Director of Strategy and Development Concessions Department
Société Générale d’Entreprise, Graduate of Ponts et Chaussées Engineering School

The population shift from the countryside to the towns has been a gradual one, taking place over the last few centuries. But in the last thirty or so years, the reduction in the cost of automobile transport, and the changing industrial scene (growth of tertiary business and the surge in new information and telecommunications technique) have led to a trend reversal. Towns have become ever more extensive, while inner city population densities have fallen. The demarcation line between city and countryside is becoming more difficult to determine. But the processes used to finance infrastructures have not changed, and hence no longer reflect social or economic realities. In particular, the continuing formal separation between the sources of finance for facilities in urban areas, and in interurban areas, has led to chronic lack of resources, both in the city itself and in its immediate suburban or periurban environment. It is time to seek a change in the way facilities are paid for out of the public or private purse, and in how resources are allocated. The combined analysis of cities, travel and current motorway pricing suggests that the approach already in used for the A14 and A46 roads might well be more widely adopted. The allocation to urban facilities of financial resources generated by the areas lying between conurbations, is consistent both with the new relationship between the city and the countryside, and with associated lifestyles.

This article follows on from articles published in Transport in 1996 and 1997, in issues 377, 379 and 385. Two years of operation of the A14 motorway, and the commissioning of two toll road facilities in Lisbon and Lyon have yielded much useful information for the analysis of toll charges in the urban environment. There is a need to reconsider the problems in Ile de France region, and this is always the subtext to this article. We begin by dealing with urban questions, before moving on to transport matters.


Personal well-being and mobility are not the same thing. Transport policy should aim at the construction of cities with proper amenities and attractive lifestyles. It should not seek to offer the maximum number of transport facilities, but rather to reduce the time spent in travel and thus the real costs of travel. It follows that what is required is actually reducing the number of population displacements over shorter or longer distances. An understanding of this point is critical to grasping the inter-relationship between city planning and transportation. Transport can be apprehended neither as a consumer good, nor as a capital expenditure. It is, rather, considered a necessary evil, the result of our industrial civilisation with its specialised division of labour. The less transportation that occurs, the better people feel. And this is the way some cities have organised their town planning (Seattle). Urban planning policy should offer the services to be provided by the city, at the point where those who really need them are located, out of a concern to reduce the population’s need to travel. The ageing of the population will certainly lead to more thinking along these lines.

This approach does, however, need to be qualified, if only because time is usefully allocated to travel. Travel acts a decompression chamber between two types of activity, between the constraints of family life and those of the world of work. The automotive manufacturers have understood this, and turned their vehicles into cocoons of comfort. Citizens are looking for an appropriately balanced quality of life, in their homes, in their place of work as adults, in the places where their children go to learn, in the amenities needed for family and social recreation. This in turn involves providing space both in and around residential areas, and the willingness of citizens and their families to spend money on the necessary travel between these various locations.

Hence an important consideration: cities are constantly changing and adapting to the existing transportation network. This network itself evolves in a continuing interaction with the city, often too late to meet its real needs (hence traffic bottlenecks), and sometimes too early (the result being that roads are underused and trains are empty).


All the parameters of urban lifestyles interact one with another. This means that it is impossible, on a doctrinal basis, to be for or against the motor car, or for or against public transport. Any urban area requires both means of transport to be able to develop.

Fortunately, there are a number of fundamental considerations which if taken into consideration can simplify the way we think about transportation.

1. Parameters of urban description

In their simplest forms, cities can be defined by four parameters:

– population densities as defined by land-use planning requirements, which brings into play the proportional relationship of built to unbuilt land, on the basis of stated levels of population density, known in French as the Coefficient d’Occupation des Sols (or COS, coefficient of land area occupancy);
– activity density, which is a sociological indicator. This is the sum of the population over a given land area, and of employment within it, divided by the land area itself. Generally speaking, this data is available commune by commune. The basic unit of measurement is the hectare. This activity density, which we shall call AD, varies from 1,200 in the densest areas of Paris, to 20 in the residential estates consisting of detached single-storey housing in the suburban areas. In the district of Paris covered by the Haussman developments of the mid to late nineteenth century, activity density is around 300;

– income density, another sociological indicator, is measured by household declared average incomes. This information is available from the files held by the tax inspectorate, on a commune by commune basis. Although adjustments are needed on the raw figures provided by the authorities, they give a reasonably accurate idea of the relative wealth of the population in urban areas, and are precise enough to be used as a basis for reflection about town and city development. Trends in these figures over time are particularly significant;

– amenity values: such values provide a means of assessment (but not a scientific measure) of the quality of life in a given area. Of course such values are not entirely independent of income density, but there are occasionally places where it is pleasant to live, where the school environment is high quality, where shopping is easy and pleasant, and where property prices remain relatively low…

In the last thirty years, there has been a twin trend towards a reduction in both population and activity densities. The fact that both go hand in hand is to be explained by the less spatially compressed lifestyles now enjoyed. As towns have become more distended, the trend is irreversible. It involves not only more comfortable living conditions, but the set-up of industrial and logistics centres in areas of low population density. However, it is also associated with lower density in the provision of services, and an increase in the operating costs of the urban environment. The trend towards expansion of the urban into the periurban environment needs to be offset by fluid traffic conditions in the Paris inner suburbs, and by improving the overall quality of services available to the population in the inner suburbs, the so-called “small crown” (jobs, retail environment, good public transport, etc). Of crucial importance here are the changes in personal income indicators, which suggest there is a growing segregation of the population by social strata or groupings.

In addition to these general parameters, logistics play a role as essential as often poorly understood by town planners. The factories where the Paris working classes used to work, the so-called “belly of Paris”, with their power stations and massive incineration units, are increasingly removing to areas further away from areas of high population and activity density. The result is urban growth in the outer suburbs (the “large crown”), associated with a continuing increase in heavy goods vehicle traffic. In the areas of dense settlement, heavy goods vehicles account for only a small portion of the traffic (5% on a daily average, and 2.5% at the rush hour). On the major ring roads, HGV traffic accounts for 20% of the whole when close to logistic hubs. While a heavy goods vehicle ,in terms of road surface occupancy, represents no more than the equivalent of two light vehicles, HGV traffic flows represent one third of total road use; and even more, if speed and transparency factors affecting light vehicles and heavy goods vehicles are taken into consideration, as these factors affect both traffic fluidity and the safety levels associated with each type of vehicle.

There naturally arises the concept of protecting the city centre (in the widest sense of the word) from traffic which has no business going there, and particularly from heavy goods vehicles when in transit. In the Ile de France region, one can conceive of a triangle of protection of the areas of dense settlement, the sides of which would link the currently developing logistics hubs located at the apexes. Inside that triangle, only private and delivery vehicles would be encouraged, while transit traffic would precisely be attracted to the outer triangle of


roads where traffic would circulate more freely, if extra facilities were built. As a result, the speed of surface public transport would be increased, and with that its customer base.

