Document 10653919

advertisement
AN APPLICATION OF MULTICRITERIA EVALUATION
METHODOLOGIES TO AUTO RESTRICTED ZONE POLICIES
by
JAI MU
WON
B.E., Han Yang University
(1974)
M.C.R.P., Seoul National University
(1976)
M.A., University of California, Los Angeles
(1979)
Submitted to the Department of
Urban Studies and Planning
in Partial Fulfillment of the
Requirements for the
Degree of
DOCTOR OF PHILOSOPHY
at the
Massachsetts Institute of Technology
November 1983
Massachusetts Institute of Technology 1983
Signature of Author
Depa3tment
z/2
of Urban Studies and Planning
No mber 9, 1983
Certified by
dd;
0alph Gakenheimer,
il
Accepted by
--
ThesIs Supervisor
q 45- - - - - =
a 4 rence Sus.Wind, Chairman,
... M
MASSACHUSETTS INSTITUTE
OF TU1014OLOY
AUG 10 1984
Ph.D. Committee
ABSTRACT
AN APPLICATION OF MULTICRITERIA EVALUATION
METHODOLOGIES TO AUTO RESTRICTED ZONE POLICIES
by
JAI MU WON
Submitted to the Department of Urban Studies
and Planning on November 9, 1983 in partial
fulfillment of the requirements for the Degree
of Doctor of Philosophy
This thesis applies the three multicriteria evlaution methods
and
Goal-Achievement-Matrix
Analysis,
- Concordance
This
policies.
zone
restricted
auto
to
Solution
Compromise
(1) to discuss the concept
thesis has two major objectives:
zone policies in general
restricted
and practice of the auto
and the rationale behind restricting auto traffic in central
(2) to
in developed and developing countries;
both
cities
the
through
compare the three multiciteria evaluation methods
in
policies
zone(ARZ)
restricted
auto
to
application
usefulnesses and
in order to assess their
specified case
limitations.
the concept and
In the first part of the thesis,
along with the
types of ARZ policies are examined,
West Europe and
experience with ARZs of the U.S.,
developing countries.
to
the
objectives
The discussion also relates
relevant
and
ARZ
schemes.
Generally, the desirability of closing off streets
underlying
the
determined by
is
areas
or
The
characteristics of a specific urban areas.
for different
objectives and ARZs are distinct
the
of
reasons
major
The
of areas.
types
has
region
each
because
be
to
differences seem
the
toward
attitudes
perceptions and
different
Thus decision
transportation problems themselves.
The key
making for ARZ selection is different.
factor is that each of the cities has different
goals which greatly affect the ARZ objectives.
the
thesis,
the
of
part
second
the
In
characteristics of transportation systems in Seoul,
policies
Seven feasible ARZ
Korea are presented.
interviews
and twelve criteria are selected through
and
For selected ARZs
ranking techniques.
and
implementation of the three methods
criteria, the
implementation
The results of the
performed.
is
2
show that the three methods are very effective in
applying to Transporation System Management types
The area license schemes turn out
of strategies.
to be the most preferred ARZ polies by the methods
examined.
The third and final part of the thesis assesses the
three multicriteria evaluation methods in terms of
The criteria seen as essential
several criteria.
include
methods
the assessment of these
in
with
interaction
of
computational burden, degree
tests,
the decision maker, ease of sensitivity
to
applicability and applicability
real-world
that
receals
The comparison
developing countries.
the results obtained from the three methds are
similar but they differ in general in terms of
have
The three methods
described.
criteria
counterbalancing strengths and weaknesses.
Thesis Supervisor:
Ralph Gakenheimer
Professor of Urban Studies
and Planning and Civil
Engineering
3
ACKNOWLEDGEMENTS
the course of this study,
Throughout
from
advice,
the
doctoral
committee
thesis
development
and
members.
my
Gakenheimer,
Ralph
Professor
encouragement
from
my
contribtions
of
assistantce
The
thesis
this study extend far
of
I have benefitted
advisor,
beyond
his
suggestions on particular phases of the research.
Alan
suggestions
provided valuable
has
Strout
that
multicriteria
aided
evaluation
ideas
and
to
the
valuable
Professor
editorial
in both the implementation
of
the
and the
of
the
methods
writing
draft.
I also wish to express sprecial thanks to
Michael
Meyer
whose earlier comments and
this thesis possible.
4
Professor
suggestions
made
TABLE OF CONTENTS
Page
1
TITLE PAGE ..............
ABSTRACT ...............
ACKNOWLEDGEMENT .....................................
.............................
.
2
4
TABLE OF CONTENTS ...................................... 5
LIST OF TABLES........................................... 9
12
LIST OF FIGURES.............................
CHAPTER I :
1.
2.
3.
4.
INTRODUCTION
Objectives of the Thesis
Auto Restricted Zone (ARZ) Schemes
Multicriteria Evaluation Methods
Organization of the Research
13
13
15
22
CHAPTER II: AUTO RESTRICTED ZONE (ARZ)
1. Introduction
2. The Kinds of ARZ Schemes
3. Traffic Restriction Experience
a. European Experience
b. American Experience
c. Experience in Developing Countries
4. General Objectives and Related Issues
of ARZ
a. Reduction of Congestion
b. Improvement of Environment
c. Reduction of Conflict between
Pedestrian and Automobile
d. Enhancement of Economy and
Activities in the City Center
e. Preservation of Cultural Attraction
and City's Unique Character
f. Encouragement of Public Transit Use
5. The Appropriate ARZ in Developed
and Developing Countries
6. Conclusion
25
27
31
31
34
37
40
41
42
44
46
49
50
52
56
CHAPTER III: EVALUATION METHODS USED IN
ARZ SCHEMES
1. Introduction
2. Assessment of ARZ Evaluation
Methods
3. Overall Assessment and Methodological
Issues
5
58
59
62
a. Methodological Issues of Cost and
Benefit Analysis
b. Methodological Issues of
Impact Analysis
c. Methodological Issues of weighting
Technique
4. Multicriteria
Evaluation
Methodologies
in Transportation Field
a. Earlier Development of the Field
b. Recent Development of the Field
5. Conclusion
63
68
70
71
72
75
82
CHAPTER IV DESCRIPTION OF TRAVEL CHARACTERISTICS
OF SEOUL
1. Introduction
2. Description of Seoul's Central
Business District
3. Characteristics of Travel
a. Trip Making Shares
b. Characteristics of Trip Makers
4. Conclusion
83
83
91
91
93
99
CHAPTER V: THE DESCRIPTION OF CRITERIA
AND ARZ ALTERNATIVES
1. Introduction
2. ARZ Alternatives
a. Off-Street Parking Fee Increase
b. Extensive Parking Meter Intallation
c. Public Transit Priority Signals
around Core Ring
d. Core Area License Scheme
e. Core
Area
License Scheme
with
Transit Improvement
f. Toll Increase for Tunnel Gates
g. Toll Charge at Bridge Entrance
3. Objectives and Criteria Proposed
(1) Capital Cost Borne by City
a. Off-Street Parking Fee Increase
b. Extensive Parking Meters in Core
c. Public Transit Priority Signal
d. Core Area License Scheme
Bus Improvement
f. Toll Increase for Tunnel Gates
g. Toll Charge at Bridge Entrances
(2) Current Operating Costs Borne by City
Fee Increase
a. Off-Street Parking
b. Parking Meters in Core
c. Public Transit Priority Signals
6
100
100
104
107
107
109
111
112
115
116
119
120
120
120
121
122
122
123
124
124
125
d. Core ALS
e. Core ALS with Bus Impeovement
f. Toll Increase for Tunnel Gates
g. Toll Charges at Bridge Entrances
(3) Revenue to the City
b. Parking Meters in the Core
d. Core ALS
e. Core ALS with Bus Improvement
f. Toll Increase for Tunnel Gates
g. Toll Charge at Bridges
(4) and (5) Travel time Saving for
and Bus
Auto, Taxi
(6) Reduction in Pollution
(7) Institutional Preference
(8) Enforceability
4. Conclusion
CHAPTER
VI:
132
140
145
147
153
EMPIRICAL ANALYSES OF THE THREE
MULTICRITERIA EVALAUTION METHOD
1. Introduction
2. Concordance Analysis
a. Description.of Theory
b. Implementation of the
3. Goal-Achievement Matrix
a. Description of Theory
b. Implementation of the
4. Compromise Solution
a. Description of Theory
b. Implementation of the
CHAPTER VII:
125
125
127
127
129
129
129
130
130
131
Concordance Analysis
(GAM)
GAM
Compromise Solution
154
154
154
165
181
181
187
196
196
207
EVALAUTION OF THE THREE METHODS
1. Introduction
2. Synopsis of the Three Methods
3. Evaluation of the GAM, CA and CS
a. Computational Burden
(1) Information Requirement
(2) Possibility of Sensitivity Tests
(2) Inclusion of Qualitative
Effects
221
221
224
227
227
229
230
b. Degree of Interaction with the
Decision Maker
(1) Understandability of
the Methods
(2) Encouragement of Public
Participation
(3) Interaction Requirement
c. Real-World Applicability
(1) Applicability to ARZ Schemes
7
232
232
233
233
237
(2) Applicability to Large-Scale
Projects
(3) Incorporation of Uncertainty
(4) Linkage to Planning Process
(5) Linkage to Decision-Making
Process
f. Applicability Transport Environment
in Seoul
(1) Adaptabilty to Scarce Information
(2) Applicability to to Rapidly
Changing Environment
4. Conclusion
238
240
243
243
248
249
250
CHAPTER VIII: SUMMARY AND CONCLUSION
1. Thesis Summary
2. Thesis Conclusion
a. Conclusion Related to
ARZ Experience
b. Conclusion Related to
the Empirical Analysis
c. Conclusion Related to
the Three Methods
3. Consideration for Further Research
4. Extension of the Research
REFERENCES
APPENDIX II-1
APPENDIX III-1
254
257
257
258
258
262
264
266
Characteristics of
Widely-Used ARZ Scehemes
Assessment of Evaluation
Methods Used in ARZ Schemes
8
271
281
LIST OF TABLES
Page
Table
II-1
ARZ Schemes
29
Table
11-2
American ARZ Schemes and Definitions
36
Table
11-3
Objectives and Relevant ARZ and
Techniques
41
Types of City Center and Relevant
ARZ Schemes
52
Table 11-4
Table
11-5
Types of City Structure
53
Table
III-1
Cases for ARZ Evaluation
62
Table
IV-1
Peak-Hour Volume/Capacity Ratio
on Five Major Arterials (1979)
89
Table
IV-2
Trip Making Shares in Seoul
92
Table
IV-3
The Amount of Daily trips Made
by Purpose
98
Parking Spaces in Seoul
105
Table V-2
Toll Gates and Lanes
113
Table V-3
Traffic Volumes on Major
Bridges
115
Summary of City-Born Capital
Costs of ARZ Schemes
123
Summary of City-Born Current Operating
and Enforcement Costs
128
Table V-6
Daily Revenues Collected at Tunnels
130
Table V-7
Revenues Resulting from Five ARZs
131
Table V-8
Logit Parameters
for Jongro Corridor
134
Table V-9
Simulated Individual Travel Times
During Peak Hours After ARZ Schemes
Using TTP.2.1 Flow Prediction Models
138
Table
V-1
Table V-4
Table V-5
9
Table V-10
Simulated Annual Time Savings for Three
Modes Using TTP.2.1 Flow Prediction Models 138
Table V-ll
Public Officials' Evaluation on
Pollution Reduction
139
The CAT for the Evaluation of
Pollution Reduction
142
Table V-13
Institutional Preferences
146
Table V-14
Normalize'd Weights for 12 Criteria
152
Table VI-1
Impact Matrix Before Normalization
167
Table VI-2
Normalized Impact Matrix
168
Table VI-3
Concordance Pairs
169
Table VI-4
Concordance Matrix
170
Table VI-5
Discordance
172
Table VI-6
Weighted Normalized Matrix
173
Table VI-7
Discordance Matrix
174
Table VI-8
Determination of G, H and T
by .700 and .300
176
Determination of G, H and T
by .500 and .500
177
Determination of G, H and T
by .200 and .700
178
Table VI-l
Normalized Impact Matrix
190
Table VI-12
Weights for Criteria
191
Table VI-13
Goal-Achievement
191
Table VI-14
Goal-Achievement Matrix
192
Table VI-15
Degree of Closeness with Respect
to i and k
193
Impact Matrix for ARZ Alternatives
and Criteria
209
Transformation of Alternatives and
Criteria
210
Table V-12
Table VI-9
Table VI-10
Table VI-16
Table VI-17
(sets)
Pairs
10
Scores
Table VI-18
Normalization
Table VI-19
Values for
212
g
w
,
*
Table
VI-20
2
Values for g
z
-
i
i
*
k
-
i
)
z )
215
217
i
Assessment of the Three Methods Against
the Criteria
11
214
k
z )
(z
Values for g
i
Table VII-1
i
*
(z -
i
Table VI-22
i
k
Values for (z
i
Table VI-21
213
and g
i
i
226
LIST OF FIGURES
Figure IV-1
Major Road Network in Seoul
85
Figure IV-2
Seoul's Central Business District
87
Figure V-1
Survey Questionnare
103
Figure V-2
Parking Meters Installed in the CBD
of Seoul
106
Proposed Locations for 13 Transit
Priority Signals
108
Area to be Affected by Core Area
License Scheme
110
Locations of Toll Gates at MajorTunnels and Bridges
114
Figure V-6
Procedure Built in T.TP 2.1
136
Figure VII-1
Major Components of the Three Methods
222
Figure VII-2
The Interaction Requirement between
the Analyst and the Decision Maker
235
Stages of Planning and Decision
Making Process
241
Figure V-3
Figure V-4
Figure V-5
Figure VII-3
12
CHAPTER I:
INTRODUCTION
1. Objectives of the Thesis
This
thesis has two major objectives:
(1) to discuss the
concept and practice of the auto restricted zone policies
general
and
traffic
in
countries;
in
restricting
automobile
central cities both in developed and
developing
the
behind
rationale
(2) to compare the three multicriteria evaluation
methods
through
policies
in
the
application to
specified
case
in
auto
order
restricted
to
zone
their
assess
usefulnesses and limitations.
2. Auto Restricted Zone(ARZ) Schemes
The
increasing
attention being paid
restriction in central cities is
ment
in
the
evolution
transportation
planning.
development
that
of
automobile
thought
in
urban
point of departure for
was characterized by an approach
automobile
the
a relatively recent develop-
conceptual
The
to
restriction would,
by
which
itself,
this
assumed
eventually
improve the traffic conditions of the urban areas.
There
has
been
a
growing
13
application
of
the
auto
and
zone policies in central cities of developed
restricted
developing countries.
The increasing interest in restricting
auto traffic within the city center is mainly fostered by the
deteriorating environmental quality in urban areas.
The con-
cept has been most significantly advanced in European cities.
of
Over 130 European citeis have implemented a large variety
ARZ policies to reduce congestion and negative effects on the
The application of this concept to U.S. cities
environment.
limited in scope.
appears
during the last two decades represent a
built
been
have
instituted
typical
some
form
of
auto
restriction.
in developing countries has been more limited
Experience
It is,
document case in the context of developing countries.
however,
in
Singapore area license scheme is the only well-
The
scope.
have
At present, more than 70 U.S. cities of varying
ARZ scemes.
size
The pedestrian malls which
fair
may
to say that various types of ARZ schemes
exist in major cities in developing countries.
Despite the growing interest and application,
still
area
urban
considerable
restricted
controversy regarding
apporpriate
what
zone schemes are appropriate
there is
in
environment and what they might accomplish.
a
given
Thus
it
seems appropriate to address several issues regarding the ARZ
schemes.
The
issues to be addressed in this thesis are
as
follows:
(1)
The
first issue to be addressed is with regard
kinds of ARZ schemes.
automobile
to
the
The schemes available for restricting
traffic are numerous and may be applied to
14
whole
of
different
situations.
It seems,
ranges
schemes
by
to
them according
measure
of
types
and
restriction
necessary to explore ARZ
therefore,
classifying
transportation
and
sizes
urban
the
exerted
degree
of
different
by
strategies.
The second issue to be addressed regards the
(2)
restricting
of
auto
traffic
developed and developing countries.
vary
The objectives of
ARZs
depending
third
issue
to be
an examination.
concerns
addressed
ARZ schemes and their associate techniques
relevant
relevant
and
ARZ
with
the
by
the
Question can be raised as to what
types of city and region.
developed
upon
The relationship between the objectives and the
ARZ schemes, therefore, deserves
The
cities
circumstances and are tailored to to the specific
urban areas.
the
both
in intensity from one city to another,
individual
(3)
in
central
in
objectives
schemes are in central
developed
countries
and
cities
of
what
kinds
both
of
supportive techniques are appropriate for these regions.
3.Multicritertia Evaluation Methods
The
transportation facilities result in multiplicity
sometimes
unidentified
transportation
arrays
of
consequences
as
facilities cut accross the many parts of
socio-economic and environmental systems.
