1
Final Report –
Port and Modal Elasticity Study, Phase II
Robert C. Leachman
Leachman & Associates LLC
245 Estates Drive
Piedmont, CA 94611
leachman@ LeachmanandAssociates. com
For
September 7, 2010
Funding: The preparation of this report was financed in part through grants from
the United States Department of Transportation ( DOT).
2
Table of Contents
Page
List of Tables ................................................................................................................... 4
List of Figures .................................................................................................................. 5
Executive Summary ............................................................................................................ 7
1. Overview ....................................................................................................................... 16
2. Outreach to Stakeholders .............................................................................................. 30
2.1. Feedback from Stakeholders – Elasticity Studies ................................................... 30
2.2. Feedback from Stakeholders – Capacity Study ...................................................... 33
3. Port and Modal Shares of Imports ................................................................................ 37
3.1. Port Shares of Containerized Trade Volumes ........................................................ 37
3.2. Landside Channel Shares of Waterborne Containerized Imports ........................... 40
4. Distribution of Imports by Commodity and Value ....................................................... 44
5. Transportation Charges ................................................................................................. 50
5.1. Alternative Ports of Entry ....................................................................................... 50
5.2. Destinations ............................................................................................................ 52
5.3. Transportation Modes ............................................................................................. 53
5.4. Components of Transportation Costs ..................................................................... 55
5.5. Transportation Unit Costs ....................................................................................... 56
5.6. Domestic Equipment Availability .......................................................................... 61
6. Impacts of Port Contracts and of Carrier and Terminal Operating Strategies .............. 63
6.1. Port Contracts ......................................................................................................... 63
6.2. Contracts Between Steamship Lines and Railroads ............................................... 65
6.3. Contracts Between Importers and Steamship Lines ............................................... 66
6.4. Carrier and Terminal Operating Strategies ............................................................. 66
7. Congestion Analysis ..................................................................................................... 70
7.1. Background on Queuing Theory ............................................................................ 70
7.2. Port Terminal Congestion Modeling ...................................................................... 72
7.3. Rail Terminal Congestion Modeling ...................................................................... 76
7.4. Rail Line- Haul Congestion Modeling .................................................................... 79
8. The Short- Run Elasticity Model ................................................................................... 88
8.1. Overview of the Short- Run Elasticity Model ......................................................... 88
8.2. Input Data ............................................................................................................... 95
8.3. Output Data ............................................................................................................. 97
8.4. The Short- Run Model: Iteration of Supply Chain Optimization and Queuing
Model Calculations ........................................................................................................ 99
8.5. Application of the Short- Run and Long- Run Models .......................................... 100
8.6. Conclusions ........................................................................................................... 113
9. Glossary ...................................................................................................................... 115
10. References ................................................................................................................. 121
Appendix A. Resume of Stakeholder Meetings .............................................................. 122
Appendix B. Asian Origin Countries for Imports Included in the Study ....................... 123
3
Appendix C. Rail Line Configuration Data .................................................................... 124
Note: The contents of this report reflect the views of the author who is responsible
for the facts and accuracy of the data presented herein. The contents do not
necessarily reflect the official views or policies of SCAG, DOT or any organization
contributing data in support of the study. This report does not constitute a
standard, specification or regulation.
4
List of Tables
Page
Table 1: Top Commodities Imported from Asia Through US West Coast Ports in 2003
and 2005 ............................................................................................................................ 45
Table 2: Top Commodities and Declared Values for Waterborne Containerized Imports
from Asia to the United States in 2005 ............................................................................. 46
Table 3: Transportation Costs – Charges Separately Billed to Customer vs. Charges
Absorbed by Carrier .......................................................................................................... 50
Table 4: Assumed Distribution of Import Volumes by Destination Region .................... 54
Table 5: Space Capacities of Containers and Trucks ........................................................ 56
Table 6: Transportation Rates Per Cubic Foot, Shenzen/ Yantian/ Chiwan – Selected North
American Destinations ...................................................................................................... 57
Table 7: Domestic Container Fleet, 1998 to 2007 ............................................................ 61
Table 8: Port Terminal Data ............................................................................................. 77
Table 9: Productivity Data for Rail Intermodal Terminals ............................................... 78
Table 10: Statistical Parameters of the Rail Line- Haul Transit Time Model ................... 84
Table 11: Assumed Shares and Declared Values for Large Importers, 2005 and 2008
Analyses .......................................................................................................................... 100
Table 12: Comparison of 2006 Actual and Model- Predicted Traffic Shares ................. 103
Table 13: Import Volumes vs. San Pedro Bay Container Fee, As Predicted by Short- Run
Elasticity Model in Base- Case Scenario ......................................................................... 103
Table 14: Import Volumes vs. San Pedro Bay Container Fee, As Predicted by Long- Run
Elasticity Model in Base- Case Scenario ......................................................................... 105
5
List of Figures
Page
S- 1. Comparative Short- Run and Long- Run Elasticities of Direct, Transloaded, and
Local Imports via San Pedro Bay in the Base Case Scenario 10
S- 2. Short- Run Elasticities of Imports via the San Pedro Bay Ports in Future
Scenarios 11
S- 3. Long- Run Elasticities of Imports via the San Pedro Bay Ports in Future
Scenarios 11
S- 4. Long- Run Elasticity of Imports Routed via San Pedro Bay, With and Without
Congestion Relief 13
1. US Port Shares of 2005 US Containerized Imports from Asia 37
2. Shares of Inbound Loaded Containers at West Coast Ports 38
3. Container Traffic Shares at West Coast Ports 39
4. Percent Intermodal Movement of Marine Containers Imported Through
US West Coast Ports 41
5. 2003 vs. 2005 Cumulative Distributions of Containerized Asia – United States
Imports 47
6. Value Distribution of 2005 Asia – United States Waterborne Containerized
Imports 48
7. Wait Time as a Function of Utilization and the Number of Servers 71
8. Import Container Dwell Time vs. Import Volume at Selected Terminals 73
9. Import Container Dwell Time vs. Import Volume at Selected Terminals
Accounting for Acreage 73
10. Modeled and Actual Import Container Dwell Time vs. Import Volume at
Selected Terminals 75
11. Predicted Port to Gate Cycle Times 76
12. Actual vs. Modeled Rail Intermodal Terminal Dwell Times 79
13. Comparison of Actual and Modeled Rail Intermodal Transit Times 85
6
List of Figures ( cont.)
Page
14. Predicted Transit Time Gains from Double- Tracking 86
15. Predicted Increases in Peak- Period Domestic Intermodal Transit Times
as a Function of Intermodal Traffic Growth 87
16. Structure of Short- Run Elasticity Model 89
17. Inputs and Outputs of Supply- Chain Optimization Model 90
18. Inputs and Outputs of Queuing Model 92
19. Interaction of Supply- Chain Optimization and Queuing Models 94
20. Short- and Long- Run Elasticity of Imports to Fees at the San Pedro Bay Ports
in the Base Case Scenario 106
21. Comparative Short- Run and Long- Run Elasticities of Direct, Transloaded, and
Local Imports via San Pedro Bay in the Base Case Scenario 106
22. Utilization of Port Terminals at Selected Ports 107
23. Short- Run Elasticities of Imports via San Pedro Bay in Future Scenarios 111
24. Long- Run Elasticities of Imports via San Pedro Bay in Future Scenarios 111
25. Long- Run Elasticity of Imports Routed via San Pedro Bay, With and Without
Congestion Relief 112
7
Executive Summary
Sponsored by the Southern California Association of Governments ( SCAG), Phase II of
the Port and Modal Elasticity Study concerns the development and application in policy
analysis of a database and analytical tools to predict flows of waterborne containerized
imports from Asia to the United States through North American ports and landside
supply- chain channels. The lead consultant performing this study was Leachman and
Associates LLC.
In August, 2005, Leachman and Associates LLC completed a long- run elasticity analysis
for SCAG. A Long- Run Elasticity Model developed by Leachman and Associates
predicts the allocation of Asia – USA waterborne containerized imports to ports and
landside channels as a function of the following input data: overall import volume;
distribution of imports by regional destination, by declared value and by size and scope
of importer; statistical distributions for container flow times from Asian origins across the
water, through ports and through landside channels; transportation rates and trans- loading
rates; and user- specified potential container fees. Repeated application of the model
enables the public policy analyst to construct an elasticity curve of import volume vs. fee
value. The Long- Run Model was used to predict import flows through the San Pedro Bay
ports as a function of potential fees at the San Pedro Bay ports and as a function of
container flow time distributions. In particular, in the case of no reduction in flow times,
a fee of $ 60 per FEU ( forty- foot equivalent unit) was predicted to cause a 6% reduction
in total import volumes handled through the San Pedro Bay ports. On the other hand, if
major improvements in infrastructure were made that enabled significant reductions in
container flow times, the analysis showed that there would be no drop in total import
volumes if fees of up to $ 200 per FEU were applied subsequent to the availability of the
new infrastructure, although the mix of importers using the ports would evolve
considerably.
The long- run elasticity analysis in Phase I generated considerable interest from
stakeholders and public policymakers ( and considerable misinterpretation of the results).
Phase II of the Elasticity Study was initiated in May, 2006. Dialogue with stakeholders
begun in the earlier study was pursued in Phase II as well, and useful feedback and more
data were obtained. The technical work in Phase II included the following elements:
 Updating the database of Asia – USA import volumes by commodity and declared
values to 2005, and updating total import volume to 2006
 Updating databases of infrastructure and container flow times by port and
landside channel to 2006- 2007
 Updating databases of transportation rates, handling rates, and fees to 2007
 The Long- Run Model was enhanced in Phase II for more accurate calculations,
and the data feeding it was updated as indicated in the preceding bullet points.
 Development of the capability to conduct “ short- run” elasticity calculations, in
which port and rail infrastructure are fixed inputs to the model, as opposed to the
assumption of the Long- Run Model taking container flow times as fixed inputs.
Container flow times in the Short- Run Model are endogenous, calculated as a
8
function of inputs for the import volume and assumed infrastructure at the various
ports and in the various landside channels. Development of the Short- Run Model
involved the formulation and calibration of queuing formulas that predict
container dwell times at port and rail terminals as a function of volume, staffing
and acreage, as well as queuing formulas to predict container transit times in rail-line
haul movement as a function of track infrastructure and rail traffic levels.
Confidential data of container flow times vs. volume and infrastructure were
received from railroads and from operators of port terminals, and these data were
used to calibrate the queuing formulas.
 Short- Run and Long- Run elasticity calculations testing the imposition of
hypothetical container fees at San Pedro Bay were made for various scenarios,
including a 2007 Base Case scenario, serving to validate the model, and four
future scenarios, serving to characterize the range of potential outcomes from
imposition of fees. The future scenarios include a Near- Term Likely scenario, two
different longer- term Optimistic scenarios ( one assuming a 10% rise in all- water
steamship line rates relative to rates via West Coast ports, the other assuming a
10% rise in the market share of large, nation- wide importers), and a longer- term
Pessimistic scenario ( assuming a 10% drop in all- water rates relative to West
Coast rates). In addition, a Long- Run elasticity calculation was made of the Near-
Term Likely scenario modified to assume a program of major infrastructure
improvements in Southern California is put in place ( the Near- Term Likely
scenario with Congestion Relief). This scenario assumes the program of
infrastructure improvements is completed and made available to importers at the
moment container fees are introduced. This scenario represents an update of the
analysis published in the 2005 Phase I report.
Total imports routed via San Pedro Bay may be broken down into three basic categories:
( 1) local imports, consisting of imports consumed within the greater region for which San
Pedro Bay serves as the closest container port ( closest in the sense of lowest landside
transportation costs), i. e., imports consumed within the region encompassing Southern
California, Southern Nevada, Arizona, New Mexico and southern portions of Utah and
Colorado; ( 2) direct- shipping imports, consisting of imports destined to other regions
which simply pass through Southern California while remaining intact in the marine box
coming from Asia; 1 and ( 3) trans- loaded imports, which are imports consumed in other
regions that are unloaded from the marine box in Southern California, perhaps stored in
an import warehouse for weeks or months, possibly receiving value- added services such
as labeling, repacking or minor final assembly, and ultimately re- loaded into domestic
containers or trailers for re- shipment to other regions. A portion of trans- loaded imports
are trans- loaded to domestic containers or trailers immediately using a cross- dock
facility, but most are warehoused in Southern California for some time before re-shipment.
2
1 Marine boxes arriving from Asia that are forwarded out of Southern California via rail move under a
single steamship- line bill of lading from Asia to the inland destination under what is termed inland point
intermodal ( IPI) service.
2 Some local imports also are trans- loaded, but for the purposes of this analysis, the trans- load category
defined herein includes only imports ultimately consumed in other regions. Also, many imports in the
9
Low- value goods imported via the San Pedro Bay ports that are consumed in other
regions, as well as goods imported by small or regional importers, typically move
through direct- shipping supply chains utilizing inland point intermodal ( IPI) services,
whereby the marine containers are loaded onto double- stack trains destined out of region.
