La lune


Documentation > Publications > Articles > Tradable permits in transportation (may 2001)

Tradable permits in transportation, a methodological discussion

French version published in 2001 in Les Anales de la Stratégie de la SNCF - translated august 2004

website of the author : - contact the author :




Reminder on tradable permits

The international negociations on greenhouse gases emissions have led certain countries to adopt the principle of a reduction of their emissions.

But once this principle is adopted by states, how is it possible to convert this wish into an effective diminution from the concerned agents, particularly industries ?

Several theoretical possibilities exist for the public authorities :

1 - it is possible to regulate : every business must diminish its own emissions by X%, and trespassers are fined or punished one way or another,

2 - it is possible to negociate : every actor, or group of actors, commits to reduce its emissions in exchange of given advantages. But if effective sanctions are enforced for those that do not respect the commitment, and that one of the parties around the negociation table is the state, this situation is equivalent to a regulation,

3 - it is possible to impose a tax : each tonne of carbon equivalent emitted will be "imposed", so that companies will have an incentive, to pay the least possible amount, to reduce their emissions as much as possible. As they stop to reduce as soon as the amount of the tax is inferior to the cost of further reduction, by fixing the amount of the tax the government chooses, in a way, the effective reduction that will be achieved.

The tradable permits are just a tool to "put oil in the wheels" to allow a group of actors to achieve a collective reduction goal that has been set by law, or through an agreement.

Let's imagine that all companies are asked, by law (or accept, through an agreement), to reduce collectively their emissions by 10% in 2 years, and that, for each company, the reference level is that of the previous year. The company A will very easily meet this objective, and actually could very easily achieve 20% : it is a biscuit factory and it had already planned to switch from fuel oil to gas to heat its ovens. The company B will have a hard time doing anything better than 5%, whatever money is invested into a reduction attempt : to achieve more than 5% it should replace all its production tool, what requires at least 3 years.

The whole idea of tradable permits is that the company A and the company B "manage together" so that the result of A+B is a reduction of 10% of their joint emissions. Practically A will "sell" to B the reductions that A can easily do and B with great difficulty. One immediately sees that the system is globally balanced by nature : a permit can only be bought from someone that did better than the global objective, and that sells the surplus to a company that has a hard time.

Practically A and B do not know each other. It is therefore necessary to organise a market where companies that easily achieve reductions can sell them to companies that don't. It is this market, with the description of the way reductions can be traded, and with what limits, that is called a "tradable permit system".

It is also possible to organize a market between countries : some countries will be able to go over the commitment they took, while some others won't be able to meet their objective. The idea is then to transfer, in conditions that remain to be set up, extra reductions from ones to others.

Some have condemned the idea and used the expression (in France) of "right to pollute", that would be unacceptable. Well, each one of us takes the "right to pollute" every single day of one's normal life :

any amount paid to a water company in exchange of the right to pour waste water in their collection pipes is a "right to pollute",

any local tax paid for waste removal is a "right to pollute",

buying gasoline is paying a "right to pollute" (gasoline pollutes !),

any administrative authorization given to a factory to emit a given substance in the environment is a "right to pollute" up to the limit of the authorization,

more generally any norm is a "right to pollute" up to the norm, any tax a "right to pollute" up to the tax, etc.

Such a system is therefore a useful way to compensate for the necesserarily arbitrary aspect of the quotas that are set up through the regulation, by allowing to everyone, within certain limits, to adjust this quota depending on one's marginal reduction costs.

The main inconvenients of tradable permits are not of moral but of practical nature :

it is necessary to be able to measure the CO2 emissions of the sellers and the buyers, that is those of everyone,

it is necessary to be able to impose sanctions on those that do not meet their objective, which requires to define penalties,

the public authorities have to control how things go,

it is necessary to define the way the initial allocations will be made (bids (2), by reference to the observed previous emissions, by reference to a production index.

Compared to permits, the tax (on the emissions themselves, or on products that lead to emissions) is much easier to handle :

it avoids the discussion on the initial allocation,

its cost is immediately proportionnal to the nuisance, and therefore a tax does not represent a bonus to "naughty ones", as is a system of permits with an assignment proportional to the past (historical) emissions,

the management costs are much lower, because the collection of the tax can be done in a very "concentrated" way (from the oil companies, or the coal fired power plants, for example),

if the tax is applied to emission precursors (like gasoline for example), it is not necessary to measure the emissions themselves for consumers taken one by one.