15
and
the
the
In recognition of
the
to
effort
environmental
devoted
planners and analysts have
transportation
factors,
much
economic and
to acount for social,
need
the
improving
methodology.
evaluation
some progress has been made in the development
Although
refinement
of
evaluation methods for dealing
economic and environmental factors,
transportation
social,
with
a common feature of
the
mostly
is
that the evaluation
is
analyses
and
based on a single criteria - economic efficiency.
Traditionally, most transportation evaluation has centered
criterion
around
travel
and
saving
time
overlooked
largely
seems to be three reasons for this tradition.
there
is
a lack of a conceptual or
theoretical
framework
which is suitable for the
transportation
involving
multicriteria.
and
or
time
This
important.
have
criteria.
decision
important
as
There
travel
cost
Criteria other than above-considerations
minimization.
been
as
such
Secondly,
cost
investment
evaluation
analysts
it was difficult to get
Lastly,
than
less
considered
was true for both transportation
decision makers.
systems
other
criteria
were
First,
the
information on various criteria.
The
multicriteria
evaluation
are
methods
generally
designed to accomodate the social, economic and environmental
factors.
an
Although some of the criteria are not measurable in
"absolute"
factors
are
sense,
entering
the very fact
the
that
evaluation
multidimensional
process
through
quantitative and qualitative analysis reflects the attraction
of these methods.
The insight provided by the
16
multicriteria
methods
evaluation
communication
can also lead to a better
between the analyst and the decision maker, thus allowing the
decision
maker a more active and understanding role
in
the
decision making process.
The emergence of multicriteria concerns in the transportation
field
multiple
attempt
The
systems.
transportation
of
impacts
of
observation
root in the
its
has
to
incorportate multicriteria into an evaluation framework was a
response
to the widespread use of conventional
analysis
in transportation project
been
a
There
has
inadequacies
of
evaluation.
of
the
analysis in
cost-benefit
conventional
range
recognition
growing
capturing
decision making.
the
Academ icians and decision makers alike have
to the need to consider intangibles such as
pointed
full
transportation
social values that are relevant in
of
cost-benefit
social,
economic and environmental concerns in choosing from a
given
set of alternatives.
Despite
the relative lack of attention,
several
writers
have advanced and applied multicriteria evaluation methods to
to
contributions
worthwhile
multicriteria
There
situations.
transportation
concern
the
with
several
of
the
evaluaiton
and
rationalization
conjuction
in
been
have
decision making.
An
early
framework
They
suggestion
regarding
the
need
for
for evaluation was made by Lang and Wohl
essentially
stressed
the need
for
a
more
a
broad
(1960).
rigorous
definition of "impact", suggesting a classification into four
17
types - economic, social, aesthetic and political.
Faced
the need for
with
which
transportation
systems
in
contained
nine
Transportation
the
of
the
considerations
are
is
the
the
which
elements
explicit
must
analysts
the
for
necessary
their
systems and
and evalaution of alternative
definition
of
Principles
oderly,
developing the information
in
consider
task
the
The major significance of
evaluation
those
of
"The
called
elements
for
of
presentation
The
analyst.
Sytems Analysis".
principles
set
delineate
clearly
considerations
a
contributed
has
(1967)
Manheim
concerns,
multidimensional
incorporating
effects.
Manheim
further
(1969)
described
as
that
transportation
socio-political
essentially
choices
can
choices
where differing interests of the various groups
appears
to
represent
models (or methods)
conventional
a
are
socio-political
The inclusion of
and balanced.
considered
choices
be
points out
clear
advance
over
the
in its abandonment of
search for only single dimensional choice.
In
goals
a
similar vein,
of
transportation
controversy,
interests
this,
Hill
widespread fuzziness about what
particularly
systems
in
should
urban
areas
or value systems often conflict.
(1967)
be
has
the
created
where
public
In response
has created a framework in which a
to
great
number of goals for transport systems can be listed and their
relationships
specified.
This
work provided a
basis
for
developing the concept of the Goals-Achievements Matrix to be
18
evaluated latter in this thesis.
The
multicriteria
method
evaluation
developed
was
(1973)
explicit form largely through the work of Morris Hill
England.
applied to a transportation plan in Cambridge,
and
evalaution
multicriteria
the
(1979)
Manheim
Later,
Rietveld
freeeways.
the
the
subjective
proposed a constraint method,
and
(1967),
Marglin
method
incorproated
which is generally
which
of Zions and
and
Wallenius
attributed
preference-
interactive
the
methods to hypothetical rural
these
applied
method,~
the
Friezs and others (1980)
preferences are gradually defined.
to
Using integer
an interactive exchange in which constraints
for
calls
programming
and provisional nature of
objective function is fully recognized in the
to
Angeles
Los
(1978) applied interactive
methods to the rerouting of commuter traffic.
programming,
the
apply
method called "impact tableau"
of alternative alignment for
effects
to
attemt
an
made
in
(1976).
They
transportation
projects.
The broadened objectives are an attempt not only to assist
decision
makers
projects
facilities.
range
evaluating
alternative
transportation
but also to educate the affected public
various
the
in
associated
effects
Perhaps
with
regarding
transportation
the most important result of a broader
of considerations
is the explicit recognition of
interrelationships between transportation and other facts
the
of
life.
Despite
several contributions to multicriteria considera-
19
evaluation
multicriteria
improve
to
either
made
been
little attempt has
in the trnasportation field,
tions
methods
available in the transportation field and other fields or
context
of
include
the
Goal-Achievement Matrix and
the
capable
of
methods
these
test
empirically
to
the
in
transportation projects.
methods chosen for this research
three
The
Concordance
Analysis,
Compromise
Solution.
and
lie
benefits
generally
are
of
these
in their ability to use many measures of
costs
Strengths
problems.
from
usually excluded
the use
are not easily quantifiable,
that
economic
conventional
economic and environmental
By reflecting social,
analysis.
factors
They
multiciteria
encompassing
methods
the
the
of
methods can provide adequate justification for transportation
projects
not
defendable
previously
the
background
of
by
economic
analysis
alone.
With
evaluation
to
paid
deficiencies
in
conventional
methods and because of the scant attention
being
in
the
multicriteria
transportation field,
evalaution
methods
the particular issues to be
addressed
in the latter part of the thesis are as follows:
(1)
which
The
first issue to be addressed regards the
ARZ schemes are currently evalauted.
manner
in
Such evaluation
It can provide a basis for not only assessing the
advantages
and
also
limitations
examining
the
multiobjective)
of
conventional
extent
to
methods
which
but
multicriteria
for
(or
are taken into account in current assessment
20
of transportation plans.
(2) The second issue to be addresed regards the way in
How are the weights
criteria are handled.
qualitative
which
for
qualitative criteria obtained?
(3)
The third issue to be addressed concerns whether
mation
on
adequate
criteria and ARZ alternatives are sufficient
for
Analysis,
This
implementing the three
Goal-Achievement-Matrix
issue
results
infor-
methods
-
Concordance
and Compromise
can be addressed by examining
of an empirical analysis.
the
and
Solution.
process
and
This serves a basis
for
whether the three methods are successfully applicable to such
transportation systems management strategies as ARZ schemes.
(4)
The
fourth issue to be addressed is concerned with
best ARZ policies in the context of Seoul.
are
to
related
be
chosen by each of
issue
selected
by
these
The best policies
methods
concerns with whether the best
the
three
methods turn
the
to
out
examined.
ARZ
be
A
policies
same
or
different.
(5) The final issue to be addressed regards the major usefulnesses and limitations inherent in the three methods in terms
of
several
criteria.
The criteria seen as
essential
for
assessing these methods include: computational burden, degree
of
interaction with the decision maker,
tests,
real-world
applicability
developing countries.
21
and
ease of sensitivity
applicability
to
4.Organization of the Research
zone
auto restricted
the
discusses
briefly
thesis
and
policies
and
the objectives of the
introduces
I
Chapter
field.
multiciteria evaluation methods in the transportation
Chapter I also presents the-organization of the thesis.
II
Chapter
the
general:
in
zone
auto
begins with a discussion of
philosophical
restricted
rationale
behind
a
cursory
restricting automobile traffic in central cities,
examination of ARZ application in the U.S., Europe and cities
in
developing countries,
other
and a taxamony of these and
This discussion continues with a comparison of the
examples.
objectives
and underlying characteristics of ARZ application
countries.
ARZ
Relevant
between
developed and developing
schemes
with regard to deveoped and developing countries are
explored.
Chapter
usually
III
examines
conventional
methods
These
include:
applied to transportation projects.
technique.
impact
analysis,
cost-benefit
These
evaluation methods are
approaches
these
associated
with
alternative
methods.
and
analysis
with their drawbacks and usefulnesses.
weighting
along
discussed,
Posing the
directs
problems
attention
to
This is followed by the review of the
multicriteria
evaluation
transportation
plans
discuss
evaluation
methods
or projects.
with
The attempt is made
to
as
an
the multicriteria evaluation methods as well
22
to
reference
transportation
to
these methods
of
application
empirical
cases.
IV
Chapter
the
of
begins with the description
central
district in Seoul and describes the characteristics
business
of the transportation systems with an emphasis on mode shares
and trip patterns in Seoul.
In
Chapter V,
Discussion
described.
to
feasible
ARZ
II
also
is
development
The
relevant ARZ schemes and
where
of
upon
the
objectives
are
built
alternatives and criteria is
are
criteria
twelve
of
made
schemes.
ARZ
feasible
relevant
Chapter
feasible auto restricted zone schemes
described in detail.
Chapter
theoretical
VI
of
structure
the three
the
of
begins with a discussion
underlying
- Concordance
methods
Analysis, Goal-Achiement Matrix and Compromise Solution - and
conducts an empirical analysis of the three methods using the
This chapter also
information on criteria and alternatives.
the
discusses
results
of the empirical
analysis
and
its
findings.
usefulnesses and limitations
Chapter VII assesses the
the three methods in terms of several criteria.
to
be considered include:
interaction
tests,
with
computational burden,
the decision maker,
applicability
real-world
ease
and
of
of
The criteria
degree
of
sensitivity
applicability
to
developing countries.
Chapter
conclusion.
VIII
This
summarizes
the
chapter
also
23
thesis
suggests
and
presents
directions
a
for
further research.
24
(ARZ)
CHAPTER II: AUTO RESTRICTED ZONE
1. Introduction
with
experience
It
schemes.
(ARZ)
zone
restricted
the concept and
presents
chapter
This
of the U.S.,
ARZs
of
auto
examines
the
types
also
and
Europe
developing
countries. The discussion also relates-the objectives and ARZ
The appropriate ARZ schemes
in these regions.
schemes
with
respect to each region are presented. This chapter provides a
for developing the feasible ARZ schemes
basis
for
criteria
ARZs
and
formulated
for Seoul's CBD serve as a vehicle
for
The feasible
Chapter V.
in
Korea,
Seoul,
conducting an empirical analysis in Chapter VI.
to
Chapter
II
criteria
and
gives
a
detailed
The Appendix
description
of
the
characteristics of widely-used ARZ schemes.
The
is
restriction
evolution
of
planning.
The
characterized
being
attention
increasing
a
relatively
conceptual
devoted
recent
thought
in
restriction could,
an
approach
which
the
development
urban
point of departure for this
by
to
auto
in
the
transportation
development
assumed
that
was
auto
by itself, eventually improve the traffic
25
conditions of the urban areas.
Despite the growing interest,
is still considerable controversy regarding
there
what
the
appropriate area restricted zone schemes are in a given urban
areas and what they might accomplish.
A review of the literature and experience with ARZ schemes
shows a wide variation of concepts and applications
ed.,
1977;
Thomson,
1972;
Koffman
and
(Herald,
Edminster,
1977;
Loudon, Pecknold and Kern, 1979; Thomson, 1977; Gomez-Ibanez,
Heaton and Goodman,
1980;
restricting
fostered
auto
The increasing interest in
1980).
traffic within the city center
is
mainly
by the deteriorating environmental quality in urban
areas, especially central business distrcits.
In
the
developed countries,
it is probably fair to say
that
interest in the ARZ schemes which developed in the
late
was
1960's
generated by the recognition
that
the
private
automobile is the primary source of traffic congestion in the
The interest in this concept was heightened in
city
center.
the
mid-1970's
consumption,
by
growing
air pollution,
concerns
regarding
energy
of
persons
and mobility needs
without access to automobiles.
This
has
concept
variety
significantly
advanced
in
Over 130 European cities have implemented a
European cities.
large
been most
of
auto restricted zone
schemes
to
reduce
congestion and negative effects on environment. Various types
of
traffic
restriction
schemes
26
are
quite
common
and
successful throughout the European countries.
contrast to European cities,
In
concept to U.S.
the
U.S.
downtown
zone
the application of
cities has been on a limited scale.
Most of
of
experience is centered around the closure
shopping street and its conversion to a
with
high degree of emphasis on
a
the
pedestrian
urban
the
this
design
element.
Some of the reasons for the current ARZ interest in
American
cities stem from the energy crisis,
and
Also
conti nued deterioration of the economic base of
of
transit
importance
and
transporta tion
successful
countries
in
has been an interest in
operational
an
districts
both
pollution
air
the
current Federal
in
of
policy
of
ARZ
schemes
awakened widespread interest
public
business
central
encouraging
system managerment (TSM) actions.
implementation
has
advantage
giving
cities.
In effect,
European
in
and
acceptance
the context of American cities and major cities
in
developing countries.
2. The Kinds of ARZ Schemes
The
are
schemes available for restricting
automobile
traffic
numerous and may be applied to whole ranges of different
urban sizes and transportation situations.
27
Many such schemes
and assignment of a price
prohibitions
represent
methods
for
turn
or
ideas such as one-way streets
time-worn
represent
Other
parking.
more radical or innovative ideas such
as
area license schemes, as yet seldom attempted. Because of the
great
of
appicability and effect
similar
it seems appropriate to begin the discussion
strategies,
strategies by classifying them according to the
such
of
of
variety
restriction
and
by
types of measure exerted
ARZ
of
degree
different
strategies.
ARZ
classifications
pattern.
restriction
to
credited
are
The first type of
Thomson(1974)
on
usually based
who
the
general
categorization
is
sets
of
identifies
three
automobile restriction methods in central cities:
l.traffic restriction, in which people are
vented by physical and/or legal means
fulfilling certain travel desires;
prefrom
2.traffic restraint, under which people are discouraged, but not prevented from fulfilling such
desires;
3.traffic avoidance, in which actions are taken
which serves to preclude the development of
travel desires in the first place.
The
second type of categorization covers a broad range of
measures.
This classification of ARZ schemes is based on the
nature of the traffic restriction.
Wagner and Gilbert(1979)
identify four groups of restriction methods:
28
l.physical
location
traffic;
which utilize the
measures,
and capacity of facilities to
design,
control
which utilize signs,
measures,
2.operational
signals, or mechanical devices to control traffic;
3.regulatory measures, which utilize
of regulations to control traffic;
enforcement
4.economic measures, which utilize monetary disincentives as tools to control traffic.
Table
II.1 shows the variety of techniques
each ARZ scheme.
relevant
Some of the techniques listed in Table II.1
play a supplementary role for ARZ schemes.
Table II-1
for
ARZ Schemes
Physical Measures
Regulatory Measures
Street closing
Street barricade
Low design speed
Cul-de-sac
Placement of parking lot
Number of parking space
Ring road/bypass
Area permit
Loading/unloading
Parking restriction
Vehicular regulation
Staggered work hours
Land use regulation
Regulation of ownership
Operational Measures
Economic Measures
Signalization system
Ramp metering
Variable route signing
Turn restriction
Special use lane/street
One-way street
Area license
Parking price/tax
Toll
Congestion pricing
Full tax
Gas rationing
Source: Wagner and Gilbert(1979)
29
In addition to these categorizations,
of
the characteristics
widely-used ARZ schemes are discussed under the following
headings in Appendix to Chapter II:
1.
2.
3.
4.
5.
6.
concept
mechanism
variations
possible changes in commuter travel
side effects
factors which may influence success
30
3. Traffic Restriction Experience
a. European Experience
Generally speaking,
in
the
the development of ARZs has its roots
fact that most European cities underwent
development
a
similar
pattern and therefore were experiencing
similar
environmental and traffic problems.
transition
by the built-up areas first reaching and
then
spilling outside.
radial
city
the
been
the
the old to the modern
Europe,
has
from
characterized
Throughout
Cities became organized around a system of
routes convering on the center.
The growing congestion in the city center caused by rapid
urbanization
reached
the point where
unquestionably necessary.
and
restricted
strategies
parking,
proved
only
Controls,
were
partially
remedial
action
was
such as one-way streets
introduced.
effective.
these
But
Thus
more
comprehensive auto restriction actions emerged along with the
above-mentioned strategies. The trend gained momentum, and by
the late 1970's, nearly every major city had banned cars from
significant
portions
of its historic central
district
and
retail areas.
The ARZ scheme was one of several actions trying to
congestion
and
conserve historic areas.
