Home / Chroniques / Freight: why trains haven’t (yet) replaced lorries
π Industry π Planet

Freight: why trains haven’t (yet) replaced lorries

Aurélien Bigo
Aurélien Bigo
Research Associate of the Energy and Prosperity Chair at Institut Louis Bachelier
Key takeaways
  • In France, since the mid-2000s, road transport has accounted for 88% of freight flows, compared with 10% for rail and 2% for waterways.
  • Rail transport is the main opportunity to reduce CO2 emissions from domestic freight in France.
  • However, France is struggling to catch up, while the modal share of rail is almost twice as high (18%) across the EU or in Germany.
  • Public policies remain insufficient, as shown by the 2021 plan to revive rail freight, which will not achieve its objectives.
  • This plan forgets three elements: considering the future demand for freight transport; the decrease in road traffic; the evolution of our economy and logistics.

Every­one agrees on the prin­ci­ple. Trans­port­ing our goods by train rather than in lor­ries is a lever to be used for the envi­ron­men­tal tran­si­tion. In addi­tion to reduc­ing CO2 emis­sions, it reduces road traf­fic, con­ges­tion and space con­sump­tion, road acci­dents and dam­age to road infrastructure.

And yet, for decades now, there has been a suc­ces­sion of plans to revive rail freight, and their tar­gets have been missed with a sta­ble share of rail freight, but this has not pre­vent­ed the announce­ment of objec­tives that are increas­ing­ly ambi­tious. The lat­est tar­get set in 2021 is to dou­ble the modal share of rail freight by 2030. Sev­er­al ques­tions arise. Where do past fail­ures come from? Is it like­ly to work this time? What should be inte­grat­ed into pub­lic think­ing and policies?

For a century, a massive modal shift towards road transport

In France, rail freight trans­port start­ed about two cen­turies ago. The first line in France was opened in 1827 and was ini­tial­ly designed to trans­port coal between Saint-Éti­enne and Andrézieux, with wag­ons pulled by hors­es or descend­ing the slope by grav­i­ty. At that time, road trans­port dom­i­nat­ed, ahead of riv­er trans­port. Then rail trans­port grew enor­mous­ly, account­ing for almost 80% of traf­fic in France at the begin­ning of the 20th Cen­tu­ry, rel­e­gat­ing road traf­fic to only 10% of the modal share1

Since the mid-2000s, road trans­port has account­ed for 88% of freight flows in France, while rail has stag­nat­ed at 10% of traf­fic, leav­ing a small 2% for riv­er freight2.

Many attempts have been made to revive the rail­ways. For exam­ple, in 2000, the SNCF aimed to dou­ble rail traf­fic by 20103. The result? Between 2000 and 2010, rail freight traf­fic was… divid­ed by 2. There are many rea­sons to explain the decline of rail and the fail­ures of past modal shift poli­cies: the lack of invest­ment in infra­struc­ture, insuf­fi­cient qual­i­ty of ser­vice4, the grad­ual aban­don­ment of cer­tain activ­i­ties (such as sin­gle wag­onloads), the de-indus­tri­al­i­sa­tion of France, and the lack of dynamism of French ports (and their poor rail con­nec­tions)5.

Freight traf­fic in France since 1960 (source CGDD-SDES).

But the rea­sons are also to be found in the incred­i­ble growth of road trans­port. The entire organ­i­sa­tion of logis­tics and the econ­o­my has been built around heavy goods vehi­cles. Thus, if the modal share of rail has fall­en from around 60% to 10% since 1960, it is cer­tain­ly because rail traf­fic has fall­en by 37%, but also and above all because total demand has increased 3.6 times over the peri­od, dri­ven by an approx­i­mate­ly ten­fold increase in road traffic.

This has rel­e­gat­ed rail freight to its main area of rel­e­vance: mass trans­port, gen­er­al­ly over long dis­tances and for large vol­umes, for which road trans­port is becom­ing less effi­cient and more expen­sive. This is reflect­ed in the dynamism of com­bined road-rail trans­port in recent years, with the open­ing of sev­er­al rail high­ways, allow­ing semi-trail­ers to cross France on adapt­ed trains6.

France is also a poor per­former com­pared to its Euro­pean neigh­bours, show­ing the room for manoeu­vre that exists in France. The modal share of rail is almost twice as high on aver­age in the EU or in neigh­bour­ing Ger­many (both at 18%). And the decline in rail traf­fic in France over the last few decades con­trasts with more dynam­ic devel­op­ments in oth­er Euro­pean coun­tries7.

