ECONOMIC THEORY AND ECONOMIC PRAGMATISM INTRANSPORT PRICING DECISIONS

NEIL J. STEEPEROFFICER IN CHARGE, ECONOMIC ANALYSISBUREAU OF TRANSPORT ECONOMICSCANBERRA

ABSTRACT:

The paper initially defines the theory of pricingdecisions and then discusses the nature of markets fortranspozt services which demonstrates that the neo­classical assumptions of pricing aIe rarely appropriateand that the theory of second best prevails Means forarriving at commercial prices for transport services

are also advanced ,.

213

TRANSPORT PRICING DECISIONS

INTRODUCTION

Goods and services a:r:'e exchanged in markets.Markets are simply venues specified in time and area atwhich buyers and sellers can meet for trade. Price andquan t,i ties exchanged in markets are simultaneously determinedby the interaction of supply and demand.

Supply levels depend upon the technology of theprocess of producing the good or service, the costs of inputsto that process and the prices of substitute inputs. On theother hand demand is affected by consumer tastes and incomesand the relative prices of otheI' goods and services whichconsumers may wish t,o purchase.

The equilibI'ation of demand and supply in mostmarkets is an aggregate process. Pr'ovided that the marketfunctions in a competitive fashion (i.e. there are many buyersand sellers and collusion between same is not possible) nobuyer or seller is able individually to influence price.Under such circumstances all buyers and sellers are pricetakers. They adjust their quantities of sales or purchasesaccording to the ruling market price" Price determination tothem in a specific sense is therefore irrelevant ..

with some added conditions which are specified laterin this paper, this freedom of competition is the cornerst,oneof neo-classical economic theory" The theo:ry of course doesnot apply to the real world but :rather, it acts as a bench­mark for economists in investigating real world phenomena ..

Nevertheless, monopolies (i"e. single sellers), andmonopsonies (i .. e. single purchasers), often exist in markets"For example take the railway commissions in most Aust,ralianstat,es and the Australian Wheat Board throughout Australia:respectively ..

When such phenomena arise, neo-classical economictheory in its strictest sense, is no longer applicable andmust be modified to take such aberr'ations into account.. Thediscussion of such occur:rences and the problems that arisewith them are the main thrust of this paper.

NEO-CLASSICAL ECONOMIC THEORY

The basis of all neo-classical economic theory isthe assertion supported by observed behaviour, that allindividuals attempt to maximise the 'utility· or satisfactionor ·welfare' that they realise from the employment of the

215

STEEPER

limited resources at their' disposal" Many caveats can beadded to this statement.. Nevertheless even Marxian economicsand altruism have bases which can be reinterpI"eted alongthese lines.

Economic theory demonstrates that there are twointerrelated aspects of optimising social welfare for anyparticular society" The first is the attainment of anefficient allocation of resources.. The second is theachievement of that distribution of welfare that is preferredby society given that par'ticular efficient resource alloc­ation"

216

the theory is static in the sense that time andhence uncertainty are not relevant

there is no interfer'ence with the free flow ofproductive reSOurces

property rights are well defined so there is noneed for an enforcement authority

there are no spillover effects or externalities inthe sense that:

the production of any good or service producedby any business or firm does not depend uponthe inputs used in some other process by someother firm

there is no government and no goods are consumedcollectively

A majo.r caveat refers to interdependent satisfactions ..Individual satisfactions obviously depend tc varying degreesupon those of others as manifest from intra-family dependen­cies within households and inter-family relationshipsbetween households" Their existence is also confirmed at. abroader' level by the incidence of social assistance andinte:.t'national aid programs. Such interdependencies aretermed externali,ties. Individuals can be expected to aspireto maximise their collective satisfaction gains as a society,given the resources at their disposal.. The level of satis­faction that they collectively achieve is usually termed"social welfare""

(i)

(ii)

(iii)

The theory is developed assuming a perfectly competi­tive economy and a lack of welfare interdependencies asdiscussed above, in order to simplify the concepts used andto clarify the type of processes involved, rather than givea prescriptive model of the workings of actual economies ..In detail the implications of these assumptions are: (Hyman1973, p .. 21)

(i v)

TRANSPORT PRICING DECISIONS

Marginal CostsAverage Costs

Ruling unit I --=:::::,=;:~;....,""'::::.------market price -r Marginal Revenue

0'-----------+----------.... OUTPUTOutput at which profitsare maximised

FIGURE 1DETERMINATION OF OUTPUT BY THE COMPETITIVE FIRM

217

STEEPER

(v) decreasing cost industries (natural monopolies)do not occur and marginal products of all inputsin all productive processes always decrease

(vi) household preferences are not satiated and marginalrates of substitution between all goods diminish