Trans- load strategies are practiced by large nation- wide importers of medium- value and
high- value goods. 3 The consultant estimates that in 2006, imports ultimately consumed
within the greater local region as defined above accounted for only 21% of all loaded
containers from Asian origins handled through the San Pedro Bay ports, IPI accounted
for 43% of these imports, and ( non- local) trans- loaded imports plus out- of- region
trucking of marine boxes accounted for the remaining 36%. By 2008, the IPI share of
Asian imports via San Pedro Bay had declined to 41%, the local share of imports rose to
23%, and the share accounted for by trans- loaded out- of- region imports and out- of- region
trucking of marine boxes held steady at 36% 4
Figure S- 1 highlights the disparate elasticities of these components of import volumes
routed via San Pedro Bay in the face of new fees assessed on imports in the Base Case
Scenario. As may be seen, for container fees of $ 200 per FEU, total imports routed via
San Pedro Bay are predicted to decline about 19% by the Short- Run Model and about
43% by the Long- Run Model. But percentage declines in the various categories of
imports are far from uniform. Local imports are predicted to decline not at all. Relatively
expensive imports ( declared values greater than $ 28 per cubic foot) that undergo
consolidation- deconsolidation and trans- loading supply- chain management practices in
Southern California en route to consumption in other regions, also are predicted to
decline not at all. Moderately- valued imports ( with declared values between $ 12 and $ 28
per cubic foot) that are consumed elsewhere and undergo consolidation- deconsolidation
and trans- loading in Southern California are predicted to exhibit some decline in volume,
down from 22% of Zero- Fee- Base- Case5 imports to 18% in the Short- Run analysis and
down from 22% to 9% in the Long- Run analysis. The largest decline is exhibited by IPI
volumes , falling from 42% of Base- Case volume to 31% in the Short- Run analysis and
from 41% to only 14% in the Long- Run analysis. 6
trans- load category change hands in Southern California, i. e., the goods are imported by an original
equipment manufacturer ( OEM) who pays for the transportation from Asia to an import warehouse in
Southern California, then purchased from the OEM by a retailer who pays for the transportation from the
import warehouse to regional distribution centers serving its retail outlets in other regions.
3 Another frequently- used name for trans- load import strategies is consolidation – de- consolidation, a name
arising because import shipments to multiple regions are consolidated as far as the port of entry before they
are broken into separate shipments to the regions.
4 The figures reported here for local and trans- loaded shares rest on the assumption that the final
consumption of imported goods in the local region is proportional to the total purchasing power of the
region relative to the total purchasing power in the Continental USA. The figures for the IPI shares are
based on the actual traffic counts.
5 Zero- Fee- Base- Case refers to the Base- Case Scenario with no new container fees.
6 Under IPI service, the importer contracts with the steamship line for door- to- door service. The steamship
line chooses the port of entry and subcontracts with railroads and draymen for landside movement. In that
sense, the port of entry is discretionary for the line, and this makes IPI traffic quite elastic to fees or other
costs imposed at one port but not at an alternative port.
10
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
$ 0 $ 50 $ 100 $ 150 $ 200
% of Zero- Fee Base- case Imports
Fee Value per FEU at San Pedro Bay
Figure S- 1. Comparative Short- run and Long- run Elasticities
of IPI, Transloaded and Local Imports via San Pedro Bay
in the Base- Case Scenario
Total - short- run
Total - long- run
IPI - short- run
IPI - long- run
Local ( goods
consumed within
region)
Transloaded < $ 28 per
cu ft - short- run
Transloaded < $ 28 per
cu ft - long- run
Transloaded > $ 28 per
cu ft - short- run and
long- run
What Figure S- 1 reveals is that local imports are totally inelastic for the fee range
depicted, trans- loaded expensive goods also are inelastic, trans- loaded moderate- value
goods are somewhat elastic, while imports utilizing IPI services are very elastic. The
trans- loaded imports generally contribute more to the local economy, providing
significant warehousing and logistics employment, but at the same time contributing
substantially more unfavorable environmental impacts in the local region ( pollution and
vehicular traffic), than the direct- shipping ( IPI) imports. Consequently, the elasticity of
trans- loaded goods is of considerable interest to policy- makers.
Figures S- 2 and S- 3 depict results of Short- Run and Long- Run analyses of the alternative
future scenarios, contrasted with the Base Case. In the Near- term Likely Scenario, total
imports via the San Pedro Bay ports exceed the Zero- Fee Base Case volume until about
$ 100 per FEU in the Short- Run analysis and about $ 75 per FEU in the Long- Run
analysis. Trans- loaded imports exceed Zero- Fee Base Case trans- loaded volumes until a
fee of about $ 350 per FEU in the Short Run, but fall below the Zero- Fee Base- Case trans-loaded
volume at about $ 150 per FEU in the Long Run. These results indicate that
adequate infrastructure and/ or staffing of that infrastructure are not yet in place at other
ports to accommodate without congestion the diversion of trans- loaded volumes away
from San Pedro Bay ports. However, the economics encouraging expansion at other ports
and their landside channels arises when fees greater than $ 150 per FEU are imposed on
imports through the San Pedro Bay ports.
11
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
110%
120%
$ 0 $ 50 $ 100 $ 150 $ 200 $ 250 $ 300 $ 350 $ 400 $ 450 $ 500
% of Zero Fee Base- case Total Imports
Fee Value per FEU at San Pedro Bay
Figure S- 2. Short- Run Elasticities of Imports via the San Pedro Bay
Ports in Future Scenarios
Total - Optimistic I
Total - Optimistic II
Total - Near- term Likely
Total - Base Case
Total - Pessimistic
Transload - Optimistic I
Transload - Optimistic II
Transload - Near- term
Likely
Transload - Base Case
Transload - Pessimistic
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
110%
120%
130%
140%
150%
$ 0 $ 50 $ 100 $ 150 $ 200 $ 250 $ 300 $ 350 $ 400 $ 450 $ 500
% of Zero Fee Base- case Total Imports
Fee Value per FEU at San Pedro Bay
Figure S- 3. Long- Run Elasticities of Imports via the San Pedro Bay
Ports in Future Scenarios
Total - Optimistic I
Total - Optimistic II
Total - Near- term Likely
Total - Base Case
Total - Pessimistic
Transload - Optimistic I
Transload - Optimistic II
Transload - Near- term
Likely
Transload - Base Case
Transload - Pessimistic
12
In Optimistic scenarios, total import volumes via San Pedro Bay exceed Zero- Fee Base
Case volume until container fees rise to about $ 125-$ 150 per FEU. In the Short Run
analysis, trans- loaded volume in the Optimistic Scenarios exceeds that for the Zero- Fee
Base Case over the entire range of container fees tested, but in the Long- Run analysis the
trans- loaded volume falls to the Zero- Fee Base Case trans- loaded volume when container
fees rise to about $ 250 per FEU. Again this is an indication that adequate infrastructure
and/ or staffing are not yet in place at other ports to accommodate diversion of trans-loaded
volumes from the San Pedro Bay ports, but economic justification to make the
needed investments or staffing additions arises once container fees imposed at San Pedro
Bay are $ 250 per FEU or more.
In the Pessimistic scenario, total volume with no container fee is 11% less than Zero- Fee
Base Case volume, and trans- loaded volume is 9% less. At a fee of $ 200 per FEU, both
total volume and trans- loaded volume in the Long- Run Pessimistic scenario are less than
half what they were in the Zero- Fee Base Case scenario. Such a volume loss would seem
to be devastating to the Southern California economy.
Figure S- 4 depicts the results of a Long- Run elasticity analysis of the Near Term Likely
scenario supplemented with a major infrastructure program offering significant
congestion relief vs. the Zero- Fee 2006 Base Case Scenario. This is an update of the
analysis in the Phase I Elasticity Study. The assumed congestion relief program is very
ambitious, including dedicated truck corridors from the ports to the major warehouse
districts permitting 40 MPH operation of double- bottom drays, major expansion of port
and rail intermodal terminals, and expansion of rail- line- haul capacity. As in the Phase I
study, the assumption underlying this congestion relief scenario is that container fees are
not assessed until after the new infrastructure is made available for use by importers. As
may be seen, for a fee value up to about $ 150 per FEU, total market share of Asian
imports at San Pedro Bay exceeds or matches that of the 2006 Zero- Fee Base- Case
scenario. Examining the components of overall imports, the market share of inland- point
intermodal imports falls below that of the Zero- Fee Base Case scenario for fees above
$ 50 per FEU, while the market share of trans- loaded imports exceeds or matches that of
the Zero- Fee Base Case scenario for fees up to about $ 200 per FEU.
13
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
110%
120%
130%
$ 0 $ 50 $ 100 $ 150 $ 185 $ 200 $ 250 $ 300 $ 350 $ 400 $ 450 $ 500
% of Zero- Fee 2006 Base Case Imports
Fee Value per FEU at San Pedro Bay
Long- Run Elasticity of Imports via San Pedro Bay Ports,
2006 Base Case Scenario vs. Major Congestion Relief
Total - Congestion Relief
Total - Base Case
Transload - Congestion
Relief
Transload - Base Case
IPI - Congestion Relief
IPI - Base Case
Not analyzed was a scenario in which major infrastructure investments are assumed to be
made in Southern California, but no investments are made at other North American ports,
i. e., container flow times via those ports would increase if substantial import flows were
diverted to them. In such a scenario, the diversion of traffic away from the San Pedro Bay
ports when container fees are assessed would be somewhat less than what is depicted in
Figure S- 4. Nonetheless, then general nature of diversion would be similar – there would
be more diversion of IPI imports than of trans- loaded imports.
A summary of the important findings of the elasticity analyses in Phase II is as follows:
 Compared to the 2005 analysis, the elasticity of imports via San Pedro Bay to
potential container fees increased markedly. This was due to unfavorable
evolutions in rail intermodal rates and dray costs. Particular changes include the
disparate evolution of domestic- container and IPI rail rates ( the former went up
more in the 2003- 2007 period than the latter), disparate evolutions of domestic-container
rail rates from Southern California vs. from other West Coast ports ( the
former went up more in the 2003- 2007 period than the latter), aggressive rate
competition for IPI business via the new Prince Rupert port from the Canadian
National Railroad, and increases in dray costs in Southern California much greater
than at Pacific Northwest ports. The resulting rate disadvantage to Southern
California ports of $ 0.05 - $ 0.10 per cubic foot of cargoes ( depending on
destination) may not seem like much, but considering the 4,000 cubic feet of
space in a domestic container, that works out to be $ 200 - $ 400 per domestic
container. And considering that a high- cube marine container accommodates
2,700 cubic feet of cargoes, such rate disadvantages work out to be $ 135 - $ 270
Figure S- 4. Long- run Elasticity of Imports Routed via San
Pedro Bay, With and Without Congestion Relief
14
per FEU. In effect, the evolutions in steamship, rail and dray rates from 2003 to
2007 eliminated about $ 200 per FEU in inelasticity to container fees at San Pedro
Bay. As embodied in the Near Term Likely scenario, about $ 150 per FEU in
inelasticity is anticipated to be restored.
 Elasticity of imports to potential fees at San Pedro Bay is a function of rail and
steamship rates, market shares of large nation- wide importers, and other factors
not under the region’s control. At issue is whether or not there are favorable
developments in such factors that offset the impact of such fees, e. g., more
competitive rail rates from Southern California, a rise in steamship line rates via
the Panama Canal, increased market share for the large, nation- wide importers,
and increased rail terminal capacity in Southern California. With such things
present, small or moderate container fees do not result in volumes less than that in
the Zero- Fee Base Case Scenario. But absent such things, or worse, juxtaposed
with unfavorable developments such as a reduction in all- water rates or increases
in rail rates out of Southern California but not elsewhere, there could be
substantial drops in volumes resulting from the imposition of major fees.
 A major program of infrastructure improvements, whose bonds are retired by
container fees not put into place until the time the infrastructure is opened for
operation, can be a value proposition for large nation- wide importers practicing
trans- load import strategies. In fact, the San Pedro Bay ports’ share of such import
traffic can be grown by a well- thought- out congestion relief program. But a major
infrastructure program funded by container fees is much less a value proposition,
or even a negative value proposition, for importers primarily using IPI services,
including importers of low- value imports and small and regional importers. To
remain competitive for the latter market, fees must be kept low or avoided
entirely.
 The author believes that the Short- Run and Long- Run Elasticity Models show
much promise for interesting policy analysis and infrastructure planning. It is
exciting to be able to capture a complete view of Asia – US imports, the
economics involved, and the limitations of current infrastructure and logistics
services. However, in the author’s opinion, the amount of data on which the
Short- Run Model was calibrated is marginally adequate; much more could be
done to refine the model as well as to facilitate wider application for improved
policymaking, strategic planning, capital budgeting and financing of
transportation infrastructure improvements. Moreover, considering the available
budget, only a limited number of scenarios have been analyzed to date. There are
no doubt other scenarios of interest to policymakers that will arise.
 Compared to the results of the Phase I Study, the Phase II results provide a
cautionary lesson that elasticity of imports can change markedly in the span of
only several years, suggesting the need for continuing analysis to keep up with the
dynamics of industry and global economics.
