emissions require limits to be specified by governments rather than traded off between polluters and pollutees as is the case in true markets.

For all other approaches, the government, at a minimum, specifies the level of tolerable emissions by examining the costs and benefits.  Costs of pollution include those impacting on health and the various sensory perceptions.  One way of avoiding imposition of officials' choices in this process is to attempt to measure individuals' subjective values.  This involves constructing shadow prices based on willingness to pay.  This, the contingent valuation method, uses market research techniques in an attempt to determine appropriate values.

Commonly, questionnaires are devised describing the goods under scrutiny and asking how much respondents would be prepared to pay.  The techniques of market research are well known and used extensively both in business and politics.

Some of the difficulties of attempting to assign values in this way are exemplified in a study by Tolley and Randall (1985).  Researchers inquiring in the Chicago area about the value of preserving air quality in the Grand Canyon expressed the question in two different ways:

1. for the Grand Canyon alone, after respondents had been shown photographs;  and
2. as part of a three part sequence which sought values for cleaner air in Chicago, in the Eastern United States and in the Grand Canyon.

In the first study, the value per head for clean air in the Grand Canyon was $90, whilst in the second it was $16.  Values were calculated by asking each respondent how much they would be willing to pay for a clear view.  Individual valuations were then summed and divided by the number of respondents to give the "average" benefit each would receive from eliminating air pollution.

The study exemplifies the pitfalls inherent in contingent valuation.  The approach's deficiencies are first, it specifies values based upon average utilities whereas demand and supply for goods in general is determined by marginal costs and marginal benefits.  Thus, a given consumer may value Bounty Bars at $20 and, at a market price of $1, obtain surplus value of $19;  but this surplus value is irrelevant to decisions about whether or not the good is produced -- or even how much of it is produced.  Supply of any good, whether it be clean air or Bounty bars, should continue up to the point where the extra marginal costs of supplying it exactly equal the extra benefits obtained.  The consequence of taking a decision based on average costs and average benefits is an all or nothing outcome;  but it is more likely that we would trade-off some benefits for some costs.

As suggested by Freeman, a more accurate replication of markets can be constructed using "choke-off" prices.  "Choke-off" prices are determined by attempting to construct a demand curve of the marginal benefits from pollution reduction.  People are asked what they would be prepared to pay for a range of marginal improvements in air quality.  Individuals' "willingness to pay" measures are summed for each marginal improvement.  The demand curve generated is a vertical aggregation of individual marginal valuations following the methodology set out by Samuelson (1954).  Vertical aggregation is required, in contrast to the normal method under which demand curves are summed horizontally, because pollution is a non-rival good -- one enjoyed or suffered by all.  The pleasure one person obtains from a clear view (in the absence of congestion) does not stop others from also obtaining satisfaction from it.

The synthetic demand curve may then be used to determine where the marginal costs equal the marginal benefit.  However, the methodology is still suspect, because respondents are not obliged to make real world choices within the constraints of their budgets.

In determining their choices of goods and services, people face a galaxy of options but are only able to satisfy a limited number of needs.  If, added to the air pollution questions in the Tolley and Randall study cited above, respondents would need to be asked about preservation of wildlife, forests, river purity, parkland in their neighbourhood and the whole host of other facets of life which could be considered externalities.  They should be confronted with the real trade-offs between these goods, and correspondingly fewer of the goods they would normally purchase.  Were such a study possible, the values measured would likely be only a tiny fraction of those assignable from seeking answers to single issues with no trade-offs involved.

Freeman expresses a scepticism about the non-use values reported in these studies.  He says:

At issue is whether the responses are measuring a true willingness to pay as defined in our basic theory of individual preferences or whether they are indicators of a general sentiment for environmental protection or preservation that is only imperfectly related to the willingness to commit resources in a true market or quasimarket setting.

These reservations have undoubted merit.  However, survey methods do allow values to be placed on particular resources by those seeking their preservation.  And, by placing upper boundaries on the amounts of resources available, more rational choices maybe possible.

APPROACHES TO REDUCING UNWANTED RESIDUALS

Almost all activities impose some cost or benefit on other parties;  and, though externalities have been the subject of a lengthy literature, the normal procedure (both for the community as a whole and in economic analysis) has been to neglect them.  To a considerable degree this corresponds with efficiency.  Arranging and monitoring contracts can be expensive;  to attempt to build ledgers of all cross-payments each of us owes and is owed would impose excessive costs and inflexibilities.

