Road pricing from theory to practice : A digest of economic, technical and political viewpoints during three decades

VTInotat

Hummer: T 97 _{Datum: 1990-10-25}

Titel: _{ROAD PRICING FROM THEORY TO PRACTICE - A digest of} economic, technical and political viewpoints during

three decades

Forfattare: _{Jan Owen Jansson, Toshinori Nemoto och}

Hans-Erik Pettersson

Avdelning: Trafikavdelningen

Projektnummer: 733 22-0 och 730 13-5

Projektnamn: _{Nya finansieringsformer och prisséttningsmetoder}

Uppdragsgivare: Végverket, VTI

Distribution: Iggi/nyf6rvérv/begrinsad/

Statens vé'g- och trafikinstitut

V" 00

a'

Pa: 531 01 Linkb'ping. Tel. 013-204000. Telex 50125 VTISGIS. Telefax 013-14 1435

Contents gig

1 BACKGROUND AND PURPOSE 1

2 ROAD USER CHARGES FOR FINANCIAL PURPOSES 3 3 ROAD USER CHARGES FOR ALLOCATIVE PURPOSES 6

3.1 The origin of the idea of urban road

pricing 7

3.2 The Smeed report 9

3.3 25 years after the Smeed report 9

3.3.1 The two faces of motorization - the

dilemma of transport policy 10

4 NEGLECTED ISSUES IN THE RECEIVED THEORY

OF PRICING OF URBAN CAR TRIPS 12

4.1 Road pricing theory 13

4.1.1 The elements of the pricing-relevant

cost of urban traffic 13

4.1.2 Cost estimation problems 15

4.1.2.1 The relevant travel time traffic

volume relationship 15

4.1.2.2 Human costs of unprotected victims of

traffic accidents attributable to cars

(OECD, 1985) 17

4.1.2.3 Pollutant shadow prices 21

4.2 Parking price theory 21

4.2.1 The parking market 22

4.2.2 Theory of optimal street parking charges 25

4.2.3 The natural division of short-term parking

and base-parking between on-street and

off-streetfacilities 28

4.3 The problem of car commuting subsidization 29

5 TECHNICAL ASPECTS OF ROAD PRICING 33

5.1 Four principal methods to collect road

pricing fees 33

5.1.1 Paying by cash on the road 33

5.1.2 Card readers 33

5.1.3 Fee stickers 34

5.1.4 Electronic methods 35

5.2 Control methods 37

5.2.1 Background ' 37

5.2.2 Manual control by traffic wardens in

a system with fee stickers attached to the

wind-screen 39

5.2.3 Video registration of license plates 40

\] m UL) m m o xo xo x . m wo o o o w > m e

Gaming simulation model of the decision-making process

The city and players

The structure of the problem The road pricing proposal The decision criteria

Framework to explain the outcomes of decision-making process

CONCLUSIONS

Road pricing schemes must be clear and simple to use, and easy to control Road pricing can give higher quality transport to everybody

Because the value of time differs greatly between commercial traffic and private car traffic, there will be no losers from optimal road pricing

The supporting role of onstreet parking price policy Concluding remark APPENDICES Appendix to chapter 2 Appendix to chapter 3 Appendix 1 to chapter Appendix 2 to chapter Appendix 3 to chapter m a m REFERENCES 53 53 53 56 58 60 64 65 66 67 68 69

My contribution (chapter 6) to this study was made while

visiting the Swedish Road and Traffic Research Institute. First

I wish to acknowledge the various services provided by the

Institute, and the financial support from three organizations,

the Royal Swedish Academy of Science, the Japan Society for

Promotion of Science, and Fukuoka University. A number of

discussions with researchers and governmental officials have

been helpful. I cannot acknowledge all their help and advice,

but particular thanks must be offered to Jan Owen Jansson,

Hans-Erik Pettersson, Hans Sandebring, P. B. Goodwin, Bo. E.

Peterson, Stig Holmstedt, Sten Wandel, Lars Hansson and Henrik Swahn.

1 BACKGROUND AND PURPOSE

The idea of urban road pricing originates from the beginning of

the 1960s. It was quickly adoptedby many transport economists,

but outside that profession it was slow in finding supporters.

With the notable exception of the area licensing system

in-troduced in Singapore in 1975, very little of substance happened for more than two decades after the theoretical breakthrough. In

the second half of the 1980s some significant events have

occurred. Now (in 1990) one has a feeling that the initial stage

of a road pricing diffusion process has started.

Time will show if this notion is right. Anyway, anticipating a

growing interest all over the world of moving from theory to

practical application of road pricing, it seems right to ask the

following questions: What is the state of the theory of road

pricing today? What useful experiences have been made in the

three decades that has passed since the origin of the idea of

urban road pricing? Not just the successes but also the more

numerous failures to reconcile theory and practice have been

instructive. What have we learnt about the decision-making

process? What difference has the technical development in the

electronics field made to the potential of the idea, and which

are the main economic and political considerations to bear in

mind in the implementation stage?

Most of the road pricing literature is British - for good

reasons. British economists were the pioneers in the area, and

the report by the Road Pricing panel (Smeed-committee) set up by

the British Ministry of Transport as early as 1962 provided a

model for the succeeding discussions. A combined Scandinavian

and Japanese perspective on received theory could be a fruitful

Large cities in Japan have been developing toll road networks to

meet rapidly rising traffic demands. The charges are relatively

high, so, although the purpose of the tolls is financial, they

are likely to restrain traffic to some extent. A more acute form

of road pricing, i.e. toll differentiation in accordance with

the degree of congestion, has been discussed, and will probably

The pricing of goods and services has two main purposes: to

generate revenue for financing the production thereof, and to

balance demand and supply (or equate marginal utility and

marginal cost as prescribed by welfare economics) which ensures

an efficient allocation of resources in the economy.

As far as road services are concerned, there are two established

"pricing" systems, which both have mainly a financial purpose:

(i) general motor car and fuel taxation, and (ii) specific

prices of particular roads ("toll roads") which come on top of

the general road user charges. An interesting extension of the

idea of toll roads is the recent "toll-ring" system of Bergen

and Oslo in Norway, where all roads leading into the central

city have toll gates. It can also be mentioned that in New York many approaches to Manhattan, and in Tokyo all major approaches to the central city are toll roads.

In all these cases road finance is the sole purpose of the

pricing. Therefore it can still (in 1990) be maintained that

Singapore is the only place, where the idea of urban road

pricing is applied in practice. The difference between the

systems of the Norwegian cities and Singapore may not seem very

great at first sight. Looking closer at the objectives that

impression does not remain: in Bergen and Oslo the idea is to

maximize revenue on the condition that car traffic should be

only minutely affected.

