demand elasticity of intermodal container transport

0 downloads 0 Views 1MB Size Report
of intermodal container transport. Bart Jourquin. Louvain School of Management, Mons Campus. Lori Tavasszy. Delft University of Technology / TNO, Delft, The ...
On the generalized cost - demand elasticity of intermodal container transport Bart Jourquin Louvain School of Management, Mons Campus

Lori Tavasszy Delft University of Technology / TNO, Delft, The Netherlands

Liwei Duan Southwest Jiaotong University, China / Delft University of Technology, The Netherlands With financial support of

Introduction

In the market area for container transport, trucks can be substitutes or complements to trains (or barges) Ø  Substitutes : when used from origin to final destination Ø  Complements : when used in the intermodal chain

Introduction Ø  Elasticities important to assess impacts of transport projects and policies e.g. infrastructure, technology (LHV), tolls etc… Ø  On some freight corridors, intermodal rail has a relatively high share. Still elasticities used concern conventional rail transport. Ø  Literature contains a wide range of estimates of freight transport elasticities (differences in applied methodologies and data). But: Ø  Only a few papers consider explicitly intermodal transport Ø  Even fewer estimate the impact of the total haul length or the preand post-haulage distances on the market. Ø  The impact of the “substitute/complement effect” has, to our knowledge, never been estimated.

Aim of the paper Ø  Question: What is the impact of complementarity of freight transport modes on substitution elasticities? Ø  Hypothesis: pre/post haulage has a dampening effect Ø  Both direct elasticity of road and cross-elasticity of rail should be smaller in intermodal case than in conventional mode choice

Ø  Analysis: Ø  A stylized, theoretical model to explore system boundaries Ø  A full empirical model of freight mode choice in Europe

versus

Synthetic model Ø  Use of generalized costs Ø  Values taken from Limbourg and Jourquin (2010) Ø  Modal split:

Dn

R

Ø  Random utility discrete choice model Ø  Logarithmic logit (Abraham’s Law), with exponent = -1

D2

ln

l2

D1 l1

O

l0

r

T

D0

O = Origin D = Destination T = Terminal

Direct elasticity road Ø  Road t.km elasticity for road costs -5% Ø  Elasticity varies between -0.1 and -0.5 Ø  Strong influence of PPH and main distance Ø  Effect smaller with high overall distance or short PPH Ø  Two factors have similar influence

Rail-Road cross elasticity Ø  Rail t.km elasticity for road costs -5% Ø  Elasticity varies between 0.4 and 0.6 Ø  Strong influence of PPH and main distance Ø  Effect smaller with high overall distance or short PPH Ø  Two factors have similar influence

Real World model

Origins - destinations = Antwerp or Rotterdam

Example

Model specifications Ø  Road competes with rail-road intermodal transport Ø  Supernetwork model with 143 terminals Ø  No pre- post-haulage at the maritime ports Ø  Same costs values as in synthetic model Ø  Logarithmic Logit (Abraham’s Law) Ø  Calibration : r2 > 0.95

Results Elasticity for total haulage distances (road cost -5%) Road Rail-Road

All Antwerp -0.14 -0.16 0.70 0.81

Rotterdam -0.10 0.56

Short haul -0.10 0.54

Medium haul -0.14 0.71

Substitution only elasticity Road Rail-Road

All -0.16 0.85

« Complement effect » = 0.85 – 0.70 ≅ 20%

Long haul -0.19 0.98

Results Elasticity per Pre- Post-Haulage (PPH) distance, road cost -5% Road Rail-Road

Short PPH -0.19 0.97

Medium PPH -0.12 0.66

Long PPH -0.05 0.29

Elasticity per Pre- Post-Haulage (PPH) distance, rail cost -5% Road Rail-Road

Short PPH 0.11 -0.54

Medium PPH 0.14 -0.70

Long PPH 0.13 -0.72

Results with IWW

Elasticity for the three modes model (-5% of trucking costs) Mode Road Rail-Road IWW-Road

ε -0.28 0.61 0.31

Main conclusions Ø  Estimations fall within broad range of published elasticity values, yet differences from conventional rail transport are significant Ø  “Complement effect” ≅ 20% Ø  Absolute value of own elasticity for trucking Ø  Increases with the total length of the haul Ø  Decreases with the length of the pre- post-haulage Ø  Both influence have similar order of magnitude Ø  IWW-Road is less sensitive than Rail-Road intermodal Ø  Impact of mode choice model?

[email protected] [email protected]