Nature or nurture: why do railroads carry greater freight share in the United States than in Europe?

Abstract

During the 1950s, the share of freight carried by railroads was similar and declining in both the United States and Europe. By 2000, the railroads’ share of freight (measured in ton–kilometers) had reached 38% in the United States while falling to 8% in Europe. This paper examines the reasons for the difference in rail’s share of freight in Europe and the United States. We find that almost 83% of the gap in 2000 is probably due to natural or inherent differences, principally geography, shipment distance, and commodity mix. However, 17% of the gap cannot be explained by these inherent differences and is presumably due to public policies including priority of passenger service, lack of interoperability at borders, service quality and rates, and incentives of the rail operators. We estimate that if that policy gap were closed, railroads’ share of freight in Europe would increase from 8% to 13%.

Appendix A: Calculation of distance and commodity-mix effects

Shipment distances. The impact of the shorter distances in the EU is estimated by comparing the effect on the railroad share when the US distance-block distribution is substituted for the EU distance-block distribution. As shipment distances are larger in the US than in Europe and railroads are more favorable for long distances than road, this calculation gives us the distance effect or, in other words, the part of the market-share difference between Europe and the United States that is explained by distance. The calculation is detailed below.

$$ \begin{aligned} r_{\rm d}=\sum\limits_{i=1}^{i=n}\sum\limits_{j=1}^{j=m}RR_{\rm om}& CM_{\rm US}^{j}\left({DB_{\rm US}^{ij}-DB_{\rm EU}^{ij}}\right)\left ({\frac{R_{\rm US}^{ij}}{RR_{\rm US}^{ij}}}\right)\\ &R_{\rm om/d}=R_{\rm om}-r_{\rm d} \end{aligned} $$

(1)

where r _d is the distance effect on rail transportation (ton–kilometers); R _om the rail transportation controlling by non-surface modes effect (ton–kilometers); RR _om the rail plus road transportation controlling by non-surface modes effect; R _om/d the rail transportation controlling by non-surface modes effect and distance effect (ton–kilometers); m the number of commodities; n the number of distance blocks; $CM_{\rm US}^j = {RR_{\rm US}^j}\sum\limits_{j=1}^{j=m}{RR_{\rm US}^j}$ is the commodity-mix distribution in the US; $DB_{\rm US}^{ij}= {RR_{\rm US}^{ij}}{\sum\limits_{i=1}^{i=m}{RR_{\rm US}^{ij}}}$ is the distance-block distribution in the US; $DB_{\rm EU}^{ij}= {RR_{\rm EU}^{ij}} {\sum\limits_{i=1}^{i=m}{RR_{\rm EU}^{ij}}}$ is the distance-block distribution in the EU. R ^ij_US is the rail transportation in the US in ton–kilometers (distance block i and commodity j); RR ^ij_US is the rail plus road transportation in the US in ton–kilometers (distance block i and commodity j)

Commodity mix. The commodity-mix effect is calculated as the difference between the ton–kilometers that the EU would move if the EU followed the pattern of the US in terms of commodity mix. The commodity-mix effect is calculated according to equation 2.

$$ \begin{aligned} r_{\rm cm}=\sum\limits_{i=1}^{i=n}\sum\limits_{j=1}^{j=m}RR_{\rm om} &\left({CM_{\rm US}^j-CM_{\rm EU}^j}\right)\left({DB_{\rm EU}^{ij}} \right)\left({\frac{R_{\rm US}^{ij}}{RR_{\rm US}^{ij}}}\right)\\ &R_{\rm om/d/cm}=R_{\rm om/d}-r_{\rm cm} \end{aligned} $$

(2)

where r _cm is the commodity-mix effect on rail transportation (ton–kilometers); R _om/d/cm the rail transportation controlling by non-surface modes effect, distance effect, and commodity-mix effect (ton–kilometers); $CM_{\rm EU}^j={RR_{\rm EU}^j}{\sum\limits_{j=1}^{j=m}{RR_{\rm EU}^j}}$ is the commodity-mix distribution in the EU.