(vii) total productive resources available for alternativeuses are fixed

(viii) information flows concerning demand, supply andprices are perfect in the sense that all detailsof market conditions including production andconsumption are instantly known to all"

These conditions define a perfectly competitivemarket without welfare interdependencies. They have beenreduced philosophically to thr'ee basic requirements. Theseare the maintenance of perfect information flows and theexistence of markets for all goods and services includingresources, along with the participation of sU£ficien t buyersand sellers in all markets to make collusion in order tointerfere with market mechanisms impracticable. Furthermore,these conditions are subject to the thorough indifferenceof any household to the welfare of any other or a completelack of externalities. (Robinson, 1934 ppl04-120)

Under these assumptions all firms set their outputaccording to price as shown in Figure I.. This diagramsimply indicates the rule: marginal cost equals marginalrevenue, if social welfare is to be maximised ..

THE NATURE OF MARKETS FOR TRANSPORT SERVICES

Transport has a special role in any economy, sinceit is the means that satisfies the demand for the spatialmobility of resources including passengers, goods andservices. Different geographical areas typically have acomparat,ive advantage over each other in the production ofgoods and services because of their proximity to resourcesand/or markets" Specialisation in production and distrib­ution by different areas and hence the resultant tradebetween them, leads to increased efficiency in the use ofthe limited resources available to society.. Consequently,trade as facilitated by transport leads to a greater volumeof production and along with it higher consumption than thatwhich could otherwise occur.. Hence, larger real incomes,lower prices and therefore enhanced social welfare resultbecause of trade" These interdependencies are indicative ofmany externalities in transport markets ..

218

TillL~SPORT PRICING DECISIONS

Official statistics consistently underestimate t,hecontribution that transport makes to the production of goodsand services in economies. Generally, ancillary transportservices in particular are ignored, their value beingassigned to other sectors., Moreover, pedestrian, privatemotor vehicle, pipeline, conveyor belt and minor modes ingeneral are also omitt.ed from official stat,istics. Theextent of errors arising from such practices are illustratedby a recent unpublished study commissioned by the BTE"Resource flows in the Australian road, rail, sea and airmodes were investigated.. It was estimated that thesecontributed about 33 per cent of all resources available inthe Australian Economy and 25 per cent of an adjust,ed GNPfigure in 1972-73. On the other hand in the AustralianNational Accounts for 1972-73, the transport and communi­cations sector was estimated to contribute only 8~ percent to the value of final goods and services at factorcost,,,

These figures amply demonstrate that tr'ansport islargely an int,ermediat,e rather than a final good or service"Demands for transport services are therefore mainly derivedfrom those for other goods and services and this in itselfhas important implications for transport policy" Transportof some sort is an essential input to all goods and servicesand repr'esents a significant, fract,ion of the tot.al cost ofmost,. Hence according t,o Marshall' s principles of deriveddemand, (Friedman, 1971, pp148-161) other things beingequal, the demand for many transpor't, services will berelatively unresponsive to changes in price especially inthe short, run" In the short run therefor'e, producers oftransport services may be able to command higher prices andprofits by restricting their output. In the longer runcompetition between transport suppliers and substitution byconswners of other inputs, say land and capital throughrelocation, will reduce their ability to act in thisfashion"

A further aspect of transport is that because of theincrease in social welfare that it generates through it,senablement of trade, defence, and emergency movements suchas ambulance services etc, non-users as well as usersbenefit from its provision., The considerable social asopposed to private benefits it provides, mean that in manyaspects it resembles a public good. Fur'thermore, because ofgeographical welfare differences, transport also haswelfare distribut,ional impact,s.,

These features of transport are o!J\'"i.::msly mostlikely to attract public attention and give rise to govern­ment intervention in transport markets for the followingreasons"

219

STEEPER

(1) Since many producers of transport services canincrease prices without commensurately changing t,helevel of demand for their services in the short run,the introduction of licensing and other regulationswhich limit entry into, or competit,ion in theirsubsector or region will be to their financialadvantage.

(2) Because of (1) above, governments are most likely toregulate or even provide and operate transportservices which are natural monopolies, in order toensure " equitable pricing" and limit lIexcess profit ll

taking"(3) The quasi-public good aspects of transport. imply that

government subsidy and/or taxation of transpor'tservices will take place. This may be achievedwithout. having direct transfer payments throughregulation, deficit financing, or payment of pI:ofitsinto consolidated revenue by government owned andoperated services.

(4) The welfare distributional aspects of transportsey-vices mentioned above, imply that gover'nments willintervene in markets in order to achieve incomedistributional goals.

(5) Natural monopolies are characteristic of many tran­sport services" Their exist,ence implies greaterintervent,ion in transport markets than in thoseof intermediate goods or services which also havepublic good aspects and distributional effects butwhich are typically provided in a competitive fashion.