Doubling the share of rail by 2030, under what conditions?

In a plan to revive rail freight, the most obvi­ous aim is to improve the sup­ply of rail freight. The 2021 plan is thus based on 72 mea­sures, grouped into three areas: mak­ing rail freight an attrac­tive, reli­able, and com­pet­i­tive mode of trans­port; act­ing on all the growth poten­tial of rail freight; and sup­port­ing the mod­erni­sa­tion and devel­op­ment of the net­work8.

While essen­tial (and with­out even going into the details and pos­si­ble short­com­ings here9), this sup­port for rail is and will be insuf­fi­cient to achieve the objec­tives, giv­en that too lit­tle account is tak­en of the fol­low­ing three elements.

What (de)growth in demand will there be in the future?

The first ele­ment is to know what the evo­lu­tion of total demand for freight trans­port will be by 2030, and thus what increase in rail traf­fic will be nec­es­sary to achieve a dou­bling of the modal share.

This pro­jec­tion is not clear­ly defined in the recov­ery plan, but the fig­ures men­tioned seem to fore­see an increase close to +10% in total demand by 2030. In such a con­text, rail traf­fic would have to increase by 120% to dou­ble its modal share. Rail would thus absorb the increase in freight traf­fic, with­out real­ly reduc­ing road traffic.

On the con­trary, in a sce­nario of strong sobri­ety on an econ­o­my-wide scale, trans­port demand could be reduced by 20% by 203010. In this case, it would be “suf­fi­cient” to increase rail traf­fic by 60% (an increase half that of the pre­vi­ous case!) to dou­ble the modal share of rail. And in this case, road trans­port traf­fic would fall by 30%, under the dual effect of the fall in total demand and the increase in the rail share.

Rail freight traf­fic 1960–2021 and tar­get 2030 (sources CGDD-SDES and Ministry).

The evo­lu­tion of total demand [the sub­ject of a forth­com­ing arti­cle, the last in the series] is there­fore fun­da­men­tal in the siz­ing of rail traf­fic to achieve the objec­tive. Past expe­ri­ence shows that the share of rail has evolved more as a result of this evo­lu­tion in total demand than in rail traf­fic. This evo­lu­tion is also fun­da­men­tal to reduce road traf­fic and thus have a down­ward effect on emis­sions, which is in line with the fol­low­ing point.

Reducing the share of road transport, an unassumed objective

Anoth­er short­com­ing of this objec­tive of dou­bling the modal share of rail is an implic­it assump­tion that is not clear­ly made. If the share of rail increas­es by 9% to 18%, this will be at the expense of road trans­port, whose share will have to fall by 9%11. To go even fur­ther, to reduce emis­sions, the objec­tive is not to increase rail freight, but to reduce road traffic.

How­ev­er, the objec­tive is nev­er pre­sent­ed in this form, and poli­cies are not ori­ent­ed in this direc­tion. For exam­ple, nei­ther this plan nor oth­er poli­cies on freight trans­port have pro­vid­ed for major changes in road infra­struc­ture charg­ing (fol­low­ing the aban­don­ment of the eco­tax in 2013–2014), the end of the par­tial reim­burse­ment of ener­gy tax­es for road trans­port, the reduc­tion in the speed of heavy goods vehi­cles or even modal shift oblig­a­tions12.

Today, the pre­dom­i­nance of road trans­port can nev­er­the­less be explained by unde­ni­able advan­tages com­pared to rail: “Roads offer flex­i­bil­i­ty, speed, reli­a­bil­i­ty, low prices, and avail­abil­i­ty of capac­i­ty”, as the pres­i­dent of the Freight Trans­port Users’ Asso­ci­a­tion13 sum­maris­es. In such a con­text, it is quite illu­so­ry to be able to reduce the share of road trans­port by 9% only by invest­ing in rail, and with­out dis­cour­ag­ing the use of road transport.

What changes in the territory, the economy and logistics?