Because of the ambitious scope of this study, this full presentation of the results is of
necessity quite long. This report provides complete documentation of the results of the
elasticity analysis, the assumptions underlying the analysis, and the development of the
methodology. To facilitate comprehension of the array of findings, new analytical
15
methodology, and applications of the methodology in policy analysis, this report includes
a nine- page Overview following this Executive Summary. Sections delving into the
details of the Study follow.
This report was prepared by Dr. Robert C. Leachman. The development of the Short- Run
Model was a fascinating and very challenging project. I would like to acknowledge the
assistance of Theodore Prince & Associates LLC, George R. Fetty & Associates, Inc., Dr.
Anne Goodchild, Mr. David Lehlbach and Arellano Associates with data collection and
stakeholder outreach efforts supporting the study. I also would like to express my
gratitude to various companies and organizations that assisted the Study with the
provision of insights or data to help calibrate the analytical models. In particular, the Port
of Long Beach graciously provided access to Customs data from its PIERS and WTA
subscriptions, and the BNSF and Union Pacific Railroads graciously supplied data on
train counts, lift counts and intermodal transit times through their networks. MARAD
also kindly provided PIERS data to the consultant. However, the Short- Run Model is an
original work of the author. None of the agencies assisting this study participated in the
development of the model, the analysis, or the formulation of findings and conclusions.
No endorsement by them of any contents of this report should be assumed.
16
1. Overview
In September, 2005, the Southern California Association of Governments made public
the “ Port and Modal Elasticity Study.” This Study developed an economic optimization
model predicting how importers would allocate Asian imports to port and landside
channels so as to minimize their total supply chain logistics costs ( considering
transportation, handling and inventory costs). Totals for all importers yield a prediction of
the overall allocations of imports to ports and channels. Repeated model calculations with
varying levels of hypothetical container or user fees and with varying assumptions about
container flow times and transportation rates enable policymakers to assess the elasticity
of imports. The Study may be down- loaded from the SCAG web site at
http:// www. scag. ca. gov/ goodsmove/ pdf/ FinalElasticityReport0905rev1105. pdf.
SCAG subsequently sponsored a Phase II of this study. In Phase II, the data and
assumptions of the model were refined, and capability was added to conduct “ short- run”
elasticity analyses whereby container flow times through ports and landside channels are
endogenous to the model. In predicting port and modal shares, the short- run analysis
accounts for congestion associated with potential shifts in port and modal allocations of
imports utilizing fixed levels of port and channel infrastructure.
This document is the Final Report for Phase II. Phase II included the following work
elements:
- Outreach to stakeholders concerning findings of the 2005 Elasticity Study ( discussed
above) and concerning Phase II elasticity research.
- Outreach to stakeholders concerning findings of a 2005 Southern California main- line
rail capacity analysis performed by the author. That study also may be down- loaded from
the SCAG web site at
http:// www. scag. ca. gov/ goodsmove/ pdf/ InlandEmpireRailStudyFinalReport. pdf .
- Updating data and trends concerning port and landside channel shares of Asia – USA
waterborne containerized trade volumes. These data are not used in elasticity
calculations, but serve as reference statistics about current practice for comparison to
results from analytical models.
- Updating the distribution of waterborne containerized imports from Asia to the United
States by commodity and value. These are important inputs to the elasticity analysis.
- Updating data concerning the transportation and handling costs for Asia – USA
waterborne containerized imports. These also are important inputs to the elasticity
17
analysis. Data on the size and composition of the fleet of domestic equipment for trans-loading
imports also was updated.
- Assessment of the impacts of port contracts and of carrier and terminal operating
strategies on the short- run elasticity of containerized imports from Asia to the United
States. The assessment of these impacts helped to shape the development of the short- run
elasticity analysis, as well as to understand limitations of the model.
- Development of analytical queuing formulas that predict container flow times as a
function of congestion in port and landside rail channels. The collection of these formulas
is termed the Queuing Model. It is the key new analytical development enabling short- run
elasticity analysis. Supporting these analytical formulas, a new database of port terminal
and rail intermodal terminal infrastructure was developed, as well as a new database of
trackage configuration of the rail line- haul network and traffic levels on the network.
- Development of a Short- Run Elasticity Model for predicting flows of waterborne
containerized imports from Asia to the United States through North American ports and
landside channels. This Model encompasses the previously- developed Long- Run
Elasticity Model, linked to the above- mentioned Queuing Model. The intent of this model
is to assess the elasticity of imports to potential container fees passing through selected
ports or landside channels assuming fixed rail line infrastructure and fixed port and rail
terminal infrastructure with fixed staffing schedules for those terminals.
Outreach to Stakeholders
During the period June 1, 2006 through July 30, 2008, meetings were held with railroads,
port terminal operators, ports, third party logistics firms, dray and trucking companies,
and major importers. The general feedback received from these stakeholders may be
summarized as follows: All stakeholders were grateful for the “ big- picture” insights
developed in the elasticity study. A typical remark: “ I am glad somebody is able to look
at the big picture.” Most stakeholders wanted to learn more about the study. All were
encouraging of continuing studies, and most were willing to provide data in support of
continuing studies. None were willing to express official support for infrastructure
improvements funded by user fees.
Additional stakeholder outreach meetings were held during the period October 2009 to
June 2010 with the San Pedro Bay ports, the Alameda Corridor Transportation Authority,
the BNSF and UP railroads, port terminal operators, dray and trucking companies, and
major importers. Their comments and feedback are reflected in this report.
As to the main line rail capacity study, all stakeholders expressed the view that plans
proposed by the study are beyond their planning horizons, typically one to five years, in
contrast to the five- to twenty- year horizons in the capacity study. For the near- term
( 2010) plans of the study, there was general acceptance, but a few objections were
expressed. BNSF and Metrolink felt that a separation of Colton Crossing was required by
2010. In contrast, the consultant found that a separation is not required for the 2010
18
forecasts of rail traffic ( assuming the BNSF main line is upgraded to have three main
tracks at the crossing), but such a separation is required at higher traffic levels and was
therefore included in the 2025 statement of requirements.
Updated Port and Model Shares of Trade Volumes
An extract of customs data for year 2005 in the PIERS database was provided to the
author by MARAD. These data specify for each US port the total volumes of imports and
exports ( measured in twenty- foot equivalent units, or TEUs). Other important data
sources examined by the author include 2005 and 2006 volumes reported by West Coast
ports and by the Pacific Maritime Association, 2005 and 2006 volumes reported by the
Intermodal Association of North America ( IANA), and the vessel strings serving Asia –
USA trade as reported by the steamship lines. The important trends that were observed
are as follows.
The share of total Asia - USA imports handled by West Coast ports continued to decrease
during the period 2003 - 2005, but the rate of decrease slowed considerably from
previous years. Considering all waterborne containerized imports from Asia to the USA
passing through US ports, in 2005, 74.5 % of total TEUs Asia – USA came through West
Coast ports, compared to 76.6% in 2003. The distribution of total Asia - USA vessel
strings by first port of call exhibits a similar trend.
The share of waterborne containerized imports from Asia to the USA passing through
West Coast ports whose landside movement was handled by rail intermodal was steady
over the period 2002 - 2006, averaging 46%. However, the shares at various West Coast
ports fluctuated significantly. During 2005, the percentage of marine containers entering
through Pacific Northwest ports that got on a train increased sharply, but then decreased
sharply in 2006. The percentage for the San Pedro Bay ports declined during 2005 but
then increased in 2006. In 2006, the figures for the Pacific Northwest ports and the San
Pedro Bay ports were 70% and 40%, respectively.
It is believed that these fluctuations are primarily due to two factors. First, the steamship
lines shifted certain vessel strings from San Pedro Bay to Puget Sound for the 2005
season, evidently in response to the summer, 2004 “ melt- down” at the San Pedro Bay
ports. After an uneventful 2005 season at San Pedro Bay, these vessel strings were shifted
back to San Pedro Bay for the 2006 season. Also in the 2006 season, several new vessel
strings serving San Pedro Bay using very large new vessels were introduced. Second, the
allocation across ports of entry by imports warehoused in the hinterlands of ports of entry
and then re- shipped to demand points in domestic vehicles has diversified. Port of entry
for certain products that formerly were mostly or fully imported through San Pedro Bay
and trans- loaded to domestic vehicles in Southern California became distributed across
several ports. For example, most large, nation- wide “ big- box” retailers practice a “ Four
Corners” policy, using two West Coast ports and two East Coast ports, each serving a
quarter of the continental United States ( and providing back- up supply to other quarters
as required), or similar policies involving 3 or 5 ports. This has resulted in a net
19
percentage increase in trans- loading activity at the Pacific Northwest ports and certain
East Coast ports and a net percentage decrease at the San Pedro Bay ports.
The reasons for the shift from the trans- loading- all- at- San- Pedro- Bay strategy to
multiple- port- trans- loading strategies are multiple, but two reasons stand out. First, with
the introduction of PierPass in Southern California and the introduction of trans- loading
facilities in the Sumner- Puyallup area relatively close to the Puget Sound ports, dray
costs faced by trans- loading importers are significantly less in the Pacific Northwest.
Second, goods that used to be imported by the manufacturer/ wholesaler to a warehouse in
Southern California and then re- sold to US retailers are increasingly purchased in Asia
from the manufacturer/ wholesaler by large “ big- box” retailers. The large retailers import
the goods themselves using “ Four Corners” or similar policies.
Combining data from multiple sources, the following break- down of 2006 containerized
imports through the San Pedro Bay ports was estimated: 21% was “ local” traffic, i. e.,
imports consumed in Southern California, Southern Nevada, Arizona, New Mexico,
Southern Utah or Southern Colorado; 43% was kept in the marine box and placed on a
double- stack train destined east of the Rockies ( this is known as inland- point- intermodal
or “ IPI” volume); and the remaining 36% was either ( a) unloaded from marine boxes in
the local region at a warehouse or trans- loading facility, re- loaded in domestic vehicles
( truck or rail) and re- shipped for consumption outside the local region, or ( b) kept in a
marine box that was trucked outside the above- defined “ local” region. The ( b) part of the
36% category is believed to be very small. Thus the amounts of traffic in IPI and trans-loading
categories at San Pedro Bay are roughly equal, and each is about double the local
traffic. For the West Coast as a whole, “ local” traffic was about 30% in 2006; IPI traffic
was about 46%; and trans- loading/ long- distance trucking was about 24%.
Since 2006, the SPB ports have lost some market share. The breakdown of 2008
containerized imports through the San Pedro Bay ports is estimated as 23% local region
traffic, 41% IPI, and 36% trans- load to domestic containers or trailers for re- shipment out
of the region plus out- of- region trucking of marine boxes.
Updated Distributions of Imports by Commodity and Value
Summaries of Customs data for year 2005 compiled by the Port Import Export Reporting
Services ( PIERS) and World Trade Atlas ( WTA) data subscription services were
provided to the author by the Port of Long Beach. These databases classify imports into
99 commodity types. The PIERS data provides volumes by commodity type ( expressed
in twenty- foot equivalent units, or TEUs). The WTA data provides total dollars of
declared values in each commodity code. The PIERS data furnished to the author spans
all waterborne containerized imports from Asia to the United States passing through West
Coast ports. The World Trade Atlas data provides summaries by West Coast, East Coast
and all USA ports. In addition, the U. S. Dept. of Transportation Maritime Administration
( MARAD) provided the author with PIERS total volumes by port for Asian imports in
2005, but no break- out by commodity type. These data enabled the author to make
20
estimates at the nation- wide level for volumes and declared values per cubic foot by
commodity type.
The author previously performed a similar analysis on 2003 Customs data for the 2005
report. Trends 2003 to 2005 in the distributions by commodity and value were therefore
assessed.
Generally, the distribution of declared values for Asia – USA waterborne containerized
imports showed little change from 2003 to 2005. The average declared value per cubic
foot of container capacity for these imports rose from $ 21.47 in 2003 to $ 21.66 in 2005.
Declared values of Asian imports routed via West Coast ports are in aggregate greater
than those routed via East and Gulf Coast ports; in 2005, the average declared value via
West Coast ports was $ 22.66, while it was $ 18.57 via East and Gulf Coast ports. Again,
this difference is little changed from that for 2003.
It is convenient to classify imports as inexpensive ( less than $ 13 per cubic foot of
container capacity), moderate ( between $ 13 and $ 26 per cubic foot), and expensive ( more
than $ 26 per cubic foot). In 2005, about 25% of imports were inexpensive, 50% were of
moderate value, and 25% were expensive. Compared to the 2003 distribution, the “ tails”
of the 2005 distribution spread out a bit, i. e., inexpensive goods became a bit cheaper and
expensive goods became a bit more expensive, but the price- points for the 25- 50- 25 split
of the distribution in 2005 remained basically unchanged from those for 2003.
To the author, this was a somewhat surprising result. During the period 2003 – 2005,
energy and transportation costs rose and there were upward pressures on Asian
currencies. But anecdotal evidence received from importers indicates there was an
increase in the number of competitive suppliers in Asia for production of certain goods.