2. Urban distortions

Two distorting factors need to be taken into consideration. The first of these is the growth of differences in the quality of life and of disposable income, as between communes. The second is the greater prevalence of congested conditions at certain points in the road network, which further contributes to intercommunal differentiation. Observations have been taken over the last eight years of average household incomes in Ile de France communes (graph 1). Each point symbolises a single commune. Annual income on the x-axis is plotted against relative growth between 1984 and 1992. When a household reaches an income level which enables it to improve its lifestyle, it moves house and into a commune of higher average disposable income. Hence a natural segregation phenomenon is at work, which increases average income discrepancies between communes. This graph shows the mechanism by which inhabitants are concentrated by homogeneous social class. The “gated communities” of Los Angeles are a caricature of this. The main function of urban policy is to reduce such differences, and the quality of the transport network is one of the conditions required for this to occur.

Pockets of low income are the result of the recent de-industrialisation of the industrial sites which came into being at the beginning of the century. These sites have now become logistics areas, and give rise to heavy traffic, making vehicle movement increasingly sluggish. There was a time when the construction of transport infrastructures was felt to be a burning obligation, under close state supervision. The current situation is totally different: decision-making is fragmented, funding is less available, the urgency less, and there is conflict between the urban environment itself and vehicle traffic. The richest areas, those which have had the most long-standing automobile tradition and associated infrastructure, suffer less from traffic congestion than the older low-income areas, where the number of cars is even now still rising very fast, indeed far faster than the road space available. To restrict vehicle traffic in these areas to public transport-only would adversely affect their chances of attracting employment. There is a considerable need for road construction in the north-east and south-east of the “small crown” and out towards Roissy. The problem most urgently requiring a solution is the road between the A4 and the A86 at Joinville.

3. The demand for transportation needs to be assessed against the yardstick of activity density.

Activity density conditions the demand for transportation, as people living the same lifestyles all travel more of less to the same degree. Hence the demand for transportation services is simply proportional to activity density. Graph 2 shows the substantial change in activity density, which decreases as with distance from the densest point in Paris (the historic quarter of the Sentier, the heart of the Middle Ages trading city). In effect this means a substantial, distance-related difference, both directly and inversely, in the level of demand for transport services, and in the space available for construction purposes. Thus solutions which are appropriate in the city centre, are not necessarily to be adopted in the outlying areas.

This subject has been studied in close detail by the General Secretariat for New Towns (V.Fouchier), a study which gives orders of magnitude, and clarifies the choices of modes of transport of inhabitants of the Ile de France region.

4. Areas appropriate for public transport

By jointly taking into consideration the factors of population density as expressed in the COS and activity density, it is possible to define with fair precision those areas where public transport would be most appropriate. In areas where densities top three hundred, the subway or RER are the clear choice. In areas where density ranges between 300 and 120, tramways and light railways are appropriate solutions. When density is less than 120, buses and minibuses are most appropriate, while the private vehicle is the appropriate means of transport in areas of detached suburban housing estates, where density levels are less than fifty.

Graph three shows that the shape of the curves which are descriptive of variations in density, is the same in a large number of historic European cities. But absolute density levels can be rather different. The city centre of Lyon barely reaches the density the of the “Paris “ inner ring.


Graph 1: Reorganisation by homogeneous group. Growth in real income 1984 – 1992
Household income in 1992 in francs 1995 annual

Department 95 Department 94 Department 93 Department 92 Department 91 Department 78 Department 77 Department 75

Graph 2: Activity Density in 1990
Density per hectare
Distance from the Central Paris Sentier quarter district in kilometres

Graph 3 Areas most appropriate to public transport Density per hectare
Distance from city centre (kms)

Light railway Tram


5. There can be no substantial tertiary sector nor high income areas without proper road connections to centres of employment

Centres of tertiary activity are not made up solely of office space! Without shops, restaurants, cinemas and play areas, la Défense would be like the City of London after 5 pm, but worse. Downtown Los Angeles, albeit well served by public transport, has not prevented there springing up other town centres nearby, which are easily accessible by car. It would be highly imprudent to give up providing proper road connections, by simply alleging that there are good public services. Regulation by traffic congestion is not regulation at all, and there should be no question of founding town planning policies on the existence of congestion. The retail business requires fluid traffic conditions on the roads, and the same applies to social relationships in the widest sense of the word (travelling for leisure purposes, meeting friends and family).

The City and Transportation Colloquium, which met between 1991 and 1994, made a number of most welcome observations on densities, on urban morphology, the organisation of space and the rationale for urban travel. The following quotations from F. Damette and P. Beckouche (Strates) are worth noting.

“The structural change (new industries in the nineteen sixties) could be summed up under two headings:

– the new industries and new industrial employment is preferably located in the older residential sector, whereas the older industrial suburbs, at a time which is increasingly critical for traditional industrial employment, are entering into a period of decline;

– in the East of Paris, on the other hand, there will be a development of storage, of goods handling, of haulier activity, of road and rail centres, dependent on wholesale commerce, linking these areas up with the European transport system, and connecting the East of Paris to the Northern and Eastern countries.”

Worth noting also is their comment that large-scale job opportunities occur in areas of high land price. However, in such space there are substantial numbers of jobs for engineers and highly paid technicians, whereas the majority of the staff may well be made up of lower income employees, office workers, and middle management and technicians. In other words, a characteristic of the new Paris is that the major part of the population is unable to find domestic accommodation “in the area where it works”.

One of the reasons for the social imbalance between the East and the West of the Paris region is the longer transport time required from people living in the East, which itself is probably related to the fact that travel distances are longer, even if the speed of travel is not slower. Travel for other purposes, requiring shorter distances that work-related travel, occurs more frequently in the West than in the East of Paris.

This analysis needs to be turned upside down. An area cannot become an area of high quality employment if it does not have the same quality of transport amenities (public and private transport) as the West of Paris. Ensuring uncongested road travel in the East of Paris is therefore a social priority.

6. Parking is a good means of acting on choices of transport modes

Urban road space, insufficient as it is to meet requirements, is in France more taken up by parked vehicles, than it is by vehicles in motion. This paradoxical situation means that in an area of dense activity there can be no differentiation between the problems of circulating and the problems of stationary vehicles. The fact that urban clearways (axes rouges) have had to be introduced in Paris, as in many other capital cities, demonstrates this.

The innovation in the charges levied on stationary (parked) vehicles have solved problems which were previously considered as due to moving traffic. The town centre of Dijon is an example. Surface parking prices are highly progressive with time, whereas they are degressive in underground or multi-storey car parks. This method leads to a good use of the road surface.

However, going too far in this direction can be harmful. Oslo, in its fight against the private car, introduced very high prices for city centre parking. Shops left the centre, and the town council had to reduce the cost of parking, to avoid fossilisation of the city centre. Manhattan and the City of London are good examples of regulation by parking prices in areas of great density, where the roads are narrow, but where traffic is reasonably fluid.