Overall therefore, the competitive theory advanced sofar is not partiCUlarly relevant to transport markets becauseof market imperfections.. Economic theory has been developedto cover such situations but will not be dealt with here indetail.. Instead a broad brush approach has been adopted ..

Government intervention in transport, markets can beexpected to be considerable.. Hence the external and distri­butional effects of intervention can be anticipated to assumeimportances which in some instances may override monetary orpecuniary considerations. The role of political assessmentsas opposed to objective economic evaluation may therefore bemore significant in determining transport policy than in manyother sectors of an economy.

220

TRANSPORT PRICING DECISIONS

THE THEORY OF SECOND BEST

Over'all, transport markets often therefore divergeconsiderably from the assumptions of classical economict,heoIy.. Hence the simple pricing rule develol?ed above i"e.set output where marginal cost equals marginal revenue, oftenshould not be followed and in fact it has been shown to bet,otally inappropriate in many instances"

The theory of second best is said to prevail when thenea-classical assumptions break down and the appropriaterules for establishing output and prices differ from thoseapplied in the classical approach.

The static economic theory advanced initially in thispaper, implies that a positive gain in social welfare couldonly arise wit,h movement from an inefficient allocation ofresources to one that was efficient. An efficient resourceallocation is one where the production of any single good orservice cannot be increased without reducing the productionof at least one other good or service. Such a resourceallocat,ion is termed Pareto optimal ..

Economists once considered that even if some of themarginal conditions for attaining an efficient resourceallocation could not be satisfied, it was advantageous toendeavour to meet as many other conditions as possible" Theemphasis in this approach was that any move towards satis­fying more of the marginal condi t,ions was desirable, as therewas an implicit belief that if conditions in an economywere more completely based upon marginal principles, the morewelfare dist,ribution could also approach Pareto optimali ty.(Nath 1969, p49)

However, more recently it has been established thateven given only one unsatisfied Pareto condition, the maximumpossible level of social welfare cannot, be at,tained withoutdepart,ing from the other Paretian efficient optima .. (Lipseyand Lancaster', 1956) Such a solution may be entitled secondbest in the sense that it is not, Pareto optimal, because itis attained given a constraint (i .. e. the initial unsatisfiedcondition) which precludes achievement, of that position.But in the light of the value judgements embodied in thePareto criterion, determination of the relative extent ofdepartures from Pareto optimality is logically inconsistent.Nevertheless theoretically, value judgements can be madeconcerning comparative desirabili ties, (Nath 1969, pp51-53)but as long as just one of the conditions for Pareto optim­ality remains unsatisfied, it is neither necessary nordesirable to satisfy the remaining conditions.

221

STEEPER

Given the existence of market imperfect.ions, problemsof measurement and the dynamic nature of economic decisionsand processes, Par-eta optimal allocations of resources anddistributions of welfare will never be attained in actualpurposeful (i.e. pOlicy implemented) economies except bychance alone and then only for short periods"

The theory of second best indicates that the elimin­ation of distortions is not necessarily desirable under suchcircumstances.. The costs of doing so may often exceed thebenefits achieved, thereby resulting in an overall welfareloss. Alternatively, at any given time, the effects of someimperfections may be offset by those of others, resulting inthe same partial resource allocative and distributionaleffects as t,hose that might pertain under perfect competition"

Those economic distortions which are not covered bythese two categories may warrant scrutiny. Their modific­ation could produce incr'eased social welfare. The generalequilibrium analysis that is involved should attempt t,oidentify all possible effects of policy changes. As such itis more realistic and potentially yields more reliableinformation than partial analysis. The lat.tt:::r is typicallyundertaken with the implicit assumption that the rest of theeconomy remains unchanged.. Consequently, general equilibriumanalyses are logically more likely to ident,ify both situationsin which market distortions offset or exacebate each otheror in which the improvement of social welfare may best beundertaken by the introduction of offsetting imperfectionsrather than "freeing ll market conditions.

The pricing rule indicated by the theory of secondbest states that prices ought to be set so that they differfrom actual marginal costs in proportion to their relevantelasticities of demand. The actual formula is not presentedhere. It is available in many economic texts" It is notparticularly relevant to this paper since it will be shownlater that it is inapplicable because of pI'oblerns of measure­ment.

A further approach has been advocated which adopts atheory of 'third best' which is really a subset of thetheory of second best. This theory centres upon informationpr'oblems. Under restrictive assumptions, the theory ofthird best implies that first best solutions are often moreappropriate than second best ones in the real world ..