The net overall effect was to leave the value distribution largely unchanged. It remains to
be seen if, in future years, currency revaluations and rising energy and transportation
costs shift upwards the value distribution curve for Asian imports.
Updated Transportation and Handling Costs
Transportation and handling costs for containerized imports from Asia to the United
States were updated to levels prevailing in April, 2007. The availability of domestic
containers for trans- loading imports out of marine containers for furtherance in domestic
vehicles also was updated.
For the purposes of elasticity studies, the continental United States is subdivided into 21
regions. Costs to ship imports from the ports of Shenzen, Yantian and Chiwan in
mainland China to selected single destinations within each region were researched. Costs
to importers for routing imports via ten alternative North American ports of entry were
developed. For each port of entry and each destination, rates were developed for two
alternative supply- chain channels: ( 1) shipping marine containers direct from China to
regional destinations, and ( 2) shipping marine containers to trans- loading warehouses in
the hinterlands of the ports of entry, thence re- loading the imports in either domestic rail
21
containers or domestic trailers for re- shipping from trans- loading warehouses to regional
destinations.
Rate quotations to various importers from steamship lines, non- vessel- operating common
carriers, intermodal marketing companies, trans- loading warehouse operators, and
trucking companies were secured by the author. Considerable variation in rates from
carrier to carrier and customer to customer was encountered. Average rates were
developed from a basket of rates for each channel.
The great majority of waterborne containerized imports from Asia to the United States
are “ cube” freight rather than “ weight” freight, in the sense that vehicles reach cubic
capacity limits before weight limits are reached. Because of the disparity in vehicle size,
it is convenient to normalize transportation and handling costs on a per- cubic- foot- of-imports
basis. Roughly speaking, the contents of three high- cube 40- foot marine boxes fit
in two 53- foot domestic vehicles, assuming the imports are “ cube” freight rather than
“ weight” freight. In general, use of the trans- loading channels requires a $ 0.00 to $ 0.20
premium per cubic foot of imports in transportation and handling charges, compared to
direct shipping. These extra transportation costs must be traded off against potential
inventory savings afforded by pooling shipments to multiple regional destinations over
the segment of the supply chain between Asia and the trans- loading warehouse. For high-value
goods, such consolidation – de- consolidation supply- chain strategies are attractive;
for low- value goods, they are not.
The viability of consolidation – de- consolidation supply- chain strategies depends upon an
adequate supply of domestic equipment. It was confirmed by the author that the
aggregate cubic capacity of domestic containers is continuing to grow at a rate
comparable to the growth in imports. Considering the increased outsourcing of
manufacturing from the United States to Asia ( and hence declining volumes of domestic
freight), this means there is sufficient equipment to expand the level of trans- loading
activity. Looking ahead, a concern for the attractiveness of the trans- loading strategy is
that decreased westbound domestic traffic from the US Midwest to the West Coast will
lead to increased westbound empty movement of domestic vehicles and upward pressure
on the eastbound domestic rates used by trans- loading importers.
Impact of Contracts and of Terminal and Carrier Operating Strategies
Steamship lines enter into long- term ( 10- 30 year) contracts with ports. Many of these
contracts involve fixed payments and/ or volume incentives. Some offer incentives for rail
intermodal movement of the marine containers ( as opposed to placement of containers on
truck chasses). These contracts limit or delay the flexibility of steamship lines in
restructuring their vessel strings or their strategies for which port to off- load cargoes
destined to inland points. The Short- Run Elasticity Model does not directly treat such
constraints, but it admits them. In making a model run, the user may input required
minimum import volumes for the ports that are respected in model calculations.
22
Steamship lines typically enter into contracts with a single western railroad ( either BNSF
or UP) to support their inland- point intermodal ( IPI) services. Before 2006, these were
typically long- term ( 8- 10 year) contracts at favorable rates. All the more recent contracts
have been year- to- year at 25- 40% higher rates. Because some lines still enjoy legacy
long- term contracts at discount rates while others pay the new higher rates, there have
recent major shifts in market shares of the steamship lines, and this in turn has resulted in
shifts in market shares between railroads, and, to a lesser extent, between ports ( the latter
because of the long- term contracts described above). Because the Short- Run Elasticity
Model is based on averages of a basket of rate quotations, it ignores differences between
lines. The last of the legacy discount contracts is set to expire in 2011, so hopefully this is
only a temporary shortcoming of the model.
Major customers of steamship lines enter into contracts each spring for shipping over the
subsequent one- year period May- to- May. Lines and major importers are loathe to make
major adjustments to vessel service and supply- chain strategies, respectively, except at
the May start of the annual shipping season. Thus changes predicted by model
calculations may take some time for the industry to implement.
Before 2006, West Coast ports had major imbalances in the counts of inbound and
outbound containers. The San Pedro Bay Ports had a surplus of inbound containers, while
Oakland and the Puget Sound Ports had a surplus of outbound containers. Beginning in
2006 the railroads changed the terms of their rates and charges for major steamship line
customers. Under the new terms, if a line’s inbound and outbound traffic to a West Coast
port area is out of balance, major penalties are imposed. ( The port areas for which this
individually applies are San Pedro Bay, Oakland and Puget Sound.) As a result, container
flows in and out of West Coast ports are much more in balance. In particular, there are
more empty containers and export loads handled through the San Pedro Bay ports than
before. In the Short- Run Elasticity Analysis Model, we only study imports and ignore
issues of imbalance in returning westbound containers. This was an important issue
among West Coast ports before flows were balanced at each port, but now that they are, it
is anticipated that this balance will persist.
Before 2005, the gate at most West Coast port terminals was open one shift per day or
perhaps two. After the institution of the PierPass program, a number of terminals on San
Pedro Bay began night- shift operations, and growth of this practice has continued. This
has a significant positive impact on terminal capacity and container flow times. In the
Short- Run Elasticity Model, we explicitly account for the number of shifts per day
terminals are operated.
A common practice among steamship lines when unloading vessels is to give preference
to IPI containers over most containers that will exit the terminal on a truck chassis. Thus
IPI containers and containers for local delivery have differing flow time statistics. These
differences are accounted for in the Short- Run Elasticity Model.
Some large importers have negotiated contracts with steamship lines allowing them extra
time to pick up inbound loaded containers before demurrage is assessed. In effect, the
23
port terminal is used as a storage area by the importer. We ignore such phenomena in the
Short- Run Elasticity Model.
Transit times for domestic- container intermodal trains tend to be shorter and more
reliable than transit times for marine- container intermodal trains. We account for such
differences in the Short- Run Elasticity Model.
Development of Queuing Formulas to Predict Container Flow Times
Analytical queuing formulas were developed for estimating import container flow times
through port terminals, rail intermodal terminals and rail line- haul channels as a function
of traffic volumes, infrastructure and staffing. Queuing theory is an area of Operations
Research pioneered by English researchers in the 1950s with continuing development by
American and international researchers up to the present day. Analytical formulas have
been developed in this research expressing the expected or average time customers wait
in a service system, as well as the total time spent in the system ( i. e., wait time plus
service time). In this report, queuing- theoretic formulas are developed to model container
flow times through port terminals, rail intermodal terminals and rail line- haul channels.
The queuing- theoretic formulas express waiting time as a non- linear function of
utilization and the number of parallel servers. As utilization is increased, waiting time
increases exponentially. For a fixed utilization, the waiting time can be mitigated by
increasing the number parallel servers ( e. g., more lift crews in an intermodal terminal or
more tracks on a rail line).
The queuing formulas developed for each of the three types of applications ( port
terminals, rail terminals, rail line hauls) were statistically fitted to 2006 industry data to
provide models of container flow time as a function of parameters for traffic volume,
infrastructure ( e. g., terminal acreage, number of rail main tracks), staffing, and hours of
operation. The analyst may employ these models to calculate predictions of changes in
container flow time as a function of changes in the parameters.
The formula developed for flow time through port terminals is as follows:
2.3
( 1 )
0.31*
2( 1) 1
  


 




 
m u
CT u
m
( S1)
where CT denotes the average cycle time ( in days) for imported containers, measured
from ship arrival until truck departure out the gate or until release of double- stack train
for pick- up by the railroad. The parameter m measures the number of loading crews
working in parallel placing containers onto truck chasses or into railroad double- stack
well cars. The parameter u measures the utilization of the loading crews and working
space at the terminal, defined as the number of import containers handled per acre per
crew- shift, divided by 4.
24
The formula developed for container flow times at rail intermodal terminals is similar in
structure:
0.334
( 1 )
0.365*
2( 1) 1
  


 




 
m u
CT u
m
( S2)
where CT expresses the average time ( in days) from truck entry of the gate of the
terminal until departure of the intermodal train. The parameter m expresses the number of
parallel loading crews while the parameter u expresses the utilization of loading crews
and working space at the terminal, defined as the total number of lifts ( both inbound and
outbound) per acre per loading crew- shift, divided by 4.
Data also was furnished by the railroads concerning 2006 average dwell times at West
Coast on- dock terminals from completion of loading of double- stack trains by the port
terminal until departure of the train. A weighted average of these data is 7.1 hours.
The development of a queuing- theoretic mathematical model to estimate intermodal line-haul
transit times ( from departure at origin terminal until arrival at destination terminal)
is summarized as follows. Data supplied by the railroads for rail corridors from West
Coast terminals ( Seattle, Tacoma, Oakland, Los Angeles – Long Beach) to major
Midwest destinations ( Chicago, Minneapolis, Kansas City, Dallas and Houston) were
analyzed by the author. It was necessary to apply the queuing- theoretic formulas to
individual segments of each of these rail corridors, whereby each corridor was broken
down into segments with constant numbers of main tracks and approximately uniform
through- train frequencies. Separate models were calibrated for transit times of
international intermodal trains and for transit times of domestic intermodal trains. The
inputs to the models include the following:
- Distance, speed, no. of main tracks for each segment of each route
- Average no. of through train movements per day on each segment
- No. of crew changes and no. of locomotive refueling stops on each route
- Extra running time for a train stopped in a siding to pass an opposing movement
on single track
The mathematical form of the model is quite involved; it is not practical to present it in
this executive summary. The interested reader is invited to review the body of this report
for complete details. The parameters of the model were fit statistically to 2006 data
provided by BNSF and Union Pacific railroads. The output of the model is the expected
( statistical average) transit times for domestic and international intermodal trains. A
database of the main- track configurations of the rail corridors, as of late 2006, was
developed by the consultant and is included as an Appendix of this report.
The Short- Run Elasticity Model
25
A particular desired enhancement to the elasticity analysis concerned the capability to
perform a “ short- run” elasticity analysis. In a short- run analysis, port and landside
infrastructure, staffing levels and operating schedules are pre- specified inputs to the
analysis, in lieu of pre- specifying statistics on container flow times. In a short- run
analysis, container flow times by port and channel are calculated by the model as a
function of traffic levels. The results of a short- run analysis predict changes in import
flows resulting from the imposition of a container fee assuming no changes in port and
channel infrastructure or in staffing levels and operating schedules of the infrastructure.
This assumption contrasts with the underlying assumption of the Long- Run Model, which
assumes that infrastructure at other ports and channels serving those ports would be
expanded as necessary to maintain current container flow times for increased shares of
imports routed through those ports and channels.
In Phase II the consultant updated the database of import distributions by region,
importer, commodity and value, as well as the database of transportation rates. A new
database was developed concerning the existing infrastructure and staffing levels of port
terminals, rail terminals, and the trackage configuration of the intermodal rail line- haul
network. New analytic queuing formulas were developed by the consultant that predict
container flow times through port terminals, rail intermodal terminals and rail line- haul
movement as a function of import volume. These formulas were statistically calibrated to
data supplied by port terminal operators and the railroads. The collection of these queuing
formulas is termed the Queuing Model.
The Long- Run Elasticity Model developed by the author in Phase I was upgraded in
Phase II and is now termed the Supply- Chain Optimization Model. Working importer by
importer, the Supply- Chain Optimization Model determines the least- cost supply chain
strategy for each importer, in terms of ports and landside channels to be used, where costs
considered include costs for transportation and handling, container fees, pipeline
inventory, and safety- stock inventory at destination regional distribution centers. The
consequent import volumes by port and channel for all importers are tallied by the model
to deduce the overall distribution of import flows.
The Short- Run Elasticity Model is an outgrowth of this Long- Run Elasticity Model. It
consists of the Supply- Chain Optimization Model and the Queuing Model working in
tandem. Iterative supply- chain optimization and queuing calculations are made within the
Short- Run Model. Starting with initial estimates of container flow times, the Supply
Chain Optimization Model selects supply- chain strategies for importers and tallies
volumes through ports and channels. The Queuing Model takes those volumes as input
and updates container flow times. Updated flow times are fed back to the Supply- Chain
Optimization Model which in turn re- selects supply- chain strategies, and so on. After a
series of iterations, the Short- Run Model converges to a stable set of import flows and
reports the result. In all test applications to date, an equilibrium solution has been reached
within ten iterations.