This explains one reason why the broad sweep of externalities has received only cursory attention in the past -- they have not much mattered.  Clean air was abundantly available.  Where externalities became important, as in the case of downstream pollution or noise, law developed to take them into account by adapting property rights.  Often the increased importance of such externalities has generated incentives for new techniques to be developed in order to control them better because of asset reductions.

Where unwanted outputs are to be reduced this can be accomplished by:

• lower production of the good of which it is a by-product (this might entail changing the composition of national income so that resources are redirected to outputs having a lower level of deleterious by-products);
• improved efficiency in producing the good so that fewer adverse side effects accompany its output;
• recovery of residual materials and recycling them;
• dilution of the "bad" so that its effect is less concentrated.  As Huber (1985) puts it "dilution may in fact be a very good control strategy.  As countless cancerous rats might attest, many things are harmful in large concentrations but innocent or even beneficial in small ones".

Bernstam (1989) demonstrates how the relationship between residuals and output is non-linear and varies over time and between economic systems.  Thus, in the US, between 1940 and 1970, prior to major efforts to reduce pollution, emissions increased by 30.1% while GNP increased by 212%.  From 1970 to 1986, total pollution declined (by one third), in part because of regulatory action.  In the main, however, both periods' trends were attributable to shifting composition of national income and technological improvement, spurred on by the ceaseless contest of competitive firms to improve their profitability -- one means to which is conserving use of material inputs.

Bernstam also shows that this same pattern is not in evidence in the Soviet Union, where emissions of air pollution in 1987 were more than twice those of the US, even though national income was probably less than one third that of the US and population only 17% higher.

In the Soviet Union, the economic system is driven by forces other than a profit based meeting of consumer needs at the lowest price.  As a result, two factors bring about a considerably higher level of waste and unwanted residues than in comparable market economies.

First, excessive inputs are allocated to capital production due, in part, to inefficient machinery.  Thus, the share of consumer goods in national expenditure has fallen from 60.5% in 1928 to 24.9% in 1987;  and capital investment's share is at least twice that of typical market economies, without growth compensating for this denial of immediate consumption.  In short, excessive resource inputs are spent on machines and residual outputs emerge while providing little contribution to well-being.

Secondly, the "command and control" mechanisms used in the absence of profit related measures must focus upon inputs:  the productive unit has its price controlled and its output levels established for it and the only way it can obtain a higher "profit" is to raise its production costs by requiring increased inputs.  In this way the firm is able to pressure the planners into lowering its production quotas and raising its output prices.

Use of market instruments to combat air pollution combines the power of individual self-interest with the best sources of information on how to reduce emissions to the levels sought at the lowest cost.  The firms and individuals who produce the emissions have the knowledge on how to reduce their outputs most economically.  They will seek to take opportunities for gain (or for reducing their losses).  In doing so, they ensure a more frugal use of resources in meeting standards than would be possible for regulatory authorities, whose information on economical means of reducing pollution cannot be as complete.  Because of their vested interest and operational familiarity, firms are likely to be much better informed about the available techniques for reducing residues in the most cost effective manner than are government officials.

PREFERRED APPROACHES

It is useful to categorise air pollution according to its three primary causes:  automotive, household energy generation and industrial facilities.  In each case the economist's solution would be to impose a tax or introduce tradeable rights.  The decisions between these and outright regulation of inputs should depend upon policing costs.  Clearly, such costs are greater with a multiplicity of sources.  Just as it makes sense for electricity authorities to strike separate deals with major users but charge generally available rates to domestic consumers, so it is appropriate for pollution controls to be tailored differently for minor, as opposed to major, source emission locations.  For the former, transaction costs of monitoring market based approaches may be prohibitive, given current technological capabilities.

Automotive and domestic sources are far more important than industrial sources, as Chart 9.6 shows with respect to Sydney (which is typical of other Australian cities).

Chart 9.6:  Sources of emissions in Sydney -- percentage of total emissions for each pollutant

INDUSTRIAL POLLUTION

In the case of industrial pollution, economies are available if trade of pollutants between different sources is allowed.  Crandell (1983), for example, found that the cost of controlling emissions from paper mills was three-fold that of controlling similar ones from metal working factories -- fewer pollutants could be achieved at the same social cost by concentrating on the latter.