The majority of revenue by far is, and always will be, generated

by the general taxation of motor cars and fuel. Motor car and

fuel taxes were originally introduced to finance road building.

With some notable exceptions - Japan and the USA in the first

place - the direct connection between the revenue from road user

taxes and road expenditure does not exist any more; the revenue

orders of magnitude. Definitions of road tax revenue as well as

road expenditure differ between countries. In an appendix to

this chapter the Japanese financial organization of the road

sector is outlined).

Table 1 Ratio of total road user taxes to total road

expendi-ture in 1982.

Country Road tax/Road exp

Spain Netherlands Great Britain New Zealand France Italy Sweden Denmark Finland West Germany Norway Switzerland USA Australia Austria Japan Argentina I H wH H va m m m wwws m N m m K O O wa m Q N m C D l -J N N K O

Source: International Road Federation.

In many countries the motor car has become a cherished milch-cow

for the Treasury, presumably on account of the proven inelastic

demand for both cars and car use.

Toll roads exist all over the world (but not in Sweden). Toll

booths are typically found when using high quality facilities

like inter-city motorways, or special links in the road network

like bridges and tunnels. In countries like France, Italy and

Spain where the total revenue from general road user taxation

well exceeds road expenditure, frequent toll roads may seem

somewhat paradoxical. If road user charges in the form of fuel

taxes etc already generates much more money than is used for

road expenditure, whyis there a need for more money? However,

creating severe bottlenecks in a mainly two-lane road network

which lacked bypasses around towns and cities. However, the

financial resources needed to complete the motorway system which the traffic forecasts at that time called for, were beyond the

means of the central government. The motorway programme (some

5000 completed kilometers up to 1990) was carried out by giving

concessions to private and public consortia of banks etc. to

build and run the motorways. This was financed by raising the

required loans against the security of future toll road revenue.

(For further details, see Quinet & Morancay, 1989).

Today, with morethan 40 cars per 100 persons in France, total

fuel and motor car tax revenue well exceeds road expenditure.

However, this money is badly needed for other public services

-education, medical care, etc. - and therefore also future motor-way expansion will probably be based on toll revenue.

In countries where there is no tradition of toll roads, the

financial problem of transport infrastructure development is

today at the top of the agenda of transport policy discussions.

This is one reason why the old idea of urban road pricing has

In Sweden, like in many other countries, the National Transport

Policy recommends that road user taxation should be designed

such that it constitutes economic charges for road services, as

far as possible. The Ministry of Transport is responsible for

the execution of this general recommendation. In the postwar

period several committees of inquiry dealing with road costs and

charges have been set up with regular intervals by the Swedish

Ministry of Transport. The terms of reference have typically

been that within the limits of a given total tax revenue - this constraint is imposed by the Treasury - the structure of charges should be such that resource allocation efficiency is obtained.

More specifically, the idea is that fuel tax should be set at

the level of the social marginal cost of road use, while car

taxes should be regarded as the fixed part in a two-part tariff,

as it were, solely serving a fiscal purpose. Two major problems

are connected with this:

(a) Besides the mentioned interest in the total revenue from

road user taxation of the Treasury, a number of other Ministries

(Trade, Industry, Energy, Environment) as well as various

pressure-groups may influence the level and structure of road

user charges. The strict economic viewpoint is easily superseded

by other considerations. In an appendix to this chapter this is

illustrated by some results from a comparative study of road

user taxation in a large number of countries. In summary it was found that

(i) the size of domestric oil supply has a strong, negative

effect on fuel tax,

(ii) the volume of domestic car production has a strong,

negative effect on the purchase tax on cars,

(iii) the importance of car export (relative to the home market) is negatively correlated to the import duty on cars,

(iv) "poor" countries tend to levy something like twice as high taxes on cars and fuel as "rich" countries.

I x. 'v: :, ': £é. ut .. in h~ v. " . L . _ «w -, ., _ ."t v. 7 * ? «3 ' 4 5 7 ur

is very difficult, and it is almost perfectly commensurate to

the amount of road use. The major problem is that the optimal

level of road user charges is very different in time and space.

A proper peak/off-peak differential is impossible to apply by

means of fuel taxation, and the practical possibilities to

differentiate the petrol tax spatially are very limited. The

social marginal cost of car traffic is much higher in urban

areas (central cities, in particular) than in rural areas, so an

undifferentiated fuel tax compromise will be far too low in the

former areas, and too high in the latter.

In the recent report of a Swedish public inquiry into urban road

user charges ("STORK"), the following levels of the

pricing-relevant cost of car traffic have been calculated:

Table 2 Road user charges for passenger cars as suggested by

STORK, US dollars per 10 km

Central city of Stockholm $ 3.33

Suburban areas of Stockholm $ 1.77

Metropolitan area of Stockholm

outside city boundaries $ 0.70

Rural areas $ 0.35

S93E93: Storstadstrafik 3 - Bilavgifter

Delbetankande av Storstadstrafikkommittén SOU 1989:43

3.1 The origin of the idea of urban road pricing

How different the received view of the matter is now compared

to the days of the road pricing pioneers! The pioneering

con-tribution to the development of the idea of urban road pricing

cal analysis and a calculation of optimal congestion tolls, based on empirical evidence from highway engineering studies. To

cite the brief summary of the paper, it attempted "to apply the

theory of marginal cost pricing to the services of the highways of the U.S.A. Empirical evidence suggests that efficient prices are generally much higher than present levels. This implies that gasoline taxes should be increased and special tolls charged in congested areas". (Walters 1961, page 676)

This was contrary to the conventional wisdom of transport

economists in those days represented e.g. by the influential

treatise by Meyer et.al. on "The economics of competition in the

transportation industries" in the Harvard Economic Studies

series published in 1959. Walters was especially critical of

their alleged marginal cost analysis. "The allocations of cost

carried out in the Harvard study have almost no relevance to

marginal cost pricing at all. And in general they give the

opposite result; on congested roads, for example, the argument

would conclude that there are more vehicles to "bear the cost"

so the tax should be lower than that on similar uncongested

roads. It seems to me that the proper application of marginal

cost pricing ... is practically unknown among highway

economists, and much need to be done to reorient research in the right direction". (Walters 1961, page 686)

Walters qualified his last statement by a reference to the now

classical, theoretical "Studies in the Economics of

Transport-ation" by Beckman, Mc Guire, and Winsten from 1955. There are

also much earlier studies of a purely theoretical nature

discussing basic reasons for private and social cost divergence, which Walters mentioned, which have generally been recognized as pioneering: Pigou (1920) and Knight (1924).