31
The
scheme
reduce
most
restriction
date has been that of cell-type
to
widely-used
generally for reasons which relate to reducing the
measures,
Cell-type
restrictin
been employed in the center of
Gothenberg,
city
the
congestion
in
techniques
have
Besancon,
Dijon,
center.
Bremen, Delft, and Groningen. Gothenberg's
CBD was divided into five cells,
Similar-ly,
with boundaries created
by
cells were created
in
four
surface
tramways.
Bremen,
Groningen and Besancon. Copenhagen, Dijon, Delft and
where
measures
essentially single-cell ARZ
adapted
Upsala
such techniques as parking restriction, turn prohibition, bus
cities
these
attempt
to
control
this
way,
traffic
and
In
and pedestrian streets were included.
street
automobile
enhance the quality of the environment.
experience with traffic cells indicates that they are
The
relatively low cost and very efective in controlling
both
city center and
the
in
situations
traffic
total
miles may be increased.
reduced
by
In
some
be
reduced
and
While through
traffic
was
may not
volume
vehicle
areas.
residential
traffic
cent
in
Gothenberg in the areas encompassed by the traffic cells,
in
most
cases
50
per
cent in Besancon and
40
per
the total traffic volume for the entire
central
area has essentially remained constant or slightly increased.
But
in
Besancon,
indications
in
both
Gothenberg
as
well
as
Geneva,
are that the redistributed traffic has
increased
efficiency
32
and
improved
the
resulted
operating
(OECD, 1979).
conditions and level of service for other modes
The other type of ARZ scheme widely implemented is that of
pedestrian streets
(or pedestrian areas).
More than a hundred
city centers have implemented a pedestrian streets.
centers
where
characterized
generally
period
the pedestrian streets had been
The city
adapted
the
by the congestion during
or most of the day.
are
peak
The reasons advanced by European
planners have ranged from strictly functional ones aiming
reducing
congestion to humanistic ones aiming at
at
protection
The broad acceptance
of the. environment in historical cores.
of pedestrian streets has been rooted primarily in addressing
these reasons.
A
rather
unique
Nottingham,
England.
implemented
in
plan
focusing
type
of ARZ scheme
was
developed
A system called "zone and collar"
Nottingham in connection with a new
on
one-way
street
was
central
designed
networks
in
to
encourage through traffic. This scheme involved imposing time
on
delays
private
automobiles during the
essentially
metering traffic passage with
signals
two locations.
at
Although
morning
peak
-
demand-responsive
intrazone
circulating
traffic was reduced 50 per cent, this scheme was abondoned in
1976
after
eleven
months because it failed
to
achieve
substantial modal shift from automobile to public transit.
33
a
b. American Experience
Experience
pedestrian
decades
in
the U.S.
appears limited
in
scope.
The
last
two
malls which have been built during the
a typical ARZ scheme.
represent
concern
Increased
the
over traffic congestion and environmental improvement in
1970s
building.
fueled an explosion in pedestrian mall
has
The concept of pedestrian malls was originated from the
to
need
revitalize the downtown area as a commercial and activity
center. The urban renewal program provided an impetus for the
downtown
newer
shopping area to become more competitive
shopping centers by offering much
suburban
convenience
and
of
variety
with
attractiveness:
business,
pedestrian
unobstructed
the
access
traffic
and
the
same
to
a
clean
surroundings.
At present, more than 70 U.S. cities of varying sizes have
instituted some form of automobile restriction (Loudon,1979).
As
in Table 11-2,
shown
implemented
street
the ARZ
schemes
most
frequently
been the closure of the downtown
have
shopping
and its conversion to either a pedestrian mall
or
a
transit mall. Most of the malls constructed to date have been
quite
average
single
modest
U.S.
examples.
in size compared with European
mall
streets.
The
is about three blocks;
effort
is generally
34
almost
to
The
all
are
improve
the
movement
pedestrian
on major streets with
a
only
minimal
effect on automobile traffic.
yet similar, type of mall in America is the
A different,
so-called "transitway" whose definition is given in Table II2.
is a compromise between a completly
Transitway
traffic-
free zone and the need for shopper's transportation system on
the
The
mall.
renewed~
transitway
interest
was generally motivated
in public transit in conjuction
by
the
with
the
effort to revitalize the city center in the early 1970's. The
Chestnut street transitway and Portland
Nicollect mall,
of
cities.
Transitways
the
planning
Chicago,
St.
the
transitway schemes
typical
or
developed
in
in several other cities are in
construction stage,
notably
in
are
American
either
Madison,
Louis, Cleveland, Denver, and Brooklyn and New
York.
35
Table
American ARZ Schemes and Definitions
II-2
Pedestrian
Pedestrian
Mall
Shopping
This is usually developed
in the core of retail and
commercial area and structured to clear separation
between traffic and pedestians.
Shopping
(More than 70 cities)
A street for which has been
restricted largely to pedestrian use
but retains road reservered
for transit vehicles intergrated
with the city-wide or regional
transit system (Koffman,1977).
Transitway
(Nicollet mall,
Chestnut street,
Transitway in
Philadelpia)
The effects of transitways have mostly been favorable. The
best
evidence
of
instituting
cities
which is both the
increasing
such schemes as well as
expansion of ongoing schemes (or plans).
case,
Similar
of
continuous
In the Nicollet mall
the
predominant
good
investment.
responses were observed at the Chestnut
transitway.
surveys
attitude
the
number
was
of
that
merchants reported that
the mall had
been
a
Accordingly, the transitway is judged to have been successful
in
accomplishing
the
goals
of
improving
the
vitality and environmental quality of the area.
36
commercial
c. Experience in Developing Countries
limited
in developing countries has been more
Experience
in scope. The Singapore Area License Scheme (ALS) is the only
case in the context of developing countries.
well-documented
ALS
Singapore
The
to
attempt
the
represents
first
use a road pricing policy
a major city.
in
congestion
to
systematic
ever
control
downtown
appropriate
Thus it seems
the experience of the Singapore ALS in terms of
discuss
to
its
characteristics and effects.
The
primary motivation for the Singapore ALS has been
tihe
relieve
hours.
Under this scheme,
certain
during
the
peak
which was put into effect in June
hours must purchase and display on their windshields
special
delineated
but
congestion occuring
vehicle entering the designated restricted zone during
1975,
a
traffic
to
area
license.
The
boundary
so as to include zones with
of
the
area
problems
congestion
exclude two circumferential bypass routes for
is
motorists
with destinations beyond the zone.
Area
passenger
persons
license
restrictions apply to all
(including taxis) with capacities of
vehicles
or less.
Residents of the central area
from the requirement,
vehicles.
To
automobiles
as, are police,
are
12
exempt
military and emergency
provide an alternative for motorists
37
and
destined
for
the
central area,
outside
a system of 19 fringe
lots
the restricted zone was implemented with shuttle bus
service between the lots and the zone.
also
parking
Parking charges
in the central area when the ALS was
raised
were
put
into
effect to provide added discouragement for automobile users.
overall
The
effect
of
the
ALS
to
seems
been
have
favorable. The government of Singapore set a goal of reducing
vehicular
the
least
traffic entering the restricted
25 per cent and,
has been success.
traffic
at least by this criterion,
was disappointing,
ring
however,
by
at
the ALS
outbound
The effect of the ALS on evening
change in vehicular counts.
outer
area
with little noticeable
Also congestion on the inner and
roads also increased
substantially
because
of
diverted through-traffic.
information
While
of
context
on traffic restriction schemes
developing countries is either nonexistent or
recent development,
fairly
of
types
in
ARZ
turn
a
that
various
restrictions,
one-way
it is fair to say
schemes such as
the
streets and parking restrcitions may exist in major cities in
developing countries.
In fact,
cities like
Lagos,
Bogota,
Teheran and Singapore, one-way streets are in abundance, with
lane
markings,
parking
restrictions
and
copious
use
of
traffic signals (Thomson,1977).
Most
Latin
American cities have instituted
streets in their old city centers.
38
For
example,
pedestrian
pedestrian
Cruz,
About 800
Porlamar and Caracas.
to
Puerto
La
meters
of streets called Sabana Grande in Caracus and
1000
Jiron
Zones such as
De La Union were colsed to vehicular traffic.
Ginza,
of
areas) were implemented in the city centers
(or
streets
and Asakusa in Tokyo were designated as
Shinzuku
an
automobile-free zones in the early 1970's. Similar scheme can
be
found
traffic
are
in
the Myungdong area in
Seoul
where
These ARZ schemes
is prohibited during the weekend.
aimed
vehicular
at relieving the congestion at
the
city
center
(informal survey conducted by auther in 1983).
The work of Migley and Watson(1981) shows that the city of
Brazil
in
Curitiba
center.through no parking areas,
waiting areas.
restricted
employed
in
regulates street parking
Similarly,
were
parking charges
parked
The city of Recife has also introduced
Recife
in Brazil.
number
of pedestrian streets in the city center to make
city
city
loading/unloading areas and
increase with the number of hours
that
the
in
a
the
center less attractive to motorists and more attractive
to pedestrians.
Apart
numbers
city
from the widely-used ARZ
odd-even
plate
of automobile are used to relieve congestion in
the
scheme,
the
odd numbers are allowed to drive into
the
center
automobiles
city
schemes,
in
with
Lagos,
Nigeria.
center on three days per week
Under
this
(Monday,
Wednesday
and
Friday) while the city center is closed for those automobiles
39
with even numbers on these days and vice versa. The same type
of
ARZ
in
implemented
is currently being
scheme
Athens,
In Mexico city, delivery vehicles are prohibited from
Greek.
operating
in
the
city center from 6:00
a.m.
to
midnight
(informal survey conducted by auther in 1983).
4. General Objectives and Related Issues of ARZ
the
discussed
Having
developing countries,
ARZ
experience
in
developed
and
this section focuses on the definition
of a set of generalized objectives and examines the
relation
of each of these objectives to various ARZ schemes.
Clearly,
different
objectives will vary in intensity from one city to
on individual circumstances and will
depending
another,
be
tailored to the specific urban areas.
Six
general objective categories have been
discussion
types
and
as
a
selected
background to the comparison
features
of
ARZ
in
developed
and
of
motives,
developing
countries. These are:
l.Reduction of congestion.
2.Improvement of noise and visual quality.
3.Reduction of conflict between pedestrian and auto.
4.Enhancement of stability of economy and activity.
5.Preservation of cultural attractiveness.
6.Encouragement of transit usage.
40
for
The objectives, relevant ARZs and techniques are presented
in
developing
is not attempted in the table
countries
the
classification
both
developed
difficulties
developed
While a distinction between
Table 11-3.
One
countries.
developing
itself,
context
is designed to focus on the
and
and
of
commonly encountered in attempting to
the
classify
the ARZ schemes and their associated techniques is that
schemes
of
many
are versatile and can be designed to
and techniques
accomplish different objectives.
for discussing the objectives,
This table provides a basis
relevant ARZs and
techniques
in detail.
a.Reduction of Congestion
This
objective
generally
appears to
a
be
fundamental
rationale behind instituting the ARZ schemes in developed and
developing
countries.
congestion
is
The
objectives of reducing
constrained by the need to
maintain
traffic
high
a
level of assessibility to ARZ areas.
The
to
major objective for cell-type traffic restrictions is
reduce the traffic congestion generated
automobiles.
The Gothenberg plan was developed to handle the
For example,
acute
by
traffic
congestion
completion
of two new
Nottingham
case
that developed in the
cross-river
links.
was based on a strong policy
41
CBD
after
Similarly,
objective
the
of
voiced
transportation"
"balanced
by
local
political
officials.
As
previously,
discussed
most American
experience
was
based on economic revitalization.
Auto restriction by itself
was not regarded as an objective.
The immediate objective of
in
a
Singapore scheme was specified in terms of
the
reduction
in
volume of traffic entering the central area
the
with
morning,
the
such a reduction expected to achieve the more
general objective of changing people's travel habits so as to
prevent
congestion
in
problem
from becoming a severe
the
future.
As
shown in Table 11-3,
reduction-of-congestion
objective
appear
cell-type and street closing.
licensing,
one-way
the effective ARZs for achieving
streets,
turn
restriction
and
to
be
area
Such techniques as
Cul-de-sac
are
generally supportive means to these ARZ schemes.
b.Improvement of Environment (Air, Noise and Visual Quality)
Untill recently,
the environmental objective has received
less attention in transportation policy both in the U.S.
in
Europe.
automobiles,
with the growth in the number
of
environmental effects had given rise
to
By the 1970s,
their
and
42
in
except
everywhere
protest
the American
pollution had become the overriding concern for
public
sems
While
similar
countries,
to
can
observed
Europe
in
in
developing
developing countries generally believe that
they
to
the
environmental
to
quality
attention
particular
pay
their
of
appears to be second order of
cities.
objective
This
priority.
the environmental objective tends to
American cities,
In
be
of
the environmental
have received more attention
conditions
afford
cannot
source
principal
Generally speaking,
complaints in Europe.
objective
the
become
noise
whereas
Air
countries.
developing
more directly addressable by the concept of ARZs than
be
of revitalization of the CBD as a
was the original objective
retail center.
In most of the recent ARZ plans in the
U.S.,
the environmental objective is included. The inclusion of the
environmental
objective
reflects the current philosophy
transportaiton policy in the U.S.
cities with regard to
of
ARZ
schemes.
As
the
environmental objective,
likely
achieving
the effectiveness of ARZ schemes toward
to
to
have
only
environmental
problems.
prohibitions,
one-way
nature
a
the cell-type of measures
limited
Many
streets
that their
effect
techniques
and
on
reducing
transitways
are
so
improvement
localized
in
potential
is even more narrowly confined to specific
43
turn
as
such
environmental
are
points
or streets.
By contrast, an ARZ scheme such as Singapore's ALS appears
to have a great potential for achieving this objective, since
it covers a wider area and induces a modal shift from auto to
other
of
mode
It was
transportation.
that
reported
the
environmental
Singapore ALS has been successful in improving
quality.
c. Reduction of Conflict between Pedestrian and Automobile
pedestrian
Improved
objective
separate
The
policy.
fact
mobility
unrelated
that
to
has
a
transportion
traditional
walk trips can be
as
regarded
been
substitutes
or
complements to vehicular trips has been largely neglected. In
cities,
American
transportation
and
thereby
transportation policy.
neglecting
of
magnitude
past
reliance
investment has forced continued
automobile,
the
nature
the
on
pedestrian-oriented
Pedestrian mobility has received more
attention in European countries than in the U.S.
In
malls
most cases,
in
volumes.
within
U.S.
the creation of pedestrian
cities has resulted in increased
For example,
the
and
Boston
shopping
pedestrian
the volume of pedestrians on
ARZ increased by 5 per cent
44
streets
overall
on
weekdays and Saturday: weeknight pedestrian traffic increased
by 17 per cent (NCHRP,
with
associated
commonly
Increased pedestrian volume is
1981).
exclusive
areas
pedestrian
or
streets which enable pedestrians to travel throughout the CBD
conflict
between
improved
travel
pedestrians
reduction
can
and vehicles
result
speed for pedestrians since delays
and
crowding
sidewalk
This
conflict with auto travel.
the
without
signals
traffic
due
the
are
of
in
to
primary
negative factors on walking speed.
cell-type ARZs seem
In terms of ARZ scheme effectiveness,
to
likely
most
reduce
originally designed on a
were
and
pedestrians
and
pedestrians
between
This is particularly true in Europe where
vehicular traffic.
cities
conflict
scale
short vehicular trips,
and
to
accomodate
central
their
areas still contain high proportion of residential commercial
uses which provide a favorable environment for
land
walking
trips.
In developing countries where the road system is
strongly
oriented toward the city center and where the road system for
heavy
absorbing
imagine
that
pedestrians
the
high
countries,
mobility
traffic is inadequate,
there
shoud
be
frequent
hard
it is not
to
between
conflicts
and vehicular traffic in the city center.
Given
in
these
pedestrian traffic in the city
some
ARZ
centers
schemes which improve
are likely to be preferable to other
alternatives.
45
the
pedestrian
transporation
d. Enhancement of Economy and Activities in the City Center
reducing
While
It is,
city center.
could
itself
is
objective
enhancement
objective
connection
with
naive to assume that the ARZ
however,
this
accomplish
of
can,
broad
urban
economic
The
objective.
in
pursued
be
course,
goals.
Other
policies,
public
development
factors may include land use
contributing
to
the
stability'of the economy and activities in
the
enhance
another important
the ARZ,
of
objective
principle
a
is
congestion
traffic
and private investment, metropolitan growth and urban renewal
policies.
terms of the effectiveness of ARZ schemes in achieving
In
the
enhanced
transit
mall
the
to
mobility
and
rban design.
and
mall
were
schemes
improving
These types of ARZs
pedestrian
been
have
in reversing the decline of economic activity
the
city center in American cities.
be
more
pedestrian
pedestrian
bear some potential since these
increase retail sales by
designed
successful
economy objective,
effective
in
fulfilling
The transit mall may
this
than
objective
mall since public transit carries a large
of passengers,
in
number
thereby increasing the potential patronage of
the downtown retail core.