Industrial pollutant trading can take a number of forms, including netting, offsets, bubbles and banking.  Netting sets emission standards for one business but allows trading within plants.  Offsets allow new pollutant sources if compensatory reductions can be obtained from other sources.  Bubbles are defined for a particular area and allow different pollutants within a given aggregate limit.  Banking enables credits to be earned for over-performance and subsequently used or traded.

Hahn and Hester (1987) explain how the EPA has allowed:

• netting since 1974;
• offsets since 1976;
• bubbles since 1981;
• and banking.

Their estimates of the effect of these measures, 1979 through 1985, are set out in Table 9.3.

Table 9.3:  Estimated effects of emission trading 1979-1985

	Bubbles	Offsets	Netting	Banking
Number	132	1000	8000	100
Cost saving ($m)	145	n.a.	4000	small
Air quality impact	neutral	neutral	slightly negative	slightly positive

They consider the effects to have been less than satisfactory because of arrangements by environmental groups which have thwarted some proposals and because of uncertainty by firms.  They attribute this latter effect to the need for EPA approval to be specifically granted and the discretion EPA has (and is thought likely to use especially with regard to bankable emissions).  Moreover.  since 1986 the EPA has insisted that bubble trades be "taxed" so that there is a net reduction in emissions of 20%.

These problems notwithstanding, the notion of emission trading as a cost effective interventionary tool is gaining increased currency.  Congress has agreed to a scheme which splits the US into two areas with unlimited trading of SO₂ emissions permitted within each of them.

Hahn and Hester (1989) also estimate there to be 132 "bubbles" within which trading takes place, and these brought a cost savings of $435 million between 1979 and 1985 while having had a neutral effect on pollution levels.  These savings are in addition to savings of up to $12 billion estimated to have accrued from "netting" -- allowing firms flexibility to over-perform in some areas of a plant to compensate for under-performance in others.

Levin (1985) quotes some specific examples where these approaches have resulted in gains.  Dupont, facing a requirement to reduce emissions by 85% in each of 119 stacks, negotiated to reduce 99% of emissions in seven stacks which proved faster and over-achieved the aggregate goal at a saving of $12 million in capital cost and $4 million a year in operating costs.  General Electric was allowed to forgo $1.5 million in capital expenditure and $300,000 in operating costs, required to meet emission controls in Louisville, by negotiating with International Harvester which was able to over-perform mandatory requirements relatively cheaply.

The Pigovian approach, which for long had been preferred by economists, is to apply pollution taxes.  Like trading, this allows greater flexibility for firms in designating the appropriate means of meeting output levels.  It also allows compensation of the community at large for residuals in excess of those considered appropriate.

Buchanan (1988) sets the strict conditions under which an externality can legitimately be countered by the imposition of a tax He suggests that:

• all persons must be equally damaged;
• all must be consumers/buyers purchasing in equal quantities;
• the revenue must be equally shared.

In such rare cases, he suggests, the price would rise so that the "bad" would be economised upon.

Buchanan maintains correctly that without his strict conditions for governmental action to combat an externality there will be distributional consequences -- consequences which will be determined by the political market, and generate social costs via lobbying and government failure.  Some redistribution is inevitable where any departure from equal usage and production occurs.

Others, for example Terkla (1984), take the view that effluent taxes improve welfare because they charge the polluter the true economic cost, and allow the replacement of other taxes which are designed to raise revenue and unintentionally distort economic choice and resource allocation.  Terkla suggests that the total revenue raised from an efficient tax, which in 1982 he estimated would optimally be set to raise between $1.8 and $8.7 billion, would generate considerable efficiency gains.  Based on Browning's estimates of the welfare losses generated by levying income tax, $0.35 per dollar collected, he estimated effluent taxes would raise welfare by $630 million - $3.05 billion.

Terkla's effluent taxis therefore seen as more efficient than the alternative means of raising revenue.  Like others, he sees merits of such a tax system in allowing the market to discover the most efficient means of adapting to a new incentive structure.  His estimates are based on effluent taxes not generating the sort of losses from work/leisure substitution which Browning's estimates project.  Hence, although many would dispute the predicted cost effectiveness, there is wide agreement that making use of the market in this way will generate economies.

In practice, taxation of residuals has not found favour.  Those facing the taxes have, of course, objected whilst environmental activists have opposed this policy approach because of the apparent endorsement it implies to the generation of pollution.

