Transcript Task 2.4: The System Dynamics Approach

Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
REVENUE Seminar 1
Brusels, June 9th 2004
The System Dynamics Approach:
Results of Scenarios for Europe
Claus Doll
Institut für Wirtschaftspolitik und
Wirtschaftsforschung (IWW)
Universität Karlsruhe (TH)
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Objectives and method of task 2.4
• Goals:
– Investigation of the dynamics behind long-term decisions in
transport network planning.
– Identification of the key drivers behind long-term optimality
decisions.
• Approach:
– Development of a small transport sector specific system
dynamics model (MARS), containing several evaluation
tools
– Application of the ASTRA model to answer general
questions concerning the link betwen transport and the rest
of economy.
• Discussion: Applicability of the framework within the case
studies.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Contents
• System Dynamics and CGE-Models
• Revenue Distribution within the Transport Sector: Structure
and Results of the MARS model
• Revenue allocation within or outside the Transport Sector:
Results of the ASTRA-Model for Europe
• Conclusions
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Task 2.4: Dual model application
MARS (Multimodal Assessment of
Revenue allocation Strategies):
ASTRA (ASsessment of TRAnsport Policies.
Partial analysis of revenue allocation
variants within the transport sector by
assuming a self-financing system of 4
transport modes.
System-Dynamics model platform developed during
several EC-funded research projects. Covers 14
countries, passenger and freight transport of all
modes, trade and production by 25 economic sectors,
government activities, environment and traffic safety.
Rough model calibration to Europe and
application to 25 combinations out of
pricing and fund allocation policies.
The model is used to investigate long-term effects of
earmarking pricing revenues in the EU Member
States.
Brief presentation of model
mechanisms and some results.
Short presentation of modular model structure and
feedback mechanisms. Detailed discussion of
scenario results for the EU-15 countries.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
MARS Model: Some feedback mechanisms
Time costs
Travel speed
Time
Welfare
measure
Modal share
Budget spending
rules
Infrastructure
capacity
Infrastructure
quality
Traffic volume
Average
infrastructure
prices
Congestion
pricing revenues
Environmental
pricing revenues
Available Budget
Fund composition
and allocation rules
Relevant feedback loops:
1. Traffic volume – travel speeds – congestion revenues – available budget – infrastructure capacity –
travel speeds – time costs – traffic volume: Negative, results in equilibrium or oscillations.
2. Time – (traffic volume) – infrastructure quality – travel speeds – traffic volume – infrastructure
quality: Slightly negative dominated by time-dependent deterioration of infrastructure.
3. Traffic volume – average infrastructure prices – traffic volume: Positive loop caused by economies
of scale in AC-Models; might lead to excessive demand or to crowding out of entire demand.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Features of the MARS model
• 4 modes and 5 transport funds (urban, inter-urban, road, P.T.
and intermodal/inter-regional).
• Pricing options: Infrastructure (AC and SMC), congestion,
accidents (SMC) and environment (SMC). plus mark-ups.
• Assessment of max. 5 revenue spending scenarios for each
of max. 5 pricing policy scenarios.
• Welfare measure = time costs valued by the „rule of half“
• Self-financing of transport sector with link to capital market.
• Stochastic deterioration of networks, by time and traffic load.
• User time costs depending of traffic load and network quality.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Definition and results of the base scenario
• Model calibration for Europe where possible
• Time horizon: 30 years.
• Results: Mode-specific revenue use recommended in 3 of 5
pricing scenarios
– Costs of fund administration and fund allocation rules to be
considered!
P1: Urban Congestion
P2: Motorway Toll
P3: Swiss case
P4: Pure SMCP
P5: SMCP+mark-ups
mill. €
mill. €
mill. €
mill. €
mill. €
R1:
No funds
-5.709
-7.984
-7.968
-3.275
6.311
R2:
Network
-10.563
-7.080
-4.151
-8.844
5.536
R3:
Area
-10.305
-4.910
-5.750
-10.070
5.359
R4:
Intermodal
-10.563
-2.447
-673
-9.922
5.739
R5:
P.T. sup.
-10.305
-2.647
-743
-10.070
5.359
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Results for pricing scenario P1:
Urban congestion charging
• Nearly / exactly identical slope of allocation schemes R2 to
R5: Litte excessive funds to distribute.