The area must be,
above all,
the
commercial center of particular city in order to sufficiently
46
achieve the economic enhancement objective.
ARZ schemes apear to
regard to European experience,
With
be less often justified in terms of the economic
objective
although
implicitly
expressed
contrast
cities
this
in
objective is both
their
ARZ
have
sufffered
explicitly
development.
sharply with American cities
economic stability of the city center.
from
enhancement
in
and
European
terms
of
While American cities
the inevitable decline of
city
their
centers, Europe's inner cities have shown a striking capacity
to
their attractiveness as commercial
retain
cultural
and
centers.
While the characteristics of the city center vary from one
countries
developing
another,
reasons
coexistence
maintained,
seem to be because
of
variety
economic
activities
of
economy
enhanced
through
vitality
has
been
always
either spontaneously or encouraged by deliberate
Where the city-center space has
government planning efforts.
been in excess demand,
for
the
may be readily achievable in developing countries.
objective
The
to
with a number of activities competing
space such as in developing countries,
ARZ schemes such
as ~pedestrian streets or traffic-free zone may be
effective
in achieving the economic enhancement objective.
Whether
or not paratransit such as jitney or taxi
should
be allowed in the city centers in developing countries depend
greatly
on the specific local
47
situations.
Paratransit
has
augmented the urban development process in the city center in
some
countries by improving access to local
some
economic
system.
Thus
rely heavily on
activities
services.
the
Also
paratransit
undermine
elimination of this paratransit may
the economic activities of the city center.
Accessibility
important
objective.
factor
to the city center has been
for
achieving
the
considered
economic
an
enhancement
The Boston ARZ, for example, shows an increase in
sales volume of 27 per cent from 1977 through 1979 (Cambridge
Systematics, Inc., 1981).
This trend can be attributed to the
existence of a transit system which draws both captive riders
and
shifted-travelers
from
automo6iles.
Thus,
with
the
reduction
of automobile accessibility to the area,
it seems
essential
that an alternative transporation system
such
public transit be available.
48
as
e.Preservation
Character
Unique
of Cultural Attractiveness and City's
This is an objective that is more closely related to urban
urban form.
design
and
rests
directly
transportation
fulfilling
in
reduction
vehicle
congestion.
cities.
sites,
historical
of
attained
success
related-objectives
Traffic
as
such
congestion
along
and noise poses a noticeable problem
pollutants
European
degree
the
on
objective
this
The achievement of
Since
town
there
seems
centers
frequently
to be a
strong
in
a
with
in
contain
desire
to
preserve attractiveness and each city's unique character.
The cell-type of ARZ appears to be productive in achieving
objective.
this
entire
city
traffic cell
The
center
scheme
as well as historical
encompasses
sites
and
the
open
spaces in some cities. For example, the traffic cells created
in Gothenberg, Bremen and Uppsala are far more extensive than
those in American cities.
While
in
the preservation objective is explicitly
several
context,
reports which study the ARZ scheme in
mentioned
the
U.S.
this objective received less attention than did the
transportation and environmental objectives.
cultural attractiveness
Preservation of
and city's unique character seems to
49
an
be
attention
in
countries although some countries have tended
to
objective
developing
that
draws
virtually
no
address this issue in recent years.
This objective can best be served by radical measures such
as congestion pricing over a larger area.
Such techniques as
one-way streets and parking restriction are likely to be less
effective in fulfilling this objective.
f.Encougement of Public Transit Use
The
encouragement
major objectives addressed in the U.S.,
cases.
This
objective
one
of transit use serves as
entails
appears
mall
to
be
best
The public
of
type
public
of
the improvement
the
the
Europe and Singapore
transit accessibility to and within the ARZ area.
transit
of
ARZ
in
accomplishing this objective.
In the U.S.,
addressable
by
this objective does tend to be more directly
downtown
auto
restriction
original objective of auto restrictionn schemes,
revitalization of the CBD as a retail center.
can
be
achieved
rather readily in
50
European
was
the
namely
the
than
This objective
cities
where
accessibility
transit
high.
already
to the city center is
Besancon experience shows that a 75 percent increase increase
ridership
bus
in
increase
This
1979).
Westland,
was observed within two
transit
in
and
(May
years
is
ridership
significant when compared to a 34 percent increase in transit
the
in
ridership
can be made for developing countries
argument
similar
A
1978-1980.
Boston ARZ during
public
where
High density cities rely more
transit use is generally high.
heavily on public transit than less densely-populated ones.
existence
The
of
needs
paratransit
mode
effectively
address
objective.
various
more
the
to
than
be
public
one
public
integrated
transit
or
order
to
in
transit
encouragement
This is particularly important for an area where
types
of
paratransit modes compete for
limited
a
In some developing countries, there are the motorized
space.
and non-motorized paratransit modes such as jitneys, trishaws
and
minibuses
which
play
relatively
providing trips over short and medium distances.
ban
give
of
roles
important
A
complete
paratransit in some urban areas may be necessary
preferential
paratransit
access to public transit given
to
slow
the
speeds and absence of schedules and fixed
stops
which impede the general traffic flow in congested areas.
51
in
5. The Appropriate ARZ in Developed Develoing Countries
Having
established the relationship between objectives
and
ARZ schemes, it seems appropriate to categorize these schemes and
their
ARZ
associate techniques by the type of city and
region.
The
listed in Table 11-4,
present
the
schemes
range
and techniques
of approaches a transportation planner might
consider
in
developing ARZ pakages in each region.
Table 11-4 Types of City Center and Relevant ARZ Schemes
Region
Techniques
Types
Schemes
Old and
small city
center
Street closure
Cell-type
Pedestrian area
Transit street
Physical barier
Traffic circulation
One-way
Parking restriction
Turn restriction
Cul-de-sac
Medium to
large city
center
Transit mall
Transit Street
Parking control
Pedesrian street
Signalization
One-way
Turn restriction
Bus priorities
Low design speed
Old and
small city
center
Pedestrian street
Parking charge
Paratransit street
street closing
One-way
Turn restriction
Low design speed
Medium to
large city
center
Area license
Parking control
Street closing
Pedesrian street
Transit street
One-way
Bus priorities
Turn restriction
Paratranist contorl
Vendor control
Traffic circulation
Developed
countries
Developing
countries
52
actions listed under techniques should be viewed
Most
or
supportive
Since
supplemental to the associated
ARZ
categorization of this
sort should be considered suggestive rather than
Likewise
implementality
conditions
including
depends
the
role
upon a
of
schemes.
countries
the cities in both developed and developing
differ considerably from one another,
as
the
host
definitive.
of
city
external
goverment,
prevailing attitudes, and existing modal shares.
The types of cities categorized by Thomson (1977) provides
a useful basis for discussing relevant ARZ schemes.
As shown
in Table 11-5, three types of categorization were made on the
basis of centralization versus decentralization.
Table 11-5
The Types of City Structure
City Structure
Cities
Centralized
Paris, Washington, London,
New York, Tokyo, Toronto,
Hamburg, Stockholm, Sydney,
Athens
Less Centralized
Boston, San Fransisco, Copenhagen,
Goteberg, Chicago, Vienna,
Melbourne
Decentralized
Baltimore, Pittsburg, Milwaukee,
Buffalo, Salt Lake City, Denver,
Detroit, Los Angeles
Source: Thomson(1977)
53
Thomson's
The
size.
employment density and city
density,
population
factors:
several
was based on
categorization
is
handling capacity of roads in centralized cities
traffic
generally
limited due to a high level of employment density.
Potential
candidate
pedestrian streets and area
traffic-cell creation,
closure,
Parking
schemes.
license
street
are:
cities
schemes for these
restrictions
more
be
also
may
applicable for centralized cities.
not be
may
cities
suitable options to consider since these
low density, grid-
are characterized by larger city centers,
and less conflict
streets
pattern
and
pedesrian streets and traffic cells
street closure,
Denver,
Detroit
Angeles,
cities like Los
decentralized
In
pedestrian
between
and
vehicular traffic. A certain zone of the city center in these
cities
may be,
however,
traffic,
closed off to automobile
depending upon the circumstances.
As
transitways
to
CBD
the
to
connecting
transit
or
seem
to
be
routes
transit
malls,
important
fairly
a
prerequisite. Such a prerequisite is most likely to be met in
older cities such as Boston,
York.
Transitways
to
applicable
Lake
or
San Fransisco,
downtown shopping
Chicago and New
malls
decentralized cities like Los
residential
distance
from
(1)
namely rapid transit connecting the
area
to
the CBD is quite poor;
(2)
residential locations to the city
54
Salt
Angeles,
City and Denver for the following reasons:
transit sytem,
hardly
are
public
suburban
the
center
trip
is
MIUMMk1w - -
(3) the downtown has not generally maintain
relatively long;
number
a
are
(4) there
as a commercial and business center;
its vitality
suburban
of
shopping
near
located
malls
residential areas.
license
Area
larger
schemes may produce potential
cities in developing countries.
developing countries,
however,
benefits
in
In smaller cities in
the license scheme would
be
much bother to introduce for the very small returns that
too
could be gained,
and similar advantages could be gained from
good traffic engineering.
The transit mall may not be relevant for larger cities
countries
activities
in a relatively limited space.
situation worse.
the
The transit
mall
make
It may even
may,
however,
to the city center in developing countries
applicable
such
various
because of the coexistence of
developing
in
be7
where
public trasit as a light rail system alteady exists and
where streets are less crowded.
the
pedestrian
street may be effective as a tool for restricting
automobile
In
a major city in developing countries,
traffic when there is
situations,
lanes
heavy pedestrian
movement. In certain
reserved for paratrasit or transit within
the pedestrian area may potentially be beneficial.
A traffic
cell type of ARZ may cause further traffic congestion due
to
the already severe traffic congestion, lack of boundary roads
and enforcement problems.
While
few would dispute that parking restrictions help to
55
traffic capacity and are essential prerequisite
maximize
many traffic management schemes,
of
limited
they may have only
success as traffic restrainer in developing countries because
enforcement problems.
of
6. Conclusion
desirability
The
or areas
closing off streets
of
generally determined by the underlying charactreristics of
urban area.
specific
this chapter must,
therefore,
countries,
to
solutions
each
has
essentially similar
be general.
the
Europe and
approaches
distinct
transportation
Grossly summarizing these experiences,
The
areas.
Examining the ARZ experience with the U.S.,
developing
a
the objectives and relevant
Likewise,
ARZ schemes are distinct for different types of
discussion in
is
and
problems.
the major reasons for
differences seem to be because each region has different
perceptions
problem
and attitudes toward the transportation
itself. Thus decision makings for ARZ selection is different.
The
key
factor is that each of the
has
cities
different
goals which greatly affect the ARZ objectives.
The
degree to which certain ARZ schemes may be useful
accomplishing
particular
objectives
is
rather
hard
in
to
determine. In most cases, no obvious matrix such as Table II3
exists between ARZ schemes and objectives.
56
Also a
direct
types
specific
with
relationship
of
schemes
ARZ
seems
to establish because of other forces operating
difficult
on
these objectives.
The
then,
conclusion,
orientations
philosophical
led
These
is that each region has
them
to
differ
transportation problems,
in
their
government.
of
traditions
and
distinct
interpretation
of
recognition of ARZ alternatives and
formulation of objectives. Obviously, the differences between
schemes
developing countries seem to imply
and
developed
the
of
exploration
these
differences
is a crucial step when applying ARZ schemes
similarities
urban areas having different characteristics.
way,
the
extent
appropriate
would
to
which
certain ARZ
depend not only on the
the
57
to
would
be
applicability
of
schemes
local constraints which would be faced
these schemes.
and
Viewed in this
these schemes and their experimental confirmation,
on
ARZ
situation.
should be tailored to the specific local
Therefore,
that
in
but
also
applying
METHODS USED IN ARZ SCHEMES
EVALUATION
CHAPTER III:
1. Introduction
the
With
background
of
schemes
ARZ
it is the purpose
objectives discussed in previous chapters,
of
chapter
this
understand
to
relevant
and
the
of
importance
the
evaluation methods through the examination
multicriteria
of
evaluation methods used in ARZ schemes.
The first part of this chapter is aimed at assessing the
case for ARZ evaluation in
the
summarizes
section.
methods
and
are explored.
methodology
some
of
disadvantages
In orther words,
logical
the
these
following
evaluation
ARZ
of
intended
it is
weaknesses
evaluation methods provides a
discussions
in
which the
58
use
relevant
A discussion
background
of
to
assessment
in
and indicate the areas in which a more
and comprehensive framework should be sought.
of
first
results of the assessment made in the
Advantages
identify
The second part
the literature.
to
the
multicriteria
evaluation
methods
approaches.
multicriteria
are
preference
justified in
to
other
This is followed by a section that discusses the
evaluation methods used in transportation sys-
tems.
2. Assessment of ARZ Evaluation Methods
A
review
of
evaluation mehtods
used
for
ARZ
schemes
reveals that there are three different methods that have been
employed in the studies conducted for cities in developed and
developing
countries.
These
methods
are:
cost-benefit
analysis, impact analysis and weighting technique.
Cost-benefit analysis employed in
toward
the
for
the
benefits only
are
issue of whether benefits exceed costs
alternatives considered.
taken
ARZ schemes is directed
into consideration.
In some cases,
Cost-benefit analyses are
mostly
used in preinvestment situations based on before-the-fact (ex
ante)
estimates of costs and benefits.
Ex ante cost-benefit
analyses are undertaken on the basis of anticipated estimated
costs
and benefits of ARZ schemes.
59
Such estimates
are
not
In
based.
the
past
of
absence
necessarily
empirically
performance
and
the cost-benefit
analysis
employed in
some ARZ schemes was based on somewhat
arbitary
empirical data,
assumptions and speculative logic.
analysis
Impact
used
in
method
is the second type of evaluation
ARZ schemes.
as
Impact analysis can be viewed
an
impact evaluation in the sense that it is undertaken after an
ARZ scheme has been shown to have a significant impact.
The
third type of evaluation method used in ARZ scheme
weighting technique.
are
not
easily
that
By reflecting values and effects
the use of the
quantified,
is
weighting
may
provide an adequate justification for projects not defendable
by the cost-benefit type of analysis.
Weighting technique is
of
ARZ
generally
used on the basis of anticipated outcomes
schemes.
The essential usefulness of weighting technique is
that
it can handle a significant number of factors including
tangibles
and
oversimplify
evaluation
This
weighting
variables or criteria.
Moreover,
intangibles.
process is suffered by subjective
interpretation.
60
tends
the
to
entire
weighting
and
Table III-1 Cases for ARZ Evaluation
Methods
Authors
Titles
Area
Gomez-Ibanez
& Fauth(1980)
Downtown Auto
Restraint
Policies
Two Restraint
Policies in
London
Income
Distributional
effects of
Road Pricing
Study of
Congestion
Boston
Thomson (1967)
Kulash(1974)
Tokunaka
(1981)
London
Boston
Washington
Los Anegeles
Los Angeles
Pricing
CostBenef it
Analysis
Peat, Morwick
Mitchell
(1974)
Bhatt(1976)
Bertrand
(1978)
Cambridge
Systematic
(1976)
Mannual for
Pedestrian
Facilities
What can we
do about
Congestion?
Congestion
Costs
Internal CBD
Travel Demand
Hypothetical
Area
Bangkok
Hypothetical
Area
Modelling
Watson&Holland Relieving
Congestion:
(1978)
Singapore ALS
ARZ in
Delaware
Valley R.P.C. Delaware
(1977)
Singapore
Chestnut St.
Trenton
Common
Boston
Nicollet
Impact
TRR(NCHRP)
Analysis
(1981)
Boston ARZ
Nicollet
Transit Mall
OECD(1979)
Managing
Transport
Gothenberg
TRR(NCHRP)
Benefits of
Separating
Pedestrians
& Vehicles
Hypothetical
Weighting
Technique
61
Data
to
order
In
in
evaluate the cases of ARZ evaluation
a
consistent fashion, five criteria are applied to each. First,
alternatives
how
Third,
Fourth,
problem?
decision
the
is
what
that
was
will
the
evaluation?
the
in
evaluation designed
implemented?
and
what were the findings? And, finally, how should the
Each case will be presented in
this
problems
have
the cases of ARZ evaluation have been included.
Each
results be interpreted?
in
manner
be considered
the
are
what
Second,
order to illustrate how different
been addressed.
All
may have certain weaknesses relative to an "ideal" study.
In
and
in
some
cases
real-world data could not be
obtained,
the
studies represent pioneering
atempts
others
However all of the the studies
substantive areas of inquiry.
a useful basis for examining how evaluation
provide
applied
be
can
full
assessments
decided
III.
to ARZ types of
and
schemes,
and weaknesses enter the discussion.
strengths
turned out to be rather
summary
of
these
assessments
methods
their
both
Because the
lengthy,
place most of their details in Appendix
The
their
in
to
it
was
Chapter
along
with
methodological issues surrounding three evaluation methods is
given in the following section.