II. —Behaviour of persons and its representation


Travel by Ile de France households has been regularly analysed by EGT surveys. The reasons, modes of transportation and distances travelled for 16,000 households (an average of fourteen per commune), are known once every eight years, and extrapolated to millions of daily movements of the Ile de France population. While a number of overall data show great stability, and are therefore appropriate to the low degree of sampling used, other data are subject to rapid change, particularly data regarding choice of mode of transport. It can be seen that the portion of travel which is work-related has reduced (from 1.4 to 1.0 trips per day between 1982 and 1990) and that travel “for other reasons” is on the rise. These reasons are not at all well understood, but they represent 1.2 trips per day, which is more than work-related travel! The trips made by people over short distances are not properly understood. However such travel is the cause of the greatest disturbance factor in urban districts and in commune centres. An attempt could be made to reduce travel such as this by town planning initiatives. More knowledge of the needs for frequent travel over short distances, as related to different jobs and activities, is in our view essential.

In Lyon, detailed knowledge of the way people travel has arisen from surveys undertaken for the construction of the northern by-pass, particularly after the new facility went into service. The same applied to Marseille (Prado- Carénage Tunnel), and at the Tagus Bridge in Lisbon. Surveys are undertaken annually, if not more frequently.

1. Flexibility

Although car drivers find dense traffic a hindrance, the quality survey in 1991 shows that drivers’ initial reactions are simply to change the time at which they travel. As the time spent at work reduces, and as jobs and times of work become more subject to flexibility, this trend cannot but continue. The second reaction from drivers is to change their itineraries, and the third is to change the day on which they choose to travel. The choice of mode of transport rates no higher than fifth, in terms of driver concerns. It is therefore a hopeless task to force people to take public transport. Rather, it is the responsibility of public transport to meet their customers’ expectations. SNCF national railways advertising “It’s up to us to make you prefer the train” is equivalent to the Renault slogan of “cars to live in”. As public transport is highly targeted to back and forth travel from home to work, it provides a poor response, except in areas of high population or activity density, to a demand for travel which varies substantially, both in time and geographically.

In some cases, car drivers can also reduce the time spent to no good purpose, because limited to transportation, by spending money “buying time”. This is the rationale behind toll roads, which are intended to be fluid, both for inter-city and for town driving. Purchasing time is close to purely commercial decision, as if the driver were buying an article off the shelf. As the intention is to save time and reduce the time spent travelling (although it can never be reduced to nil), people’s willingness to pay must be considered in marginal and not in average terms, and will be highly variable, depending on individuals and circumstances. The travel decision is based on the time which cannot be spent otherwise except travelling, and does not necessarily involve the total time spent travelling.

Quite clearly, households’ willingness to pay for time savings, has nothing to do with the value of time used by the authorities when making economic calculations. The choice of mode of transport (whether or not to use a powered vehicle) for short trips depends on many parameters, among others, the nearness of services and the quality of the urban environment. A single trip can be undertaken either on foot, by bike, or using a powered vehicle, depending on the “quality” of the pavements, the weight which needs to be transported, safety, ambient noise, or air quality. The alternative between public and private transport will also depend on such parameters. A further factor is seasonality. The problem is indeed a complex one.

2. Rigidity

In contrast to the point of view presented above, many people are not in a position to choose their timetables, mode of transportation or destination. These are captive travellers, in thrall to their lifestyles. Amongst such persons are those with children, who are prisoners to their timetables. Or again, these may be inhabitants of suburbs with poor public transport services: they are captive to their cars, in the sense that they waste less time if they use a car, than if they take public transport. Finally, there are those who cannot drive, or do not wish to drive, who are captive to public transport or to the goodwill of other drivers (car-sharing).


All such people exert considerable pressure on political decision-making, as equality of treatment of citizens forces decision-makers to take into account their requirements, and hence to make choices which are outside the market, in the sense that decisions will protect those who have the least choice or the least financial resources, in a manner which is independent of supply and demand. To summarise, a distinction has to be made between forced transportation, and unforced or chosen transportation. In both cases, there is little in common between the two types of travel.

At the supermarket I buy transportation services to my holiday destination. This is an example of unforced transportation.

I am obliged to travel at a fixed and given hour, and pay for my Orange Card in the Ile de France region, or for my travel in a private vehicle. The lifestyle which I require, the way the city is organised and my own financial resources make it impossible for me to have a residence close to the centres of interest in my life (workplace, shopping district, etc.). This is an example of forced transportation.

I may pay for a trip when I go out to do the shopping, and whether this is a forced or unforced action will depend on people, the shopping they do and the day on which they do it…

Up to a certain point, in transportation accounting systems, it might be wise to calculate as a negative GDP value, people’s “forced travel”, and to count as a positive item of GDP, the travel undertaken by people involved in “unforced transportation”. Indeed, “forced transportation” does not contribute to creating value within a city environment, although it is considered on a par with means of transportation reflected in administrative accounting procedures. This methodological thinking could well be pursued to good purpose, and take us out of the institutional framework into a possible quantification of magnitudes.

At all events, it is certain that there are very substantial distortions between the micro-economic and political approaches. Examples can be found, in the crossings of the Tagus in Lisbon, and in the outer Paris ring-road.

3. Practical example: Lisbon

Lisbon is an historic city located on the north bank of the Tagus, against a backdrop of wonderful hills. The north bank is home to nearly two million inhabitants, and the south bank, nine hundred thousand. Lisbon is an estuary city, not a river city, as the distance between the two banks at this point of the Tagus is more than two kilometres, which prevented crossing by land until 1966. In 1966, the first bridge was put up, and in 1998, the second. The south bank was the object of development in the 1960s, particularly its heavy industrial sector. The sector is now in decline, and the tertiary sector, being rich in both employment and employee wages, is expanding mainly from the north bank. Hence, there is a very substantial flow of transportation from south to north. There is, therefore, a pressing need to create a massive flow of south-north transportation. This means a growing need for forced trips; getting a job, crossing the Tagus in a car, by train of ferry.

The current pricing system on the old bridge (crossing which costs 150 Escudos for the round trip), is 5 Francs per day, or 2.5 Francs per trip. Portugal’s purchasing power parity is low (0.7 in relation to France), giving the equivalent of FRF 3.5 in France. This is very low and public transport prices for crossing the Tagus are higher. It had been planned to increase prices on the old bridge to 320 Escudos, to line them up with the new when it was opened. However, politically it was decided otherwise after the demonstrations which came in the wake of the last major increase in the toll, in 1994. The new Vasco de Gama bridge was opened at a crossing price of 320 Escudos, without any change to the price on the old bridge. The extremely rapid rise in traffic on the new bridge (40,000 vehicles a day since April 1998), whereas crossing the old bridge was free, showed that at an absolute value, the price of 320 Escudos was appropriate.