222

TRANSPORT PRICING DECISIONS

However the theory is not as yet sufficiently welldeveloped to be discussed in detail at a general forum suchas this" However it does highlight the problem of informationand in particular that aspect of it called mensuration whichis the main basis for this paper"

PROBLEMS OF MEASUREMENT

It has been demonstrated above that under nea-classicaltheory, it is necessary to set prices according to specificrules concerning the relationship between costs and returns,if resource allocative efficiency and welfare rnaximisationare to be achieved.. This obviously cannot be done if costsand returns cannot be measured.

In addition to the impossibility of objectivelydetermining the relative merits of different distributions ofsocial welfa:r:e, insu:r:mountable difficulties are also en­countered in assessing the real costs and benefits associatedwith monopolies, externalities and public goods. Fu:r:'therrnore I

even pecuniary costs and benefits cannot always be measuredsufficiently 'well to allow the application of economic theoryin policy problems because of the difficulties associatedwith joint products and common costs. Finally, non-pecuniaryand information benefits and costs also usually cannot bereadily measureda

Joint Products and Common Costs(l)

Many goods and services are not distinct entitiessince they fulfil a number of functions simultaneously.. Forinstance vehicles of every mode are capable of carrying bothfreight and passengers.. Under such circumstances freight andpassenger services are often joint products and the costs ofsay vehicle operation and providing infrastructure areshared in common between these services.

Economic theory demonst:r:ates that for resourceallocative efficiency to be attained under the assumptions ofperfect competition, the level of output of every good andservice must be determined so that the margin:ll cost of itsproduction equals its ruling market price or average revenue.Determination of marginal costs for each joint productrequires that all common costs be allocated in some waybetween them. When joint products are produced in fixedproportions it is usual and aesthetically pleasing to allo­cate COIIUllon costs according to the product mix ratio.However such costing rules are open to question.

I As indicated below joint and common are used inter­changeably to describe the costs of producing a numberof goods or services simultaneously in the one processa

223

STEEPER

Five general types of production of multiple servicesor goods are cited in economic literature .. (Bressler and King,1970, pp176-181) These are:

(1) Joint services in fixed proportion

(2) Outputs with varying composition

(3) Independent se'I'vices from a single plant

(4) Single seIvices that may be applied to several physicalcommodities and hence may be regarded as differentproducts

(5) Joint services with variable proportions ..

With the exception of the very short run when fixedp:roportions may apply 1 only the variable proportions categoryis relevant to transport services in a pragmatic way -.. Inthe long run, input mixes, technologies of production andhence product mixes all change ovex time"

In the variable proportions case, the levels ofcommon costs attributable to each good or service cannot bedetermined on any sound basis.. The economic contribution ofeach input to each product is always variable depending uponchanges in input, prices and technology.. Allocations ofcorrunon costs in such inst,ances must necessarily be completelyarbitrary" Furthermore, even in the short, run fixed propor­tions case, the outputs of many processes are classified intomajor and minor or "by" products by producers and consumersalike. The marginal cost of producing the latter is usuallyregarded as being relatively low and actual market pricesconsistently confirm this view.. The use of a product mixratio rule for" allocating corrnnon costs is t.herefore probablyinappropriate in such instances"

It is widely accepted in the literature that themost, pleasing way of determining the costs of jointly pro­duced services is to calculate the sum of two components ..Marginal direct costs, that is those which can be separatelyallocated to specific goods or services, should be added topro-rated marginal common costs.. The basis of pro-rating isnecessarily purely subjective and hence implies potentialhuman error. The consequence is necessarily sub-optimal:resource allocation if pricing or taxing strategies are basedupon such values, unless the " r ightll prices are chosen bychance alone.

224

TRANSPORT PRICING DECISIONS

Marginal cost pricing of jointly produced goods andservices and hence the achievement of overall resourceallocative efficiency in an economy with joint product,S, istherefore usually not possible, even if all the assumptionsrequired for perfect compet,ition previously listed hold ..

Non-pecuniary Benefits and Costs

The implicit assumption common to most empiricaleconomic research and nea-classical theory, is that realincome levels and satisfaction gains achieved from thatincome are indicative of each other. This assumption isgenerally considered to be reasonable for aggregates ofhouseholds, however it is untenable for comparisons of thesatisfactions of individual households because of theirunique nature s "

Apart from the invalidity of inter-h0usehold utilitycomparisons, both pecuniary and non-pecuniary elements giverise t,o satisfaction and the relative importances of thesevary markedly from household to household" Because t,heycont,ribute to satisfactions, both pecuniaries and non­pecuniaries have costs and benefits associated with them"Pecuniary elements are traded in markets and can be accountedfor financially by their ruling market prices.. Such valuesare appropriate because t,hey are taken into account byhouseholds when they are faced with decisions concerning theemployment of their limited resources and incomes"

On the other hand, non-pecuniary components per seare not traded in market.s and therefore cannot be readilyvalued in the same way as ordinary goods or services.. Theyar'e generally associated with externalit,ies or public goods"Some such as noise or air pollution can be measured inphysical units.. If they are capitalised in the values ofother resources such as real estate, attempts may be made todetermine their values by economic analysis. Econometricanalyses may also be used to est,imate shadow prices for suchcomponents e .. g" time.