3. Overall Assessment and Methodological Issues
The
primary emphasis of this
62
section is
to highlight
the
important
methodological
analysis,
impact
discussion
is
analysis
based
and
cost-benefit
surrounding
issues
technique.
weighting
evaluation
ARZ
of
on the assesments
The
listed in Table III-1 and discribed further in this section.
a. Methodological Issues of C/B Analysis
are
There
with
seven major methodological
deserve
analysis that
cost-benefit
associated
issues
discussion.
some
First, the length of project life and choice of discount rate
were
largely neglected in most ARZ evaluation cases.
such
capital intensive projects as rail lines
these
variables
evaluation
imply
assume
malls
changes
may
not be particularly relevant
of ARZ type of projects.
ARZ
that an ARZ such as
in
pedestrian streets or
have value forever:
land
use,
schemes,
highways,
for
density
to
shopping
it is always possible
population
the
however,
It is rather naive
capital and maintenance costs.
will
or
Unlike
that
commercial
and
activities can make the ARZ obsolete.
With regard to discount rates,
it seems necessary to take
the discount rate into account in evaluating an ARZ scheme in
conjunction
reasons
with
discussed
the
length of the project
above.
A discount
considering the concept of opprotunity.
incorporate
the
level of uncertainty
expected returns on investment.
63
rate is
life
for
useful
the
when
Also dicounting can
associated
with
the
However, the selection of an
.............
and subjective.
biased
long-term
with
High discount rates
periods
planning
aften
very
and project life is
rate
discount
appropriate
conjunction
in
the
nullify
effectively
consideration of long-term effects.
Second,
in
while distributional effects were briefly touched
of the ARZ studies,
some
distributional
scheme
is
selected
or
where the ARZ
is
scheme
ARZ
which
affect
can considerably
weighting benefits to pedestrians can push
Heavily
toward
weights
of
choice
The
calculation.
cost-benefit
actual
the
in
they were not embeded
applied.
ARZ
the
The discussion
pedestrian-oriented ARZ schemes.
on
the way weighing can be done with respect to income groups is
beyond the scope of the present study.
Most
statement
about whose interests the ARZ schemes are intended
the
important
There are two main reasons why this is
to serve.
to
explicit
no
of the cases of ARZ evaluation contain
Firstly,
know.
decision-makers and others concerned in
whose
process of decision-making will wish to know from
viewpoint the ARZ schemes are being evaluated:
are merchants and new users included?
possible
to
determine
for instance,
it is
Secondly,
the incidence of
gains
not
losses
and
between different groups unless one knows whose interests are
involved.
Generally
speaking,
cost-benefit analysis looks at costs
and benefits in efficiency terms,
or benefit for all users.
treating equally the
Different ARZ projects,
64
cost
however,
needs of those groups,
contribution
each
a project with
less
of
its
toward an equitable distribution of income
and
actually be more desirable
may
efficiency
in
may be quite different
effects are included,
such
If
case.
and
groups to which the cost and benefit accure,
population
the
because
effects,
distributive
different
vastly
have
may
because
opportunities.
surprising that even in
is
it
Third,
used for transportation evaluation was not
surplus in
consumer
of
measure
out
widely
surplus which has been
consumer
analysis,
benefit
cost-
systematic
terms
money
assumes,
of
description
in his original
The
considered.
as
the
Marshall
pointed
concept,
that the marginal utility of money varies depending
upon the income of users.
when
Furthermore,
consumer
the
of
income,
maximize
to
equitable
a tradeoff has to be made
between
This is because projects
these two often conflicting goals.
less
only
surplus (efficiency) but also to achieve
distribution
that
not
is
goal
are most efficient in increasing total benefits may
efficient
distribution
and
in
reducing
vice
the
dispersion
resolution
A
versa.
of
income
between
conflicting value judgements then will remain a task for
decision-maker.
value
all
maker
What is important,
judgements be made clear,
in
is that all
and that the
a position to make his own
65
the
that the value positions of
interested groups be elaborated,
be
therefore,
be
judgement
decision
on
all
relevant issues although in practice this is often difficult.
Fourth,
benefits
Boston
the
is
value placed upon time in the evaluation
often treated quite loosely (see the
(Gomez-Ibanez and Fauth) and London
of
cases
of
In
(Thomson)).
the two cited cases, an average time value was applied to all
users affected by
transportaiton
ARZs.
the
the
However,
value placed on users' time can vary from person to person in
such as time of
a number of factors,
with
accordance
length of trip and transportation mode used.
in
assumption
homogeneous
purpose
and
countries
users,
mode
behavior
Viewed in
used.
of
this
way,
sensitive
to
countries.
A
and c harateristics.
t ime
va lue
as
those
a
developing
taste
of
They do not appear as
users
in
developed
quesiton can be raised as to
similar
trip
placing
raises more s erious question s as to the
trip mode
are
made
users,
on time saving in the context of
value
monetary
respect to the
with
The underlying
trips
that
was
evaluations
the
day,
whether
very small time savings have any value at all.
Fifth,
not
with o nly a f ew exceptions,
the alternatives were
evaluated di rectly in relation to the objectives of
proposed
ARZ
general
terms.
alternatives,
sch emes.
it
Objectives were
specified
in
the
very
effects
of
was therefore necessary to break down
the
order
In
objectives
into
objectives
provided
more detail.
an
to
The
assess- the
presence
obvious problem
in
of
this
intangible
respect.
There is little doubt that nearly all transportation projects
66
Intangible
involve both tangible and intangible objectives.
objectives were largly neglected in most ARZ evaluations.
mention has been paid to the secondary effects
no
Sixth,
have
Economists or public policy analysts
of ARZ projects.
to really come to grips with the question of how to deal
yet
correctly with secondary effects.
Keynesian
multiplier
stemming
from
These basically refer
and
expenditures
activity
economic
ARZ
these secondary effects might largely concern the
They present not only difficulties
local economy in general.
in correct treatment from a theoretical standpoint,
entail
level
In our
changes surrounding the transportation project.
situation,
negative)
and
effects (both positive
to
mbasuremen.t
tremendous
problems
because
but also
of
the
diversity of effects involved.
Seventh,
the
way most studies classify the benefits
different users appears to be rather naive.
Classes of users
for the ARZ are usually not explicitly considered.
there
are
taken
into
three different classes of users that
account
last
- diverted,
for
gernerated,
and
In fact,
should
be
long-run
category often being assimilated into
the
The distinction between the
two
users,
the
second
in other writings.
made here is that while both groups are undertaking new trips
including modal shifts,
job
This
location.
All
long-run users have changed home
other trips are
considered
or
generated.
distinction is made because benefit measurements to the
two groups can be made somewhat differently.
67
b.
Methodological Issues of Impact Analysis
also
has
analysis
impact
analysis,
been
have
for
used
largely
method seems particularly suitable for
which
cost-benefit
and pilot studies in the transportation field.
demonstration
This
In contrast to
deserve some discussion.
analysis
impact
issues surrounding the
Methodological
noticeable
impacts within a
those
projects
relatively
short
period of time.
Singapore),
With one exception (i.e.,
many studies that
have been examined are mostly based on the ad hoc survey
and
A major weakness of many studies was the failure
interview.
to systematically examine the various indirect impacts of ARZ
schemes.
Because of the limited range of effects considered,
most studies do not provide much insight.
affected
were
mostly ignored by all
the
The social groups
studies
examined
consideration
of
environmental effects was also excluded by most studies.
The
except
that
The
Singapore.
for
Singapore and Gothenberg studies however,
made an attempt to
evaluate pollution and noise.
With regard to the consideration of the sales volume, no
formal
evaluation
of
change was attempted beyond
comparison of sales figures after implementation.
a
rough
It appears
hard to evaluate the degree to which economic viability is to
be
achieved.
Inevitabily,
68
a
considerable
amount
of
the
evaluating
in
unavoidable
was
judgement
subjective
objective of downtown revitalization.
data
Appropriate
cases.
some
in
serious
be
to
seem
implementation
ARZ
data and measurement problems before
the
First,
analysis.
impact
in
several drawbacks inherent
thus
are
There
ARZ
of
describing conditions at the time
initiation were not available in sufficient detail at a later
For
the
Nicollect Transit Mall and Chestnut
Transitway,
for
example,
date.
collection
information
exceptional
of
case
The
schemes.
no attempt was made at
prior
to
ALS
base-line information before implementation.
information
of
Gothenberg's
of
collected
of
the
to
collect
to
Also the amount
of
implementation
the
be
traffic cell was not eqivalent in scope to that
information
the
before
and
appears
a deliberate effort was made
since
data
initiation
the
the Singapore
Street
after
gathered
Thus
implementation.
impacts on pedestrian activity and behavior were not able
to
be ascertained.
Second, the before and after situation poses a different
of
type
methodological
difficulty.
There is
a
risk
of
confusing the effects of the project with those effects which
would
of
the
have occured in any case as a result of the
evolution
It is
extremely
transportation and urban systems.
difficult to distinguish other effects which might have
bearing
on
development
the
ARZ
policy
itself.
scheme
related
to housing
69
For
and
example,
some
urban
infrastructure
environmental
might have had an effect on transportation and
inception
city center between the
the
in
systems
of
the
traffic cell scheme and the opportunity to observe results.
as
Finally,
non-market
external
Similarly,
it
from
primary
recognized in cost-benefit
was
of
effects cause problem
analysis,
measurement.
is difficult for analysts to distinguish
appropriately
to
and
secondary effects
the
the
account for the latter.
c.
Methodological Issues of Weighing Technique
weighting
The
systematic.
arbitary
it
intangibles.
is
diverse
The
costs
and
benefits
and
The weighting scheme employed in the ARZ scheme
a kind of comprehensive weighting technique in
faulted
systematic
is comprehensive in the sense
variables were identified and rated.
be
to
The simple weighting is often based on
on the other hand,
identifies
simple
from
varies
scores and subjective judgements.
weighting,
that
method.
technique
for
70
36
However, this study can
the arbitariness with
assigned and scores derived.
which
which
weights
were
4. Multicriteria Evaluation Methodologies in Transportation
At the outset, it is necessary to discuss the definition
of
multicriteria.
measuring
Multicriteria can be used for
Multiobjectives
the degree to which multiobjectives are met.
are multiple values, multiple preferences or multiple 'desires
of
problems involve multiple objectives.
choose among several alternatives,
decision
Many complex
makers or individuals.
decision
A decision maker must
in
each of which results
impacts describable in terms of criteria.
naive
is
It
evaluated in terms of a single
and
stated
as transportation systems can
such
project
between
imply a complexity
systems
transportation
environment,
the
transportation
of
to assume that objectives
system
economics
a
complex
be
objectively
since
criterion
of
interactions
Yet,
people.
and
objective is defined in
of
terms
a
single'dimension in many cases.
The
multiple criteria approach accomodates
criteria
(environment
recognition
vs.
economics
vs.
of relevant trade-off issues.
social)
that
fact
decision
making
sense,
the
entering
the
quantitative
or
qualitative
multidemensional factors are
process
evaluation
through
either
through
Although some of
the criteria are not measurable in an "absolute"
very
conflicting
reflects the attractiveness of
71
this
approach.
In
contrast
methods,
to
Probably
multiobjective
transportation
strategies
function
reason
major
a
has
been
effort
methods
evaluation
received
have
the
transportation
for
the
relative
and
date involve
in
the
single
a
field.
scarcity
evaluations
is that almost all
literature
to
and
in
formulations
developed
resource
mulicriteria
evaluation
attention
scant
water
of
of
development
the
multicriteria
relatively
field
the
where a great deal of
science
management
devoted
to
of
the
solution
objective
such as the minimization of transportation cost
or
time.
Despite the generally scarce attention,
several writers
have advanced and applied multicriteria evaluation methods to
transportation
worthwhile
contributions
multicriteria
There
situations.
concern
to
the
have been serveral
very
of
the
rationalization
conjunction with
in
evaluation
and
decision-making.
a. Earlier Development of the Field
The
emergence
transportation
field
multiple impacts which
for
judging
the
of
multicriteria
has
its root in
concerns
the
observation
The
72
attempt
the
of
multicriteria
leads to the notion of
impacts.
in
to
incorporate
multicriteria into an evaluation framework was a response
to
widespread use of conventional cost-benefit analysis
in
the
transportation project evaluation.
analysis
cost-benefit
focuses
As was stated previously,
There has been a growing
of economic efficiency.
criterion
cost-benefit
recognition of the inadequacies of conventional
Academicians
are relevant in transportation decision making.
and
decision-makers
alike
intagibles
consider
that
capturing the full range of social values
in
analysis
single
the
on
primarily
concerns
environmental
such
in
have
as
pointed
a
set
given
to
and
social
economic,
choosing from
need
the
to
of
alternatives.
concern for the multicriteria nature of the
In general,
transportation
system emerged in the early 60's in the
U.S.
although there were some implications relating to the impacts
associated
This
1960.
with transportation planning before
concern mainly came from the methodological problems inherent
in the Red Book (mannual for transportation evaluation).
Red
Book
emphasizing
focuses
primarily
on
ratios,
benefit-cost
capital
road user benefits related to
the
The
and
maintenance costs.
An
framework
They
early
a
broader
Wohl
(1960).
suggestion regarding the need for
for evaluation was made by Lang and
essentially
stressed
the
need for
a
more
rigorous
definiiton of "impact", suggesting a classification into four
types - economic,
social,
aesthetic and
73
political.
While
emphasizing the need to identify correctly the differences in
was
foundation
will
laid for estimating just who
in
The
efficiency.
required sacrifices in economic
of
terms
gains
basis for the evaluation of social
a
established
also
they
alternatives,
between
efficiency
economic
the
get
benefits of a given project and who will incur its costs.
A suggestion by Kuhn (1961;
various
objectives
as
such
the
1962) is more broad in that
intangible
and
tangible
effects of potential investments in transportation facilities
should
construction
Tangible items include
be aggregated.
and vehicle operating costs.
Intangibles include the quality
noise,
of transportation service,
dirt,
aesthetics and the
separation of industrial and residential land uses.
he
does
not
analytical
it
include
technique,
Kuhn
specifically
in
Although
proposed
his
the
also advocates determining
distribution or incidence of benefits and costs.
Another trend of this time (i.e.
the
concern
early 1960's) has been
investments in transportation facilities.
of
economic
Institution
effects
for the broad spectrum of economic
objectives set forth by Fromm
study
list
of
The following set
in
a
this spectrum of effects
Brookings
in
terms
appropriate to developed as well as to developing countries.
income
national
aggregate
in
(1) growth
distribution
by an equitable
accompanied
between and within population, business and
regional groups;
(2) increase in the kinds and amounts of final
goods and services available to consumers,
industry and government;
74
(3) development of a national industrial structure
capable of earning foreign exchange and of
supplying domestic markets;
(4) establishment and maintenance of a high level
of employment (Fromm, 1965).
The
implication
of
investments
in
multidimensional
for
call
facilities
transportation
that
these effects is
considerations.
A wide spectrum of objectives was also considered in the
design of a new expressway in Chicago
and
location
late 60's.
in
the
The objectives included decent housing for every
family, elimination of poverty, elimination of crime, removal
of
slums,
Some. of the aspects of the expressway location which
1968).
were
(Pikarsky,
facilities
and adequate recreational
therfore
reduction
are
considered
in the value of adjacent
dislocation
of
people,
division
property,
and
disruption of neighborhoods, and visual blight.
b. Recent Development of the Field
Faced
concerns,
with the need for incorporating
Manheim
which
transportation
systems
in
nine
analyst.
The
of
of
the
considerations
are
a
task
Principles
of
The major significance of
the
called
elements
Transport Systems Analysis".
set
contributed
delineates the
clearly
considerations
contained
has
(1967)
multidimensional
"The
principles to evalution is the orderly, explicit presentation
75
of
elements
those
the
which
consider
must
analysts
and
the information necessary for the definition
developing
in
evaluation of alternative systems and their effects.
be
transportation
further points out that
(1969)
Manheim
as
described
socio-political
essentially
choices
can
choices
where the differing interests of the various
are
is viewed in context with a number
transportation
where
supplementary programs,
of
evolved
concerted strategies can be
The
penalized.
ensure that no single group is unduly
which
that,
suggested
It has been
considered and balanced.
groups
need for socio-political choices appears to represent a clear
over
advance
conventional
models
its
in
methods)
(or
abandonment of the search for only single dimensional choice.
a similar vein,
In
goals
transportation
of
should
systems
urban
in
particularly
controversy,
created
has
be
where
areas
public
In response
or value systems often conflict.
interests
this,
widespread fuzziness about what the
great
(1967) has created a framework in which a
Hill
to
number of goals for transport systems can be listed and their
specified.
relationships
major
hirarchical
levels.
the community are striving.
an
environment
The
highest
The
three
into
level,
"ideals",
those qualities of life for which the members
represents
provides
groups the goals
He
operational
of
The middle level, "objectives",
list
of
elements
those
of
the
which are affected by transportation facilities.
objectives
are
expressed
76
in
such
a
form
that
the
facility
given
a
of
contribution
lists
those
which
can
transportation
elements of the
altered
be
produce
to
their
to
project
"policies"
The lowest level,
can be measured.
achievement
or
itself
system
different
of
levels
achievement of the objectives.