Results for policy scenario 1
Urban congestion pricing
R1: No Funds
R2: Network
R3: Area
R4: Intermodal
R5: P.T. Support
5'000
Present value of user time costs (mio.
Euro)
0
0
10
20
30
40
50
-5'000
-10'000
-15'000
-20'000
-25'000
Time horizon (years)
60
70
80
90
100
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Details for pricing scenario P1
R2: Network funds
R3: Area funds
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of congested traffic
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of congested traffic
14,0%
14,0%
12,0%
12,0%
10,0%
10,0%
8,0%
8,0%
6,0%
6,0%
4,0%
4,0%
2,0%
2,0%
0,0%
0,0%
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
Time (years)
50
60
70
80
90
100
Time (years)
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Network quality standard index
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Network quality standard index
80%
100%
90%
70%
80%
60%
70%
50%
60%
40%
50%
40%
30%
30%
20%
20%
10%
10%
0%
0%
0
10
20
30
40
50
Time (years)
60
70
80
90
100
0
10
20
30
40
50
Time (years)
60
70
80
90
100
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Results for pricing scenario P2: Average
infrastructrue cost charging on motorways
• Much more dynamic than P1 due to more stable excess funds
available for redistribution.
Results for policy scenario 2
Motorway tolling
10.000
R1: No Funds
R2: Network
R3: Area
R4: Intermodal
R5: P.T. Support
Present value of user time costs (mio. Euro)
5.000
0
0
10
20
30
40
50
-5.000
-10.000
-15.000
-20.000
-25.000
Time horizon (years)
60
70
80
90
100
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Some details fore pricing scenario P2
R3: Area funds
R4: Intermodal fund
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of congested traffic
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of congested traffic
12,0%
14,0%
12,0%
10,0%
10,0%
8,0%
8,0%
6,0%
6,0%
4,0%
4,0%
2,0%
2,0%
0,0%
0,0%
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
Time (years)
50
60
70
80
90
Time (years)
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of interest payments at annual
expenditures (excluding credit pay-back)
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of interest payments at annual
expenditures (excluding credit pay-back)
100%
120%
90%
100%
80%
70%
80%
60%
50%
60%
40%
40%
30%
20%
20%
10%
0%
0%
0
10
20
30
40
50
Time (years)
60
70
80
90
100
0
10
20
30
40
50
Time (years)
60
70
80
90
100
100
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Results for pricing scenario P5: Full SRMC +
mark-ups
• Due to high and stable revenues in each mode no transfer
required and positive welfare until year +75
Results for policy scenario 5
SRMC pricing with mark-ups
10.000
R1: No Funds
R2: Network
R3: Area
R4: Intermodal
R5: P.T. Support
Present value of user time costs (mio. Euro)
8.000
6.000
4.000
2.000
0
0
10
20
30
40
50
-2.000
-4.000
Time horizon (years)
60
70
80
90
100
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Some details for pricing scenario P5
R1: No fund allocation
R4: Intermodal fund
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of congested traffic
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Share of congested traffic
12,0%
12,0%
10,0%
10,0%
8,0%
8,0%
6,0%
6,0%
4,0%
4,0%
2,0%
2,0%
0,0%
0,0%
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
Time (years)
50
60
70
80
90
100
Time (years)
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Network quality standard index
90%
Mode 1 Road Inter-urban
Mode 2 Road Urban
Mode 3 P.T. Inter-Urban
Mode 4 P.T. Urban
Network quality standard index
90%
80%
80%
70%
70%
60%
60%
50%
50%
40%
40%
30%
30%
20%
20%
10%
10%
0%
0%
0
10
20
30
40
50
Time (years)
60
70
80
90
100
0
10
20
30
40
50
Time (years)
60
70
80
90
100
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Sensitivity analysis for selected key
variables
• Negative performance of all pricing scenarios in the long run
due to the ambitious definition of the reference case.
• Time is less critical for the optimality ranking of the revenue
allocation schemes than expected.
• In general, the model is rather stable against changes of
parameters. one of the most sensitive ones is the influence of
road quality on speed.
• The sensitivity results are to be considered in front of the
specific calibration fo the model and might be totally different
for other constellations.