The Short- Run Elasticity Model calculates import volumes by port and landside channel
as a function of given infrastructure and operating hours for port and rail terminals, given
26
trackage configurations of the rail network and given levels of non- import rail traffic,
given transportation rates, given contractual volume requirements at ports, given import
volumes and a given value distribution for those imports. Like the Long- Run Elasticity
Model developed before it, the Short- Run Model assumes a given distribution of imports
among 83 large, nation- wide importers and 19 generic importers acting as proxies for
small and regional importers, tailored to match the overall declared- value distribution of
imports reflected in customs data. The continental United States is divided into 21
regions, with the entire import demand for each region concentrated at a single location.
The geographical distribution of import destinations is assumed to be the same for all
importers. At present, this distribution is set to be proportional to purchasing power in the
regions, but other distributions could be input to the model. At present, eleven alternative
ports of entry in Canada, the United States and Mexico are considered. Like the Long-
Run Model, the Short- Run Model performs the Supply- Chain Optimization calculations
to select the least- cost supply- chain strategy for each type of importer, considering total
transportation and inventory costs borne by the importer.
The intent of the Long- Run Model is to assess the wisdom of potential long- term
investments in port and landside transportation infrastructure, as well as to assess the
impact of user fees to recover costs of such improvements. In the Long- Run Model,
container flow times by channel are fixed, reflecting an assumption that over the long
term the various ports and transportation carriers would make investments to maintain
existing service quality and thereby protect market share. This conservative assumption is
suitable for assessing the merits of potential investments with 25- 50- year payback
periods, as the intent is to evaluate potential investments assuming competing ports and
competing channels may make the necessary investments to maintain their current
service quality in the face of growing volume or growing competition.
In contrast, the Short- Run Model assumes the infrastructure of the entire transportation
network is pre- specified and fixed. 7 It also observes minimum volumes that must be
channeled through various ports, reflecting the requirements of prevailing contracts.
Container flow times are endogenous to the Short- Run Model, responding to congestion
( or lack thereof) in various ports and channels. The Short- Run Model is thus useful for
projecting more near- term responses of importers to changes in fees, rates or
infrastructure.
Tandem calculations of the two models provide a range for the diversion of import
cargoes resulting from imposition of container fees. A conservative, short- term estimate
stems from the short- run calculation, while a liberal, long- run- potential estimate stems
from the long- run calculation. The Models may be used to predict changes in import
traffic flows in response to not just potential fess, but also to changes in port and rail
terminal infrastructure, staffing or operating hours; changes in rail network configuration
or non- import traffic levels; changes in transportation rates; changes in the distribution of
imports by value and by importer type; changes in the geographical distribution of import
destinations; or changes in overall import volumes.
7 Although the infrastructure and operating schedules input to the model need not be the same as current
actual conditions, i. e., future scenarios can be analyzed.
27
Elasticity Analyses
Applications of the Long- Run and Short- Run Models were made to analyze hypothetical
user fees at the San Pedro Bay ports in several scenarios, including a 2007 Base Case, a
Near- term Likely scenario, an Optimistic I scenario ( in which all- water rates rise by
10%), and Optimistic II scenario ( in which the share of total imports for large, nation-wide
importers rises from 40% to 50%), and a Pessimistic scenario ( in which all- water
rates fall by 10%).
Potential container fees in increments of $ 50 per FEU up to $ 500 per FEU were tested in
model runs, and changes in the distribution of import flows were observed.
The Base Case scenario has the following features: 2006 total volume of Asia – USA
waterborne containerized imports, 2005 distribution by declared value, 2007
transportation and handling rates, and mid- 2006 infrastructure at ports and in landside
channels. Large, nation- wide importers with average declared values for imports as
specified in the consultant’s Phase I ( 2005) report are assumed to have a 40% share of
total imports. This Base Case represents the consultant’s best estimate of conditions
prevailing in 2007. Solutions of the Short and Long- Run Models for the Base Case
Scenario match actual import flows in 2006- 2007 very well.
The four future scenarios incorporate the same total volume of imports and the same
distribution of imports by declared values as in Base Case Scenario, but vary assumptions
about the evolutions of rail and steamship line rates and about future terminal
infrastructure and staffing. One near- term future scenario, termed the Near- term Likely
Scenario, and three longer- term future scenarios were formulated.
In terms of infrastructure, the Near- term Likely scenario is the same as the Base Case
Scenario except a domestic intermodal rail terminal that was opened in 2009 at the Port
of Tacoma is included in the scenario. Compared to the Base Case, significant
adjustments were made to rail rates in this scenario: ( 1) Domestic rail container rates
were adjusted to reduce the gap between rates via West Coast ports for inland point
intermodal ( IPI) movement of marine boxes and rates for reshipment in domestic rail
containers after trans- loading. The gap was reduced by $ 0.10 per cubic foot of imported
goods to Eastern destinations and by $ 0.05 per cubic foot to Midwestern destinations. ( 2)
IPI and domestic container rail rates via San Pedro Bay Ports were adjusted to be more
competitive with other USA West Coast ports to all Midwestern and Eastern destinations
except Minneapolis. ( Seattle- Tacoma has a rate advantage for imports destined to the
Minneapolis region that is retained in this scenario.) After the adjustments described in
( 1), the total transportation and handling cost per cubic foot for the trans- loading channels
via West Coast ports are $ 0.00 - $ 0.12 more per cubic foot than direct inland movement
of marine boxes using IPI service, depending on the destination region. The rationale for
( 1) is that the gap between domestic- box and marine- box rail rates widened considerably
during the period 2004 – 2008 because of fuel recovery surcharges placed on domestic
rates while no fuel recovery surcharges were placed on the international “ all- in” IPI rates.
28
Moreover, enough steamship lines continued to enjoy long- term legacy contract rates
from railroads so as to keep IPI rates low. As the legacy contracts expire, the lines are
forced into shorter- term contracts for IPI service from the railroads that feature steep rate
increases, ranging 25% - 40%. The last of the legacy contracts will expire in 2011.
Finally, the decline of the domestic economy has made the supply of domestic rail
containers plentiful and placed downward pressure on domestic rates. The rationale for
( 2) is as follows: The 2007 rail rate quotations secured by the consultant favor Pacific
Northwest ports over Southern California ports to a number of destinations. This made
sense, perhaps, at a time when rail lines serving Southern California were more congested
than lines serving the other West Coast ports, and when westbound was the head- haul
direction for domestic boxes to/ from the Pacific Northwest while eastbound was the
head- haul direction to/ from California. Starting in 2006 and continuing to the present, the
railroads have made large investments to double- track their transcontinental main lines
serving Southern California. The consultant expects the railroads to adjust their rates so
as to insure utilization of that investment in lieu of encouraging traffic to use other West
Coast ports served by rail lines with less capacity. The consultant believes this scenario is
likely in the near term.
Beyond the near- term, it is difficult to forecast transportation rates and services and the
shares of imports by large, nation- wide importers vs. small, regional ones. Accordingly,
the consultant prepared several alternative scenarios illustrating the range of outcomes
that are plausible. One crucial variable is what will happen to so- called “ all- water” rates
charged by steamship lines for container shipment via the Panama Canal to East and Gulf
Coast ports. An optimistic scenario tested by the consultant features such rates rising by
10%. A pessimistic scenario features such rates falling by 10%. Another crucial variable
concerns the share of total imports in the hands of large, nation- wide importers vs. that in
the hands of small and regional importers. Accordingly, another optimistic scenario is
formulated in which the total import share in the hands of large, nation- wide importers
rises from 40% to 50%. A final important variable concerns the available terminal
capacity and crew- shifts at port and rail terminals serving the various West Coast ports.
Accordingly, the optimistic scenarios assume the BNSF railroad’s proposed Southern
California Intermodal Gateway ( SCIG) terminal is opened. The pessimistic scenario
features increased terminal capacity at other West Coast ports but no increase at San
Pedro Bay ports. Summary descriptions of the two optimistic and one pessimistic
scenario are as follows:
Optimistic I: Includes all features of the Near- term Likely Scenario. In addition: assumes
that the proposed BNSF SCIG rail terminal is opened, all- water steamship line rates via
the Panama Canal are raised by 10%, and there are increased crew- shifts at certain
Southern California rail terminals.
Optimistic II: Includes all features of the Near- term Likely Scenario. In addition: assumes
that the proposed BNSF SCIG rail terminal is opened, the share of total imports for large,
nation- wide importers rises to 50%, and there are increased crew- shifts at certain
Southern California rail terminals.
29
Pessimistic: Includes all features of the Base Case Scenario. In addition: assumes all-water
steamship rates via the Panama Canal are lowered by 10%, a new domestic
intermodal rail terminal that was opened in 2009 at the Port of Tacoma is included, and
there are increased crew- shifts of operation at Oakland and Pacific Northwest rail
terminals.
For the Base Case Scenario, the Short- Run Elasticity Model predicts the imposition of a
$ 100 per FEU container fee on imports via San Pedro Bay would result in a 10% drop in
the market share of the San Pedro Bay Ports. The Long- Run Elasticity Model predicts a
23% drop for the same fee. Most of the diverted volume would move to the Puget Sound
and Canadian West Coast ports. The specific amount of traffic loss from the San Pedro
Bay ports would depend on the extent to which those ports increase operating hours,
crews on duty, and/ or acreage of their port terminals. It also would depend on potential
responses of the railroads, who might be incentivized to adjust the transportation rates
that they charge steamship lines for imports routed via Puget Sound ports vs. rates
charged for imports routed via San Pedro Bay.
For the future scenarios, the elasticity results vary widely. In the Near- Term Likely
scenario, total imports exceed Zero- Fee Base Case imports up to $ 100 per FEU in the
Short- Run calculation and $ 75 per FEU in the Long- Run calculation. In Optimistic
scenarios, total imports exceed Zero- Fee Base Case imports up to about $ 125 - $ 150 per
FEU in both the Short- Run and Long- Run calculations. In contrast, in the Pessimistic
scenario, total imports via San Pedro Bay fall sharply with fees. For a fee of $ 200 per
FEU, total imports via San Pedro Bay fall by about 30% in the Short- Run calculation and
50% in the Long- Run calculation.
A Long- Run Elasticity calculation also was made of the Near- Term Likely scenario
assuming a major program of congestion relief is in place before fees are assessed. This is
the same program that was analyzed in the Phase I study. The results are somewhat
different this time around. For container fees uniformly assessed on all imports, a fee of
$ 150 per FEU results in the same market share for the San Pedro Bay ports as in the
Zero- Fee 2007 Base Case scenario. For higher fees, total market share falls below of the
Zero- Fee Base Case. Considering the components of overall imports, the share of IPI
imports begins to fall below the Zero- Fee Base Case share once fees greater than $ 50 per
FEU are assessed, while the San Pedro Bay ports’ share of imports managed under the
trans- load strategies would be higher than in the Zero- Fee Base Case only for fee values
up to $ 200 per forty- foot equivalent unit ( FEU).
The contents and conclusions of this report reflect solely the views of the author, and not
those of the ports, terminal operators, the railroads, dray and trucking companies,
logistics providers, SCAG, DOT, MARAD, or any other agency assisting this study.
Although various importers, logistics firms, port terminal operators, Union Pacific and
BNSF graciously supplied raw data and qualitative insights aiding the development of the
Queuing Model, these parties were not involved in model development, analysis or
conclusions; and, therefore, they should not be considered to have endorsed any findings
in this report.
30
2. Outreach to Stakeholders
The consultant met with importers, transportation and logistics service providers, ports,
and port and terminal operators. Feedback concerning the methodology and findings of
the prior studies was requested, as well as data and guidance for performing the analysis
in Phase II. During the period June 1, 2006 through June 30, 2008, meetings were held
with the following stakeholders:
Railroads ( BNSF and Union Pacific)
Port Terminal Operators ( SSA Marine and MTC)
Ports ( Tacoma, Seattle, Vancouver, Long Beach, Los Angeles)
Third Party Logistics Firms ( Expeditors, Cal Cartage, APL Logistics, NFI National
Distribution Centers, American Port Services)
Dray companies and associations ( Container Freight EIT, Premier Transport, Washington
Trucking Associations)
Major importers ( Target, Toys ‘ R Us, Toyo Tires, Sony)
In addition, presentations were made at stakeholder forums sponsored by the following
agencies:
Distribution Managers Association ( Southern California Chapter)
SCAG ( Goods Movement Task Force)
Appendix A of this report lists the specific stakeholder meetings that were held. The
following sections summarize the feedback received by the consultant at those meetings.
2.1. Feedback from Stakeholders – Elasticity Studies
Feedback from Railroads
Both BNSF and Union Pacific expressed anxiety about user fee mechanisms. There is the
fear that funds so collected might be diverted to pay for other, unrelated purposes. There
is also the fear that fees might not have an appropriate “ sunset” provision, or that
legislation might be enacted extending fees indefinitely, i. e., the charges might continue
even after the infrastructure bonds are retired, again resulting in funds diverted to pay for
unrelated or unapproved purposes. They expressed concern that there is the prospect of
this in the case of the Alameda Corridor fees. 8
Some railroad managers expressed skepticism of continued growth in market share for
imports trans- loaded at West Coast ports. They have experienced strong demand for
intermodal movement of marine boxes to new inland distribution centers such as at
Logistics Park, IL.