In addition, there are several practical difficulties in devising a workable taxation regime.  Buchanan (1988) points to one major difficulty.  The various parties are likely to have different interests.  Those producing, or using most intensively, the output of the polluting facilities are likely to wish to see the tax levied at as low a rate as possible;  those on whom the impacts of the residues fall most heavily are likely to favour a prohibitive level of tax;  taxpayers who are relatively indifferent to the pollution and the output the facilities' produce are likely to favour a tax which maximises the revenue raised so that other taxes might be reduced.  How are these differences to be reconciled?  The obvious gains will lead the parties to engage in wasteful lobbying exercises to promote their particular interests.  Can we be confident that governments will arbitrate dispassionately?

Pollution taxes are based upon the fundamental principle that the polluter pays.  Yet this tends to treat the polluter as the malefactor and the pollutee as the victim.  In fact there is a mutuality of interest.  The unowned resource did not belong to the pollutee in the first instance.  As soon as mankind breathes air some impression is made on the natural environment.  It is no more certain that the population surrounding a polluting factory has the rights to clean air than the owners of the factory have the right to soil it.  This statement is even more graphic in situations where the factory owner was there first and the "victims" moved in later (perhaps to take advantage of the opportunities to improve their well-being offered by locating close to the factory).

Industrial air pollution is more easily combated by granting tradeable rights to pollute.  Means of monitoring major sources of pollution are readily available as is the technology to effect this.  Hartley and Porter (1990) draw attention to the application of deuterated methane, a chemical tracer which mimics SO₂ to detect sources of pollution in southern Utah.  Tietenberg (1990) assembles eleven empirical studies of market based approaches to pollution control compared with the "command and control" approach.  Each study estimates the cost of a "command and control" strategy limiting pollutants in a localised area.  These approaches are contrasted with the least cost allocative method involving either trading or taxes.  Table 9.4 gives the ratios of the "command and control" outcome to the estimated least cost allocative mechanism.

Table 9.4:  Empirical studies of air pollution control

Study	Pollutants Covered	Geographic Area	CAC Benchmark	Ratio of CAC Cost to Least Cost
Atkinson and Lewis	Particulates	St Louis	SIP regulations	6.00 a
Roach et al	Sulphur dioxide	Four corners in Utah	SIP regulations Colorado, Arizona, and New Mexico	4.25
Hahn and Noll	Sulphates standards	Los Angeles	California emission	1.07
Krupnick	Nitrogen dioxide regulations	Baltimore	Proposed RACT	5.96 b
Seskin et al.	Nitrogen dioxide regulations	Baltimore	Proposed RACT	14.40 b
McGartland	Particulates	Baltimore	SIP regulations	4.18
Spofford	Sulphur Dioxide	Lower Delaware Valley	Uniform percentage	1.78
	Particulates	Lower Delaware Valley	Uniform percentage regulations	22.00
Harrison	Airport noise	United States	Mandatory retrofit	1.72 c
Maloney and Yandle	Hydrocarbons DuPont plants	All domestic reduction	Uniform percentage	4.15 d
Palmer et al.	CFC emissions from non-aerosol applications	United States standards	Proposed emission	1.96

Notes:

CAC = command and control, the traditional regulatory approach.

SIP = state implementation plan.

RACT = reasonably available control technologies, a set of standards imposed on existing sources in non-attainment areas.

^a Based on a 40 µg/m³ at worst receptor.

^b Based on a short-term, one-hour average of 250 µg/m³.

^c Because it is a benefit-cost study instead of a cost-effectiveness study, the Harrison comparison of the command-and-control approach with the least-cost allocation involves different benefit levels.  Specifically, the benefit levels associated with the least-cost allocation are only 82% of those associated with the command-and-control allocation.  To produce cost estimates based on more comparable benefits, as a first approximation, the least-cost allocation was divided by 0.82 and the resulting number was compared with the command-and-control cost.

As Tietenberg points out, the estimated savings are theoretical -- they are gains achievable on the basis that sunk costs have not been incurred, perfect information is available and multilateral trades take place.  Moreover, if emission credits are traded on a pollutant-by-pollutant basis, rather than on an amalgam of pollutants, the trades themselves are rendered considerably more complex Nonetheless, the wide number of studies (each of which demonstrates considerable gains from applying market principles) present powerful evidence against "command and control" methods.