In parallel, the government began gradually regulating the use of cars with a policy on parking. The introduction of metered parking in the centre of Lisbon, at around 150 to 250 Escudos an hour, or 1,200 to 1,500 Escudos a day, was acceptable in social terms, whereas the mere idea of putting up crossing prices on the old bridge, by 170 Escudo per day to 320 Escudos, caused strong reactions. This suggests the fact that people have consciously understood that a car which is not in use and which takes up space on the roadway or pavement should pay (this is a city-dweller response), whereas putting taxes on travel, particularly when the travel is forced, is felt either to be an extra tax, or an impediment to economic activity.


On the old bridge, it was the lower income households, forced into travel, that would have suffered the most if the bridge toll fee had gone up, particularly if it was levied at the same rate throughout the day. Hence the political decision not to do so. From this it follows that pricing should be on the basis of services rendered. The price charged should be low, to match the long time taken to cross the bridge, therefore there would be under- pricing at peak hours for frequent and forced crossings from people on the south bank. Rather than set prices for the whole of the day at the lowest level people would be willing to pay, and hence lose toll revenues (although these could be offset by taxes elsewhere), it would be better to raise prices to the upper toll levels, and make reductions for the small section of the population which is poor and which is forced to cross the bridge at peak travelling times. Studies on this pricing principle are now underway. We do not know which solution will finally be adopted. The position of cars in the city is handled in another way, by means of parking meter policy.

4. Another example: Lyon

The problem in Lyon was how to construct and finance a toll ring-road at a time when the access radial roads into the city centre were free, while at the same time forcing people to take the toll motorway, for the twin purpose of being able to upgrade the radial roads and to increase revenues from road usage. This initial decision was to set up high pricing at around FRF 15 for going through the Caluire tunnel. To force people to use the toll road, the capacity of the free road, whose construction was financed out of general taxation, was substantially reduced, but without any compensatory improvement in the quality of the urban environment. Public reaction was extremely negative, particularly as traffic restrictions covered journeys in the morning from the east of the city, where incomes were lowest, and favoured those who lived in the west. The problem was further amplified by the political questions of the toll road from Roques to Toulouse, which had been settled in favour of road users.

However, once the travelling public had some degree of freedom of choice, and were offered appropriate tolls (FRF 7 for a subscriber and FRF 10 for occasional use), better suited to household expenditure patterns, and to the forced travelling behaviour in the Lyon region, the ring-road to the north became, as it was intended to be, an acceptable alternative. On the basis of two earlier surveys, and of the subsequent change in tolls, an accurate behavioural analysis was undertaken. This threw light on three major factors.

For frequent journeys, people’s willingness to pay (expressed in FRF/hour), is lower as the distances are shorter. It is estimated that this willingness ranges from FRF 20/hour to FRF 60/hour over distances around 20 kilometres. In terms of toll charges, this means that the politically acceptable price varies exponentially and not in a straight line. For journeys of less than two kilometres, the travelling public will not pay for a toll motorway and will stay on the conventional road, irrespective of its condition or size. Here the data from the INRETS Matisse model are apposite. It is estimated that willingness to pay varies from FRF 50/hour for distances of less than eighty kilometres, to FRF 120/hour for distances greater than three hundred kilometres. These observations are corroborated by studies in the United States, where the exponential shape of the curve is regarded as the received wisdom.

The O/D matrix in the evening is very different from the morning’s, because at this time there are many triangular journeys (home-work-other activity-home). In the morning, people go out to work, and have no intention of paying for the right to get up a little bit later. In the evening, the reasons for travel are extremely diverse, and in many cases the public are willing to pay for a faster road facility in order to save on time.

Travel for professional reasons is highly sensitive to toll pricing, at least as much as for private driving. Hence a breakdown of behaviour by travel intentions is not appropriate when studies of this kind are undertaken. Rather journeys need to be reassigned into two categories, those where the cost is felt to be nil, and those where the cost is not felt to be nil.

It is striking to see that cutting the toll charges in half has not changed the level of revenues at peak travel times and has considerably increased revenues during times of average traffic. This proves that the initial price was well in excess of the optimum revenue level. The anti-toll campaign was an emotive response, which quite certainly had a negative effect on the accuracy of measurements taken, but not in any fundamental sense with regard to its overall analysis of the situation.

5. Traffic models in urban areas are improving, but only slowly

The example of the Prado-Carénage tunnel in Marseille, where revenues were substantially less than expected, has caused specialists to take a new look at their models. The information given by the cities of Lisbon and Lyon have provided a much needed extra input to analysis. The logic of current models used to analyse traffic in urban areas is economic and not behavioural. The predominant feature of decisions taken by users and customers of


toll roads is, however, far more behavioural than it is economic. This new understanding will make it possible to improve the old model. Among the reasons why they did not work efficiently are:

– willingness to pay was on occasion not properly distinguished from the values as perceived by road-users or the value set on time-savings; the result was a considerable over-estimate of users’ willingness to pay.

– willingness to pay above all depends on the amount required relative to monthly, annual, household and company expenditures; the models at the same time paid little attention to the frequency-of-use parameter.

– willingness to pay was considered to be a constant, varying in direct proportion to distance, whereas it in practice increases exponentially with the distance of travel;

– over time, willingness to pay grows less fast than changes in gross domestic income. The reasons for this are as follows:

– most of the growth in wealth of the country has been redistibuted by the indirect wealth redistribution circuits, in a way which is not reflected in the employees’ personal back account, and hence is not felt to be a real growth in personal income;

– improvement in the comfort levels of public and private transport, like the increasing use of the Walkman, of the mobile telephone and of in-car radios, has considerably increased levels of comfort while travelling, and reduced the feeling of waste of time, both for forced and unforced travel. In perception terms, the time spent travelling is no longer felt to be completely wasted.

– the algorithm used in the models is not a correct representation of the queuing effect, nor of the impact of queues on user behaviour; Michel Rousselot’s article in the current edition of Transports exposes some of the weaknesses of these models.

– there is insufficient knowledge about the relative number of people who are refunded their travel expenses;

– there is also insufficient knowledge of the relative number of people whose travel time-slots are the result of forced travel, and those who have a degree of flexibility when deciding to journey.

In other words, the transport models used up until very recently were unsuited to the problems actually encountered. This was the case both in terms of the social and economic assessment of transport projects, and in terms of the representation of travellers’ behaviour, at the time of their journey decision. It was therefore only to be expected that elected members had a perception of these matters which was quite other than those of the technicians! The most recent studies of the Centre d’études sur les réseaux, le transports, l’urbanisme et les constructions publiques (Study Centre on Networks, Transportation, Town Planning and Public Constructions) shows that our understanding is going in the right direction, and that discrete choice models are a more precise description than was earlier the case of travel behaviour in the urban environment.


In France as a whole, in 1996, car-related expenses amounted to FRF 665 billion/year, breaking down as FRF 200 billion on the purchase of new cars (1.8 million units), FRF 25 billion purchase of second-hand cars (4.5 million units) and FRF 440 billion in operating costs. This breakdown by type of expenditure has remained unchanged in the last ten years.