An additional consideration is that the tax applies to all the emissions while permits only apply to traded emissions (3).

But a tax requires that the situation is about the same worldwide, otherwise there is a high risk of fiscal "dumping" : the big greenhouse gases emitters move their factories or subcontract - in countries that do not have or have low taxes, and as the place of emission is of no importance regarding greenhouse gases, the goal - a global decrease of the emissions - is not achieved.

In particular, tradable permits will be very uneasy to use for low and very scattered unitary sources, such as for raod transport and, to a lesser extent, for air transport. This does not prevent the national Council of Transportation to evaluate these tools. It then seemed opportunate to examine how the SNCF [National French Railways] would be positionned in such a system, in order to help the discussion to come.


Reference level chosen

Two possibilities - at least - exist to set the reference level that will be used to define the goals of each company :

prorata the past emissions (also called historical emissions),

prorata the goods produced or the service made.

The first case corresponds to the following context : for a given collective assignment (that is that concerns all companies) of a decrease of X%, the objective of each company corresponds to its emissions of a given year minus X%. Each company must then manage to reduce its emissions by X%, and if it can't do it but that the sector, taken as a whole, does manage to achieve the collective goal, the company buys permits on the market from those that did better than -X%.

The advantage of this way to proceed is that is creates no winner and no loser in the short term, as X is generally small. The major inconvenient is that it gives de facto a bonus to "bad ones"(those that emit a lot) by allowing them to last as such and to keep their market shares. Thus, there is a risk to keep unchanged a situation that one would want to change.

The second situation corresponds to the case where the reference level of each company is not set on the basis of its past emissions but on the basis of its production. In this case, the total emissions of the sector are divided by the total number of production units of the sector (tonnes, or litres, or, here, or Each company then gets an objective based on its output (each tonne produced "is worth" so much carbon, or each litre "is worth" so much carbon, etc), and not on its previous emissions.

It is possible, for example, to calculate the average "carbon content" of all French pen manufacturers and decide that it is this content, minus X%, that is used to define te the objectives for all French pen manufacturers : each one must manage to do less than the "average greenhouse content per pen", and if he cannot he will buy permits from those that do better, because with this system some do better than the average from the beginning.

Indeed, those that have factories that use less fossil fuels per pen sold do better than the average as soon as the system is set up. They hence benefit of a "rent" by selling immediately, on the market, what corresponds to the difference between their "greenhouse gas content" per pen and the average, to those that have factories that use more fossil fuel by pen manufactured, and that will pay to buy permits.

In the case of transportation, the initial allocation depending on the service done would be an allocation per or per

The advantage of this way to attribute permits is that it allows those that have been vertuous before the others - even if it was not the aim at the time - to get an instant benefit out of it : such a way to proceed is therefore a bonus to virtue.

A second advantage is that it will not incitate the "good ones" to become "bad" - particularly in the case of transportation, where many investments involve the long term - but it will strongly incitate the "bad ones" to become good, as the effort relies almost totally on them. We thus get very close to a tax in this case.

Its inconvenient is that it creates short term losers : all those that emit more than the average per production unit, that will have to buy immediately permits on the market to compensate their low energy efficiency. However we will see that with the amounts envisionned the "loss" remains perfectly bearable by the losers.

For our reflexion regarding transports, we will suppose that it is the second way to allocate which is chosen : the reference level is set prorata the service done, which means accordingly to the number of and of done by each mode.

In this case, the SNCF, which has emissions per and by much lower than those of other modes (road, planes) will then benefit of a rent for which the present note tries to assess the magnitude.



Transportation of people

In France, transportation of people have totalled 792 billion, as follows :



Passengers per véhicule


Consumption (litres) per or vé

Fuel consumed (million litres)




0,0752 (v.k)








Planes (domestic)




0,11 (p.k)








Figures have been communicated by the SNCF or drawn from the "compte des transports de l'INSEE".