Population Change
Institut für Wirtschaftspolitik
und Wirtschaftsforschung
Universität Karlsruhe (TH)
Potential Labour Force
Exports, Imports
Sectoral Goods Flows
REM
FOT
Generalized Cost OD
Transport Expenditure,
Performance, Time
Transport Demand OD
TRA Transport Cost, Time OD
VAT Revenue
Fuel Tax Revenue
GDP, Employment, ....
Fuel Price
Fleet Structure
ENV
Emissions, Noise,
Accidents
WEM
Car Fleet
VFT
GDP, Productivity
VKT
POP: Population
MAC: Macroeconomics
REM: Regional economics
FOT: Foreign trade
TRA: Transport
ENV: Environment
VFT: Vehicle fleet
WEM: Welfare
MAC
Fuel
Price
Modules:
Population Structure
GDP, (Un-)Employment, Sectoral Output
Consumption, Investment in Vehicles, VAT
Disposable Income
ASTRA
modules
and main
interfaces
POP
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Impact
chains
and
their
time
structure
Abbreviations:
GDP: Gross
domestic product
GVA: Gross
value added
TPF: Total factor
productivity
FD: Freight demand
PO: Production
output
IO: Input-output
Pricing
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
ASTRA-T Scenario Definition
Charging regime
Revenue allocation policy
Refund by
direct tax
reduction
Refund by
labour cost
reduction
Reinvestment in
in mode of charge
collection
Reinvestment
by crosssubsidisation
Congestion charging
in urban areas
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban road
user tolls
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
SMCP in all moces
(TIPMAC scenarios)
• Fixed allocation of reinvestments to road types (single carriageway roads, motorways) or to
rail facilities (network, terminals, rolling stock).
• Refund via tax increases: No price increases assumed as indicated by IASON model
applications of CGEurope and E3ME).
• Refund via social contributions: 50% employers (partly increase of GVA) and 50% for
consumers (partial use for increased consumption).
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Development of total revenues
450
400
Total revenues
350
bill. Euro
300
250
200
150
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
100
50
0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
• Outstanding level of TIPMAC SMC-revenues against partial toll regimes.
• Lowest level by urban congestion revenues.
• No great impact of transport-specific feedback loops on level of revenues.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Overview of results for 2020
Percent change from BAU to policy
Scenario
GDP
Employment
Con- Invest- Exports
sump- ments
tion
TFP
Tons
Tkm
Trips
Pkm
CO2
Congestion-DT
1.52
-0.03
2.45
7.53
0.16
0.84
0.14
16.12
-0.04
0.24
1.52
Congestion-LC
0.87
-0.12
1.05
4.30
0.11
0.58
0.05
16.10
-0.05
0.21
0.87
Congestion-Road
1.42
0.48
1.08
3.96
0.19
0.76
0.69
16.45
-0.01
0.35
1.42
Congestion-Cross
1.33
0.42
0.95
3.91
0.18
0.66
0.58
16.42
-0.02
0.37
1.33
Interurban-DT
-1.19
-0.78
-0.11
0.10
-2.17
-1.79
-2.18
-0.53
0.01
-0.49
-1.19
Interurban-LC
-1.79
-0.84
-1.59
-2.78
-2.21
-2.03
-2.18
-0.50
-0.02
-0.54
-1.79
Interurban-Road
0.35
0.47
0.61
2.34
-1.77
-0.70
-0.85
0.20
-0.04
-0.31
0.35
Interurban-Cross
0.17
0.39
0.39
2.06
-1.81
-0.91
-0.98
0.23
-0.04
-0.31
0.17
SMCP-DT
1.13
-0.45
2.99
7.44
-3.03
0.07
-1.91
17.28
-0.42
-2.37
1.13
SMCP-LC
-0.62
-0.77
-1.21
-0.57
-3.04
-0.59
-2.08
17.44
-0.44
-2.50
-0.62
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Development of GDP (leading indicator)
2,0
1,5
GDP against BAU
1,0
[%]
0,5
0,0
-0,5
-1,0
-1,5
-2,0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Explanation
• Congestion charge: Generally positive as stimulation of consumption and
investments are not deemed by the decrease of exports
• Inter-urban toll: First negative development as exports get more
expensive. Positive development of reinvestment scenarios due to
increased investments and stimulated TFP. No recreation of refundalternatives.