1. The automobile park

The automobile park is gradually stabilising (graph 4). No-one knows the future limits to that park, but observation of levels reached in the United States and in Germany show what might happen in France. After rapid expansion between 1955 and 1990, the growth in the automobile park is levelling off asymptotically. In terms of growth in national toll revenues, little can be expected from growth in the vehicle park, this at a time when new charging structures for motorway tolls, more appropriate to the services rendered by the motorway itself, would bring in extra financial resources. Note that the first motorways to be built were on the historic Paris-Lyon-Marseille road, and that second generation motorways (in the Alps, the Pyrenees and the East-West


axes) have laboured under some financial difficulties, affecting all operators, with the exception of Cofiroute last year, which scraped into the black.

VAT, TIPP, customs duties, and the vignette road fund licence are the largest sources of financial income for the local authorities, and raise more than FRF 200 billion/year. In addition there are specific toll charges, which are initially assigned to investment, and gradually reassigned into the creation of “funds” which recycle the revenues from road tolls into the more fashionable investments of the moment. Examples of this are the motorway toll (FRF 26 billion/year in 1997, parking fees (approximately FRF 1.7 billion) and to complete the arrangement, the facilities toll (FRF 0.6 billion approximately).

GRAPH 4: Number of private vehicles and commercial vehicles per thousand inhabitants in France



Urban projects Forecasts

Total forecast
Low total forecast
Number of private vehicles Number of commercial vehicles High total forecast

GRAPH 5: Kilometre revenues – 1996 Millions of Francs including VAT per km Distance involved in km

To represent the financial flows arising from toll charges, it makes practical sense to classify motorway sections by decreasing order of revenue (Graph 5). Motorway sections are clearly profitable when they serve areas of dense population (left portion of the graph). The extension of the motorway networks into increasingly low population density regions is equivalent to extending the curves to the right. Revenues from sections already amortised are reallocated to cover the expenditure on motorway sections where traffic is low. The financial situation of motorway companies is broadly healthy. Although they are burdened by substantial debt (FRF 160 billion today), revenues at current levels mean that they can service their debt. The currently envisaged extension of the term of the motorway licences would give breathing room for a number of extra investments, which could not otherwise be financed by pure project financing. The question then arises: should further investments be made on motorways extending into the countryside, and increasingly further away from conurbations?

There is another way to look at the financial balance. The sections already amortised could provide support to the construction of motorway sections where the investment outlay is extremely high, in particular the near suburban sections. It is here that the future of the motorway construction companies will be decided, depending on whether or not there is a solution to the chronic lack of funding, and hence lack of investment in the transport field, in Ile de France.


In 1996, in the Ile de France region, expenditure on transport amounted to FRF 201 billion. Car-related expenditures amounted to FRF 150 billion, of which FRF 138 billion was borne by drivers, and FRF 12 billion by employers. Expenditure on public transport amounted to FRF 42 billion, of which FRF 10 billion were directly paid for by road users, FRF 14 billion paid for by employers, and FRF 18 billion out of general taxation. Taxi-related expenditure was FRF 5 billion, half of which was refunded by employers. The FRF 4 billion spent by two-wheeled vehicles rounds off the total.

Investment expenditure in Ile de France was FRF 2 billion/year and per road, and FRF 3 billion on public transport. Funds are available and make it possible every year to build a 1km section of motorway , 1km of metro (the equivalent of 5 km of tramway), and to improve tracks and trains on the SNCF French national railway carrier. This means accepting the unpalatable fact that piecemeal initiatives are in many cases too weak to act directly on the shape and population density of the different portions of the city.

The challenge in political and financial terms facing the current masterplan is rapidly to gain access to funding of FRF 2 billion extra/year, or 1.5% of the household transportation budget in France. The percentage is a low one, and revenues in such an amount could be raised, provided there were a new service, which would meet the expectations both of households and of transportation companies, provided that such revenues were not considered as an extra tax, which people would not understand.

Note also that the consumer price index is no longer the main guide for public decisions in our field, and that customers are increasingly sensitive to the improvement of their city. If they can have it explained to them the reasons for which some prices are increased, they will be ready to accept them.

2. Cost indicators

Note that the term “transport cost” should be defined with precision, as it in fact covers a number of very different notions:

– the monetary cost paid by the traveller, which is the equivalent (for public transport) to the price charged by the carrier, and which should (for car journeys) include the cost of parking.

– the framework of travel charges as perceived by the traveller, as a factor in the decision to repeat this mode of displacement (mode, timetable, group travel with other people, cancellation). This framework cost is very different from the actual financial charge paid by the travelling public, as it includes reductions, direct reimbursements by the employer, and reimbursement of indirect taxes by the tax department.

– the perceived monetary cost, increased by a factor representative of the difficulty of transport, is assessed by time value, itself a multiplicand of transport time. The perceived monetary cost is believed to represent a real assessment of the cost of travel to the traveller.

– the monetary cost to the collective interest, which is an aggregate of the monetary outlays of the travelling public (making allowance for refunds of travel expenses), and the costs borne by investors through investor savings collection subsidies on, among others, SNCF, RATP, private lines, road construction and maintenance.

The overall cost to the collective interest, which is the aggregate of the above-mentioned costs and the “time cost” spent in transport, is assessed by the conventional value set by the government, in its statistical estimates of time spent in travel and of the costs incurred thereby.

It is also clear that travellers’ decisions are taken in the light of the relative (and not absolute) values set on the perceived costs, and that, so long as such relative values are at variance with the cost to the collective interest, individual behaviour will not spontaneously go in the direction of overall cost optimisation (Graph 6).

3. Non-monetary parameters

Inhabitants of the Ile de France region take 22 million trips/day in powered transport, the average length of each journey being around 30 minutes. If the conventional value of FRF 75 is applied to estimate “time spent in transport”, the cost is FRF 300 billion, which is one and half times the total real monetary outlays, and equivalent to 14% of the Ile de France region GDP! It would of course, be ridiculous to use this estimate as a basis for concrete decision-making.


A recent Anglo-Saxon study, covering and in part confirming the study of P. Bovy and I. Salomon, evaluates the economic loss (= “time wasted”) in Europe, due to road congestion, at 0.25% of GDP. If the figure for Ile de France is considered to be around this average, the cost of road congestion might be assessed at FRF 5 billion annually. Given the fact that the Ile de France region is the area where most road congestion in France is concentrated, we have no hesitation in doubling the cost to FRF 10 billion annually. This is equivalent to assigning a cost of FRF 40/hour to time spent unproductively in traffic congestion.


Table 6


Starting point – Destination


Car (1)

Public transport

Distance as the crow flies

Coût ressenti

6 km

46 km

Home – shopping/leisure – home

Using the French state planning discount rate, economic considerations alone would justify tangible investments of up to FRF 120 billion – a figure exclusive of the associated operating costs – in order to eliminate road congestion as a cause of recurrent economic loss. The unavoidable operating costs related to such extra investment would limit the amount of economic outlay in immediate investment terms to around FRF 100 billion. However, no such investment outlay is or has been forthcoming. This shows how difficult it is to allocate public monies to the reduction of non-monetary costs, which are mainly incurred in the pursuit of private interests.