What distinguished Walters contribution from earlier work was

3.2 The Smeed report

Walters had many kindred souls among transport economists on

this side of the Atlantic. Interest in the new idea was also

evoked among some traffic planners and engineers - and

politicians. In Britain the Ministry of Transport set up a

panel, as early as 1962 under the chairmanship of R.J._Smeed,

Director of TRRL with the purpose of examining the economic and

technical possibilities of introducing road pricing. The members of the panel included i.a. M.E. Beesley, C.D. Foster, G.J. Roth,

A.A. Walters, J.G. Wardrop, and J.M. Thomson who was acting as

secretary to the panel. The report by the Smeed-committee

(Ministry of Transport, HMSO, 1964) was the most important work

on the subject for a long time: almost every economic and

technical point made in extensive later literature is found, or

at least foreshadowed in the Smeed-report.

From a technical/economic point of view, it is a masterpiece,

and yet, so far, it has turned out to be "a voice in the

wilderness". The political mood, which in the beginning of the

sixties seemed rather favourable, has not been right for road

pricing anywhere apart from the city-state of Singapore.

3.3 25 years after the Smeed report

25 years later (after the Smeed report) the mood appears quite

different. Now we have a better perspective on motorization in

urban areas, and global environmental problems are looming

large. Many observers predict that the 19905 will be the decade

for a road pricing breakthrough. Why is that? Why is a new mood

3.3.1 The two faces of motorization - the dilemma of

transport policy

In rural areas the car is almost indispensible. In the richer

European countries, three out of four adults living in the

country will soon own a car. Just as in North America, the car

has become an individual rather than household good.

Motoriza-tion is not a problem in rural areas, at least it will not be by the time clean engines/fuels become a standard.

The car is certainly of great benefit for many urban dwellers

too and freight transport by road is a must. In big cities and

central parts of medium-sized and smaller towns, however,

un-restricted car use seems incompatible with urban amenity. The

relative area of central cities is minute. Nevertheless, the

concentration there of human acitivity exposes large numbers of dwellers, workers and visitors to central cities to an unhealthy

environment which is paradoxical in rich countries, and tragic

in poor countries - the ugly face of motorization. During this

century development has been positive in some respects. Heavy

industry as well as wholesale trade has been relocated to city

outskirts. In seaports the original port around which the old

city has grown up, has been replaced by larger seaport terminals

closer to the sea. Two reasons are that modern ships require

deeper water, and modern cargo handling requires much larger

berth back-up areas than is available in city centres. In

addition, nowadays heavy truck traffic through central cities

to/from the port is unacceptable. In cities with early airports,

international air traffic is no longer allowed there, but has

moved far out to airports in more or less rural areas.

The accomodation of car traffic right into the CBD is the excep-tion from the rule. Today, many people are disappointed with the

result of rebuilding the old central cities to make room for an

increasing number of cars. Will this development continue, or

are there better alternatives? Is a transport system based on

the private car inadequate when the population density exceeds a

These are very difficult questions. Economists prefer not to

take a direct stand, but are inclined to answer: Let market

forces determine the direction of urban transport development!

The problem is, that so far, market forces have hardly been

utilized at all. In retrospect it is easy to say that economic

charges for car traffic and car parking should have been

introduced from the very beginning of motorization in urban

areas, so that the demand for road and parking space would have

been properly checked. Thereby, the pressure to pull down old

buildings, and demolish an urban life-style that we now are

missing, would have been eased, and the both literal and

metaphorical "levelling out" of city character in the name of accessability for car travellers could have been avoided.

Is it too late to do something today? The clock cannot be turned

backwards, but by correcting the existing gross market failure,

4 NEGLECTED ISSUES IN THE RECEIVED THEORY OF PRICING OF URBAN CAR TRIPS

The preceding definition of the problem is the general motive

for separate charges (on top of the fuel tax) for urban road

space. There is, of course, a much more specific normative

theory of road pricing, to which we now turn. A more narrow view is taken, when it comes to the specific theory: Therefore, it is

good to bear the wider issue in mind, because without the

popular frustration with the existing traffic conditions and

quality of urban life, road pricing would have little chance of

actual implementation.

Many surveys of road pricing theory have been made over the

years; a recent, good one is Goodwin and Jones (1989). We will

discuss some neglected points first by stressing the importance

of parking fees. As a complement to received road pricing

theory, we will outline a parking price theory. Secondly, and

partly as a consequence of discussing the parking issue, the

fairly well-known, but nevertheless neglected fact that large

numbers of central-city car commuters are heavily subsidized in

a more conventional sense of the word will be focused on. Free

or subsidized parking is one form of subsidization. Another is

company cars used for journeys to work as well as other private

travel. In terms of car traffic reduction, it would probably

have a greater effect to make car commuters pay the total

average costs of work trips by car out of their own pockets,

than raising the private cost of car travel from the average to

the social marginal cost, as prescribed by the theory of road

4.1 Road pricing theory

4.1.1 The elements of the pricing-relevant cost of urban

traffic

The pricing relevant cost that should equal price in optimum

consists basically of three main elements: the cost inflicted on the producer of road services, the cost inflicted on the fellow

road users, and the cost inflicted on "third parties", or the

rest of Society.

_ dTCprod+ Q . dACuser + dTCrest

P .. dQ user

= MCprOd + Q -

d§g + MCrest

(1)

-TC the total cost (as a function of Q) of the

producer

MCprOd = the marginal cost for the producer

user _ .

AC the average cost for the users (as a function of Q)

rest _ . .

TC the total spillover cost for the rest of SOC1ety

(as a function of Q)

rest _ . .

MC the marginal cost for the rest of SOCiety

This formulation of the pricing-relevant cost components applies

generally to all transport services - in fact to all services,

i.e. non-storable goods - where the users take part in the

pro-duction process, as it were, because propro-duction and consumption

are simultaneous. The three components can be of very different

relative importance, depending (i) on what should be assumed

about production factor fixity, and (ii) on the character of the

As far as urban road services for car traffic are concerned, all attention was originally focused on the middle term of (l) - the "congestion cost". This component is, as shown, a product of the

traffic volume and the change (= increase, in this particular

case) in the user cost caused by an additional unit of traffic.

However, it is the existence of the third component MCrest

-that, at least in Sweden, has brought the old idea of road

pricing to the top of the agenda. It is made up of costs caused

by air pollution, noise, vibration, etc., and traffic accident

spillover effects, i.e. accident costs not borne by the

motorists themselves. The relative order of magnitude of the

components of the pricing-relevant cost can only be guessed at,

given the present uncertainty regarding the harmful effects of

air pollution, the difficulties of calculating "warm-blooded" accident costs, etc. In the background material prepared for the

Swedish Transport Policy Bill 1988/89, and in the report by the

committee of inquiry into urban road user charges cited above

(page 7), some tentative figures are put forward, which suggest

that for cars without catalytic converters (exhaust fume

cleaners) the congestion, accident, and environmental cost

com-ponents make up almost one third each of the total

pricing-relevant cost. For cars with catalytic converters the

environ-mental cost component falls drastically to something like a

sixth of the value applying to cars without.