However, such efforts are time consuming and expen­sive, and are also subject to the errors common to allstatistical inference" The costs of such studies areusually considered by Governments to outweigh the potentialbenefits to be gained from the information tlley produce.Subjective judgements by pOlicy makers and conclusionsreached through t,he political process are used in t,heirstead in order to assess the optimum level of output andconsurnptionof non-pecuniaries, ext,er'nali ties and public

225

STEEPER

goods. Regulations rather' than pricing or taxing strategiesare often applied to directly ensure that such levels areachieved, probably because selection between alternativesusually involves making further' j udgements concerning marketresponses.

Determination of out,put levels by objective marginalcost pricing of non-pecuniaries or goods and services withnon-pecuniary elements, is therefoI'e not possible.. Pricingsuch occurrences is essentially a subjective process andtherefore subject to human error.. Shadow fJI'ices must be setby judgement rather than measurement.. Therefo:re, as long asnon-pecuniaries arise, resource allocative efficiency isunlikely to be achieved in any economy except by chancealone"

Information Costs and Benefits

The existence of perfect flows of informationconcerning all market conditions has been shown to be a mostimportant condit,ion for the_ achievement of resource alloc­ative efficiency in both perfectly competitive and fullyplanned economies.. However, in practice, such perfectiondoes not occur and information must be assembled.. In thereal world the costs of achieving perfect information flowswould be infinite.. A trade off therefore exists betweenthe quality of infor"mation obtained and the possible futurebenefits that it engenders and the costs of Obtaining thatinformation"

Costs are incurred in collecting, collating andanalysing data because of the complexity of markets andgovernment directives and administration" In many instances,the costs of assembling additional information on whichdecisions can be based often appear to outweigh the benefit,sthat can be achieved from these extra data. On such occasionsthe decision maker has two alternative options available ..Either information services can be purchasE:d or he must basehis decisions upon subjective assessments from the limiteddata already at his disposal ..

Information service agencies such as brokers andother professional advisers often enjoy greater efficiencyin data assembly than decision makers because of theirspecialisation and experience.. Such specialists are alsooften able to spread the costs of each unit of informationover a number of clients.. Information purchased from suchsources is likely to be more reliable than that assembled bya decision maker in isolation and can be obtained at lowercosta However, such services increase the costs of goods andservices since they increase transaction costs" For thisreason, many follow the second course of action prescribedabove ..

226

TRANSPORT PRICING DECISIONS

The situation is analagolis t,o the trade offs involvedin attempts to value non-pecuniary costs and benefits" Bynot assembling extra information, the decision makerdemonstrates that he is prepared to bear the risk of lossesarising from his relat,ively lI uninformed ll decision" Presumablyhe considers that- such losses will probably be no more thanthose that might. arise if he based his decision upon eitherext,ra self assembled or purchased information.

The larger and more complex the market in which thedecision maker act,s, the more import,ant information costs arelikely to be relative to other transaction costs.. The costsof bargaining depend upon the degree of difficulty thatpurchasers and sellers have in reaching a concensus aboutmar'ket price, which in tUIn is dependent upon the quali tyof information available to them and the size of the market.

In the perfectly competitive model, information isinstantaneously known to all, concensus concerning prices isreadily reached and marketing costs are known with cert,ainty"When information costs are incurred however, the resourcesavailable for other uses are r'educed, given the limitedbudgets of market, operators.. Buyers and sellers will at,temptt,o pass these costs on to each other in order to minimisesuch losses, through buyer and seller resistance to prices,substitution of similar goods or services and by curbing thephysical size of their transactions" Their success in doingso depends upon the responsiveness of quantities supplied anddemanded in the market to price"

The effect of information and bargaining coststherefore, is that consumers and producers do not attain thesatisfactions and profits respectively that they would in aperfectly competitive market. In addition, since the infor­mation available to them is imperfect and unevenly distri­buted, most incur greater' or smaller marketing costs thanthey anticipat,ed. Following each transaction, therefore,some gain more satisfaction or profits than they expectedwhile others are in the opposing position. Hence the tech­nical criterion for resource allocative efficiency is not, metunder real circumstances except by chance alone ..

Other Cost Problems

Several particular aspect.s of imperfect informationwarrant specific discussion since they are most relevant toeconomic questions concerning transport. These are differ­ences between perceived and resource costs and private andsocial costs.