Coût < Pour la collectivité

Coût ressenti

FRF 19 (bus + RER + metro (5) FRF 38 (VP + RER + metro (6)

Coût pour la collectivité

Paris – Paris

Home – work – home

FRF 3 (2) FRF 75 (3)

FRF 93

FRF 4 (4)

FRF 13.5 (metro) FRF 23 (bus)

Home – shopping/ leisure –home

FRF 22 (3)

FRF 48.5

FRF 8 (5) FRF 8 (6)

FRF 13.5 (metro) FRF 24 (bus)

Inner “crown” – Paris

Home – work – home

FRF 8 (2) FRF 80 (3)

FRF 113.5

FRF 5.5 (4)

FRF 29 to 30
(metro or bus + RER)

16 km

Home – shopping/leisure – home

FRF 25 (3)

FRF 62.5

FRF 11 (5) FRF 8 (metro); FRF 30 (bus + RER) (6)

FRF 29 to 30
(metro or bus + )RER

Outer “crown” – Paris

Home – work – home

FRF 20.5 (2) FRF 92.5 (3)

FRF 175.5

FRF 9.5 (bus + RER + metro (4) FRF 11 (VP + RER + metro (4)

FRF 64 (bus + RER + metro) FRF 79 (VP + RER + metro) (7)

FRF 34 (3)

FRF 103.5

FRF 64 (bus + RER + metro) FRF 70 (VP + RER + metro) (7)

Paris – suburbs

Home – work- home

FRF 12.5 (2)

FRF 122

FRF 7.5 (4)

FRF 42 (metro + RER + bus)

25 km

Home – shopping/leisure – home

FRF 20 (3)

FRF 76.5

FRF 15 (5) FRF 40 (6)

FRF 42 (metro + RER + bus)

Inner “crown” – Inner “crown”

Home – work – home

FRF 4 (2) FRF 44 (3)

FRF 65

FRF 4.5 (4)

FRF 28 (bus)

8 km

Home – shopping/leisure – home

FRF 15 (3)

FRF 38.5

FRF 9 (5) FRF 16 (6)

FRF 28.5 (bus)

Outer “crown” _ Outer “crown”

Home – work – home


FRF 37

FRF 4.5 (4)

FRF 18 (bus)

10 km

Home – shopping/leisure – home

FRF 3.5

FRF 23.5

FRF 9 (5) FRF 16 (6)

(1) Vehicle occupancy rate: For home-work travel 1: for home shopping or leisure travel 1.5 (2) Free parking in a space provided by employer
(3) Paid car parking
(4) Orange Card reimbursed up to 50% by employer

(5) Orange Card;
(6) Public transport tickets in books of ten
(7) Taking the RER express subway from private vehicle parked at take-on points

FRF 18 (bus)


The general context today

The conurbation at the heart of the Ile de France region needs to be understood within the framework of European competition among capital cities, which provides a sharp incentive to the construction of transport infrastructures. Berlin, Milan, Lisbon and Barcelona among others, have recently been transformed, and have concentrated on the area or areas linking their airports to their historic city centres, and to urban districts which are now undergoing profound change. We need to act like them!

What are the dominant features of the present system? Technical aspect

Traffic in the east of the region is far more affected by heavy goods vehicle traffic than the west, and closer to saturation levels at all times of the day;

The major road problem in the east is the A4 to A6 link, as was forecast twenty years ago;

Saturation levels on the major axes are considerably closer to being reached when the same road system carries highly differentiated traffic (HGVs and private vehicles). This is the result of differential speeds at any one time, and the accident risk arising from this;

The HGV tunnel on the A86 (west) is part of the rationale of by-passing the conurbation itself (HGV triangle). The light vehicle (small tunnel) could justifiably be a toll road, as this area is one of high personal income. If there is no direct payment made by road users at this point, none will be made at any other point. Road services in the East of Paris would continue to be free of direct charges, and hence this region would receive favourable treatment, relative to the West of Paris;

The A4-A86 link in the Bois de Vincennes-Joinville-Nogent sector is virtually unavoidable for HGVs, as the A86 between the A6 and A1 operates as a radial road through the industrial and commercial access roads into the Paris region. The only alternative in the densely built-up areas is the Eastern ring-road between A6b (Porte d’Italie) and A3 (Porte de Bagnolet), whereas the first motorway to be saturated in the morning, the Franciliene, which is a major thoroughfare for light vehicles, should be upgraded.

Financial impact: financing, tolls and regulations

It is impossible to finance the necessary new road infrastructures out of public funding alone. The requirements for secure rates of return from private funding require the existence of private companies in excellent financial health looking for long term returns (pension funds), which could be attracted to investments of this kind. Such pension funds are very few and far between.

Politically speaking, it is impossible over the short term to set up toll motorways in the East of Paris, without having first demonstrated, in psychological and behavioural terms, how a toll road would work. Such an approach should be tested in the West, on the A86 after the A14.

Generalised urban toll roads would in technical terms, be monstrously unwieldy, and would have very little chance of rapid deployment, due to practical, legal and political difficulties. The only alternative might be the creation of a special vignette (road fund licence) for the Ile de France region.

In the East of Paris, the only toll system that would be politically acceptable, and have a future ahead of it in terms of raising revenues, would be that levied on heavy goods vehicles. Urban toll roads become a considerably simplified solution if the number of vehicles affected is reduced, and form a limited and homogenous category.


Class 1 Class 2


Class 3 Class 4 Class 5 T otal Friday

Road hauliers are part of a sensitive category, with its own requirements and constraints. Among these are the time factor (just-in-time deliveries), as well as cost factors. Sensitivity on the part of the industry arises because road haulage is excessively competitive, not so much against other means of transportation, but within its own industry. For this reason, the rationale behind any agreement to pay for road usage should be the subject of extremely detailed study, in association with the road haulage industry itself.

A rule such as levying a toll, or enforcing prohibitions on traffic, has more likelihood of being accepted if it is applied to all without exception. Modulating per-kilometre pricing as a function of distance would make it possible to segment the customer base. This would go a step in the direction of lightening the load of heavy- goods vehicles on the road infrastructure, and be a step to the introduction of more equitable tax levies on the long-distance hauliers, who are heavier users of services than short-distance carriers.

The existing proposed measures for heavy vehicles are prohibitions. But these are not attractive to the industry, nor indicative that any value is set upon it. The introduction of specialist carriageways reserved to HGVs, on the other hand, would be psychologically beneficial, even if such carriageways were subject to toll charges.

2. Some ideas for sourcing new revenues in the future

As a means of raising financial revenues, subjecting vehicle traffic to financial levies is a classic approach. The whole problem is to avoid the money so raised being allocated to the general government budget, which is not in a position to provide proper protection to the funds raised, supposedly for “ investment purposes ”, given that the demands made upon that budget for operating costs are unlimited.