In view of its ambi­tion, dou­bling the share of rail trans­port must final­ly be part of a wider evo­lu­tion of the econ­o­my and logis­tics. First of all, it is a ques­tion of mak­ing land-use plan­ning poli­cies con­sis­tent, while motor­way projects are still under­way and logis­tics ware­hous­es con­tin­ue to be built on the out­skirts of towns with­out being con­nect­ed to the rail­ways. This main­tains and even strength­ens the cur­rent depen­dence on road trans­port. The pos­si­bil­i­ties of shift­ing to rail will also depend on changes in the types of goods trans­port­ed and over what dis­tances. This rais­es ques­tions about indus­tri­al pol­i­cy, the tran­si­tion of the agri-food or con­struc­tion sec­tors, which account for a large pro­por­tion of freight traf­fic in France14.

Final­ly, in view of the con­straints inher­ent in rail, the logis­tics organ­i­sa­tion must also change to make more use of the train. In par­tic­u­lar, it is a ques­tion of loos­en­ing time con­straints, avoid­ing just-in-time and tight logis­tics chains, and instead man­ag­ing stocks in order to mas­si­fy trans­port flows. This is, for exam­ple, con­trary to the cur­rent trend towards the growth of e‑commerce and rapid deliv­ery, which are struc­tural­ly unfavourable to rail and for which there is a lack of reg­u­la­tion15.

Oth­er modal shift opportunities

Rail trans­port is the main oppor­tu­ni­ty for modal shift to reduce CO2 emis­sions from domes­tic freight in France. But oth­er pos­si­bil­i­ties exist in the glob­al, nation­al and local logis­tics segments.

On an inter­con­ti­nen­tal scale, mar­itime trans­port should be pri­ori­tised as much as pos­si­ble over air freight, which emits around 100 times more CO2 per tonne trans­port­ed16. Giv­en its much high­er cost, air trans­port is used more for high val­ue-added prod­ucts or when time is of the essence, such as for per­ish­able food­stuffs. Increas­ing­ly, e‑commerce play­ers are also invest­ing in air freight for fast deliv­er­ies, a trend that should be limited.

For nation­al trans­port and in addi­tion to rail freight, modal shift is also pos­si­ble via riv­er trans­port. Its devel­op­ment is essen­tial­ly lim­it­ed by geo­graph­i­cal con­straints and by the avail­able water­ways, which explain its cur­rent modal share of 2%. There is, how­ev­er, poten­tial for devel­op­ment and trans­fer from road trans­port on major routes, inter-basin links or for deliv­ery in the cen­tre of conur­ba­tions con­nect­ed to a nav­i­ga­ble riv­er17. Final­ly, for last-mile deliv­ery, car­go bikes can replace a large pro­por­tion of light com­mer­cial vehi­cles. Per unit trans­port­ed, the cli­mat­ic ben­e­fit of this trans­fer is very high, espe­cial­ly if the deliv­er­ies were pre­vi­ous­ly rapid and not very opti­mised and made by com­bus­tion engines. How­ev­er, over­all, the gain in freight trans­port emis­sions is lim­it­ed, as the demand that can be addressed remains lim­it­ed (dis­tances and vol­umes being rel­a­tive­ly small for car­go bikes). More sig­nif­i­cant ben­e­fits will be felt in terms of reduc­ing air pol­lu­tion, space con­sump­tion, resources, noise and dan­ger to oth­er users.


The modal shift objec­tives have cer­tain­ly been too ambi­tious in rela­tion to the trans­for­ma­tion efforts required to achieve them and the iner­tia that exists in freight trans­port. How­ev­er, break­throughs are nec­es­sary to achieve the decar­bon­i­sa­tion objec­tives, jus­ti­fy­ing ambi­tious modal shift objec­tives. Pub­lic poli­cies have been and still are far from com­mit­ting to strong shifts, both because of insuf­fi­cient sup­port for rail trans­port, in terms of invest­ment or sup­port for activ­i­ty. But also, and above all, because the sup­ply (or sup­port) pol­i­cy for rail is insuf­fi­cient to seri­ous­ly chal­lenge the cur­rent hege­mo­ny of road trans­port18. Yet it is almost exclu­sive­ly with this bias that modal shift pol­i­cy is still con­sid­ered in France, despite the fail­ures of past policies.