The
they
which
The objectives are in turn connected to those
affect.
ideals,
objectives
are connected to those
policies
It
the level of achievement of which they express.
should be noted that in this framework, benefits are positive
goal achievement,
to
contributions
negative contributions.
while costs
as
appear
The measure of effectiveness for any
plan or project is thus expressed as the level of achievement
of each objective.
presentation
Explicit
operating
definition.
which
in
of
policy decisions on goal achievement in
much
remove
can
the effects
of
of
the
fuzziness
design
and
this
way
surrounding
goal
It provides an especially beneficial
framework
the
expected
to
present
to
decision-makers
consequences of different decisions.
One can envision trade-
off funcitons showing the relationship between the levels
of
of two objectives as a function of the.value
of
achievement
some policy variable.
Also,
the incidence of various costs
and benefits can easily be shown.
a
The work of Hill (1967) is
very significant step in these regards.
achievement
this
thesis,
Since this "goal
matrix' is one of the methods to be examined
the method will be discussed in detail in
77
in
the
next chapter.
to
contribution
valuable
Another
Balance
in the transportation field is the planning
concern
Lichfield
1968).
Sheet Method developed by Lichfield (1964,
center
this method to road proposals for a shopping
applied
of
multicriteria
the
a
Edgware,
England.
suburb of North-West London in
the
asset of the method is that it does not insist on
major
translation
of
recognizing
that
unquantified
terms,
monetary
into
items
relevant
this
It
time.
present
all
attempt to
however,
the
at
is not technically possible
does,
A
take
these
items into account in order that they might
be
viewed alongside the quantified items.
Whereas
such
ignore
conventional
cost-benefit
Planning
the
unquantifiables,
analysis
tends
to
Balance 'Sheet
method at least incorporates them formally into the analysis,
thus
making
the analyst or decision-maker
aware
of
their
and relevance and allowing for a trade-off between
existence
these items and those which are more easily quantified.
large,
pointing
the need to quantify as many variables as
possible
the number of intangibles is generally very
both
to
and
for
a
intangibles
that
a
Yet
more
rigorous
method
into the analysis.
certain
degree of value
of
incorporating
However,
judgement
it is
will
the
inevitable
still
be
necessary in reaching a decision from such an analysis and it
seems
that the Planning Balance Sheet at least narrows
the area in which value judgements will be necessary.
78
down
is
method
recognizes
also
that different
This
involved.
and benefits upon the various groups
costs
of
incidence
explicitly attempts to trace the
it
that
method
advantage of the Planning Balance Sheet
Another
groups
within
the
community are bound to have different aspirations with regard
to particular plans.
advanced
(1975)
Manheim
methodological
of
similar line
a
Along
evaluation
an
development,
methodology
to the current transportation planning
appropriate
process.
The puppose of this evaluation approach is not to choose
best alternative thorugh a rational process,
issues
concern,
of
the concerns of affected
hurt from each action considered.
weighting
are
not
used.
but to identify
parties,
the
be
Elaborate techniques, for
Trade-offs are not made
rather the trade-offs involved are made
evaluator,
the
and those who will gain or
of each alternative,
impacts
more
by
the
explicit
by the evaluator.
The
evaluation approach recommended by Manheim consists
of four activities.
basic
data
which
Prior to commencing the four activities,
include
information
from
citizen
participation in early planning as well as impact information
are organized .and reviewed:
Activity I -
the issues are viewed from the perspective
of the affected parties. Affected parties
is
identified and impact information
are
which
reviewed to determine the extent to
When
affected.
are
parties
these
mitigation
and
compensation
possible,
The opinions of
measures are proposed.
79
affected parties are also reviewed and
information is
when necessary additional
provided to the public.
II - each action is examined and issues of
community concerns, feasibility, equity,
and acceptability to the public are exaAt this point modifications are
mined.
sought to increase desirability.
Activity
Activity III - issues are viewed from the perspective
Compromises
of the process as a whole.
sought and
are
groups
between conflicting
between
consensus
patterns of conflict or
groups are examined.
previous
results of the three
IV - the
activities are documented and reported to
interested
makers and other
decision
parties.
Activity
The
results
purpose of this approach is to
disaggregated form and to
a
in
evaluation
present
community
solicit
reaction to proposed actions as a whole rather than simply
set
The evaluation system is also ment to be
of goals.
active
component of planning.
alternatives,
a
an
evaluate
Rather than merely
the evaluation process is used to generate new
ideas and possible modifications.
Since then,
a
particular
muticriteria
Angeles freeway.
actions.
qualitative
saving
evaluation
impact
an
method,
to the effects of alternative alignment for the Los
tableau,
and
Manheim (1979) has made an attempt to apply
in
The impact tableau contains goal variables
Goal
variables
can
be
expressed
and quantitative terms (for example,
travel time) .
The actions are
transportation plans being considered.
80
the
in
both
30 minutes
alternative
Rietveld (1980) applied an interative programming method
to
the
rerouting
programming,
of
the
commuter
subjective
Using
traffic.
and provisional nature of
objective function is fully recognized in the
calls
for
method,
the
which
an interactive exchange in which constraints
preferences
are
abstraction
planning,
In contrast to
gradually refined.
goal-achievement
matrix,
from
the
however,
Rietveld's
heterogeous elements
thought
tifiable,
and
process,
impact
of
Hill's
real
by
world
the decision
objectives are not
predictions
and
suffers
in which concrete alternatives shape
makers's
integer
fully
quan-
are
often
themselves
uncertain.
Friezs
(1980)
proposes a constraint method,
which
is
generally attributed to Marglin (1967) and to the interactive
method
preference-incorporation
(1976).
Friezs
(1980)
of
applied
and
Zionts
these
two
Wallenius
methods
to
hypothetical rural transport projects.
The
assist
broadened
decision-makers
transportation
public
objectives
in
projects
regarding
but
the
are an attempt not
evalualing
also to educate
various
effects
only
to
alternative
the
affected
associated
with
transportation facilities.
Perhaps the most important result
of
considerations
a
broader
range
of
is
recognition of the interrelationships between
and other facets of life.
81
the
explicit
transportation
5. Conclusion
This
issues
and
chapter
surrounding
weighting
evaluations.
need
has summarized the
major
cost-benefit analysis,
technique
In so doing,
through
the
mehtodological
impact
examination
analysis
of
ARZ
this chapter has demonstrated the
for a multidemensional consideration in
transportation
field in general and ARZ evaluations in particular.
82
DESCRIPTION OF TRAVEL CHARACTERISTICS OF SEOUL
CHAPTER IV
1. Introduction
discussed
Having
the need and
attractiveness
multicriteria evaluation methods, this chapter
1)
in
which
ARZ alternatives
Business
applied;
be
will
the travel characteristics with an emphasis on
discuss
the
is intended to
describe the characteristics of Seoul's Central
District
of
2)
mode
shares and trip patterns.
2.
Description of Seoul's Central Business District
section describes the characteristics of the Seoul
This
environment
the ARZ
of the
schemes
Central
are
applied.
Business
a
information
However,
for
basis
on
a
exploring
the
To
situation,
present
variety of characteristics
The
District
are described followed by travel characteristics.
obtain
Thus
which
characteristics
general
(CBD)
in
is
data are scarce and often inadequate if
required.
available.
description about transportation characteristics has to
be based on the scanty data available.
83
it is located on both
Seoul is the capital city of Korea;
Its central business district (CBD)
sides of the Han River.
area
for
mostly all-purpose roadways used
has 6,249 km of road,
local
through traffic,
shown , in
As
transport.
metropolitan
Seoul
lies in the northern part of Han River.
VI-1,
figure
public
and
freight,
traffic,
is
network
the
essentially radial, with 13 major corridors converging on the
central area.
Since World War II,
and
Seoul has rapidly grown in both size
The large population of the city and
population.
high
rate of growth create a dense and centralized distribution of
fact,
In
population.
social
and
progress,
significantly
alter
transformation
the
on
to
combined
have
environment
the structure of the urban
and transportation patterns.
occurring
technological
growth,
economic
Major new development has been
south of the Han River and
a
of
policy
decentralizing both residerntial and employment opportunities
has been followed.
activity is heavily concentrated in
Economic
Seoul's
banking,
manufacturing, and education.
The
In
the
is shown in the upper-left map in Figure
metropolitan
center
is
VI-2.
region of nearly 8.5 million people,
the principal-urban area,
center for a variety of services.
served
CBD.
city center is a strong regional and national center
for government,
CBD
the
serving as a
city
regional
Traditionally the CBD has
as center for five specific classes of
84
the
activities
-
Figure
VI-1
Major Road Network in Seoul Metropolitan Area
N
River
R1
-
R3
Ring Roads
Ring Roads Planned to be built
Metropolitan Boundary
0
85
1
3
5 km
(both
employment
office
and
entertainment,
professional services,
retail shopping,
sectors),
private
government and
schools.
Government offices are heavily concentrated on the North
Western part of the CBD and are also scattered throughout the
and finance in the middle part of the CBD.
stock market,
and south of the CBD,
east
and
wholesale
approximately
200 m.
lie
area
the
within
flows.
pedestrian
and bakery shop are typical example of this type
Jewelry
of
There are lots of small shops located
the corridors to take advantage of
along
retail
where two major enclosed
markets
shopping
In
the
in
the focal points lie
retail shopping category,
the
insurance,
is a cluster of offices relating to
There
CBD.
of
shop.
and
Dining
entertainment
throughout the area,
dispersed
Myong-dong,
Sogong-dong
particularly
prominent
as
occupying
a
Lotte Hotel,
located near the Seoul city hall, has shown new
number
The
and
significant
various
space
occupiers
including elementary,
entertainments.
types of
within
the
CBD
highschool and college.
86
of
These areas
and shops are located in these areas.
shopping,
are heavily used by different groups of people for
working
near
The area
in these categories in recent years.
buidings
the
in
The same holds true
role.
industry with several major hotels.
lodging
vigor
with the core areas such
Moogyo-dong
and
widely
are
establishments
other
The
are
schools
Many of
the
Cent ral Bus3in 11es
D1S t r i
Figure VI -2
It
Central Business District in Seoul
SCALE
I
kj
A
1:4000
4
J
n
o
iIl
nw
viol~~
on.o"fDn
7
14
Don
oni
GayII'
<
n4n
cng
Don
igD
-
E
I
w
o
ordo
NAan
co
unY
Na by,
it9
s
.
-NN
tIl
a
,Aenn
an
o
-II
D gA;
D
the
of
located in the Northwest and the North
are
schools
CBD.
form
Apart
nature
of other uses that reflect both its historic
variety
wide
the CBD contains a
these functions,
midestern
The
and its location in the heart of Seoul city.
part of the CBD is the major foci of theaters.
the
While
road
to
leads
accidents,
trip
combined with very high
users
considerable
and
substantial
various
the conflict between the
wide streets,
of
network
a
central area of the city has
problems
of
These
environment.
rates
generation
traffic
problems
congestion,
growing
are
due to the increase in vehicle ownership and
rapidly
Conflict
usage.
exists between pedestrians and vehicles and between
vehicles and vehicles.
As
access to the CBD from most areas of Seoul.
these
radial
addition
along
roads
are
the
major
trip
The areas along
attractors.
to the concentration of important commercial
these corriders,
mostly
provide
the main radial routes
shown in Figure IV-1,
In
areas
government offices and buildings are
located within the CBD.
Among these
radial
roads,
Yulgokro, Jongro, Ulgiro, and Thegero are the major east-west
arteries entering into the CBD.
pattern
in the CBD reflects these
The existing street network
major flows of
movement.
central business district is a ring road
that
has continued to function as an important cicular road.
This
Circling
the
ring extends approximately 2.5 km
88
from the center of
Seoul.
There is also an outer ring road, bordering the Han River and
which
extends 5 km from the center of Seoul.
trips
are
street
directed
toward the city
center,
network has become heavily congested.
most
Because
the
In
downtown
addition,
the Han River which is crossed by few major arterials creates
pockets of highly congested entrances to the CBD area.
As shown in Table IV-1, the five major east-west arterials
in
CBD
roads
suffer from
serious congestion.
Volume
far exceeds the design capacity except
Therefore,
traffic
volumes
for
approaching the core
these
on
Yulgokro.
all
from
directions are exceeding capacity limitations.
Table
IV-1
Name of road
Peak-hour Volume/Capacity Ratio on
Arterials (1979)
Volume per hour
Capacity
per hour
Five
Major
V/C
Yulgokro
2,928 (pcu)
3,600
0.81
Jongro
3,649
3,600
1.01
Chunggero
6,227
6,000
1.04
Ulgiro
4,010
3,600
1.11
Thegero
4,945
3,600
1.37
Note: Capacity is the maximum number of vehicles which has a
reasonable expectation of passing over a given section of a
89
Both volume and
Capacity is an hourly volume.
roadway.
capacity
are measured in passenger car unit (pcu).
Seoul City Government, Long-Term and Mid-term
Source:
Planning for Seoul: Transportation Sector, 1980.
There are six main overpass roads
also
connecting
area.
While
each
(elevated roads)
the CBD with the rest of
metropolitan
the
overpass
two-way serivce is possible on these
roads, the width of these roads is narrow and the maintenance
These
poor.
roads are heavily used by taxi and automobible
traffic.
The current traffic situation is typified by dense traffic
flows
on
Vehicles
behavior.
driver
congested
low
the major corridors,
areas
are
and
standards
sound
most
in
illegally parked
Car horns
in the CBD.
through frustration.
driving
incessantly
Air becomes extremely polluted due
to
the heavy emission level generated by vehicles.
In
an
subway
effort to improve present
traffic
By 1980, subway line 1 was in operation
and subway line 2 was under construction.
4 were to be completed by 1986.
total
expected
length
of
a
areas
system was planned in Seoul conneting suburban
to the city center.
a
congestion,
Subway lines 3 and
Ultimately, nine lines with
250 km are envisaged to
cater
increase in urban population and to relieve
traffic congestion occurring throughout the day.
90
to
the
severe
The subway
system
will
Seoul
2001,
by 9 subway lines.
cities
in
role
urban
most
to
eventually be linked
will
riders.
bus
then absorb most of existing
satellite
Hence subways will play the
in
transportation
Seoul
key
be
of
about
11
will
and
By
paramount importance for work trips in Seoul.
3.
Characteristics of Travel
a. Trip Making Shares
total
The
million per day.
destination
in
number of -trips made in Seoul
is
The significant portion of trips have their
CBD,
to the north of the
Han
River.
The
average person trips in Seoul in 1977 was 1.4 trip per person
per day with 73% by public transit,
19% by taxi,
4% by auto
and 4% by other modes.
morning
The
between 6:30 A.M.
trips.
P.M.
peak hours generally occur in
the
period
and 9:30 A.M., representing 37.8% of total
The off-peak trips which occur between noon and 5:00
also
appears to be significant.
occupy 24.4% of total trips.
These off-peak trips
Twenty-one point seven percent
of total trip is made in the evening peak hours between
P.M. and 8:00 P.M.
5:00
The remaining trips which represent 16.1%
of the total occur between 8:00 P.M. and midnight.
91
Table
modal
67.7
IV-2
shows that the role of the bus in terms
shares has been reduced from 88.7 percent in
percent
in 1977.
Taxi's share has rapidly
during the same period.
percent
of
In 1977,
total modal shares,
passengers per day.
1970
to
increased
the taxi represents
and each taxi
of
carries
18.9
130
The information on walking trisp is not
available.
Table IV-2
Trip Making Shares in Seoul
Year
1977
1970
1973
88.7
85.5
67.7
Taxi
3.3
4.6
18.9
Private car
3.2
4.2
4.0
Gov't or company
Provided car
-
3.4
3.7
Subway
-
-
4.8
Other
4.8
2.3
1.9
Total (%)
100
100
100
Modes
(%)
Bus
Walking-trips are not included (no data are available).
Note:
Source: Korean Institute of Science and Technology (KIST) Survey
Report for Comprehensive Transport Planning in Seoul, 1978.
92
b.
Characteristics of Trip Makers
In
for
provide
approximately
are used to a greater extent
Seoul
than any other mode.
two
serving
to
intra-city
an
In the region as a
the subway system plays a relatively small role as
whole,
commuting
mode except for the Jongro
primarily
the
services
extended
The
metropolitan
The subway has been limited
(mostly between city centers) operation.
serve
trips.
urban
throughout
corridors and is largely
major
typically
These modes
two-thirds of all
Buses
prevalent.
are
public transit modes
subway and microbus.
are bus,
They
three
Seoul,
same routes as
corridor.
Microbuses
do
with
buses,
between the CBD and the
outlying
some
areas.
role of buses is principally to provide services to
middle
and low income residents.
a
These groups regresent
the
a
large captive market (groups have no mode choice alternatives
In spite of high coverage and an
than bus).
other
average
headway of around 4-6 minutes, the quality of service is very
poor in terms of comfort and speed.