227

STEEPER

Perceived Costs

Households and firms have different bUdgets orreSOurce portfolios and differ'ent preferences for satis­factions derived from the range of reSOurces available tothem. Furthermore, they have different levels of informationat their disposal and the time horizons that they considerwhen making investment or consumer decisions also varymarkedly" For this combination of complex reasons, manyconsumers and producers appear to disregar'd or place zerovalue on many of the true costs and benefits of their acti-vi ties I especially when long term and non-pecuniary costs andbenefi ts are involved ..

For example, they may consider that the costs ofrunning a motor vehicle include only out-of-pocket runningexpenses and may disregard the value of time used in travel­ling,and the costs of maintenance, depreciation, interest oncapital, and the effects of pollution produc~d by theirvehicle a Such costs are not borne immediately even thoughthey must be met ultimately in one way or another ..

Because of their uniqueness, each individual, house­hold and firm perceives each available good or service andthe costs and benefits pertaining to it, at different levelseven though they may be overall the same a Hence, theoreticalexplanations of social, political and economic behaviour areoften not confirmed by statistical tests using data at theindividual or disaggr:egated level ..

The wide variation between individual perceptions andthe inherent imperfections of data and mathematical formula­tions of such theories, often combine rather than offset eachother to invalidate otherwise legitimate tests. On the otherhand, for the same reasons, spurious theories may be con­firmed by statistical testing of aggregated data, because ofthe manner in which the distributions of perceptions anddata and model errors interact ..

Resource Costs

Many of the costs faced by consumers and producersare not resource costs but are simply money transfers withinthe economy concerned.. For example, income, sales andexcise taxes are not resource costs to a whole economy butthey may appear to be to a firm.. In considering the valuesof resource flows, care must therefore be exercised to ensurethat such transfers are not included in economic analyses ofresource use.. This is despite the fact that they influencesupply and demand and therefore effect the levels of flows

228

TRANSPORT PRICING DECISIONS

of resources. The dut,ermination of those costs which arerelevant to an analysis depends upon the level at which thest,udy in the economy is bcl.n r

J conductpd. Those appropriatefor a corporation as compared to thosp per'tinent to anational government for say the same investment project,would most probably be quite different.. (llishan, 1972)

Social and Private Costs

Differences bet,ween social and private cost,s arisebecause groupS or corrununities have prefer'ences that differfrom t,hose of particular individuals and hou;::;eholds. Thecosts and benefits that they face therefore also differ.. Forinstance, while some individuals may not value soil conser­vation works along transport inf:r:astructures, society as awhole could have strong preferences foT. them ..

In a purely competitive market economy, divergencesbetween social and pIivate costs do not occur" They ariseonly because of the existence of externalities, quasi-publicgoods, and welfare distribution problems, since if all goodsand services are traded in markets, all sat,isfaction gainsand losses aI'e fully compensated for in the market mechanism"(Stigler, 1967)

As in the case of externalities and public goods,unresolved differences between social and private costs mayalso be attributed to infoImation problems.. The existence ofsuch divergences could be attributed to the costs of bargain­ing in order to reach a concensus, being considered to exceedthe potential benefits" Otherwise private or politicalaction could be expected to be undertaken to internalise suchdifferences ..

Dynamic Considerations

The discussion above has been based upon a staticworld" Tastes, resource availabilities and technology havebasically been assumed to be fixed and botli the partial andgeneral economic equilibria have been discussed at particularpoint,s in time.

In the real world however, unpredict,able changes inresource availabilities, physical and economic environments,human tastes and wants, and technologies of production andmarketing occur" These ensure that the operation and theattainment of reSOUIce allocative efficiency are muchmore complex issues in practice than the neo-classicalstatic theoI"y prescribes"

229

STEEPER

Unless economic equilibria tend to be stable, statictheory is only I'elevant for short periods of time.. Ifequilibria shift in a consistent fashion, static theoreticalpositions can be compared in sequence in order to give anindication of dynamic economic conditions.. Furthermore, ifsuch movements are of regular form, continuous formulationsmay be used to indicate the dynamics of such equilibria" Thelatter approach has particularly been applied to businesscycles and economic growth.

The major difference between static and dynamicsituations is the extent of risk and uncer'tainty concerningall economic factors. Risk and uncertainty was shown t,o beassociated with problems of measurement and information atparticular times in preceding sections. In dynamic theory,these factors also extend into the future.. Hence while thestatic theory related so far incorporates r:isk and uncertaintyas viewed at a particular point of time, dynamic theoryincorporates continually occurring charges in risk anduncertainty over time" Dynamics therefor:'e add a further:dimension to economic theory.

Households, firms and governments may be consideredt,o hold portfolios of assets including human ones" Thecomposition of these can be changed by investment, so thatthe flows arising from those portfolios are expected tomaximise future yields of satisfaction to their owners.ACcording to this approach, economic act,ions are based uponexpectations of future demands, supplies, costs, returns andlife expectancies, taking into account expectations ofprobabilities that those expected levels will actually beachieved ..