There is little use made of economic instruments to control Australian industrial emissions.  This may change.  Both the New South Wales Government and the Commonwealth Treasury have placed on record their favouring of market based measures where these are possible.  Indeed, the New South Wales Government has announced its intention to place a greater priority on "pollution taxes and charges, pricing of services based on true costs, tradeable emission rights and government subsidies" (Greiner, 1990).

At the present time, however, pollution in Australia is combated only by "command and control" methods.  Some flexibility is provided for in certain circumstances.  Thus, the Victorian Government, in introducing more stringent requirements for the control of conveyor equipped coating lines in 1988, specifies input controls in detail but allows firms to meet the standard in other ways providing they are able to demonstrate that these achieve equivalent results.  Even in these cases, however, the regulations contain rigidities over and above the absence of provision for trading.  These include grandfathering provisions which discourage the replacement of equipment.

AUTOMOTIVE AND HOUSEHOLDS

It has previously been suggested that opportunities to make use of market mechanisms are limited.  Monitoring and other transaction costs may make even these partial market type approaches inapplicable for determining efficient household and vehicle pollution behaviour.

In the case of motor vehicles, governments the world over have introduced standards for emissions, more recently by seeking the use of lead-free petrol.  A general approach may be the rational solution, notwithstanding that in Australia, citizens of places like Albury Wodonga with little pollution would not obtain value from the increased capital and operating costs involved (the latter partly hidden by the governmental requirement that lead free petrol be cross subsidised by leaded petrol). ⁽¹⁾

The move to lead free petrol has contributed to a marked reduction in the level of lead in urban areas.  Thus, both in the centre of Melbourne (where it was previously at double the level set as acceptable) and in the suburbs, lead levels have exhibited a considerable decline.

Grenning (1985) is critical of Australian mandatory emission control standards adopted in Australia Design Rule 37 (ADR 37) during 1986.  He favours emission charges over the technical solutions introduced.  Grenning maintains that the standards adopted were overkill because:

• any problem which occurs is confined to Sydney, and, to a lesser degree, Melbourne (which together on the widest interpretation, might account for 30% of the vehicle population);
• pollution levels in these cities had began to decline anyway as a result of industry restructuring and relocation.

ADR 37 meant a cost per vehicle, at 1985 prices, of $70-160 which is in addition to a slightly higher impost introduced by the previous standard.  Important shortcomings of a standard like this are that they apply only to new vehicles -- and perhaps, therefore, to only 12% of the annual stock.  In addition, the increased cost (and reduced performance) creates disincentives to replace existing vehicles and therefore, to some extent at least, has perverse effects.  Furthermore, achieving the targeted output of emissions by using a "command and control" approach is far from certain.  It depends crucially upon the vehicles being properly maintained and is totally negated if owners disconnect the control mechanism.

Grenning favours a charge based on the outcome of emissions as measured at the annual vehicle test.  Although this would make use of more direct and effective controls, it would also have shortcomings:

• the annual inspection is only a once per year measure and there will be ways discovered which would allow vehicles to demonstrate a short term measured acceptability in emissions;
• it may entail greater costs if owners are required to have modifications undertaken retrospectively,
• it entails some administrative costs, and if cars can be registered in areas where these additional costs are not required there will be considerable incentives for evasion.

Leaving aside the issue of whether or not mitigation of emissions specified for Australia was necessary, it is not clear that the "command and control" solution is inferior to the generally preferred output based solution in this particular case.

In the case of households, many locales have banned coal and wood burning, though the latter somewhat ironically has more recently shown an increase in popularity because it is thought to be a more natural fuel.  Prohibition would not be the preferred solution of most economists yet it may well be more efficient than imposing an easily evadable tax.  As with garbage, it is not easy to see how contractual difficulties can be overcome to allow tradeable rights to operate with respect to this source of air pollution.

RURAL POLLUTION

Aside from the issue of urban air pollution -- one which largely involves health and unpleasant smells -- there are issues of rural air pollution.  In the main these involve maintaining a pristine air quality.  Such issue shave not assumed any importance in Australia to date and major industrial sources of air pollution facilities in areas of high natural value are most unlikely to be economically justified.  A contemporary exception is the controversy over the location of a high temperature incinerator to service the southeast corner of the continent.