8 In fact, there is a sunset provision in the case of the Alameda Corridor.
31
Feedback from Ports
The Port of Long Beach has been extremely helpful with data for the elasticity study,
sharing Customs data secured under their subscriptions to PIERS and the World Trade
Atlas. A meeting was held with Long Beach staff who explained the terms of port leases
and operating agreements. Long Beach staff also provided the consultant with statistics
they collected concerning the fraction of imported marine boxes moving inland on rail.
The Pacific Northwest ports were envious of the analyses the Ports of Long Beach and
Los Angeles have received from the SCAG studies. They would like similar analyses
performed for potential improvements in the access infrastructure to their ports.
The Port of Vancouver explained that imports to the USA via Canadian ports are
practical in the case of direct shipping of marine containers. The marine boxes may move
in bond from the Canadian port to the USA border, so that Canadian duties do not need to
be paid by importers. Trans- loading of imports destined to the USA also can avoid
Canadian duties if the entire contents of the marine boxes are going to the USA and if the
trans- loading is carried out in a bonded warehouse. But de- consolidation and trans-loading
at a third- party or importer- owned facility cannot avoid Canadian duties. And if
inventory is to be held in Canada for some time and its final destination is not yet known,
or if the contents of the marine box have mixed US and Canadian destinations, again
Canadian duties on the entire contents cannot be avoided. Thus, it is uneconomic for
importers to develop supply chains involving de- consolidation for the USA market using
a Canadian port. Large importers distributing across both USA and Canada are forced to
maintain separate supply chains for the Canadian and American markets. This is
inefficient for them; typically, the Canadian market is one tenth the size of the USA
market.
Feedback from Port Terminal Operators
Port terminal operators explained that the throughput capability of port terminals is
determined by available acreage, staffing hours and staffing levels. If more working shifts
are added and more space is provided, then more volume can be handled. If space is
available, the port terminal operators believe existing port terminals have the capability to
handle much more volume. They simply add more shifts when volume requires it
( presuming they can obtain the workers).
Both SSA Marine and MTC, Inc. provided the consultant with statistical data on
container dwell time vs. terminal utilization.
Feedback from Third- Party Logistics Providers
32
Third- party logistics providers observed a sharp up- tick in trans- loaded volumes in the
spring of 2006 after rate increases were announced for direct inland shipping of marine
boxes. They expect the trans- loaded market share to continue to improve. They note that
railroads now have pricing power they did not have before. As long- term contracts with
steamship lines for inland movement of marine boxes expire, the railroads are sharply
raising the rates and demanding short contract durations ( e. g., 1 year). Some lines
experienced 30- 40% rate increases from the railroads in the last couple of years. This is
driving more import volume towards trans- loading strategies.
The primary opportunity area for increased trans- loaded volumes concerns cases where
imports by a wholesaler are sold “ on the water” to retailers, and the goods are trans-loaded
into shipments to retailers from a de- consolidation warehouse in the hinterland of
the port of entry. “ The big- box- store companies already finished their transition to de-consolidation
import strategies. Now the forefront of activity concerns integration of
wholesaler and retailer supply chains by using de- consolidation in the hinterland of the
port of entry and elimination of stationary inventory.”
Another area of trans- loading growth concerns importers with a mixture of “ weight
freight” and “ cube freight”, such as a merchant of home improvement products ( nuts and
bolts are weight freight while furniture and cabinetry are cube freight). By suitably
mixing weight freight and cube freight at a de- consolidation center, landside transport
costs can be significantly reduced.
Trans- loading and de- consolidation is growing rapidly at the Pacific Northwest ports.
Feedback from Dray Companies
Driver shortages are a great challenge, yet dray companies still are not paid well. In
Southern California, some importers and dray companies have organized to conduct all
operations in and out of the port terminals at night to avoid the PierPass fee. In the Pacific
Northwest, there has been considerable growth of trans- loading and de- consolidation
facilities in Sumner, Puyallup and other municipalities in the Kent Valley. Dray operators
report that they are often able to complete four import box movements from the Port of
Tacoma to these de- consolidation facilities within a single driver shift.
Feedback from Large Importers
After the Summer, 2004 “ meltdown” at the San Pedro Bay ports, a number of importers
chose to diversify their supply chains. Most operating on a nation- wide scale with
sufficient volume for de- consolidation strategies ( e. g., the bog- box retailers) now practice
a “ four corners” strategy whereby nation- wide imports are allocated among four ports of
entry, and then de- consolidation is carried out at each port of entry. 9 Thus, distribution in
Southern California has evolved from distributing nation- wide to more focus on
distributing for consumption only in the Southwest. The number of firms practicing de-
9 Some importers practice two- corner, three- corner or five- corner strategies, but the basic concept is the
same.
33
consolidation went up, and the total import volume went up, so the total trans- loaded
volume through the San Pedro Bay ports did not decrease in spite of the increased
adoption of “ four corners” supply- chain strategies.
Another aspect of the evolution in supply- chain strategy among “ big- box” importers was
to erect large “ import warehouses” in the hinterlands of the ports of entry selected for de-consolidation.
Using “ pull system” logic, only imports demanded by regional distribution
centers are trans- loaded and shipped immediately, the rest are held in the import
warehouse to wait and see where demands materialize. Once demand at a regional
distribution center develops sufficiently to make a request to the import warehouse, if that
warehouse is out of stock, the other import warehouses ( at the other three corners) are
checked for stock before a replenishment order is placed with the factory in Asia. Thus
there is a nation- wide pooling of inventory, even though it is physically distributed across
four import warehouses.
2.2. Feedback from Stakeholders – Capacity Study
Concerning the rail capacity planning study, the primary reaction of the railroads was that
the time horizon of the SCAG study is beyond their planning horizon. They have not
developed capacity plans within the Los Angeles Basin that far out in time ( 2025), so it is
difficult for them to comment on the plans. Even 2010 is a stretch for them. There was
general concern expressed about the difficulty in securing regulatory approval for
capacity expansion projects in California compared to elsewhere. Union Pacific
commented that they have no near- term plans for capacity expansion in the Los Angeles
Basin, their priorities are elsewhere for the next couple of years. BNSF commented that
they believed a grade separation of Colton Crossing would be required by 2010, slightly
sooner than in the consultant’s report. ( More discussion of this point is provided below.)
They also commented that, though the pooling of Union Pacific and BNSF trackage over
Cajon Pass would be beneficial and would push out the need for large capital
expenditures, it was not possible for them to negotiate an acceptable deal with Union
Pacific. So the BNSF moved forward with building on its own a third main track over
Cajon Pass.
Some of the importers were very concerned about potential shortfalls in rail intermodal
capacity out of the Los Angeles Basin in particular and out of West Coast ports in
general. They asked for a copy of the consultant’s long- term capacity plan, and were very
glad to receive it. They wondered why the railroads are not planning out to the horizon
studied by SCAG. They perceive the growing capacity shortage and declining rail service
quality as a serious problem for them.
During the spring of 2005, a number of meetings were held with stakeholders of the
main- line rail capacity study. To more fully appreciate the perspectives of the
stakeholders, the feedback received in these meetings is included in this report.
34
The preliminary findings of the rail main- line capacity study were presented to Metrolink
and BNSF in Los Angeles on April 25, 2005. ( BNSF serves as a strategic partner to
Metrolink in planning track capacity for joint passenger and freight operations between
Hobart and Colton Crossing.) Generally, improvements within the Los Angeles Basin
planned on BNSF Lines by BNSF on behalf of Metrolink and those planned by the
SCAG- sponsored study were in agreement. The only exceptions were at Colton Crossing.
Metrolink and BNSF are projecting that a grade separation of Colton Crossing will be
required by 2010 ( whereas the SCAG- sponsored study finds it unnecessary for the 2010
traffic levels but required for higher traffic levels, the 2025 traffic level in particular).
Metrolink and BNSF also were planning for a flying junction connection with the UP
Yuma Line in 2010 ( again not required in the rail main- line capacity study’s Status Quo
Alternative for 2010, but required in the Status Quo Alternative for 2025). As of the date
of that meeting, Metrolink and BNSF had not analyzed traffic levels beyond those
forecasted for year 2010.
Given this concern, the consultant retrieved the specific simulation results for Colton
Crossing. These results are summarized as follows:
Scenario/ Train Type Fraction of Average Delay ( minutes)
Trains Stopped ( including trains not delayed)
2000 Base Case
UP trains across crossing 31.4% 1.7
BNSF trains across crossing 36.6% 3.1
2010 Status Quo
UP trains across crossing 63.6% 6.5
BNSF trains across crossing 26.6% 2.0
2010 Alts. to Status Quo
UP trains across crossing 64.1% 11.2
BNSF trains across crossing 30.1% 2.6
Note that the simulated stoppages of BNSF trains decline slightly in 2010. This is
because the BNSF Line is planned to have three main tracks across Colton Crossing in
2010, but it had only two main tracks in the Year 2000 Base Case. Note also that delays
at Colton Crossing in 2010 are higher for UP trains and slightly higher for BNSF trains
under the Alternatives to the Status Quo than under the Status Quo. This is evidently
because of congestion at West Colton backing up along the UP Line to Colton Crossing.
( The junction with the UP Palmdale Line at West Colton is planned to remain as is in the
2010 scenarios but is planned become a full flying junction in the 2025 scenarios.)
The consultant’s conclusion is that an at- grade crossing at Colton is feasible for the 2010
traffic levels assumed in this study, provided the BNSF Line is equipped with three main
tracks. However, this configuration has little capacity to spare. With almost one third of
35
BNSF trains getting stopped and almost two thirds of UP trains getting stopped at the
crossing, the 2010 traffic levels are close to the maximums that can be accommodated
without grade separation. So the BNSF – Metrolink proposal to implement the separation
in 2010 is not many years early compared to the time when the consultant believes it
would be truly required.
A letter dated April 28, 2005 was received from Metrolink indicating that levels of
passenger service in 2010 and 2025 different than assumed in the rail main- line capacity
study were being evaluated. These levels of service are as follows. ( Figures include both
Amtrak and Metrolink services. New Metrolink figures are cited first, assumptions of this
study are second.)
Line Segment 2010 2025
BNSF Hobart – Fullerton 72 compared with 96 118 compared with 106
BNSF Atwood – Riverside 42 compared with 38 82 compared with 62
BNSF Riverside – Colton 24 compared with 24 40 compared with 36
Considering the time and budget limitations of the study, the consultant was unable to re-do
the operational analysis for these new passenger train frequencies.
A second presentation of the preliminary findings of the rail main- line capacity study was
made to BNSF management in Fort Worth, TX on May 5, 2005. This time, the discussion
was focused on track capacity improvements between San Bernardino and Barstow ( i. e.,
BNSF main lines outside the Metrolink service territory). BNSF’s plans for 2010 call for
three main tracks on their line between those points. The SCAG- sponsored study plans
for three main tracks Summit – Barstow, but it plans a fourth main track San Bernardino
– Summit ( Status Quo Alternative), and it plans three main tracks San Bernardino –
Summit plus integration of the UP Palmdale Line with the BNSF Line between Devore
Road and Silverwood and a fourth main track Silverwood – Summit ( Alternatives to the
Status Quo). As discussed in Section 7 of the Main- Line Rail report ( Leachman [ 2005b]),
while the three- main- track configuration proposed by BNSF is indeed feasible, average
freight train running times are predicted to be about 15 minutes longer than in the Year
2000 Base Case. The increased levels of improvements planned in this study are believed
to be necessary to achieve Year 2000 transit times for the Year 2010 forecasts. As of the
date of that meeting, BNSF had not analyzed 2025 traffic levels.
BNSF management also remarked that productivity improvements they are striving to
achieve may temper train movement growth. They indicated that in 2004, BNSF
intermodal unit volume ( trailers and containers) to and from Southern California
increased by about 14%, yet the number of intermodal trains operated increased by less
than 4%.
A presentation of the preliminary findings of the Main- Line Rail study was made to UP
management in Omaha, NE on May 6, 2005. In general, UP management concurred with
the planned improvements. UP indicated that a similar plan had been jointly presented by
36
UP and BNSF two years ago to MTA, with copy to SCAG. ( The consultant has not seen
that plan.) UP also indicated that their plans for accommodating 2010 traffic levels call
for increasing the percentage of UP train movements routed via the Alhambra Line
between Colton and Pomona, and decreasing the percentage routed via the San Gabriel
Line. This is consistent with the Alternatives to the Status Quo formulated in the SCAG-sponsored
study. Complete double- tracking of the San Gabriel Line between West
Riverside and Pomona was seen by UP management as unrealistic, whereas double-tracking
the Alhambra Line between West Colton and Pomona was more practical and
part of their plan, again consistent with the SCAG- sponsored study. UP management
indicated that, in general, making capacity improvements in Southern California is much
more difficult than elsewhere on their system, given the environmental reports and other
requirements. As a result, no near- term track capacity improvements were planned by UP
for the Los Angeles Basin, and their near- term capacity improvement projects were being
undertaken elsewhere.