1See his­tor­i­cal fig­ures in Neiertz, 1999.
2For graphs and oth­er traf­fic fig­ures in the arti­cle, see CGDD-SDES.
3See Les Echos, 2000. Traf­fic has fall­en from 57.7 bil­lion tonne-km (G t.km) in 2000 to 30 G t.km in 2010, i.e. a divi­sion by 1.92 in 10 years; CGDD-SDES fig­ures.
4On this sub­ject, see the barom­e­ter of ship­pers’ per­cep­tions of rail trans­port, EuroGroup Con­sult­ing.
5We could also men­tion the effects of open­ing up to com­pe­ti­tion since 2005, which remain sub­ject to debate. For its pro­po­nents, com­pe­ti­tion has made it pos­si­ble to low­er prices and sta­bilise the modal share, com­pared to the decline in pre­vi­ous years. For its detrac­tors, it has weak­ened Fret SNCF, faced with major dif­fi­cul­ties pre­vent­ing the his­toric oper­a­tor from hav­ing a strong ambi­tion in freight and ben­e­fit­ing from the stronger scale and net­work effects of a sin­gle oper­a­tor.
6See in par­tic­u­lar the page on the MTECT-MTE web­site on the sub­ject of rail motor­ways.
7See Euro­stat data, about 18% modal share in 2019 for the EU-27 and Ger­many, for trans­port with­in the coun­tries, exclud­ing pipelines; see also evo­lu­tion graphs in LVSL (2021) for traf­fic since 2003, or in Insti­tut Mon­taigne (2022) for traf­fic since 1970.
8The edi­to­r­i­al of the Min­istry’s 2021 recov­ery plan men­tions that “The State has cho­sen to take on board the ambi­tion of the sec­tor’s play­ers to dou­ble the modal share of rail freight over the decade, from 9% in 2019 to 18% in 2030, i.e. around 65 bil­lion t.km. In the longer term, the State has set itself the objec­tive of achiev­ing a modal share for rail freight of 25% by 2050. There have been recent breaks in the freight trans­port sta­tis­tics, and not all the ref­er­ence and pro­jec­tion fig­ures to 2030 are detailed in the strat­e­gy, which some­times makes it dif­fi­cult to relate them to past devel­op­ment sta­tis­tics. The fig­ures should there­fore be tak­en as orders of mag­ni­tude.”
9The financ­ing needs for rail infra­struc­ture are sig­nif­i­cant, and recent announce­ments do not yet spec­i­fy the share that will ben­e­fit freight (see the arti­cle by the Banque des ter­ri­toires). Also, the play­ers in the sec­tor (exclud­ing SNCF, rep­re­sent­ed by AFRA) do not seem to be very con­fi­dent that the objec­tive will be achieved, par­tic­u­lar­ly because of the short­com­ings of the net­work (Con­nais­sance des Ener­gies).
10See for exam­ple the ADE­ME’s Transition(s) 2050 sce­nar­ios. Mod­er­a­tion of trans­port demand will be the sub­ject of the 4th and last arti­cle in this series on decar­bon­i­sa­tion of freight trans­port.
11If the share of water­ways remains at 2%, which in any case should change lit­tle by 2030.
12For the par­tial reim­burse­ment of the TICPE, see the page on the Entreprendre.Service-Public web­site; for mea­sures to reduce speeds or impose modal shift oblig­a­tions, see the Shift Pro­jec­t’s PTEF freight report.
13State­ment by Denis Choumert of AUTF, an asso­ci­a­tion rep­re­sent­ing ship­pers, in Le Monde in May 2022. More­over, the rail freight recov­ery plan also men­tions on page 11 that road trans­port has « become the ref­er­ence mode in terms of com­pet­i­tive­ness, flex­i­bil­i­ty and qual­i­ty of ser­vice (punc­tu­al­i­ty) ».
14See the IDDRI sce­nar­ios pub­lished in 2019 (or those of ADEME and the Shift Project, cit­ed above) break­ing down the evo­lu­tion of demand for 6 cat­e­gories of goods.
15See in this respect the rec­om­men­da­tions of the France Stratégie report in 2021 (Pour un développe­ment durable du com­merce en ligne), which have not been fol­lowed much to date.
16See in par­tic­u­lar the fig­ures from EEA, 2022; CE Delft, 2021.
17See in par­tic­u­lar the ADEME study in 2022 on inter-basin links. Or VNF in 2021, which men­tions a pos­si­ble increase in trans­port­ed vol­umes by half by 2030.
18See the arti­cles by François Combes and Patrick Niérat on the sub­ject, par­tic­u­lar­ly their arti­cle in TI&M in 2020.

Our world explained with science. Every week, in your inbox.

Get the newsletter