The consumption per passenger is a personnal estimate :

on domestic flights, the calculation has been made from the consumption of short range planes, taking into account the average filling rate,

for cars, we have the number of, of (therefore of an average number of passengers per car) and an average consumption per

As a result a, all modes, consumes about 0,0388 litre of fuel and leads to the emission of 26,7 g carbon equivalent.

A by train generates only 2,3 g carbon equivalent. An initial allocation proportional to the service done would allow the SNCF to valorize on the market a savings of 24,4 g per, for a total of 1,9 million tonnes carbon equivalent (that is the equivalent of the emissions of a little less than 1 million French people).

With a tonne of carbon worth 500 F, the SNCF could sell on the market about 1 billion Francs per year.

With a tonne of carbon worth 1.000 F, this amount would become 2 billion Francs



Just as above, the situation can be described as below




Average load (tonnes)


Consumption (litres) per or vé

Fuel consumed (million litres)
Light utility vehicles




0,097 (v.k)


Heavy duty vehicles




0,3734 (v.k)








Planes (domestic)



1.000 (*)

1,1 (t.k)








Figures have been communicated by the SNCF or drawn from the "compte des transports de l'INSEE", except (*) which is only a rough figure.


Just as above, a, all modes included, corresponds to the emission of 37,8 g carbon equivalent, when a transported by the SNCF corresponds to 3 g carbon equivalent.

The difference with the average is worth here 1,8 million tonnes carbon equivalent. With a tonne carbon equivalent worth 500 F, the SNCF could sell on the market a little less than 1 billion Francs per year.



The above rough figures suggest that, without taking the transaction costs into account, and supposing that a tradable permits system is set up in the transportation sector with an initial allocation proportionnal to the service done, the SNCF could valorize on the market about 2 billion Francs per year with a tonne carbon equivalent worth 500 F.

This amount is only related to the railways activity.

Let's recall that a tonne carbon equivalent worth 500 F represents about 35 cts per litre of gasoline (it requires about 1600 litres of gasoline to emit a tonne carbon equivalent) and that it will not have any significant influence on the modal choices.

With a tonne carbon equivalent worth 1.000 F, the "rent" goes up to 4 billion Francs per year.

With a tonne carbon equivalent worth 10.000 F, which represents a doubling of the price of gasoline (about 6 F extra per litre), and that would come closer to what would be necessary to get a significant modal transfer, the rent linked to the sale of permits by the SNCF would amount to 40 billion Francs per year.

It would be interesting to compare the above sums to the investments that should be done to allow railways to become the dominant mode of freight (for example to go from a 25% to a 75% market share in France).

It should be noted that, furthermore, the time horizon which is concerned (a railway lasts one or two centuries) includes the probable disappearing of commercial fossil fuels (4) and that it is not absurd to think about modal transfers of this magnitude. In particular, for the long term (50 years or more), long distance road transport will probably be condemned, either because of a decrease of the fossil fuel supply, or because of a real will to fight against climate change.

At last, whatever commitment to reduce greenhouse gases will be taken, it will not be possible to achieve it with an everlasting growth of transportation. The context in which we should situate ourselves therefore also includes a stabilization or a decrease of the volume of physical exchanges.

A couple of remarks to end :

part of the permits sold by the SNCF would be bought by the subsidiaries (Calberson, etc). The net result for the SNCF group would thus be diminished by as much,

this note just aims at clarifying the discussion, with hypotheses that will probably not be easily met. If transports were to be submitted to a tax and not be part of a tradable permits system (what seems wishable from any perspective otherwise), this calculation can still be used to guide negociations between the government and the railways representatives, as follows : allocate to the modal transfer from road to train the carbon tax collected from other actors prorata what goes above the mean value.



(1) Let's recall that the Kyoto Protocol allows the setting up of a tradable permits system only between Annex I countries - those that commited to a reduction, basically the industrialized countries - and that this system can in no case be used to transfer the effort from rich countries to poor ones.

(2) which is equivalent, in a way, to set up a tax

(3) as a result the financial impact is higher in the case of a tax, and thus there is a superior efficiency regarding modal transfers

(4) The disparition of fossil fuels in the economy will happen much before the physical disparition of these fuels : the remainder will be "preempted" for military usages and various indispensable basic services (food, health, etc).

Back to index
Back to the top