ENVIRONMENTAL REGULATORY COSTS

It is often pointed out that ecology and economics have much in common in so far as both start from the premise that everything is interconnected.  Many point to a comity of interests between the two frameworks.  Some, like Hamrin (1981) suggest that the application of environmental standards on emissions required by government regulations will actually benefit both the environment and the economy by saving energy, virgin resources, and so on.  Such assessments glide over the costs in terms of resources ards.

Others, more conventionally, suggest that such a comity crisis, and would be the natural outcome of market forces if property ownership rights could be adequately defined to prevent excessive use of "unowned" resources and the consequent externalities generated.  Fred Smith (1989) goes further than this and maintains that modern technology can allow all externalities to be internalised.

The mainstream view is that for some goods the externality looms so large and the difficulties of internalising it are so great that interventions by government are essential.  Such interventions can only be legitimate where they are based upon the construction of shadowmarkets for evaluating the worth of those activities where externalities inhibit provision by natural markets.  Although there have been no analyses of the aggregate costs of environmental regulation in Australia, a number of studies of the costs of air and water pollution control have been conducted in the US.  These include many estimates of the costs of environmental protection on economic growth by Crandell;  Christainson and Haveman;  and Conrad and Morrison.  But perhaps the most rigorous has been that of Hazilla and Kopp (1989).  Taking the Environmental Protection Agency's (EPA) cost estimates of federally mandated pollution controls ($425 billion in 1981 dollars), $648 billion in 1981-1990 current dollars), the authors apply elasticities of substitution, both to the economy's outputs and inputs.  Because of substitution, the initial estimates are lower than those derived from the EPA's engineering based estimates.  Both consumers and producers take actions to alleviate the cost burden which regulation imposes, for example, by switching purchases to goods which do not have additional cost requirements.  In this way the aggregate cost imposition is muted.  But the dynamic, secondary impacts of these costs must also be factored in.  In addition, the effects of the intervention cannot be confined to one particular time period but will flow on into subsequent periods.  The resulting costs from making these adjustments are estimated at $977 billion, which by 1990 translates into a diminution of real GNP of 5.9% and of investment by 8.4%.

This approach, however, tends to take preferences and technical capabilities as given.  In fact, both consumers and producers can make rapid adjustments.  And over time alternative needs and new means of meeting them are found while, both supply and demand curves for a particular good tend to become flatter.

In the case of consumers, for example, the introduction on congested bridges of express lanes which may only be used by multiple occupancy cars brought considerable behavioural changes both in Sydney and in San Francisco.  Consumer adjustments to picking up or accepting rides from total strangers were remarkably swift and it is difficult to argue that the costs estimated at the outset prevail in anything like their original magnitude after a short transition period.  On a larger scale, adjustments following the implementation of major infrastructural changes within cities -- changes which were envisaged to affect property values markedly -- have been absorbed without lasting declines in these values.  For example, converting vehicular roads within cities to pedestrian malls has often brought very rapid behavioural changes on the part of shoppers which were unanticipated and which totally negated the adverse effects previously expected.

For producers, the very rapid adjustment of some industries to the four-fold increase in oil prices, which took place in the 1970s, reveals great flexibility.  The Japanese steel industry converted from oil firing to making more efficient use of coal, a formidable energy-saving innovation.  Entrepreneurial reactions like this cannot be incorporated into general equilibrium models except by using non-scientific "fudge factors".  Indeed, the inability to account for the role of the entrepreneur in seeking out opportunities constitutes perhaps the greatest shortcoming of all economic modelling.

Because of compensatory variation, it is probable that the economic analysis of environmental quality regulation overstates the costs to society.

There are, however, other factors which would tend to operate in the opposite direction.  One is that entrepreneurship itself is a scarce resource and energies directed at ameliorating an intervention might be energies which would otherwise be directed at discovering new means of adding value.  In addition, modelling typically assumes zero transaction costs, perfect factor mobility and other notional attributes which we tend to group under the heading of perfect markets.

The work of Hazilla and Kopp, originally commissioned by the EPA, has a strong following within the agency, even though its findings have not been formally endorsed.  Publicly the EPA quotes a more conservative, less comprehensive cost of environmental interventions which amounts to only 1.7% of GNP.  Nonetheless, the Hazilla and Kopp work constitutes the state-of-the-art in estimating environmental costs and, because it examines the total picture, is superior to those estimates which confine their impacts to specific sectors.

The costs of environmental regulation estimated by Hazilla and Kopp incorporate only the costs of those regulations falling under the control of the US Environmental Protection Agency.  These cover air and water pollution and waste disposal.  They do not include other regulatory interventions which fall within the environmental embrace, such as use of forests and wilderness, protection of flora and fauna and measures to combat soil erosion.