Except for the subway,
are
Seoul
all public transport services in
provided by the private sector.
Present
public
transport services are offered by registered companies owning
a number of vehicles.
are
determined
issued
to
The official public transport
by the municipal
operate
government.
specific routes in
93
the
routes
Permits
city,
and
are
the
vehicles
of
allocation
The taxi
used for shared rides mostly during the peak hours.
the peak hours.
during
Each passenger is
persons
3
been estimated to be 2.5 to
has
rate
occupancy
widely
Taxis are
assessment of the needs of each corridor.
crude
a
to each route is based on
for
responsible
paying his share of the taxi fare.
of service:
kinds
distinct
taxis in Seoul are providing two
metered
The
their
to
are carried directly from their origin
passengers
more
exclusive service for which one or
destination without intermediate passenger pick-ups and dropand shared service for which the driver picks up three
offs;
or four passengers along the route.
The fare for the exclusive service is typically based on
the
taxi meter while the fare for a shared ride is generally
In most cases, a rate sightly above the
between passengers.
equivalent
which
rate is considered the
metered
driver and the passengers.
taxis
driver often intervene to
a reasonable fare for different
contribute
nevertherless
congestion
passengers.
many
who
not
Though
passengers.
problem
play an important role in
hour
do
to
the
between
Because.different passengers may
be going very different distances,
charge
In some cases,
appear to have been worked out
rules
from
fare
base
each passenger's share is calculated.
informal
divided
fare for a shared ride is equally
The
negotiable.
Seoul,
they
transporting
peak-
in
This service is particulary attractive to
own
a car
94
and
whose
income
level
is
relatively high (from upper middle to high income group).
The bus, in more than one sense, is the "common carrier"
of the Seoul.
Every income group except the lowest (who are
employer-
poor) and the highest (who use autos or
extremely
are
its
among
numbers
found in large
provided
vehicles)
riders.
Those in the middle-income and low-income
brackets
and while the use
form a significant part of the bus riders,
of the bus by the high income group is not as frequent as for
the
middle-income
often used.
groups,
In fact,
the mode is still the
one
most
the largest part of all income groups
ride the bus to work.
The
majority of transit riders in Seoul,
use bus or subway, own no autos.
they
Auto ownership is extremely
The income of those
low even among those with high incomes.
who
whether
use the subway for work trips is relatively higher
for those who use buses.
workers
the median income of
who rode the bus in the Jongro corridor in 1977
was
it
was
w18,500
about
For example,
than
per
month while among subway riders
about w27.646 per month.
The
much
higher
reasons
the
high
income level of riders using the
than
those using either bus
microbus
or
in
currently three times
as
The level of
for this mode is much better than for the buses
subways in terms of comfort and time delay.
95
The
lies
as the fare structure of bus and subway.
service
subway.
for this high income among microbus riders
high level of fare structure,
appears
and
in
This is indicated by the large differential
is nominal.
exist
must
a large population
Given this situation,
automobile.
an
operating
and
transit
public
riding
between
currently
fare
of bus trips is made by people for whom the
proportion
cost
large
the
cost,
significant
a
as
fares
bus
existing
the
perceive
a segment of the polulation may
Although
bus
which can afford more than the cost of
For these people, the
transit but which cannot afford a car.
costs of riding public transit are probably small compared to
the
(waiting,
cost of the time involved in making the trips
transfer andd travel time) and the discomfort of the crowding
on buses.
As
Seoul,
stated
previously,
there
is
transport,
private
and
are
Travelling
by
is
taxi
transit,
particullarly
requires
very
proportion
minibus
taxi,
walking,
walking
and
of the market for-public transit
large
a
result,
As a
public
than
fares,
at the current
travel times.
long
motorcycle.
and
expensive
more
much
by
modes
the
trips,
For the remaining
car in 1977.
available
made
than 5 percent of trips were
less
of
public
to
alternative
comparable
no
population
for most of the
essentially
is
captive.
The
ridership
combined
is
effect
composed
these
of
of two
roughly
other
necessary
trips
made
96
is
defined
that
types.
including work,
certain share of trips is essential,
and
factors
by
people
who
the
A
school
have
no
alternative means of transport.
in
trips
These people will make their
spite of long waiting times and extreme
The remaining trips are optional,
often shopping or
crowding.
recrea-
tional trips which could be foregone, short trips which could
be
made
on foot,
The number of optional trips
etc.
will
decrease as waiting times and crowding increase.
important characteristic of travel is the
trip
The amount of trips made varies according to
trip
Another
purpose.
purposes.
purpose.
Next
are
percent
2.6
trips
Table
shows average percent of
IV-3
trips
by
Work trips account for 69.1 percent of total trips.
journey
to and
of all trips.
from
school,
represent.ing
Recreation and shopping account
and 1.2 percent respectively.
Recreation and
for
shopping
made in Seoul appear to be far smaller than those
north American cities.
19.1
in
The remaining three purposes account
for less than ten percent.
97
Table
(unit:1000)
The Amount of Daily Trips made by Purpose
IV-3
1977
1973
1970
%
%Trips
Trips
Trips
%
School
1,636
24.3
1,872
24.5
2.194
19.1
Work
1.137
16.9
1,563
20.5
2,586
22.4
Shopping
127
1.9
64
0.8
133
1.2
Business
362
5.4
343
4.5
696
6.1
64
0.9
113
1.5
295
2.6
Work-Home 3,279
48.6
3,624
47.0
5,367
46.7
26
0.4
1
15
0.1
109
1.6
55
0.7
208
1.8
6,740
100.0
7,635
100.0
11,496
100.0
Recreation
Travel
etc.
Total
0.02
*Business trips represent trips made for business purposes during
the working hours.
Source: KIST, Survey Report for Comprehensive Transport
Planning in Seoul, 1978.
Work
trips have continuously increased since 1970.
contrast to work trips,
only
slightly
approximately
from
the share of school trips
1970
to
1973,
and
5 percentage points from 1973
98
increased
decreased
to
In
1977.
by
The
in the share of school trips during this period
decrease
to
attributable
the fact that
is
middle
(1) the relocation of
and high schools from the CBD to outlying areas, south of Han
River, as part of government's decentralization policies;
of
all
categories
increase
percent
The
have steadily
trips
Total
the absolute number of school trips increased.
fifty-one
A
increased.
1977.
and
has been observed between 1973
(2)
increase in total trips is due largely to a rise in
car
ownership.
4. ConclUsion
chapter
This
Seoul's
are
to
has
described
the
characteristics
Central Business District in which ARZ
be applied and discused the
travel
of
alternatives
characteristics
with an emphasis on mode shares and trip patterns.
can
It
be
seen from the above
discussion
that
the
in
the
transportation problems are characterized by:
(1)
heavy
concentration of a variety of activities
CBD;
(2)
severe congestion occuring in the major arterials in the
CBD;
(3)
a
high degree of dependency on public transit
buses and subway and other public transit;
(4)
rapid increase in taxi ownership.
99
such
as
CHAPTER V: THE DESCRIPTION OF CRITERIA AND ARZ ALTERNATIVES
1. Introduction
District,
Business
Central
Seoul's
described the travel characteristics of
Having
the
primary
this
of
purpose
the
chapter is to determine the set of criteria and ARZs for
Korea.
The
criteria and ARZ alternatives is built
upon
multicriteria analysis in the context of Seoul,
of
advancement
the
with
regard
this
in
ARZ alternatives to be advanced
and
criteria
to
detail.
and developing countries are described in
developed
The
where objectives and ARZs
II
chapter
chapter will be used for empirical anayses in chapter VI.
2. ARZ alternatives
order
In
schemes,
a
these
public
officials
survey of objectives and ARZ alternatives in the
Urban
of the CBD was conducted within the Bureau of
Transportation,
Center)
of
in transportation planning concerning alternative ARZ
engaged
context
to obtain the opinions
Urban
Planning and Police
(Traffic
A questionnare was distributed among the
bureaus.
Sixteen
Control
staff
public officials responded to
100
in
the
survey.
In an attempt to give decision making groups a sense
the
benefits
and
costs associated with the
criteria
reference to each of the proposed ARZ alternative,
were
effforts
dealing
made.
with
First,
have
Bureau
of
Transportation
were
interviews
as
feasibility
implementation with regard to
the
ARZs
of
chosen
for
and the magnitude of the benefits and costs.
respondents
understood
and
They
also asked to provide the author
were
were willing to devote the
been
Urban
to
the
proposed
The
fairly well what was intended
discussion
with
series of
six officials who
ARZ types of actions in the
of
necessary
with
six
after
time.
currently
available data on all criteria.
Second,
after receiving the currently available data on
all criteria, author throughly analyzed the data
and came up
with approximate figures for quantifiable criteria and simple
ranking
for
non-quantifiable
implementability
A
criteria.
Also
the
of each of the proposed ARZs were examined.
number of telephone calls were made and
several
meetings
were held for this purpose.
Third,
it
was
deemed
necessary
interview all sixteen public officials.
them
the
at
this
point
to
The author presented
characteristics of each of the proposed
ARZs
and
magnitude of benefits and costs associated with all criteria.
A
week
designed
later,
to
the author prepared a questionnare that
determine the set of criteria
101
that
should
was
be
considered
in
the multicriteria analyses in Chapter VI
and
the establishment of the relative scorings or ratings for all
alternative ARZ schemes and weights for all the criteria. The
questionnare
shown in Figure V-1 identifies
and ARZ alternatives.
102
the
objectives
Figure V-1
Survey
Questionnare
I am interested
in obtaining opinions of city
officials
concerning their perception of
the
transportation objectives and automobile restricted
zone schemes within the context of the centeral
business district.
Appearing below is a list of
set of objectives and ARZ alternatives. Should you
objectives
or
ARZ
find
that
additional
alternatives are necessary, please include them
under "other."
The list of objectives and ARZ
alternatives is to be ranked in terms of importance
to the transportation situations in the CBD. The
ranking can be done on a scale between 0 and
10
where 10 indicates the most desirable objectives or
alternatives.
Rank
Objectives
1. Minimize capital cost
2. Minimize current operating cost
3.. Maximize revenue
4. Reduce travel time for auto and taxi
5. Reduce travel time for public transit
6. Reduce pollution
7. Minimize traffic accident
8. Minimize instituitonal obstacle
9. Minimize inforcement problem
Other: (please specify)
ARZ alternatives
----------------------------------------------
Rank
On-street parking fee increase
Extensive parking meter installation
Public transit priority signals
Core area license scheme (ALS)
Core ALS with transit improvement
Toll increase for tunnel gates
Toll charges at bridge entrances
Outer area license scheme
Toll charge at bridges with parking
lots
j. Pedestrian streets
k. Entrance ramp metering at bridges
1. Outer ALS with transit improvement
m. Outer ALS with transit improvement
and fringe parking lots
a.
b.
c.
d.
e.
f.
g.
h.
i.
103
------------------------------------------------------------------
The
ranked
respondants
alternatives in order of
basis
ARZ
and
objectives
the seven-
Using unweighted average ranking,
alternatives.
ARZ
The results provided a
importance.
important
selecting
for
and
objectives
the
top-ranked ARZs were selected.
a. Off-street Parking Fee Increase
b. Extensive Parking Meter Installation
c. Public Transit Priority Signals around Core Ring
d. Core Area License Scheme
e. Core Area License with Transit Improvement
f. Toll
Increase for Tunnel Gates
g. Toll Charges at Bridge Entrances
Since the underlying concepts, mechanisms and effects of
ARZs have been explored in detail in Chapter II,
the special
of each alternative in the context of
characteristics
Seoul
are briefly discussed.
a. Off-street Parking Fee Increase
(from W500 to W1000 per hour)
Parking
arrangement,
users.
The
controlling
traffic
in
Seoul
is
characterized
spacing
city
government
has
these parking problems.
some
disorganized
and inconsideration
close
flows,
by
not
been
104
and
other
effective
Where there
degree of observance
for
are
in
high
enforcement
exists,
but
generally the majority of the existing
parking
restrictions are ignored.
Parking
in the CBD is currently available at both
street
and on-street locations.
number
of
These
demand.
The
CBD
and
parking
commercial purpose.
Table V-1
to
spaces
On-street parking spaces are controlled by the
200 Won (.25 U.S.
Dollar) is
charged for every 30 minutes for these parking
spaces is
amounts
total demand was estimated to be 38,947
government,
off-street
the
spaces apprear to be far fewer than the total
for the CBD.
city
As shown in Table V-1,
parking spaces available in the
16,306.
off-
lots
areas.
are privately owned and
The standard rate for these
500 Won per one hour.
used
CBD
No. of
facilities
.
Other Areas
No. of
spaces
No. of
No. of
facilities spaces
On-street
108
(public spaces)
3346
28
835
Off-street
482
(parking lots)
12960
2693
56725
Total
16306
2721
57560
590
Source: Division of Parking Facilities in Transportation
Bureau of Seoul, 1983
105
Most
for
off-street
Parking Spaces in Seoul
Type
being
The
off-street
parking
fee
increase
would
vehicles.
company-provided
This
an
government
additional charge only on private automobiles and
and
put
be
would
policy
effective for morning peak hours between 7:00 A.M.
and 10:00
A.M. which represent the most congested hours of the day.
A
terms
rate of 500 Won per hour appears to be rather low
in
rate
is
of
the
restricting
equivalent to .625 American dollar.
city
dollar per hour.
not
seem
U.S.
A parking charge of 500 Won per hour
does
a burden for those in high income
be
price
Thus
bracket
that auto owners typically fall in
high income category in Korea,
rather
Most parking lots in the
one
Considering
Seoul.
This
the equivalent of more than
charge
centers
demand.
parking
inelastic
the
these auto owners seem to
to an increase in
in
be
charge.
parking
small increse in the parking charge would probably
not
induce them to switch to other modes.
The impact of a parking fee increase would be felt
strongly
between
Most
by
long-term
7:00
A.M.
commuters who arrive
and 10:00 A.M.
the
core
and park there all
day.
non-work trips are made after 10:00 A.M.
increase would no longer be in effect.
would
not
be
increase
might
shifts.
Auto
affected
significantly.
in
when the
Thus non-work
The
parking
have impacts beyond those related
owners
most
to
or hired drivers might avoid the
fee
trips
fee
modal
fee
increase at private commercial parking facilities by shifting
to street spaces, thereby creating congestion on the streets.
106
b. Extensive Parking Meter Installation (500 Parking meters)
second ARZ alternative considered is the
The
The amount of
installation.
meter
parking
would be increased by 15 percent.
meters
parking
extensive
Five hundreads additional
are to be installed under this
alternative.
This on-street parking restriction is designed to
short-term
parking
travelled
heavily
by
auto
spaces
public
drivers or
hired
or congested roadways in
discourage
drivers
the
CBD.
on
One
hundread and fifty parking meters have been installed in four
locations
since
April
1981.
The
parking
were
meters
purchased from the VDO company of West Germany at 157,670 Won
Seoul,
1983).
Much
of the on-street parking is now being filled by
public
lots
scattered around the CBD.
($197)
each
(Department of
Transportation,
Most cars parked
on
these
lots are also short-term.
c.
Public Transit Priorty Signals around Core Ring
This
and
taxis
hours.
in
scheme involves time delays on private automobiles
willing to enter the core area
during
the
peak
A priority would be given to buses at entrance points
the rings around the CBD as shown in Figure V-3,
responsive
signals would be installed at
107
thirteen
Demandentrance
Figure V
-
2
Proposed Locations for 13 Transit Priority Signals
R1
C D
-
0
3
Proposed Locations for Signals
R1 - R3 Ring Roads
108
Han Rib
points
to
automobiles
the core, area.
and
the core area.
20
minutes.
responsive
Time delays imposed
on
private
taxis depend on the level of congestion
in
Time delays generally range from 5 minutes to
There
are presently
131
electronic
signals in Seoul metropolitan area.
demand-
The demand-
responsive signals installed at 13 entrance points in the CBD
can
be
converted
Demand-responsive
into
public
signals
transit
require
priority
capital
signals.
investment
and
impose ongoing costs associated with enforcement.
Enforcing
the
expensive
time delay on automobiles and taxis would
because
of
the large man-power
requirement.
be
This
scheme
would reduce travel times for all transit vehicles, including
bus and microbus.
transit
would
be
A modal shift from auto and taxi to public
expected
to occur as a
result
of
this
scheme.
d.
Core Area License Scheme (w 1,000)
This alternative is designed to increase traffic
capacities
and
four
handling
of the core area by imposing restrictions on auto
taxi modes.
Automobiles and taxis carrying fewer
passengers would be required to display
during morning peak hours.
area
than
licenses
A license would be necessary for
operating such a vehicle within the core area.
109
The core area
sinless
Figure V
DIslrict
-
4
Area for Area License Scheme
SCALE
i4o00
19~
-
'r
-
Am I
-. Ga
\ay
Don.
NOR
~~
~
"IRD11
~
~
g
"4r1.f
al
Gn
a. )o
lly aF)
-1
4.)h.
j0jj
H-4.