The first term of (l) - the producer marginal cost - is normally trifling: the reason for this is, of course, that the basic

pro-duction factor of road capacity is assumed to be given, so the

only producer costs that are assumed to be variable in the short (pricing-relevant) run are a few minor items of traffic control.

(Road repair and maintenance costs are virtually independent of

the volume of car traffic). In the case of parking fees, it can

be argued that under certain circumstances, the opportunity cost

of on-street parking spaces can be regarded as a

pricing-relevant cost item. Is it not possible to argue in the same way

about road space used by moving cars? Parking space can be

in-creased and dein-creased in a piece-meal fashion, which is not true about road space for traffic. The car traffic capacity of a road

can be changed fairly quickly, at least it can be decreased rapidly, by giving away a whole lane to parking, or to buses, or

part of a lane to bicyclists and/or pedestrians. Anyway, very

lumpy capacity changes are involved, which makes it unsuitable

for marginal cost calculations. It is nevertheless worth

pointing out that the value of the best alternative use of road

space now used for car traffic is significant for the level of

the optimal road traffic charges. The congestion cost component

of these charges should neither be substantially higher nor

sub-stantially lower than the value of road space in the best

alternative use per unit of car traffic forced out by a capacity reduction, or taken aboard by a capacity addition.

4.1.2 Cost estimation problems

When it comes to cost estimation it is right to concentrate on

the main components, i.e. on the congestion, accident, and

en-vironmental costs. We will discuss them in turn and make one or

two points on each.

4.1.2.1 The relevant travel time traffic volume relationship

The congestion cost component is obtained by (i) ascertaining

the appropriate relationship between traffic volume and speed,

and (ii) determining the applicable value of time. A large

amount of literature exists on both topics, neither of which we

will go into. A problem that has intrigued some authors on road

pricing theory and which still seems to be a puzzle, is

connected with the backwards-bending shape of the relationship

between travel time and traffic flow [see Jansson (1969), Else (1981, 1982, 1986), Nash (1982), Button and Pearman (1983) and Kawashima (1988)].

(bl

'1 number of car trips in the morning

Figure l (a) travel time - traffic flow relationship, and (b)

When traffic density in a city road network is very high,

a point like A in figure la is obtained: it is a very bad

situation because the production of car-kilometers per unit of

time is below capacity, and yet travel time per kilometer is

substantially higher than at full capacity. Starting from a

moderate flow level, where speed limits rather than traffic

volume determine travel time, it seems that the marginal cost

goes towards infinity as the capacity flow is approached, and at some stage it appears that somehow the marginal cost will become negative.

This intricacy is hair-splitting. High capasity utilization

occurs in peak-hours, and what happens, as the demand for road

space goes up, is that the peak period becomes wider and wider.

It is not the relationship between travel time and

traffic-volume per hour that should be the basis for the congestion cost

calculation, but the relationship between travel time and the

number of car travellers completing their trips in the morning

peak, taking into account that the morning peak is not of fixed

duration, but will expand into the previous off-peak time. The

pricing-relevant relationship between travel time and the number

of car trips in the morning rises more gently - like the one

depicted in figure lb - than the relationship between travel

time and flow per hour. (For further explanation, see Jansson

[1969].)

4.1.2.2 Human costs of unprotected victims of traffic

accidents attributable to cars (OECD, 1985)

It has been observed in Sweden as well as elsewhere that

traffic-accident risk stays constant with respect to traffic

volume at least up to a level of the volume/capacity ratio which

rarely is reached on most roads. In other words: The number of

accidents is roughly proportional to the traffic volume. This

means that, in a statistical sense, an additional car on a road does not change the accident risk for the other (original)

is equal to the accident risk facing the additional car. This

fact is the basic rationale for the current treatment of

accident costs in calculations of optimal road user charges,

which in a briefest possible summary is this:

The expected accident cost per car kilometre is b - c, where

b is the constant risk and c is the expected cost per

accident. The latter unit cost is divided into three main

components: c = C1 + c2 + c3 where c1 is the human

"warm-blooded" costs of death and injury and c2 + c3 represent the

external "cold-blooded" resource costs, of which c2 is the

expected value of the difference1 between future production

and consumption of victims of an accident, and c3 is the

cost of hospital treatment, etc.

Since the idea of optimal road user charges is to make road

users aware also of the external costs of road use to the

rest of society, the accident cost component in the charge

is normally obtained by taking b(c2 + c3) minus the possible

part of the total traffic insurance premium intended to

cover some external costs.

This principle of calculating the pricing-relevant accident cost

seems wrong when taking into account that different categories

of road users constitute very different threats to each other.

The main difference is between unprotected road users, i.e.

pedestrians, cyclists and motorcyclists, and those travelling in

cars and buses. It must be borne in mind that the accident risk

per car-kilometre, b, is the sum of the accident risk of car

(and bus) users, b1, and the accident risk of unprotected road

users, bII. Therefore, there is also a difference between

private and social costs as regards the human cost component;

the car user is facing a private human cost = bIcl, but the

charge for automobile use should also include the expected human

cost of death and injury suffered by unprotected road users =

In cost/benefit-analyses of road investments in many countries

the human costs of traffic accidents (i.e. the values of lives

saved and injuries prevented by new or better roads) play an

important role on the benefit-side. In the name of efficiency in l.Sometimes the "gross method" is applied, implying that no cons is deducted. This is illogical as the purpose is to calculate th

the resource allocation, human costs should play a fully

con-gruous role in the road user charges; lives of unprotected road

users will be saved by car traffic reduction. If the human cost

values applied by the Swedish National Road Administration for

investment appraisals also were applied for the calculation of

road user charges, an appreciable rise of these charges in urban areas would be indicated, even on uncongested roads.

A large majority of accidents involving unprotected road users

occur on urban roads. In a relatively low risk country like

Sweden, about one unprotected road user is killed, and 10 are

seriously injured per 100 million car-kilometres in urban areas.

These figures are twice to three times as high in some European

countries, and still higher risks apply in other parts of the

world. In a number of developing countries, figures more than

twenty times as high are reported. At present the Swedish

National Road Administration applies human cost values of about US$ 1 000 000 per fatality and US$ 200 000 per seriously injured

person. In the numerical example given in Table 3, it is shown

in the left-hand column what the addition to the road user

charge should be with these human cost values under different

risk assumptions.