The preliminary findings of the Main- Line Rail study also were presented to SCAG’s
Goods Movement Task Force on April 20, 2005. In attendance were representatives of
the Alameda Corridor – East Joint Powers Authority. ACE representatives remarked that
they had approached the Union Pacific with a proposal to buy the former Southern
Pacific main line west of Pomona ( via Alhambra) and to buy the historical Union Pacific
main line between Riverside and Pomona ( via Pedley) from Union Pacific in order to
make this route an exclusive passenger train route, leaving the former Southern Pacific
main line east of Pomona and the historical Union Pacific main line west of Pomona as
an exclusive freight route. ACE representatives related that Union Pacific refused this
offer. The conclusion of ACE representatives was that efforts to re- route freight and
passenger trains as proposed in this study were hopeless, because freight railroad
agreement to do so is lacking.
What ACE proposed is roughly like Alternative 1( b) of the study, except much more
extreme – mandating completely disjoint freight and passenger ownership and operation
of lines. In the consultant’s view, the ACE proposal represents too much of a hardship on
Union Pacific, as certain freight trains need to run via the lines proposed by ACE for
exclusive passenger use, e. g., intermodal trains to/ from the City of Industry and the Los
Angeles Transportation Center terminals, and carload freights to/ from the Coast Line
route to Northern California. The consultant believes Alternative 1( b) as presented is still
quite viable and in the best interests of all concerned – public agencies, passenger service
operators and freight railroads. The reception received by the consultant from Union
Pacific is indicative of this. 10
10 Alternative 1( b) involves shifting all Metrolink operation between Pomona and Los Angeles off the line
via East Los Angeles and onto the line via Alhambra. It also involves shifting Union Pacific through freight
train operation between Colton and Ponoma off the line via Mira Loma and onto the line via West Colton,
except for unit auto trains to/ from the Mira Loma auto terminal. This separation of most freight and
passenger operations reduces the capital investment requirements for high levels of both passenger and
freight traffic as well as increases safety.
37
3. Port and Modal Shares of Imports
3.1. Port Shares of Containerized Trade Volumes
Figure 1 displays the 2005 shares of waterborne containerized imports from Asia to the
USA as a percentage of the total passing through US ports. ( Not included in the 100%
total are imports from Asia to the USA that passed through Canadian or Mexican ports,
and then came into the USA using landside border crossings.) These are figures on a TEU
basis. As may be seen, 24.2% came through East Coast ports, 1.3% through Gulf Coast
ports, 55.6% through the San Pedro Bay ports, and 18.9% through other West Coast
ports.
Figure 1. US Port Shares of 2005 US
Containerized Imports from Asia ( TEU basis)
55.6%
18.9%
1.3%
24.2%
San Pedro Bay
Other West Coast
Gulf Coast
East Coast
Source: PIERS, courtesy of MARAD
In 2003, the East Coast share was 23.4%; in 2002 it was 21.0%; and in 2001 it was
18.6%. Thus the rate of growth in East Coast share slowed during 2003 – 2005.
Focusing on West Coast ports, Figure 2 displays the trend in shares of total inbound
loaded containers at major West Coast ports. As may be seen, the San Pedro Bay ports
dominate other ports by a wide margin. The SPB ports’ share has been eroding gradually.
It took a sharp drop in 2005 ( mostly diverted to the Pacific Northwest ports), then
38
recovered most but not all of this in 2006. It is believed the sharp drop in 2005 reflected
decisions by certain steamship lines to shift certain vessel strings from the San Pedro Bay
ports to the Pacific Northwest ports for the 2005 season. These shifts were a response to
the “ melt- down” at the San Pedro Bay ports during the late summer of 2004. With the
introduction of additional terminal shifts funded by the PierPass program, the 2005
season was handled smoothly at the San Pedro Bay ports. This encouraged the lines to
shift the strings back to the San Pedro Bay ports for the 2006 season. In addition, more
capacity in new strings serving San Pedro Bay was added for the 2006 season than was
added serving the Pacific Northwest ports. Another drop occurred in 2008, mostly due to
the opening of Prince Rupert and the shifting of a vessel string there from San Pedro Bay.
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
50.00%
55.00%
60.00%
65.00%
70.00%
75.00%
2001 2002 2003 2004 2005 2006 2007 2008
Figure 2. Shares of Inbound Loaded Containers at
West Coast Ports ( TEU basis)
San Pedro Bay
Sea/ Tac/ Port
Vanc/ P Rupert
Oakland
Source: Port web sites.
Figure 3 displays trends in shares of total container movements, both inbound and
outbound, both loaded and empty, at the West Coast ports. As may be seen, the share for
San Pedro Bay was stable for the years 2001 – 2005, but then increased sharply in 2006
and has been fairly stable after that. Comparing to Figure 3, only a portion of this trend is
explained by imports. The sharper rise in shares of total container movements stems more
from increased outbound movement of containers. Indeed, outbound empty containers
39
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
50.00%
55.00%
60.00%
65.00%
70.00%
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
% of Total West Coast TEUs
Year
Figure 3. Container Traffic Shares
at West Coast Ports ( international loads and empties,
inbound and outbound, TEU basis)
SP Bay Ports
SF Bay Ports
Sea/ Tac/ Port
Vanc/ P Rupert
Source: Port web sites.
were the fastest growing segment of container movements at the San Pedro Bay ports
during 2006, albeit exports grew as well. 11
It is believed that the most important reason for this trend stems from changes in business
terms between railroads and the steamship lines. In 2006, the railroads initiated financial
penalties on the lines that apply if their eastbound and westbound container flows to
individual West Coast ports are out of balance. Before these penalties were instituted, the
lines commonly brought most imports in through San Pedro Bay, but returned most
westbound empties and export loads from interior points through Pacific Northwest ports.
The lines would operate their vessels in strings that would call at San Pedro Bay first,
then move up the Coast and call at Puget Sound last before returning to Asia. To save
time on westbound transit and to reduce vessel loads, the westbound containers from
interior points would be routed via the Pacific Northwest ports. This forced the railroads
to absorb expenses for re- positioning empty well cars from the Pacific Northwest ports
and Oakland down to the San Pedro Bay ports. The trains re- positioning empty
intermodal equipment are known as “ bare- table” trains. With the changed financial terms,
the flow of bare- table trains down the West Coast is now much less, and the outbound
flows of containers handled through the San Pedro Bay ports are higher.
11 As of this writing, 2006 was the last year of growth for the San Pedro Bay ports in aggregate.
40
3.2. Landside Channel Shares of Waterborne Containerized
Imports
Figure 4 displays trends in the fraction of containers imported through West Coast ports
that are placed on double- stack trains for inland movement. ( Not included in the rail
movement shares are cargoes that were trans- loaded to domestic containers.) Statistics on
rail movement of individual types of marine containers ( 20s, 40s and 45s) came from the
Intermodal Association of North America ( IANA). These figures were compared to
statistics on total inbound loaded containers ( TEUs) furnished by the ports. Also
contributing to this analysis are statistics on the mix of marine box types ( 20s, 40s, 45s)
handled through the West Coast ports, obtained from the Pacific Maritime Association.
IANA does not break out figures for Oakland from those for San Pedro Bay; they are
aggregated in the “ Cal” category. The “ PNW” category includes Portland, Tacoma and
Seattle.
As may be seen, the fraction of total inbound loaded international marine containers via
West Coast ports that got on a train declined from almost 50% in 2000- 2001 to a little
over 41% in 2008. Statistics by region are revealing. In 2005, there was a sharp increase
in the Pacific Northwest, followed by a sharp drop in 2006. Concurrently, 2005 saw a
slight drop in California, followed by a larger increase in 2006. In 2006, the PNW
fraction stood at 70%, while California was at 42%. After 2006 there was a steep drop in
the PNW and a smaller drop in California, bringing the PNW fraction down to about
54%, the California average below 40%, and the overall West Coast fraction down to a
little over 41% in 2008.
The gradual decline of the inland point intermodal ( IPI) share of inbound containers at
the West Coast ports is believed to be primarily due to the increasing market share of
large, nation- wide retailers who practice consolidation – de- consolidation inventory
management strategies, and increasing adoption of such strategies by wholesalers of
moderate- valued and expensive goods who import their products from Asia and sell to
retailers in the USA.
The more turbulent trend in the IPI share of imports at the PNW ports is explained as
follows. First, in 2005, much discretionary inland- point- intermodal ( IPI) traffic was
shifted from routing via San Pedro Bay to routing via the Pacific Northwest as a response
to the “ melt- down” during the 2004 peak season at the San Pedro Bay ports. As noted
above, this volume shifted back in 2006. Second, there also were changes made to
consolidation – de- consolidation strategies. Under such strategies, goods are stripped out
of marine containers, sorted and re- loaded in domestic vehicles in the hinterlands of the
ports of entry before movement inland. If not immediately required, the goods may be
seasonally stored at warehouses in the hinterlands of the ports. Many goods that
previously were imported solely or mostly through San Pedro Bay under such strategies
began being imported using a set of 4- 5 ports.
41
35.0%
40.0%
45.0%
50.0%
55.0%
60.0%
65.0%
70.0%
75.0%
80.0%
85.0%
2000 2001 2002 2003 2004 2005 2006 2007 2008
Figure 4. Percent Intermodal Movement of
Marine Containers Imported Through US West
Coast Ports ( TEU Basis)
PNW
Total West Coast
Cal
Source: IANA, Port web sites, PMA.
For example, a popular import distribution strategy among “ big- box” retailers that has
evolved is the so- called “ Four Corners” strategy. An importer might use, say, San Pedro
Bay, the Pacific Northwest, Savannah and New York – New Jersey as ports of entry for
its imports. Warehouses and trans- load facilities located in the hinterlands of each of
these four ports primarily serve one fourth of the continental United States, but can serve
as back- up sources for serving any of the retailer’s regional- distribution- center demand
points. This strategy reduces transportation costs compared to an all- trans- loaded-through-
San- Pedro- Bay strategy. Moreover, because all four ports can supply any
location, safety stocks are effectively pooled nationwide, economizing on total inventory
almost as much as if a single warehousing and trans- loading port were used. ( On the
downside, there is an increase in total inventory associated with the extra pipeline and
safety stock inventory required by all- water transit to the East Coast. The Four Corners
strategy is this better suited to moderate- valued goods than to expensive goods.) The
strategy also serves to provide important diversification of the risk that trouble might
develop in one of the port channels, such as what happened at San Pedro Bay in 2004. If
there is surge capacity in the other channels, then the retailer’s supply- chain strategy is
made more robust under Four Corners than under a policy of bringing in all imports
through one port.
The trend from single- port distribution strategy to multi- port distribution strategy also is
influenced by the increasing market shares of the “ big- box” retailers and their changing
terms of business with their suppliers. Traditionally, these retailers bought goods from
42
manufacturers here in the USA, even when the goods were manufactured in Asia. The
manufacturer made the goods in Asia, then brought them into the USA to its warehouse
in the hinterland of the San Pedro Bay ports. When the goods were sold to retailers, the
retailers paid for the domestic freight from this warehouse to their USA locations. The
“ big- box” retailers increasingly negotiate with their suppliers to buy the goods in Asia at
a reduced price, then handle the distribution themselves. Because of their large scale, the
retailers may be able to achieve reduced total costs. In terms of trade impacts, this shifts
goods from all- trans- loaded- through- San- Pedro- Bay to trans- loaded- through- four- ports.
Very expensive goods with rapidly declining prices ( e. g., electronics) and goods with
very uncertain demands ( e. g., style goods or new toys) are still imported by
manufacturers/ wholesalers using an all- through- San- Pedro- Bay strategy and then re- sold
here in the USA ( because the retailers do not want to risk the inventory investment), but
everything else sold by the large, nation- wide retailers has moved to the multi- port
strategies.
The net impact of increased use of multi- port trans- load import strategies is a sharp
increase in the percentage trans- loading of imports and a sharp decrease in percentage IPI
inland movement of marine containers at the Pacific Northwest ports, as depicted in
Figure 4.
The customs data sources do not provide reliable data on the distribution of destinations
for waterborne containerized imports. However, estimates may be developed from the
data described above for the relative shares of three broad categories of imports for a
single port or a group of West Coast ports: ( 1) Imports consumed in the general region
that is “ local” to the ports; ( 2) Imports for which the marine container containing them is
placed on a double- stack train for movement east of the Rockies; and ( 3) Imports whose
marine container was trucked out of the local region, plus imports which were unloaded
from marine containers in the local region, then subsequently re- shipped out of the region
in domestic vehicles, either rail or truck (“ trans- loaded” imports). For the last category, it
is believed that trans- loaded imports comprise a much larger volume than that for long-distance
trucking of marine containers.