The statement that all will attempt to maximisetheir expected future net satisfaction streams given theirpreferences for risk is the basis of this approach.. Under:such circumstances, a high risk prefer:'er will adopt a highlyvariable but potentially high yielding portfolio, while alow risk preferer will select a portfolio with a safer andhence potentially more stable but lower expected return ..(Tobin, 1958)

In the dynamic situation of the real world, decisionmakers therefore cannot act with certainty, but must formu­late decisions on the basis of expectations. Attempts toevaluate theories concerning t,he formation of expectationshave not been sucessful" Nevertheless it is obvious thatthey must be based to some extent upon past experiencesmodified by predelictions of future disasters, wars, insti­tutional changes and pOlitical and industr:ial activity.

230

TRANSPORT PRICING DECISIONS

Expectations of households will only be perfect bychance alone because of the existence of such events"Therefore, householdr, firms and governments will rarelyactually achieve their economic goals in the ways thatstatic theory indicates, even though they may attempt todo so ..

PRAGMATISM AND PRICING DECISIONS

The Paper so Far

The discussion so far in this paper' initiallydefined perfect competition and showed that pricingdecisions are irrelevant in such an environment" It wasnext demonstrated t,hat transport markets are highlyimperfect and that the simple rules of elementary economicsare not usually applicable to transport circumstances ..Thirdly, the theory of second best was discussed whichfundamentally simply advocates price discrimination as t,heoptimal basis for establishing prices" This is the approachreconunended by professional economists since it encouragesstructural adjustment in an economy.. However as we knowwell, its distributive impacts often lead to politicalaction to control "unfair pricing" practices ..

The second part of the paper reviewed problems ofmeasuring marginal costs and returns.. It, was shown t.hat fora large variety of reasons they cannot be measured accurately ..Furt,hermore even if t,hey could be determined, the costs ofdoing so will often outweigh the benefits achieved from theextra information made available. Moreover, there is theproblem that prices must be set for the future, not for thepast" Hence expectations must. be formulated before pricescan be established and t,hey a.re inher'ently inaccurate"

One is therefore tempted to conclude that marginalcost based pricing determinations are practical impossibili­ties. It must be emphasised that perfect objective infor­mation on any subject has an infinite cost.. Nevertheless,in some instances, a little knowledge may have markedbenefits.

Commercial marginal cost curves can be determined byeconometricians from past data.. It is a very expensiveprocess which is inherently inaccurate but in some instancesmay be worthwhile as an aid to forming expectations onwhich pr'icing decisions can be based.. The first and secondbest pricing rules should not be forgotten as they form auseful basis for understanding the underlying considerationsthat must be taken into account in a pricing decision.. Allthis paper says is that they cannot be practically appliedin most situations" '

231

STEEPER

Charging what the market will bear

The establishment of prices on the basis of costs,whether assessed subjectively or objectively, simply doesnot make sense philosophically from a political or an economicpoint of view.. Nevertheless, costs are a constraint that anorganisation must cover in the long run in order to remainviable ..

Such action will impede structural adjustment in thesector and ultimately within the economy.. If marginal costsare below marginal revenues then it is indicated thatsociety demands more than is being produced" Excess pI'ofi tswill encourage other firms to produce either that particulargood Or service or a substitute for it, t,hereby increasingoutput, decreasing price and enhancing social welfare ..

Alternatively if marginal costs are above marginalrevenue, excess supply prevails and firms will exit from theindustry" The exodus will make resources available tosatisfy society's other demands thereby also usually enhan­cing welfare.. However there are usually costs associatedwith firm failure which may be sufficient to warrant govern­ment intervention.. Nevertheless the adjustment process isusually so powerful that such interventions are temporary"

On the basis of this socially desirable need forcontinuing structural adjustment in the economy, the obviouspricing rule is to charge what the market will bear"Essentially this means equilibrating supply with demand.Supply should be adjusted so that there is either no deficientor excess demand ..

For the operator in a competitive market segment,the relevant price is dictated by market pressures.. He muststrive to keep his costs down below that price in order toremain viable. For him the process of pricing is relativelysimple.. He simply needs to observe the actions of hiscompetitors and remain flexible enough to adapt to changingcircumstances in a timely fashion ..