Table 3 Additions (dollars per kilometre) to road user

charges on account of human costs of accidents

involving unprotected road users US DOLLARS

Number of respectively and and and

killed and $ 200 000 $ 500 000 $ 1 000 000

seriously

injured per 100

million car kilometres,

Human cost per killed and seriously injured persons $ 1 000 000 $ 1 000 000 $ 1 000 000 respectively 1 and 10 .03 .06 .ll 2 and 20 .06 .12 .22 3 and 30 .09 .18 .33

In Sweden as well as in a number of other countries, human cos:

values have recently been increasing rapidly. There is a growing feeling that serious injuries are valued too modestly; the human

cost of a serious injury, causing life-long disablement should

be looked upon in the same way as fatalities as far as human

cost assessments are concerned. The last column in Table 3

illustrates the costs that would be applicable in that case. It

should be remembered that the current operating costs of cars

are something like US$ .2 per kilometre, so the issue of the

cost responsibility for the sufferings of unprotected victims of

traffic accidents is not just ofacademic interest. In many

cities in the developing world, road user charges calculated in

accordance with the principle suggested here and on the basis of

the life and limb evaluations applied in OECD countries would

make car travel very expensive.

As a final remark, it is noted that a low number of accidents

involving unprotected road users is achieved if "conflicts"

between car traffic on one hand and pedestrians and cyclists on

the other hand, are avoided. What is remarkable is that such

conflicts can be reduced if car drivers show little

consider-ation for unprotected road users. A rather perverse relconsider-ationship

is consequently conceivable: The less consideration drivers

show, the less the component of road user charges will be due to

human costs of unprotected victims of traffic accident: And

conversely, the more inviting car traffic conditions are to

pedestrians and bicyclists "to mix", thanks to thoughtful

driving behaviour, the greater the number of potentially

dangerous conflicts will be as well as the accident costs

attributable to car traffic. It would obviously be absurd if, in

fact, drivers were penalised for good behaviour. This dilemma

points to an important missing element in accident cost

cal-culations - the cost of insecurity felt by unprotected road

4.1.2.3 Pollutant shadow prices

At the moment, the only possible way of obtaining cost figures

for CO, NOx, HC, etc. is as follows: (1) given certain targets

(set by Parliament) concerning total emissions, an efficiency

condition is that the measures taken to reach each target should

be applied to such an extent that the ratio of the reduction

effect to the measure cost is on the margin the same for all.

Pricing with a view to reducing traffic is one measure, and con-sequently the environmental cost component in road user charges

is determined by the cost of various devices to check harmful

polutants relative to their emission-reducing effects.

4.2 Parking price theory

The theory of urban road user charges can be divided between the

theory of optimal road user charges for car traffic, and the

theory of optimal charges for on-street parking. Curbstone

parking charges have existed for decades, and their potential

importance (relative to possible road traffic charges) is by no

means secondary. For a car commuter working in the CBD of a

million-city the parking cost is the dominant component in the

daily travel cost provided that he has to pay the full

market-price for a parking space. The daily parking cost can be

anything from $ 6 and upwards depending on city size, whereas

the private car operating cost of the journey itself is normally

below $ 6. As is shown in table 4 on page 24 in Stockholm the

cost of renting a parking space on a monthly basis is about $

6-10 per day in the CBD. (In central Tokyo for instance, or on

lower Manhattan it is many times more). The pertinent additional

fact, which we will comment on later, is, however, that in most

cities only a (small) minority of car commuters pay the full

market-price for parking.

An important difference between the price theory for road

traffic, and the price theory for parking is that the latter

off-street parking is mainly in private hands. There is a strong interdependency between public parking policy primarily directed to the on-street market, and the private sector parking market.

Before going into prescriptive. price theory for the public

sector parking, it is necessary to briefly describe the entire parking market.

4.2.1 The parking market

The total parking supply canbe subdivided in a number of ways:

on-street and off-street parking outdoor and indoor parking

reserved (private) and public parking

O

O

O

O

municipal and privately owned parking

Furthermore, off-street indoor parking can be divided into

multi-storey parking garages ("parking houses") and garages for one or a couple of cars. The latter can be divided still further into basement and ground floor garages.

This is not an exhaustive list of parking supply classification

criteria, but it should be sufficient to suggest that parking is a rather heterogeneous service. From an economic point of view -in order to understand how the park-ing market works - the most

important distinction is that between the parking facilities

used for "parking at the base" and the parking facilities used for "parking away from the base". Like its owner, every car has

a home, or "base", where the car is kept at night-time, etc. In

the case of a pure company car the base is not a home, but the

car is night-parked at the premises of the company. A private

car (or company car disposable for private travel) used for the

journeys to and from work has - like its owner - a second

"base": the place of work. The characteristics of the demand for base-parking is (i) that each parking period is fairly long-term and (ii) that it occurs frequently and at the same place.

When the car is used for various trips to destinations away from

the base - to the shops, to visit customers, to do different

errands - short-term, occasional parking in many different

places is needed.

Two fairly separate segments of the parking market cater for

parking and occasional parking, respectively. In the base-parking market segment, parkers typically own or long-term hire

parking space, and in the other market segment, i.e. the market

for mostly short-term occasional parking, parkers pay per hour

of parking space occupation (unless they are invited by the

person/company/institution they are visiting to a parking space reserved for visitors).

The following basic economic principle is at the bottom of the

parking market segmentation: If you have a place - a room, a

garage, etc. - to let you will accept a substantially lower

price per day by a tenant who rents it for a whole year than by

a tenant who rents it just for a day. There are two causes of

the economies of long-term letting: The administrative costs

rise, and, more important, the non-paying, unoccupied time will

increase, the shorter the mean period of tenancy is. From a

tenant s point of view, facing a rent structure that tapers off

quite markedly, the question is, whether his use of the place

concerned will be sufficiently frequent to justify long-term

renting. For example, a travelling salesman who visits a

particular city once a year will stay at an hotel, and pay, say

$ 100 per night; if he is such a frequent visitor that he spends

at least one night per week in the city concerned, it may be

profitable to rent a small furnished flat on an annual basis. In

the commercial, off-street sector of the parking market this

natural relationship between prices for short-term parking space

As an example of the price structure in the parking market, the current situation in the central city of Stockholm is summarized

in table 4 below. In the commercial sector parking prices vary

continuously with the distance from the city centre of

Stockholm. The intervals given in the table are mainly explained

by systematic spatial price differences within the two areas

distinguished - the CBD and the rest of the central city. In

contrast, on-street parking prices are perfectly uniform in each

of these two areas.