Such estimates were developed as follows for the West Coast ports as a group as well as
for just the San Pedro Bay ports. To do this, it was assumed that geographical distribution
of the consumption of Asian imports is proportional to the geographical distribution of
purchasing power ( population multiplied by income per capita). This is believed to be a
reasonable assumption because the lion’s share of imports to the USA from Asia are
retail goods or goods that are very close to ready for retail sale. Imports of raw materials
or inputs to manufacturing are much less.
Next, the region “ local” to the West Coast ports was defined to include the following
states: CA, OR, WA, ID, NV, UT, AZ, NM, CO, WY and MT. (“ Local” is meant in the
sense that containers of Asian imports with destinations in these states are most cheaply
routed through West Coast ports. According the US Census web site, collectively these
states account for 22.4% on the total purchasing power in the continental United States.
According to Figure 1, about 74.5% of total imports Asia – USA came through West
43
Coast ports. Thus “ local” traffic handled through the West Coast ports was
( 0.224)/( 0.745) = 30%. According to Figure 4, direct inland rail intermodal movement of
marine boxes from West Coast ports accounted for 46%. This leaves 24% in the third
category ( trans- loading plus any across- the- Rockies long- distance trucking of marine
boxes.
Now consider the San Pedro Bay ports. Suppose we define the region “ local” to the San
Pedro Bay ports to include Southern California ( 62% of California purchasing power),
Southern Nevada ( 67% of Nevada purchasing power), Arizona, New Mexico, Southern
Utah ( 33% of Utah purchasing power), and Southern Colorado ( 50% of Colorado
purchasing power). Again, “ local’ is used in the sense of containers of Asian imports with
destinations in this region are most cheaply routed through the San Pedro Bay ports.
According the US Census web site, this region accounts for 11.8% of the total purchasing
power in the continental United States. According to Figure 1, about 55.6% of total
imports Asia – USA came through the San Pedro Bay ports. Thus approximately
( 0.118)/( 0.556) = 21% of the 2006 imports through the San Pedro Bay ports were “ local”.
Now let us assume 35% of the marine boxes imported through the Port of Oakland in
2006 got on a double- stack train. This is a judgment; the chosen figure is much less than
the figure the Pacific Northwest ports ( 70%), and somewhat less than the figure for
California as a whole ( 35%) and hence less than the ( unknown) figure for Southern
California. The rationale for this judgment is as follows: Oakland is more of a “ local”
port for imports than either the Pacific Northwest ports or the San Pedro ports,
originating many less marine stack trains. On the other hand, trans- loading activity is
much less than at either San Pedro Bay or Puget Sound. As noted in Figure 4, direct
inland rail intermodal movement of marine boxes from all California ports accounted for
42%. Considering the relative import volumes at Oakland and at San Pedro Bay depicted
in Figure 2 and the assumption of 40% rail intermodal movement from Oakland, the 42%
figure for California translates into a figure of about 43% of the 2006 loaded marine
boxes entering San Pedro Bay got on a train. This leaves 36% in the third category ( trans-loading
plus any across- the- Rockies long- distance trucking of marine boxes. In sum,
inland- point- intermodal movement of marine containers and trans- loading of imports for
re- shipment in domestic vehicles are roughly equal in Southern California ( IPI is a bit
larger), and each of those categories is about twice as large as “ local” imports.
From the 2008 data point in Figure 4, it is estimated that in 2008 the IPI share of San
Pedro Bay imports from Asia had fallen to 41%. Elasticity calculations discussed in
section 8 estimate that total market share of the San Pedro Bay ports fell from 2006, such
that local- region imports rose from 21% to 23% of the total Asian imports via San pedro
Bay. This leaves the 2008 share accounted for by trans- loaded imports and marine boxes
trucked out- of- region holding steady at 36%.
44
4. Distribution of Imports by Commodity and Value
US Customs defines 99 commodity types for classifying waterborne containerized
imports to the United States. Two commercial subscription services are available for
analyzing customs data. The PIERS database provides TEU volumes by commodity type
and port. The World Trade Atlas ( WTA) database provides breakdowns by total declared
value by commodity type for groups of US ports ( e. g., West Coast, South Atlantic and
Gulf Coast, North Atlantic). PIERS totals by commodity type of imports from selected
Asian countries ( see Appendix B for a specific list) for calendar 2005 imported through
West Coast ports were furnished to the author by the Port of Long Beach. The Port of
Long Beach also furnished the author with nationwide WTA totals as well as West Coast
totals for 2005. In addition, MARAD furnished the author with PIERS summaries of total
TEUs of 2005 Asian imports by US port, but indicated it was not able to provide the
author with a break- out by commodity type.
The author joined these data to develop statistics on the average declared value per TEU
and the average declared value per cubic foot of imports for each commodity type. To
compute values per cubic foot, the mix of loaded 20s, 40s and 45s imported through West
Coast ports was secured from the Pacific Maritime Association’s database. An
assumption was added concerning the mix of standard- size ( ISO) versus high- cube 40s.
PMA data for 2005 indicates that total TEUs of inbound loaded containers at West Coast
ports were 12.74% 20s, 80.51% 40s, and 6.75% 45s. It was assumed that 40% of the 40s
were ISO boxes and 60% were high- cube ( 9 ½ feet high). Usable cubic feet of capacity
were assumed for the various box types as follows: 1,169 for 20s, 2,395 for ISO 40s,
2,684 for high- cube 40s, and 3,026 for 45s. This made for a weighted- average cubic
capacity per TEU of 1,274.73 cubic feet.
The first step in the analysis was to take West Coast PIERS and WTA data and join them
to obtain declared values per TEU and per cubic foot for each commodity type. 12 Table 1
summarizes the top twelve commodity types ( by TEU volume) imported from Asia
through West Coast ports for 2005. Shown are average declared values for each
commodity type as well as for all imports. The top fifteen commodities account for more
than 80% of the volume of imports. Furniture and bedding is the largest- volume
commodity by a wide margin. The average declared value on the West Coast is $ 22.66
per cubic foot, which moved up only slightly from 2003. However, the value varies
widely by commodity: For Furniture and Bedding, it is less than $ 8.00, while for
Electronics and Electrical Equipment, it is almost $ 40.
12 A subtle difficulty in doing this join is that the commodity types in PIERS and WTA extracts match in all
but one category. PIERS includes a “ Miscellaneous Manufactured Articles” commodity type, but WTA
does not; WTA includes “ Special Other” but PIERS does not. A weighted- average declared value per TEU
developed from other manufactured commodity types was applied to the “ Miscellaneous Manufactured
Articles” commodity type in order to join the data.
45
Table 1: Top Commodities Imported from Asia Through US West Coast Ports in
2003 and 2005
Commodity Type Estimated
2005
Volume
( TEUs)
Estimated
2005 Avg.
Declared
Value
($ per cu. ft.)
Estimated
2003
Volume
( TEUs)
Estimated
2003 Avg.
Declared
Value
($ per cu. ft.)
Furniture and Bedding 1,489,050 7.87
1,014,304 8.27
Electronics and Electrical
Equipment 876,972 39.55
749,301 37.46
Machinery 838,461 51.40 660,809 50.23
Toys, Games & Sports
Equipment 700,228 17.02
662,977 16.56
Motorcycles and Auto
Parts 591,753 24.65
480,347 20.19
Plastic Goods 446,087 14.63 352,676 13.18
Apparel not knitted 407,402 26.30 329,477 27.93
Steel Goods 362,630 15.43 265,186 14.13
Footwear 357,244 24.91 318,032 24.37
Rubber goods 303,114 14.37 197,900 14.63
Miscellaneous
Manufactured Articles 252,590 22.94
273,785 23.42
Leather Goods 228,805 16.14 199,295 18.05
Wooden Goods 209,892 8.24 104,707 10.91
Apparel knitted 171,525 51.71 149,591 53.81
Ceramic Goods 156,602 6.34 108,646 8.38
Subtotal, Top 15 Types 7,392,356 5,867,036
All Commodities 9,134,672 22.66 7,222,099 22.32
Source: PIERS, WTA and PMA data.
Next, East Coast ( including Gulf Coast) WTA data was analyzed. Because total TEUs of
Asian goods imported through East and Gulf Coast ports are known from MARAD-supplied
data, and total declared value of such goods is known from the WTA data, the
average declared value for Asian goods imported through East Coast and Gulf Coast
ports in 2005 could be deduced. This figure is $ 18.57, or about $ 4 less per cubic foot than
the average figure for imports brought in through West Coast ports. This is in line with
the expectation that high- value commodities are imported through the West Coast,
typically only San Pedro Bay, in order to most tightly control inventory costs.
As a trial, West Coast average declared values per TEU were divided into the WTA-reported
total declared values for each commodity type imported through East Coast and
46
Gulf Coast ports. The resulting TEU figures were summed across all commodity types to
obtain an estimate of total TEUs imported through East Coast and Gulf Coast ports, and
this figure was compared to the actual PIERS total for the East and Gulf Coasts. This
resulted in too small a sum. The author then judged down the declared values of certain
high- value commodities in order to get TEU totals to match. Next, the revised figures for
West Coast and for East Coast/ Gulf Coast were combined to develop nation- wide figures.
Table 2 displays the author’s estimates of TEU volumes and average declared values for
waterborne containerized imports from Asia to the United States in 2005. The top fifteen
commodities account for about 78% of the total volume of imports, and the overall
average declared value is $ 21.66 per cubic foot, only slightly higher than that for 2003.
Table 2: Top Commodities and Declared Values for Waterborne Containerized
Imports from Asia to the United States in 2005
Commodity Type Estimated
2005
Volume
( TEUs)
Estimated
2005 Avg.
Declared
Value
($ per cu. ft.)
Estimated
2003
Volume
( TEUs)
Estimated
2003 Avg.
Declared
Value
($ per cu. ft.)
Furniture and Bedding 2,069,444 7.87 1,484,160 7.80
Electronics and Electrical
Equipment 1,000,598 39.24
847,223 36.60
Machinery 969,789 51.08 48.97
Toys, Games & Sports
Equipment 902,120 16.57
855,301 15.54
Motorcycles and Auto
Parts 733,930 24.65
524,777 20.00
Plastic Goods 599,505 14.63 492,552 12.28
Apparel not knitted 585,670 25.60 451,775 25.78
Steel Goods 471,354 15.43 344,088 13.68
Footwear 425,897 24.91 370,784 24.37
Rubber goods 399,432 14.37 279,014 13.86
Miscellaneous
Manufactured Articles 290,276 22.94
273,785 23.42
Leather Goods 280,131 16.14 237,649 17.72
Wooden Goods 252,590 8.24 146,437 10.08
Apparel knitted 240,721 50.93 195,839 50.55
Ceramic Goods 214,542 6.34 145,123 8.38
Subtotal, Top 15 Types 9,436,000 7,401,863
All Commodities 12,104,795 21.66 9,370,896 21.47
Source: PIERS, WTA and PMA data.
A distribution of import volumes by declared value was developed by sorting the
commodity types in increasing order of value. Cumulative distribution curves for 2003
47
and 2005 Asia – USA waterborne containerized imports are displayed in Figure 5. In the
Figure, one can observe the declared values at which certain percentiles of total import
volume are reached. Note that about 25% of imports have declared values of $ 13 or less,
25% have declared values of $ 26 or more, and 50% have declared values in between. We
designate these three declared- value ranges as inexpensive, expensive and moderate,
respectively. In the 2005 elasticity study by the author, it was found that inexpensive
imports are most efficiently handled by direct inland shipment in marine containers via
the closest port. Moderate- value imports, if distributed nation- wide, are most efficiently
handled by a consolidation – de- consolidation strategy ( such as the “ Four Corners”
strategy) using multiple ports with warehousing and trans- loading facilities in the
hinterlands of the selected ports of entry, while expensive goods, if distributed nation-wide,
are most efficiently handled by a consolidation – de- consolidation strategy using a
single port of entry ( most commonly San Pedro Bay).
Figure 1. 2003 vs. 2005 Distributions of
Containerized Asia - US Imports
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56
Declared Value ($ per Cubic Foot)
Percent of Total Imports( TEU Basis)
2005
2003
Source: PIERS, WTA and PMA data.
As may be seen in Figure 5, there were no great changes in the value distribution for
Asian imports from 2003 to 2005. Where the blue ( 2005) curve is above the red ( 2003)
curve, imports are getting cheaper; where the red curve is above the blue curve, imports
are getting more expensive. There seems to be some spreading of the distribution, i. e.,
inexpensive imports are getting a bit cheaper while expensive imports are getting a bit
Figure 5. 2003 vs. 2005 Cumulative Distributions of
Containerized Asia – USA Imports
48
pricier, but the inexpensive – moderate – expensive break points in the distribution are
basically unchanged. The implication is that the overall, nation- wide allocation of
imports to landside channels was basically unchanged from 2003 to 2005. ( That
implication is confirmed in Section 3 above.)
Given rising energy and transportation costs, and given upward pressure on Asian
currencies, the author had anticipated some upward shift in the value distribution curve
from 2003 to 2005. But basically this did not happen. Anecdotal evidence received from
importers suggests that for some c