• SMCP and inter-urban tolling show, that the consumption impulse caused
by the reduction of direct taxes is superior to the stimulation of
employment via the reduction of labour costs.
• Road investments seem to perform slightly better than cross-funding,
caused by higher time savings achievable in road.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Employment effects
0,8
0,6
Employment against BAU
0,4
[%]
0,2
0,0
-0,2
-0,4
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
-0,6
-0,8
-1,0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
• Diffuse picture: most positive development of reinvestment scenarios.
• Initial peak in SMCP-LC due to high income and consequently high potential to
reduce labour costs. But this is not sustainable due to generally high extra load on
production costs.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Effects on total consumption
4,0
Consumption against BAU
3,0
[%]
2,0
1,0
0,0
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
-1,0
-2,0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
• Most significant stimulation by refund via direct tax reduction
• Effect is neutralised in iter-urban tolls due to the reduction of exports
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Effects on exports
0,5
Exports against BAU
0,0
-0,5
[%]
-1,0
-1,5
-2,0
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
-2,5
-3,0
-3,5
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
• Clear picture: inter-urban road tolls and SMCP on all modes increase production
costs in export-oriented industries and thus reduce the productivity in this sector.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Investment effects
10,0
Investments against BAU
8,0
6,0
[%]
4,0
2,0
0,0
Congestion-DT
Congestion-LC
Congestion-Road
Congestion-Cross
Interurban-DT
Interurban-LC
Interurban-Road
Interurban-Cross
SMCP-DT
SMCP-LC
-2,0
-4,0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
• Short-run: Positive impulses from direct use of revenues for reinvestment.
• Long-run: Better performance of investment stimulation by refunding
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Sensitivity analysis
• Method: Switching the link of transport to particular measures off.
• Performed for three scenarios:
Congestion-DT, Congestion-Cross and Inter-urban-cross.
Example:
Interurban-Cross
Impact on variable
Link transport on
Consumption
Employment
Export
Intermediates
Investment
Productivity
GDP
-0.381
0.267
-0.439
-0.005
3.031
-2.675
Employment Consumption
-0.296
0.386
-0.297
-0.011
1.509
-0.680
-0.558
0.625
-0.576
0.199
3.252
-3.003
Investment
-1.106
0.022
-0.765
0.050
7.925
-6.081
Export
-0.076
-0.059
-1.720
-0.088
0.588
-0.723
TFP
tons
tkm
-0.203
0.134
-0.268
-0.113
1.810
-2.398
-0.344
0.235
-1.035
-0.166
1.809
-1.622
-0.211
0.108
-1.742
-0.311
1.293
-1.287
trips
pkm
0.057 0.015
0.043 0.121
0.072 0.019
0.108 0.076
0.036 0.369
0.091 -0.120
CO2
-0.230
0.159
-0.751
-0.003
1.016
-0.790
• Most significant influence of transport on investments
• in case of strong modal shifts in long-distance transport strong influence on TFP.
• Strong impact on exports in case of high price increases in long-distance transport.
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Development of sensitivities over time
• Example: Influence on GDP in Inter-urban-cross scenario
Base Interurban Charge
Change of GDP against BAU
Percentage change to BAU [%]
4,000
3,000
2,000
1,000
0,000
-1,000
2005
2010
-2,000
-3,000
2020
Exclude Transport
Consumption
Exclude Transport
Employment
Exclude Transport
Export
Exclude Transport
Intermediates
Exclude Transport
Investment
Exclude Transport
Influence on
Influence on
Influence on
Influence on
Influence on
Influence on TFP
Exclude Gov Debt Influence on
Investment
-4,000
Year
Institut für Wirtschaftspolitik und Wirtschaftsforschung
Universität Karlsruhe (TH)
Conclusions
• Considering revenue spending alternatives short- and long-term
developments are to be distinguished.
• The optimality of particular allocation schemes is driven by the indicators
considered and thus by policy preferences.
• In general the reinvestment of revenues in the transport sector seems to
crease most positive effects via its stimulating impact on investments and
factor productivity.
• The ASTRA model indicates a better performance of investments in
roads compared to rail when considering economic indicators However,
ASTRA does not contain a sophisticated capacity model, taking into
account local network conditions. This information is to be contributed
from the case study level.