Toll payment in return for uncongested traffic conditions, charges raised by motorway vignettes (road fund licence), and tolls levied on rapid urban thoroughfares, are alternative toll concepts. A detailed analysis of the different potential sources of finance shows that there are few hopes of raising substantial revenues from these sources. Irrespective of the principle adopted, the area subjected to taxation would be suburban, and the beneficiary area would be the areas of dense activity or population, these areas already being the best endowed in terms of income levels and of transportation resources. To adopt such an approach would be politically very difficult.

The area toll system in Oslo has often been cited as an example of a toll system up which acts out of a combined concern to improve town planning procedures, to encourage local transfers, and keep general traffic levels down. In practice, however, the opposite has occurred. After a decision on town planning and travel approaches (the main aim was to ensure a multiple lane road connection from the western residential area to the eastern industrial area, to avoid crossing the town centre by private and heavy goods vehicles), the city of Oslo then sought financial resources and went to the central government and to the population.

The Norwegian government subordinated its agreement to participate to the introduction of a toll system, and the population accepted this. The toll access set up covers the poorer areas of the city (which could not occur in Paris). Tolls in the area are only raising money for financing purposes and the funds so raised have been 70% allocated to the road infrastructure itself, the remainder being allocated to public transfer infrastructures. The cost of toll collection is relatively substantial, at 12% revenues.

If area or rapid urban thoroughfare toll systems were to be introduced, they would at all events raise their own financial resources. There then exist two possibilities. Either prices charged are high, and tolls are dissuasive in respect of journey or modal transport decisions. Alternatively, road charges are low, and here the toll is only used for the raising of finance. This begs the question as to who would dare set up a toll road which would be dissuasive in effect? This is the problem Dennis Package encountered in Stockholm, in spite of the high degree of sensitivity to these questions of the socialist government of the time, and a low toll charge (FRF 8). It should also be added that steps should be taken to ensure that, if traffic is displaced, this occurs on a modal basis, without rerouting traffic into secondary routes. It would be illogical and unacceptable for this to occur, as the city


centre would be even more congested, buses slower and the whole scheme would contribute to making the city more extensive!

3. A difficulty shared by all urban toll roads: charges on rush hour traffic

In the rush hour, there is head-on opposition between two concepts.

The economic concept of scarcity of supply is an incentive to make the car driver who takes his or her car out at rush hours, pay the top rate (particularly if the driver is a captive one). Under this system, taxation and the social security system should take care of the economics of the equitable redistribution to the population of the revenues raised by road charging.

The political approach, however, looks in detail at the sort of people who take their cars out in the rush hour, and ascertains if they are really forced to do so, whether they actually can pay, or change their travel time-table, or go over to public transport, if any. Under this approach, people are regarded as captive, but not thereby financially penalised.

Under the first approach, there is a tendency to require too high a price from rush hour travel. Under the second, too low a price, making the service rendered less good, the users being captive. On Paris suburban roads, a road tax is equivalent to making people who live far away from the centre, often for reasons of low income, pay for keeping the road network uncongested, for the benefit of those living closer in.

As a result of the observations of a large number of existing toll roads, we identify two fundamental parameters to be applied when charging for road use. The first is the frequency of use of the facility, and the second, the degree to which its use is forced. If there is a low degree of forced use of the road, the principle of higher charges should prevail. If there is a high degree of forced road use, the under-charging principle should be given priority, (see Transports n°385, September-October 1997).

4. An unexploited source of revenues: Ramsey-Boiteux tarification on motorways

The approach to toll motorway charging in France is still based on a straight forward per-kilometre pricing. The only original feature in the motorway toll system at present is in the HGV sector, where a subscription rate for frequent users applies. This rather basic approach to charging is probably the result of a long-standing phobia of motorway operators regarding heavy goods vehicle traffic, due to the extra road maintenance costs caused by high axle weights.

However, by keeping a large part of heavy goods vehicle traffic on toll free roads, the current policy is not only very far from optimising the revenues which could be expected from HGV traffic, but also makes the burden of maintenance fall on the shoulders of the central government and local departments, although this could be avoided, if HGV traffic was more concentrated on the main thoroughways, namely on the motorways themselves.

If there were a proper knowledge of the variations in the willingness to pay by HGV operators, as dependent on the distance travelled, it would be possible to set up incentive commercial policies for motorway charging, particularly for unforced users. Allowance could be made for distance travelled on a non-distance-aggregative basis. This could be massively and sustainably attractive to heavy goods vehicles, and is possible with the use of today’s modern technology.

As far as the licensed motorway management companies are concerned, the motorway network is their sole asset, and their fixed costs are very much larger than their marginal operating and traffic-related maintenance costs. Also, their only competition is the toll-free road network, in particular on short and medium distance journeys. This is because competition, at best theoretical, from other means of transport (rail, river and air), is in practice extremely low, and affects only marginal segments of the market.

As toll motorways are not, with a few exceptions, saturated, Ramsey-Boiteux tarification is appropriate. This involves charging by customer segment, in amounts equal to variable costs and a portion of fixed costs, allocated in a manner inversely proportional to demand elasticity for each customer segment.

Current technology also makes it possible to charge individually for each heavy goods vehicle, even in urban areas where there are no toll gates. Whereas levying tolls on all of the ten million vehicles which are regularly on


the roads in the Ile de France represents a technical challenge which, at present, cannot be met, the application of toll technology to the heavy goods vehicle fleet is now possible.

What are likely reactions from the road haulage business?

This is a hyper-competitive and extremely fragmented environment, in whose eyes motorway tolls are extra costs to be avoided if possible (except when there are prohibitions forcing motorway use). Payment of motorway toll charges is accepted by hauliers, but on a case by case basis, depending on the time factor (reliability of the route chosen (toll or toll-free) and time taken to travel it, as opposed to its alternative).

The general aversion to tolls in the haulage industry is probably due to the fact that, as excess charges, they are generally greater than the net profit that carriers can expect from conducting their business. Competitive advantage is likely to be the only effective incentive to the industry to accept the tolls which it at present rejects. But the motorway usage pricing policy at present in force is trending in the direction of equal charging, which is contrary to customers’ expectations. The notable exception among the French motorway licensed operators is the Fréjus tunnel, which is in competition with the Mont Blanc Tunnel and the Mont Cenis Pass. Fréjus has introduced an aggressive commercial policy to increase its market share. This has involved it in a suitable flat- rate frequent user pricing system which is in the main targeted to specific customer segments, and offers them the loyalty bonuses, on a basis quite different from that of the standard frequent user bonus.

5. New principles for road charging

The first principle is thinking about potential customers in segments, and devising a pricing policy suited to each. The local carriers using short distances should be given an incentive, as the extra costs over short distances (of for example driving to connect to the national motorway network) are proportionally greater and the time savings lower, than on very long distance journeys. Such measures would also make national companies more competitive, insofar as they are geographically well-represented in France.