At the government level however the pricing processis often much more complex" If government operates acompetitive business such as say an airline, it simply needsto behave as the private entrepreneur described above" As amonopolistic operator, regulator or' provider of infrastructurehowever, its goals cover wider areas than mere monetaryprofits" Equilibration of demand and supply may ofteninclude not only user pricing but the imposition of approp­riate combinations of taxes, subsidies and regulations ineither a direct or indirect fashion"

232

TRANSPORT PRICING DECISIONS

In effect a level of output must be establishedwhich society is prepared to pay for ei theI' direct,ly as usercharges or indirectly through fiscal or other internal transfers"In establishing the appropriate level of transfers they arefurther faced with the standard budgetary problem encounteredby all governments - how much can we afford politically andeconomically t,o collect and how do we distribute it, betweenboth the vast number of society's demands for welfareredistribution and the needs of economic management?

As discussed earlier, the process is further complic­ated by such decisions being made now for the future" Theymust be based upon expectations and past performances oroccurrences may be almost irrelevant" Budgetary performanceof Australian Governments over the past five years bearwitness to the great extent to which future economic variablescan be under or over estimat,ed even in the short run"

The process is therefore not, only highly subjective,but also appears to require a sixth sense.. Perhaps we oughtto employ clairvoyants as mat,ter of routine.. A more detaileddiscussion of these factors is given in a recent,ly released

report. (BTE, 1977)

An ideal aim of pricing that is oft,en advanced, is tocover at least joint operating costs in the short, run, butthis may be negated by social demands which require thepayment of subsidies that enable operation at prices whichare insufficient to cover these costs. There are plenty ofexamples of long standing deficit situations in whichoperating costs were obviously not met by revenue alone, butwere covered by governments or by the public at large if notovertly, covertly through overall deficit financing orhighly protective regulations"

The process of pricing obviously requires experimen­tation which can be done on a partial rather than overallbasis, but it must obviously require considerable flexi­bility so that decisions may be reversed if it is found outthat expectations have been wrong.. The economy is continuingto change its structure at, an ever increasing rate" As suchthe stability that enabled successful rate hikes by theapplication of an overall percentage increase no longerprevails.. Increases in real terms are no longer necessarilyuniform acrosS the demands and supplies for different goodsand services or for the same service or good in relat,ion t.oall other goods or services over time"

Overall therefore, pricing of government services isa very complex quest,ion" Fiscal limitations, compensationfor external effects, the provision of public goods,

233

STEEPER

subsidisation of monopolies and welfare distribut,ion questions,must all be traded off against each other.. It thereforeinvolves political judgements which can only legitimately bemade by politicians since they alone are held accountablefor the outcomes of their determinations.. Nevertheless,much can be gained from the experiences of other likeinstrumentalities both domestic and foreign, in a mannersimilar to the approach advocated above for competitiveoperations"

CONCLUSION

This paper suggests that prices can not be specific­ally set according to the simple nea-classical marginal costpricing rule because t,he costs themselves are nearly alwaysindeterminate. It further states that. in the r:eal dynamicimperfect world, such actions would be inappropriate and notaid movement towards maximising society's welfare.. Pricediscrimination should be employed.

However:, this paper does not purport the naive viewthat costing exercises are therefore wholly irr:elevant t,ooper:ating and investment pOlicies of firms and instrument­alities.. Nevertheless, it is argued that they simply mayprovide useful input to managerial decisions.. They cannotbe used in a deterministic fashion as a rational means forprice determination" The later process is necessarilysubjective, futuristic and requires a great deal of flexi­bility if marked resource misallocations and welfare lossesare to be avoided.

234

235

REFERENCES

TRANSPORT PRICING DECISIONS

Cost-Benefit Analysis, George AlIen andMishan, E,J" (1972)Unwin Ltd" London"

Lipsey, R"G" and Lancaster, K" (1956) "The General Theoryof Second Best 11 f ~w of Economic Studies"

Hyman, D"N" (197.3) The Economics of Government Activity,Halt Rinehart, Winston Inc", USA ..

Nerlove, M" (1958) The~namics of Supply: Estimation ofFarmers' R~~se_!:~~~ce, John Hopkins Press, Baltimore,Maryland ..

Nath, S"K" (1969) ~~~raisal of Welfare Economics, A"M"Kelly 1 New York"

ABS, (1975) Australian National ACcounts, 1972-1973,Canberra"

Bressler, R"G" and King, R,A" (1970) Markets, Prices andInterregional Trade, Wiley and Sons, USA"

Freidrnan, M. (1971) Price Theory - a Provisional Text,Aldine PUblishing Co .. , Chicago, 8th Printing"

Ng Y-K (1977) "Towards a Theory of Third Best li ,_PU__bl_l'C_Finance, No" 1"

Robinson, Joan (1934) "What is perfect competit,ion?" TheQuaI·terly Journal of Economics (November),

Stigler, G.J. (1967) The-!heory of Price, Macmillan, NewYork, Third Edition"

Tobin, J" (1958) "Liquidity Preference as Behaviour TowardsRisk", Revi.ew of Economic Studies, Vol" 25 (February)"