Other general information can be obtained from these analyses. Thus, United States experience
(Deakin in Barde and Button, 1992) indicates that the effectiveness of regulatory measures to reduce
emissions, the opposition to control by taxation and, finally, the regulatory measures themselves often have
unexpected consequences on other factors of which they modify the conditions of equilibrium in an
unpredictable way; for example, the style of development or use of land.
However, apart from the few cases quoted, there are not many countries or cities where an
integrated transport and environmental policy approach is followed. This is particularly regrettable, since
improvements in these areas cannot possibly result from a single instrument, but only from the combination
of a large number of tools, owing to the many conflicting objectives involved, and to the complexity of their
interrelationships.
Conclusions: future ways forward
Knowledge of transport sector social costs, of prime importance in the definition of broader
environmental policies, is gradually improving as more and better research is done. The uncertainties that
remain have many causes, most of these being related to the difficulty of calculating monetary values in the
absence of markets, and to our imperfect understanding of the harmful effects of transport in certain fields,
such as noise or pollution.
Subject to these reservations, the social costs of transport seem to be in the neighbourhood of the
following values, as a percentage of GNP:
• Accidents: 2 per cent;
• Noise: 0.3 per cent;
• Local pollution: 0.4 per cent;
• Congestion:
-- Total time: 8.5 per cent;
-- Supplement compared with free-flowing traffic: 2 per cent;
• Overall pollution: 1 to 10 per cent in the long term.
The greater part of these costs stem from road transport. Costs per traffic unit are highly variable,
depending in particular on the extent of urbanisation concerned. The following ranges and orders of
magnitude can be proposed: one passenger-km has a social cost about the same as one tonne-km; the cost by
road (car or truck) is about 0.02 ECU per unit-km; the cost by rail is about 10 times less, and of the same
order of magnitude as the cost by air. These values are too imprecise for it to be possible to distinguish any
time trends or differences between countries.
Social costs carry very low weights in the overall internalisation of environmental effects,
seemingly because of the resistance generated -- a confrontation in which intuition and guesswork tend to
play a greater role than any economic calculation. In this context, it is possible to pick out certain aspects for
which internalisation has made relatively little progress:
• The environment should be more fully integrated into investment decisions at the Development
Plan stage.
• Improvements in safety require training and information programmes and greater use of
cost-benefit studies.
• The taxation instrument is used little (or poorly) in controlling noise or road pollution; in fact, it
can be a useful complement to a regulatory policy, and its introduction should be promoted.
Research and information programmes will probably be necessary for this purpose.