Table 4 Parking prices in 1990 in Stockholm central city, US

dollars

PARKING Central Business Central City

MARKET District outside CBD

$ per month Long-term "hire" Parking garage 200-300 100-200 Outdoor, off-street I 100-200 50-100 $ per day Medium-term "hire" Parking garage 15-25 1 5-15 Outdoor, off-street - 5 10 $ per hour Paying by the hour

Parking garage 2-3.5 1-2

Outdoor, offstreet 1.67 0.83

Street parking 1.67 0.83

A lot of overlapping between base-parking and short-term parking

markets is often a mark of inefficiency. The most common,

distortive feature is free or very low-priced curbstone parking

in central cities, which results in the streets being filled up by long-term parking avoiding the appropriate but more expensive

As seen from the table, in Stockholm central city a car commuter

will pay more or less the same for on-street parking during

eight hours as for off-street parking on a per day basis. This

is bad, but not as bad as in many other cities, where it often

is much cheaper to park in the street if you can find a space

- than in an off-street parking garage. One common aim of

parking policy is to reserve the street spaces for short-term

parking. However, rather than raising the curbstone parking

price well above the off-street price level, there seems to be a

preference for supplementing time limits, in spite of the fact

that it is more difficult to control that parkers do not exceed the appliable time limit than merely controlling that they have paid.

4.2.2 Theory of optimal on-street parking charges

We will deal with this issue under the assumption that proper

road pricing in one form or another is applied. Parking pricing

policy is very difficult when it is burdened by the task of

restraining traffic as well as balancing demand and supply of

parking_ space. Another basic assumption is that off-street

parking is adequately regulated by market forces. Off-street

parking facilities in the open as well as indoors must compete

for scarce urban land with a host of other possible land uses

-housing, offices, shops, restaurants, etc., as well as

plantations and parks. In the CBD of large cities the land value

is the main off-street parking cost component even in

multi-storey parking garages.

When it comes to on-street parking charges, two possible

approaches are (1) to pursue the analogy with the off-street

parking market, arguing that street spaces should also be priced

such that the ruling land value is included, or (2) to take a

short-run view, assuming that total on-street parking space is

given, and arguing that prices should be set to adjust the

(1) The former approach is all right if the on-street parking

supply can be changed relatively quickly both upwards and

down-wards and the opportunity cost of the land in question is

cal-culable. Then the quantity supplied and the price charged

should be determined simultaneously. The solution will yield an optimal on-street parking price basically equal to the value, or

opportunity cost of the land taken up by a parking space. The

best alternative use of the land could be, for example, to

extend the area of an adjacent pavement restaurant, or to plant

trees.

(2) In case there is no short-term alternative use of the

existing curbstone parking spaces in a particular street,

parking pricing is solely a matter of obtaining the optimal rate of capacity utilization. This is obtained when the price equals

the short-run marginal cost (SRMC), defined, as usual, as the

marginal cost associated with a production facility of fixed

capacity. When calculating SRMC, the all-important fact to bear

in mind is that the demand for the non storable goods

repre-sented by parking has an important stochastic element in it.

The parking demand fluctuates at random (in addition to the

predictable cyclical variations in the demand) which makes a

conventional deterministic approach to costing and pricing

im-possible. For one thing, in a hypothetical deterministic

setting, one would aim at 100 per cent capacity utilization all

the time (Only when unoccupied parking spaces would remain with

a price = O, a capacity utilization less than 100 % could be

justified in the hypothetical deterministic case.) In reality

the mean capacity utilization should be well below 100 per cent,

even in streets where the parking demand is very intense. A

queuing theoretical approach is fitting with the qualification

that motorists who find all parking spaces occupied in the

streets where they prefer to park,_do not literally form a queue

to wait for a space to become free, but look elsewhere for a

place to park. The expected short-run marginal cost consists of

the time and effort consumed by frustrated parkers on the "fully occupied" occasions, which are bound to occur even when the mean

occupancy rate is well below 100 per cent. (For further

discussion, see Jansson [1984] chapter 3.6).

Between these extreme cases the most common cases are to be

found. Rarely, there is no alternative use of curb-stone parking

space. One obvious use is, of course, for traffic. When taking

into account that a whole additional lane could be gained by

eliminating all curb-stone parking along a particular street,

and, conversely, that a whole lane would be lost by setting up

just one parking space in a no-parking street, it appears that

the marginal opportunity cost of parking spaces is likely to

fall initially very steeply. As is illustrated in fig. 2, this

can easily result in two points of intersection between the

marginal cost and the marginal benefit of curb-stone parking.

Generalized cost of parking

A

_{SRMC mm}

Figure 2 Long-run and short-run marginal costs (LRMC and SRMC)

and the marginal benefit of curb-stone parking in a

particular street,.where n is the maximum number of

parking spaces

A particular short-run marginal cost presupposes that the number

of spaces is given. In the illustrated case of fig. 2 it so

happens that the demand curve and SRMC(n), i.e. the SRMC

intersection point 1 gives the maximum net benefit (provided

that at least one parking space is to be provided in the street

in question) the right policy in this hypothetical example is to

provide curbstone parking space along the whole block.

Inter-section point 2 gives the minimum net benefit (that is, a

situation that really should be avoided). Point 3 must also be

considered. Such a corner solution can be superior to point 1.

This is a possibility that has to be examined in each particular case. Some streets should have no parking. On the other hand, in

most cases where parking is allowed, a half-measure should

probably be avoided, and the maximum curbstone parking space

should be provided.

4.2.3 The natural division of short-term parking and

base-parking between street and off-street

facilities

Optimal street parking charging is a fairly complex matter, when

it comes to the detailed structure of the charges. A basic

feature is all the same, that the optimal level of charges

cannot deviate too much from the value of the land used. Were

the average level of on-street parking charges much higher,

more curbstone spaces should be provided, and were it much

lower, better use of some of the existing curbstone parking

space could most likely be found.

Bearing this in mind, the appropriate relative prices of

on-street and off-street parking can be reflected upon: normally

on-street parking should be substantially more expensive than

off-street parking, (a) because the former is provided on a per

hour-basis, whereas garage space etc can be rented on a monthly

basis, and (b), because the cost of the land taken up can be

distributed between several floors in the multi-storey parking

garages typical of central cities.

With such a price structure, which unfortunately is reversed in

long-term parking would arise: long-term parkers could simply not afford to park in the streets, where practically only

short-term parkers would be found. Today, a major problem is that

streets are filled up by long-term parked vehicles, which

reduces the capacity, and obstructs the smooth working of the

road transport system. In some cities car commuters still can

leave their cars in the streets for a whole day free of charge,

'or for a total charge which is many times less than the

comparable land value: What an anomaly!

4.3 The problem.of car commuting subsidization

Parking-policy makers say that a major problem is the fact that

something like half the total supply of city parking space is

"private", i.e. not offered on the open market for parking. In

central cities the majority of private parking space exists on

the premises of private firms as well as public organizations,

and is used by employees (for free or for a nominal charge). A

second category of private parking space exists on the premises

of private houses and blocks of flats. The existence of this

non-market sector is a problem for parking policy aiming at

traffic restraint, since the city government can (so far)

neither restrict the volume supplied, nor put a price on this

kind of parking space. However, the present discussion of

parking pricing assumes that adequate road pricing is applied,

which, of course, means that traffic restraint is not a direct

object of the parking policy. The question is, if the existence

of non-market (privately and publicly owned) parking supply is

still a problem?