CONCLUDING COMMENTS

The magnitude of the costs involved in effecting pollution control makes the means by which this is undertaken of considerable importance to general well-being and not least to industrial competitiveness.  Compared with other countries, Australian levels of pollution are low -- in part because our cities tend to have fewer industrial facilities and their populations are less concentrated.

In addition, pollution levels have been reduced, notwithstanding industrial growth and far greater numbers of automobiles.  This outcome, welcome as it is, is the result of "command and control" policies.  Such approaches may well be unavoidable and the most effective means of combating the minor source domestic and automobile emissions which together account for the preponderance of urban pollution.  They have been demonstrated not to offer the lowest cost strategies for control of emissions from major industrial sources.  Australian authorities, however, have not sought to apply market-based solutions, except in limited cases where offsets within major sources have been negotiated.

---

REFERENCES

Bernstam, M.S. (1989) "Productivity of Resources, Economic Systems Population and the Environment":  Is the Invisible Hand Too Short or Crippled?, Centre of Policy Studies, Monash University.  To be included in Davis, K. and Bernstam, M.S. (1990) (eds.).  "The Endless Frontier and Resources", Cambridge Univ. Press, New York.

Buchanan, J.M. (1988) "Market Failure and Political Failure", Cato Journal 8, 1: 1-14.

Coase, R.H. (1960) "The Problem of Social Cost", Journal of Law and Economics, 3: 1-44.

Crandell, R.W. (1983) "Controlling Industrial Pollution", Brookings, Washington.

Freeman, M., Haveman, R.H. and Kneese, A.V. (1973) "The Economics of Environmental Policy", Wiley, John, N.Y.

Greiner, N.F. (1990) "The New Environmentalism".

Grenning, M. (1985) "Australian Motor Vehicle Emission Policy A Costly Mistake", CEDA Monography, 80, Melbourne.

Hahn, R.W. and Hester, G.L. (1987) "The Market for Bads", Regulation, 3, 4, pp48-53.

Hahn, R.W. and Hester, G.N. (1989) "Marketable Permits", Ecology Law Review, 16, 2, pp361-406.

Hamrin, R. (1981) "Environmental Quality and Economic Growth", Council of State Planning Agencies, Washington.

Hartley, P.R. and Porter, M.G. (1990) A Green Thumb for the Invisible Hand, Tasman Institute, Melbourne.

Hazilla, M. and Kopp, R.J. (1989) The Societal Cost of Environmental Quality Regulations:  A General Equilibrium Analysis, Resources for the Future.

Huber, P. (1985) "The I Ching of Acid Rain", Regulation, September-November.

Levin, M.H. (1985) "Building a Better Bubble at EPA", Regulation, March-April pp33-42.

Pole, N. (1973) "An Interview with Paul Ehrlich", The Ecologist, 3, l: 18-24.

Samuelson, P. (1954) "The Pure Theory of Public Expenditure", Review of Economics and Statistics, 36: 387-9.

Smith, F. (1989) Environmental Policy:  A Free Market Proposal, p.32-37 Tulanian.

Terkla, D. (1984) "The Efficiency Value of Effluent Tax Revenues:, Journal of Environmental Economics and Management, 2, pp107-123.

Tietenberg, T.H. (1990) "Economic Instruments for Environmental Regulation", Oxford Review of Economic Policy, 6, l, pp17-33.

Tolley, G.S. and Randall, A., Establishing and Valuing the Effects of Improved Visibility in the Eastern United States, Report to the US EPA.

Verall, F.N. and Simpson R.W. (1988) "Trends in Ambient Air Quality in Brisbane", paper presented to ANZAAS.

---

ENDNOTE

1.  Interestingly, however, a survey about environmental concern conducted by the Australian Bureau of Statistics (Cat. No. 4115.0) listed concern about pollution as being highest in the two Australian regions, the Australian Capital Territory and the Northern Territory, where problems in this regard would be much less evident than elsewhere.  This may reflect the preferences of people living in those two regions.  It may also reflect a heightened awareness about environmental matters generally, a possibility which is perhaps corroborated by generally enhanced levels of concern registered about other environmental issues including, nuclear power, nature conservation, old growth and rain forests, soil erosion, and water salinity, etc.