Another factor is that, due to the extreme sensitivity of the haulage industry to price factors (which in particular affect the smaller carriers, witness the regular demonstrations and protests of the French haulage industry), consideration might be given to setting up a more socially-oriented tariff system, to capture a non-captive customer base with a little disposable revenue.

The second principle should be to create customer loyalty. A good way of doing this is, as the airlines do, to provide loyalty discounts to drivers who frequently decide to use a paying section of the motorway, by offering mileage bonuses. But the haulage company itself should also receive a loyalty pay-out, in the form of bonuses dependent on real loyalty to the motorway, or if not (and this is an easier option) on the basis of regular use of that motorway.

The third principle is to encourage the use of the network by creating flat fee charges over distance-related segments of the road, which give entitlement to reductions on the basic charge.

The fourth principle would be to set up a bonus system, which could be tied into other discounts, and which would be triggered by compliance with traffic rules and the adoption of proper driving behaviour on motorways by all the vehicles in the company fleet.

A fifth principle would be to make a connection between the charges payable and the quality of service really offered to the haulage company, for example in terms of travel times, with the allocation of bonus points in the event of unpredictably poor traffic conditions.

Such a commercial policy should make it possible for motorway companies to increase their revenues by several (possibly some 3 or 4) billion French francs. A glance at heavy goods vehicle traffic on the A14 (Graph 7) bears this out. Traffic is no more than 1,000 HGVs per day, as against 22,000 private vehicles. The overwhelming majority of HGV traffic remains on the A13. Is the current charging system particularly intelligent, for a road link between Orgeval and La Défense, the construction of which cost FRF 10 billion?

Map 1: Expressways in Ile-de-France


Areas of change
Set-up of interchanges to be specified Short-term planned projects
Primary networks
Primary networks envisaged in longer term Saturated main routes

Map 2: Logistics hubs
Public transport
HGV itineraries in orbital dense areas

Areas of change
Set-up of interchanges to be specified Short-term planned projects
Public orbital transport
HGV itineraries
Primary networks
Primary networks envisaged in longer term Saturated main routes

6. Unexploited possibilities for cost reduction

Although every project studied in this field seeks to apply the above principles, few studies look at the question from the overall perspective. A fundamental approach would be to look at the scale of the new facilities to be constructed. The small-scale underpasses of the 70s as well as the small tunnel on the A86 West have shown the way forward. With rare exceptions only, these infrastructures are in the heavily built-up areas (inside the heavy goods vehicle triangle), and here there is no need for roads to be built to take all types of vehicles, including the largest: the existing road facilities are robust enough.



Given what we have just seen, our proposal is extremely simple. We believe that the heavy goods vehicle triangle should be set up as shown in map number 1:

North-South side: creating a common truck road linking A86-A;
South side: complete the as-yet unfinished construction of the A86 at Fresnes and Antony;

Link between Roissy and St Quentin: extend as far as possible the Boulevard du Parisis, go underground from Groslay to Epinay, then to Gennevilliers, and complete the A86 up to St Quentin.

Such an approach would have the advantage of settling the problem of the exit to the A16, with traffic separating out into the Parisis Boulevard Intercommunal, and would also considerably improve saturation around the Epinay zone and at the Défense and Roissy junction.

Within this triangle, the necessary public transport infrastructures would be set up (tramways, metros, bus lanes, etc.), as well as cycle tracks (map number 2).

Over time, the number of lanes on the A86 could be doubled, for a reasonable cost, in the east of Paris, from the A6 to the A1, using a light vehicle carriageway which would in major part be underground (except when it is possible to use existing facilities). Existing surface roads would be totally or partially (depending on the existing construction limits) dedicated in the day to heavy goods vehicles and would be subject to tolls. The underground carriageway, dedicated to light vehicles would be charge-free. At night, the underground carriageway would be closed for most of the time, for reasons of economy, or open during maintenance periods of the surface carriageway. At night, circulation on the surface carriageway would be free, which would encourage heavy goods vehicles to shift their times of travel (at least for those who are able to do so). Part of the construction limits recovered at the surface could be used to set up sound screens specifically designed to cope with heavy goods vehicle traffic.

The final objective would be to prevent HGV access to the ring road (“ périphérique ”) when in transit, and more specifically to the internal rapid carriageways of the A86, until the completion of the A86 PL Est. This would make it unnecessary to double up the ring road (“ périphérique ”) in its southern part, and would thereby increase the rate of the toll returns on the A86 PL.

Finally, the ideal would be to associate the road haulage companies (through their federations and cooperatives), with the licensed road operators for the undertaking of the necessary works, in order to ensure that some of the revenues (from the tolls paid by those not members of the haulage associations) could in part offset the tolls paid by those within the associations. The reason for this is that, like the vehicle manufacturers, the hauliers should be recognised as having a main interest in being able to control their future.


Ile-de-France still needs rather less than 100 kilometres of high-speed roads, most of them of HGV grade, the total cost of construction being in the order of FRF 75 billion. If the construction of these roads takes too long, they will lose their town planning rationale: the conurbation will change without them, and move towards increasingly low densities, on the American model. Urban motorway tolls took ten years to be politically acceptable in Norway, prior to the undertaking of the construction works financed by it. Urban motorway tolls were rejected in Sweden after years of discussion, and Paris is not Singapore. The Ile-de-France region is not ready to put up with the firm hand of management of public spaces which is necessary, given the tight physical constraints on this region. The overall planning process at present appears to be too drawn-out, and as a result, the urban area will continue to become ever-more extensive, all the time there is no resolution to the black spots of traffic congestion in areas of dense occupancy. The continued higher rate of growth of new individual housing, as compared to apartment blocks, shows the urgent need to act now!

In our opinion, FRF 25 billion finance could be raised by setting tolls on new roads. This would be equivalent to raising revenues on the toll stretch of FRF 1.7 billion. At all events, public funds or guaranteed revenues in the


amount of FRF 50 billion would be essential. Some particularly robust private companies could become players, by providing long-term corporate finance over a part of the network.

The situation of the mixed public/private companies who are the licensed operators of motorways in the open countryside is overall a healthy one. Provided there is an end to the construction of high-cost motorways in areas of low population density, and provided new charging systems are introduced, associated with an extension of motorway operators’ licensing periods (now under study), the SEMCA system could be the way to raise additional finance. Either revenues would be paid in the form of dividends to the Ministry of Finance, or they could be used to set up a guarantee or other special fund allocated to the construction of urban roads. A dedicated public sector company on the Norwegian model could be set up, if it were wished to ensure the allocation of these revenues remained solely the transportation networks themselves.

From a political point of view, the allocation of revenues, from open country stretches of motorways which have already reached payback, to the urban network, is no more problematic than the allocation of such revenues to unprofitable stretches of countryside motorway. To do so, would only be to repeat what has already been done with the A14 and A46. All that is needed is an vision, for France as a whole, of how road toll revenues might be used. There is no need to be hemmed in by old-style laws and procedures, set up some 50 years ago, which are no longer necessarily appropriate.