It is definitely a problem on account of imperfections in the

rules of income taxation. Free employee parking in the central

city is a huge fringe benefit, which, if it remains untaxed,

(it does in all countries we know of), distorts land use patterns.

Secondly, the owners of the land in question - it can be anything from one or two spaces on "left-over" land in an odd

corner to a large courtyard used for hundreds of parked cars

-have a wrong view on the opportunity cost. The owner may think

that the land in question cannot be used for anything but what

it is currently used for, that is car parking. However, this is

seldom true in a central city, where concrete and asphalt are

too dominating. Some trees or a plantation of bushes and flowers can make a large difference to a dull place. The problem is that the enjoyment of vegetation is a "collective want" to a large

extent, i.e. a public goodz.

Quite a different matter is that the employer provides parking

space for those cars that are part of the production factors of

the enterprise concerned. A large number of jobs in cities

requires a lot of car driving to carry out various tasks. What

should be done with such a car (required during work) outside

work-hours? It can be argued that rather than letting it stand

idle, it should be used for the journeys to/from work as well as

for other private travel in evenings and at week-ends. This

rationale double-use of cars is widespred. Nothing could be said

against this practice, provided that the employer and the

employee using a company car both pay the whole cost involved.

Often this is not so: taxpayers at large are an involuntary

third party in the financing of this fringe benefit.

In an investigation some five years ago, reported in Jansson and

Swahn (1987), we took a reasonably representative sample of car

commutors to the CBD of Stockholm, and asked what they really

paid out of their own pockets for their journeys to work. The

2.The theoretical solution to this problem is to raise the opportunity cost of private parking land. The collective of people in a city (or a neighbourhood in case of markedly "local public goods") should via its government make a permanent offer to buy back "useless" (for anything but parking) land for the going price of adjacent land. The land-owner may not accept the offer. The parking use may be more valuable for him than the money offered. However, as long as the offer exists, there is no economic problem of land allocation.

result which is illustrated in table 5 was, first, quite

un-expected.

Table 5 Typical costs of CBD car commuting 25 km per day in

Stockholm in 1985

Category Private

Category of proportion cost per

car commutor (rough workday,

figures) US $

1 Using the car at

work

1.1 The private car is

used both for the journeys to work and for business

trips 27 % $ 0

1.2 A company car is

disposable for

private travel 13 % $ 1

2 Not using the car

at work

2.1 The full price of

travel and parking

is paid 5 % $ 8

2.2 Parking is

sub-sidized 35 % $ 3

2.3 Parking is

sub-sidized, and the work-trip costs are deducted from income in the

income-tax report 20 % $ 1

As few as five per cent of the car commuters to the CBD in

Stockholm paid the full market-price of travel and parking. At

second thought, the low figure was not very surprising: the

typical "market-price" in 1985 was $ 8 per workday, which was

far more than most commuters were prepared to pay. More

sur-prising was that, for various reasons, some 60% of the commutors

paid very little for their car trips to/from work, less than

In larger cities than Stockholm the full market-price including parking was and is now, substantially higher than the Stockholm

price, so a good guess is that full-price payers constitute a

comparably small proportion of CBD car commuters in all the

great cities of the world.

On the basis of the figures in table 5 above, it can easily be

imagined that if all car commuters in the Stockholm CBD case

were required to pay the going market-price of car travel and

parking out of their own pockets, a dramatic change in the modal

split would occur. So in conclusion, we think that when the

subject of road pricing is up for discussion, one should also

address the question why the subsidization of car travel to work

In this chapter a brief presentation of the technical

methods available for collecting road pricing fees is made and two studies relevant to the problem of con

trolling road pricing systems are presented.

5.1 Four principal methods to collect road pric-ing fees.

5.1.1 Payingvby cash on the road.

This is probably the most common fee collecting tech nique. Besides ordinary selling of tickets that often can be used as a complement to other technically more

sophisticated methods, the method of throwing a

cor-rect amount of coins into a coin box could be seen as an example of this payment technic,

One advantage with this method of fee collecting is

that it is easily understood by the public.

Furthermore, paying by cash on the road makes it easy

to control that the fee really is paid, as it in

prin-ciple is ijSSible* to physically prevent the driver

from passing until he has paid. The most important

disadvantage is that this technique is both time and

space demanding, from which follows that it is citen

impossible to use for capacity reasons. 5.1.2 Card readers

This technic can be used for paying by ordinary credit

cards, but there are also other types _of magnetic

cal s that could be used. There is a punch ticket likerd

rd.vdurxi is valid for Si given number CM. passages

rt

'1 W

11 _{Fv 1}

- time the {maxi is used ii: is receded so tine cor

~44

(D k\ "i

. . . q .

,1 ' .' r. 3'- r 3-K :- * y- Ir .~ :- :. C. :- ' f . :~ w c. . r. *1 1' kl g a r. 1". v f- _

iambei of ixaniining passages .n. poSsibic .n; leg

.+-/-26}: p i F n

could feel insecure about not getting a receipt when using a credit card in this way.

5.1.3£e§__s_tick_era

This method means that the driver buys some sort of

periodical ticket that is visibly attached to the ve hicle. As with paying by cash it s an easily under

stood method and apart from that, the use of a fee

sticker is neither time nor space demanding,

Important disadvantages with this method are that it is difficult and expensive to control and it is almost necessary to make use of some sort of time discount. This system is used in Singapore and planned to be

used in Stockholm. In Singapore, they have a manual

control and the intention is that the control in

Stockholm also should be manual. A study of the capac ity of traffic wardens to do this type of control has been done by the Swedish Road and Traffic Research Institute (VTI). A summery of the result of the study is presented below.

Fee stickers is used also in Bergen in Norway but the

control here is made by Video technique. In spite of

this the control procedure seems to be rather labour

demanding. The procedure is as follows. Time samples

of traffic is video filmed every week-day (Sundays and

Saturdays are free of charge in Bergen). The licence

plates of all registered vehicles are transformed into a computer file.This file is matched to a register of

the licence plates of the vehicles for which fee

r

1

stickers were ooughf 1e actual month Beside the man

(1 #

FL

!

uai - --,- A - f. r '- ' g . . .1 a!" ,- ~ F - - r. .

Lately a technically more advanced systems have been

used. These systems demand that the vehicles are

equipped with a small electronic device that either

actively sends transponder or passively reflects

-responder a particular signal for each vehicle.

The technique is best described in connection to fig-ure XXXX. The electronic device mentioned above is in the figure XXX labelled identification card. This card could be attached to the wind screen inside the car in front of the inner rear mirror. When the vehicle ap

proaches the 'toll station' it is

Ide ntitication

card

National

identification

Video

Monitoring and

of registered

-

registering

car

camera

computer

l

Antenna or

T

Collection of

radar

firstly registered by an inductive transmitter so the

system is informed that a vehicle is on its way. The

vehicle then passes a video camera which at this mo

ment still is passive. Next an antenna in the case

of transponders or a radar device - in the case of

responders , is passed which records the code of the

electronic identity card. The controlling computer

system now make use of this unique code and charges

the road user for the tripp. This charging can be made in several different ways. The road user could allow a direct charging up of his account, a certain number of trips could bee paid for in advance and this number could be reduced in the same way as with a punch tick

et when passing. It is possible to pay for a certain

period of time in advance during which the electronic

identity card permits passages, the driver could be

charged subsequently and so (Hi. If the vehicle lacks

an identity card or the vehicle for some other reason

isn't allowed to pass then the computer starts the

video camera and the vehicle is registered from behind

in such a way that the licence plate can be read. The vehicle owner then can be traced by the national vehi

cle register and then charged a penalty or fee. The

latter is interesting as, if the video registration

can be made reliable enough, there are plans to make

use of the registrations as an alternative mode of

payment for those road users who haven t got an elec

tronic identity card The road user is finaly given

a receipt for the passage by a signal indicating if

the passage was approved or if it was video

regis-tered. Special signals can be given, if for instance a subscription related to the identity card is about to

finish. (J :1 H ") r -T p C) :3 07 o o , :5 U (I) {L H O :5 (D

One disadvantage might be that the technique is rather

difficult to understand although it is easy to use.

Furthermore, there is a suspicion towards computer

technique as it is thought of as a threat to personal

integrity

EiZ Control methods

5.2.1 Background

Usually, we ask in a moralising way why road users

don't obey rules and regulations of the traffic sys

tem. Now let us look at the problem the other way

around and ask what is it that makes nmst road users

behave in accordance to the law. From this point of

view the following ought to be stressed.

- One important reason to follow the regulations is

that we realise that we otherwise would run a serious risk of being involved in an accident. For instance, there are very few people who break rules like driving

on the wrong side of the street or in the wrong direc tion in one way streets and so on.

There is another type of regulation that we try to obey in order to show solidarity with other road users hoping that they will do the same to us. Examples of this type of rules is parking regulations, regulations about the use of silencers on motor vehicles and so on .

A third reason for being law-abiding as a road usera

r

[.

1

.

3 simply that we don't want tt gain criminal records

e we demonstrate tur solid:;ity With the system by

F4

.

CiiCV a lI lg 131311.83- Eil'ld IECUlatlQJS EVER 111101le11 WE 3.1-8? 7;

H.

cori v- inced C): tl iElI USEIU 'I'1 SE? . f "if instance, man '

. r. 1". 1-.'_. .~ P.. - " * m.t'. v " I. r.t . . - pr. r. r' l 1v". A tPp...

-{I -.z~-« -.zI ; '~_L.L 1 ~_. t- I .J Al'4'4 [v.1 '-u" _4\ .4 J '

..

U) ?

-Finaly, we should not disregard that the risk as well as the amount of punishment can be of importance to our willingness to be law abiding.

These causes for being law abiding certainly interact and the importance of the different points to differ-ent types of regulations probably varies considerably. If the utility gained by breaking a rule is marginal and the important reason for following the rule is se curity, then we can expect a high degree of law-abid ing and no special control is probably needed. On the other hand, if the most important reason is solidarity

with the fellow road users or with the system and the

profit to gain is of no importance, traffic-checks are probably necessary.

In order to be obedient to a rule for reasons of soli darity it is probably also important that the

regula-tion is regarded as just, i.e. that everyone loyally

shares the "sufferings" demanded. This is once again a reason to have an effective control of the obedience

to rules depending on the solidarity argument. If the

road users experience that a lot of people break the

rules without. getting punished it would be expected

that the rule will be regarded as unjust and the obe dience of the rule will decrease.

One conclusion from the above is that the quality of

the control method is of essential importance to a

road pricing system. The quality of a control method

to see that the laws of traffic are observed can be

described in many cbfferent aspects depending on the

tytm: of method, In 51 general sense, independent of

(1ype of method, it seems reasonable to coacentrate on

The capacity of the control method, i.e. the propor

tion of the violators that actually is detected and

punished.

The legal security that the method offers, i.e. the risk of wrongly accusing a person of violating the rule.

- The threat to the personal integrity by the control

method.

The first of these aspects is of course the most im portant one as it refers to the general aim of the control. The other three aspects constitute important preconditions for this superior aspect. It ought to be

noticed that the cost effectiveness of the method is

dependent on a number of factors that aren't directly

related to the control technique. For instance the

road users opinion about the relevancy and efficiency

of the control and about the sanctions for violating the controlled rule. It seems to be important to study the road users adaption to a road pricing system

care-fully so that improvements according to information

and control could be dens. This would be necessary to keep the obedience to the rules regulating the system on such a high level that would be demanded to get the system to work in the long run.

Because of the plans to introduce a road pricing sys tem in Stockholm, VTI has studied some aspects of the two available methods for controlling the traffic in

road pricing systems, i.e. manual control by traffic

wardens and control by video technic A summery of the

.d below

results of these studies is present

\T

I

['

impossible to study the detectability of a few vehi-cles missing the sticker in a traffic flow were the

majority of the vehicles had such stickers. The task

of the traffic wardens was therefore transformed to

detecting a few vehicles having stickers among a use jority of vehicles lacking them. As long as the stick ers can be supposed to be attached to the wind screen on a standardised place and the task includes identi fying if the stickers are valid or not, this should be possible. The proportion of vehicles with stickers was

about 10% out of a traffic flow of 500 600 vehicles

an hour during day time conditions and about 25% out

of a: flow of 200 - 500 vehicles during night condi tions. The sticker was 100 x 100 mm with a yellow

flu-orescent background on which the figure 3 or 4 was

printed in 75 mm high signs. The study was run as a

daylight and a night time condition of 30 minutes each a day for four days al togetherTen wardens working independently of each other were instructed, when de tecting a car having a sticker, to read the figure on the sticker and the license plate and the make of the car into a tape recorder.

The wardens managed, as a mean, to identify about 70%

of the vehicles during daylight and about 55% during

night time. How efficient these checks will be in re ality is of course a question of how much staff one is prepared to engage for the task. A serious problem was that quite a lot of incorrect registrations were made,

which means that for reasons of justice, two indepen

dent registrations of a car should be made before any one is fined or punished.

-\ " / vv~ - -.w . -,-4. « K-\ V . a. 5" - «ta v - N H