1.R
It
U .
o
for
area license scheme (ALS) is shown in
the
Figure
Vehicles
Options include daily, monthly and yearly licenses.
display a special sticker when entering the core
would
The license
to divert auto traffic to other modes or times.
in
only
the
morning peak would
affect
policy would have an impact not only on the
this
peak-hour
Shifting
or
area
core
whole.
work trips from auto and taxi to
either
the
would
subway
80
a
on
also
but
itself
about
the area that could be covered,
Given
percent of commuters.
Pricing
only.
would be required for the morning peak hours
applied
is
A charge this high is necessary
Won ($1.25) per day.
1,000
area
The price level considered
13 entrance points.
through
bus
V-4.
metropolitan
increase
public
area
as
transit
ridership,
thereby making waiting times longer.
e.
Core Area License Scheme with Transit Improvement
This
Area
alternative
License
Scheme
is essentially the same as
except that public
transit
the
would
Core
be
Before considering area license policy, it must be
improved.
recognized
that when charges are imposed for auto
and
taxi
use, alternative services must be offered to accomodate those
who
do
These
improved
not wish to pay or who find the
alternative
public
services
transit
would
service
111
charges
consist
and
excessive.
primarily
increased
of
transit
capacity.
service improvements in core areas would
Transit
include increased frequency of service on existing corridors.
Sensitivity
analysis
will be performed to
investigate
the
effect of changes in the frequency of public transit service.
f.
Toll Increase for Tunnel Gates
alternative consists of increasing tolls from
This
Won to 500 Won at three tunnel gates.
shown in Figure V-5.
200
Won (one way).
Seoul
metropolitan
These
toll
gates
Transportation
installed
to
construction.
southern
and
in
The toll for automobiles is
Namsan,
area including five tunnel
is
gates.
toll
of
Bureau
Seoul.
originally
The toll gates
cover the capital costs
were
incurred
tunnel
for
These tunnels provides access to the CBD from
northern Seoul.
3rd
Some of these
tunnels
also
Among five tunnels, three tunnels
Namsan and Geomwha) are the most
heavily
Hence, peak-period congestion on
tunnels is substantial.
situations
currently
are owned and operated by the
traveled tunne ls in Seoul.
these
are
There are presently eight toll gates in
provide for th rough trips.
(1st
Three toll gates
200
Evidence of these
shown in Table V-2.
tunnels ranges from one to five lanes.
congested
The number of lanes
in
The other tunnels and
roadways where tolls are installed are usually not
congested
during peak hours, unless accident or construction delays are
112
encountered.
It should be mentioned here that there woild be
traffic impacts from the toll increase.
may decrease as a result of toll increase,
of
A volume of traffic
but the magnitude
relatively
volume decrease may not be significant due to
little
in
increase
toll charge and no
viable
alternative
roads.
Toll Gates and Lanes
Table V-2
Toll -gates
Number of
Machines
Number of lanes
Outbound
Inbound
Daily
traffic
volumes
lst Namsan
4
2
3
37,600
2nd Namsan
2
1
1
10,000
3rd Namsan
10
5
5
60,100
Bookak
4
2
2
25,100
Geomwha
8
4
4
40,500
Yongbigyo
4
2
2
23,800
Nambusoonwhan
8
4
4
19,300
1
1
Not
available
1st Gangnamro
Source:
Division
Seoul.
Collected
by persons
of
Toll Gates in
113
Transportation
Bereau
of
Cen tr a I Bt s in e ss D i striect
Figure V-5 Locations of Toll Gates at Major Tdnnels and Br idges
SCALE 1:40000
Sag
19
a
-fwh
-I
-n
L
&YU o.
,K, -"OR 1s
W."Ig I It
-It
do
olltt
-i
R
:1
Gae
-y
O
Locations of Tc
at Major Tunne
a,(
IA
Locations of T
at Bridges
7NIar.nIf
Gay
toll
qf
rt/IoA
a
s
C
g.
Toll Charge at Bridge Entrances
This alternative involves installation of toll gates
the five bridge entrances.
be
charged
morning
for
peak
connecting
bridges,
500 Won (62.5 U.S.
each vehicle entering
hours.
There
are
the
cents) would
bridges
during
major
bridges
thirteen
Among these thirteen
suburban areas to the CBD.
five
bridges play an important role
access to the CBD during the morning peak hours.
Table V-3,
at
in
providing
As shown in
these five bridges are extremely congested during
the morning peak hours.
Table V-3
Traffic volumes on Major Bridges
Average daily
traffic
Bridges
Hourly traffic per one lane
during peak hrs.
Jesam
119,745
1,248
Seoul
105,119
1,167
Banpo
80,587
*
Seongsoo
80 , 291
1,388
Jamsoo
76,423
1,150
Source:
Josun Newspaper, January 29, 1983.
Automobiles amount to 74 percent of all traffic in these
115
five bridges, and trucks and buses represent 15.5 percent and
10.5
percent
respectively.
Since
toll
gates
only in Seongsoo bridge to cover the
installed
cost incurred when the bridge was completed,
is
3.
have
been
construction
Seongsoo bridge
excluded for the toll charge.
Objectives and Criteria Proposed
While the previous section describes the ARZ alternatives,
the
purpose
objectives.
within
of this section is to identify
the
underlying
The set of objectives should provide a framework
which
the
The
evaluated.
performance of the
alternatives
can
be
objectives selected should be defined so as
to represent all those charateristics of an alternative which
may be considered desirable or undesirable.
The objectives should also include those characteristics
of
a project which may improve or lessen the
approval
and implementation.
criteria
for
Performance measures
evaluating the success with which
ARZ schemes meet the objectives.
of
these
likelihood
measures
include the
reflect the relevant objectives.
116
of
provide
alternative
The criteria for selection
accuracy
with
which
they
In formulating the problem
objectives should be
in a multicriteria context,
the
to
extent possible in the form
fullest
benefits.
A
method
each
express
quantified
of
costs
to achieve this quantification
objective
in
terms of a
set
of
is
and
to
measurable
criteria.
As was stated previously,
the development of objectives
for the evaluation of ARZ schemes is based on the surveys and
conducted with city officials in
interviews
to
decided
exclude
objectives
also
difficult to
are
were not considered here.
one
costs.
energy
deals
with
was
minimization
estimate
due
lowest
to
the
Goals related to secondary effects
scarcity of information.
Goal
and
It
as these objectives not only received the
but
ranking
the accident
Seoul.
Four major goals are
identified.
and
maintenance
minimizing
capital
Goal two implies reducing travel time for auto, taxi,
bus and subway.
Goal three focuses on minimizing undesirable
impacts of the transportation system such as pollution.
The
fourth
the
goal involves objectives which are responsive to
political
and
implemented.
institutional
These
system within which
goals,
along
criteria, are described below.
a
The
with
they
objectives
are
and
The goals here can be used as
general and abstract of desired transportation situations.
objectives
are specific and operational
goals.
117
statements
of
(i)
Minimize the net costs of improving the
of transport service
Goal:
quality
Objectives
Performance Measures
Minimize capital costs
Cost (monetary)
Minimize current operating
Costs
Maximize Revenues
Cost (monetary)
(ii).
Goal:
Monetary value
Improve the quality of transport service
Performance Measures
Objectives
OVTT saved for auto
OVTT saved for taxi
OVTT saved for bus
IVTT saved for auto
IVTT saved for taxi
IVTT saved for bus
(OVTT=Out-of-vehicletravel time
IVTT=In-vehicle-
Reduce travel time
for auto & taxi
Reduce travel time
for public transit
(iii).
Goal:
Minimize the undesi rable impacts
transport system
of
the
Objectives
Performance Measures
Reduce pollution
Pollution level
measured on ordinal
scale for auto, taxi
and bus.
118
Goal:
(iv).
Minimize
political
alternatives
obstacles
to
the
Objectives
Performance Measures
Minimize
institutional
obstacles
Institujtional
preference index
Minimize
problems
Enforceability
index
enforcement
(1) Capital Cost Borne by City
For most alternatives,
there are capital costs involved
include
may
fixed
and
the
alternatives
Since
transportation
system managemnt (TSM) type
major
cost
capital
alternatives
is
examined,
by
variable costs borne
government.
likely to
be
of
are
Purchase of parking meters
Purchase of public transit priority signals
Construction cost of toll gates at bridges
Cost associated with signs and markings.
119
actions,
involved.
the following costs
the
examined
capital cost:
-
components
The
constructing and purchasing the system.
in
For
city
are
no
the
considered
Having
ARZs,
discussed the types of capital costs relevant for
it seems appropriate to describe the capital costs for
each scheme in detail.
a. Off-Street Parking Fee Increase
No capital cost is necesary.
b. Extensive Parking Meters in Core
Parking meters would be purchased at W 160,000 ($200)
meter. five hundreads parking meters are needed.
per
Marking and
installation costs would be W 10,000 per meter.
c. Public Transit Priority Signals
The existing 13 electric signals would be converted into
the
public transit priority signals.
also needed.
Signs and markings are
City officials estimated the capital cost to be
W 5000,000 per priority signal.
d. Core Area License Scheme
Signs
and Markings would cost W 1,0000000
each entrance point.
entrance points.
($1,250)
at
Signs and markings are necessary for 13
Police staff estimated that five additional
120
patrol
cars
scheme
within
be needed to
would
effectively
enforce
this
Each vehicle would cost W
the core area.
7
million (Hyundai Automobile Manufacturing Company, 1983).
e. Core Area License Scheme with Bus Improvement
The area extent of bus improvement generally corresponds
to
that for area license within the core area.
from various parts of the metropolitan area.
core
transit
to
also be provided on routes leading
can
improvement
The
the
The
main
bus
type of service improvement considered is an increase in
a reduction in wait times
and
capacities
can
improvement
be
number of buses are needed.
This
and
Accordingly,
an
increased
Service would be provided every
This is a significant improvement given
five minutes.
the
buses.
"headways"
achieved by decreasing
the route structure.
improving
for
that
headway in the core area was between 10 and
average
minutes (Korean Institute of Science and
Technology,
15
1978).
The number of buses which would be required is estimated from
these
routes
peak-hour
obsorb
require
modal
a
associated
private
relevant
frequency
on
in the core area were doubled in order
to
characteristics.
service
If service
shifts from automobiles and
doubling
with
of
the
bus
fleet.
taxis,
Capital
a doubling of bus fleet would be
bus companies.
it
would
costs
borne
by
Capital costs in this sense are not
on the part of city government.
121
Put
differently,
costs to society as a whole are to be ignored.
f. Toll Increase for Tunnel Gates
The
capital costs to be considered include posting
sign boards.
tunnel.
New sign boards would cost W 800,000 per
new
each
Labor costs involved in erecting new sign boards are
included in this amount.
g. Toll Charge at Bridge Entrances
It
gates
was
would
previously,
estimated
by public officials that
be required for each
Seongsoo
machine
were
a
toll
As
was
bridge is excluded because toll
are already in operation.
constructing
bridge.
four
to
and purchsing a
be W
10,000,000
automatic
(based
discussion with officials in Transportation Division,
Overall,
stated
gates
The capital costs associated with
gate
estimated
toll
toll
on
a
1983).
the capital costs associated seven ARZ alternatives
are summarized as Table V-4.
122
Table V-4
Summary of City-Born Capital Costs of ARZs
ARZs
Capital
Costs (million won)
a. On-Street Parking Fee
0
b. Parking Meters in Core
80
*c. Public Transit Signals
65
d. Core ALS
48
e. Core ALS with Bus Improvement
41.5
f. Toll Increase for Tunnel Gates
4
g. Toll Charge at Bridge Entrances
160
(2) Current Operating Costs Borne by City
Associated
with each of ARZ alternative discussed above
are costs of maintenance.
current
In contrast to capital costs, the
operating costs are incurred after the system is
operation.
in
The actual costs would depend on how each alternative
implemented
and
structure.
For the sake of calculation, current operating costs
the way it fits in the
123
current
administrative
is
would be expressed on an annual basis.
a.
Off Street Parking Fee increase
Administrative Cost
administrative
Some
monitoring
this
cost includes
policy.
However,
the
notifying
these
and
types
of
administrative work can readily be handled by existing staff.
Thus administrative cost is not required.
Enforcement Cost
No
major
enforcement cost would be
Given
necessary.
that higher fees tend to put more press'ure on illegal parking
places elsewhere, some enforcement st-aff might be required to
It is assumed that enforcement
control the illegal parking.
costs due to the illegal parking are negligible.
b.
Parking Meters in Core Area
Administrative Cost
Two additional staff:
w 3 mil.
per person-year.
Enforcement Cost
It
was
enforcement
City
estimated by transportation
division
staff would be required to enforce this
government currently pays w 200,000
It would cost w 2.4 mil. per person-year.
124
per
that
30
scheme.
person-month.
it should be noted
the
are presently 10 enforcement staff in
there
that
core
area to control parking meters.
c.
Public Transit Priority Signals
Administrative Cost
City
signals
officials
can
be
noted
controlled
that
and
public
transit
monitored
by
priority
the
staff
currently available in the Traffic/Control Center in Seoul.
Enforcement Cost
Five enforcemnt staff for each system would be needed to
enforce this scheme.
Total of 65 staff for entire locations
would be paid at w 200,000 per person-month.
d.
Core ALS
Administrative Cost
(i)
Two additional staff would be required centrally to
control and monitor this scheme.
It would cost w 3 mil. per
person-year.
(ii) License printing also involves administrative cost.
Total
trips
that are likely to be affected by the
estimated to be 321,747 (see Appendix).
was
Since a modal shift
is most likely to occur after implementaiton,
125
ALS
it is
assumed
only 75 percent of total peak hour trips is taken
that
A licence would be sold every six month.
consideration.
multicolor
licence
1983).
into
A
license with adhesive backing would cost w 30 per
(interviewed with two printing companies
in
Seoul,
An estimated sale rate of 90% is assumed.
Enforcement Cost
person-year.
patrol
per
additional staff would be paid at w 2.4 mil.
Fifty
This
scheme also would require
cars each operating 100 km per day
5
additional
at w 100 per
km.
(This estimate is based on the information provided by police
Bureau
of Seoul).
needed
to
patrols.
effectively perform the
The
person-year.
e.
Also 10 additional police staff would be
duties
associated
with
police staff would be paid at w 3.6 mil.
per
315 working days are assumed.
Core ALS with Bus Improvement
Administratic Cost
It
was
necessary.
estimated that four additional staff
to
The Public
Transport
in Transportation Bureau needs two additional
control
be
As with the core ALS, two additional staff would
be used centrally to control the ALS.
Section
would
and monitor the bus improvement
staff would be paid at w 3 mil. per year.
126
program.
staff
Each
Enforcment Cost
Twould
that
estimated
officials
City
80
staff
enforcement
be necessary at the minimum level.
The duty
of
includes enforcing the ALS and the remaining 30
staff
involves
50
staff
enforcement related to the bus improvement program.
These staff would be paid at w 2.4 mil. per person-year.
enforcement
cost
essentially
the
concerning
same
as
patrol
for
the
car
Core
The
operations
ALS.
The
is
costs
associated with patrols are the same as for the Core ALS.
f.
Toll Increase for Tunnel Gates
Administrative Cost
No administrative cost would be necessary.
Enforcement Cost
No enforcement cost would be necessary
g.
Toll Charge at Bridge Entrances
Administrative Cost
Three
primary
staff
duty
would be
involve
needed for each
controlling toll
associated with toll collecitons.
at w 2.4 mil. per year.
127
bridge.
machines
Their
and
work
Each staff would be
paid
enforcement Cost
No enforcement cost would be necessary.
Table V-5
Summary of City Born Current Operating and
Enforcement Costs (million Won)
ARZs
Administrative
Cost
Enforcement
Cost
Total
Cost
Parking Fee
Increase
0
0
0
b. Parking Meter
6
72
78
c. Pulbic Transit
0
156
156
a.
Signals
d. Core ALS
19
171.5
190.5
e. Core ALS with Bus
25
243.5
268.5
f. Toll Increase for
Tunnels
0
0
0
g. Toll Charge at Bridges
7.2
0
7.2
128
... ....
c. Revenue to the City
A
that
major
impact of the pricing-related ARZ
they generate more revenues than costs.
schemes
Because
is
large
numbers of cars remain on the road, paying the charged price,
revenues
are quite large compared to capital and
costs.
The revenues can be generated from
proposed.
should
In
calculating
the
be taken into account:
revenues,
maintenace
five ARZ schemes
several
factors
the number of modal
shifts,
the number of time shifts and peak-hour factors.
b. Parking Meters in the Core
The
daily
currently
revenues collected from the
installed
in
parking
the core area amount to
w
meters
258,954
which is equivalent to w 1811 per meter (Parking Division
in
transportation
be
Bureau,
1983).
The annual revenues can
derived on the basis of these daily revenues.
d. Core ALS
The
likely
CBD bound trips during morning peak hours that
are
to be affected by the ALS was estimated to be 321,747
per day.
Since some drivers will switch modes,
change
the
time period of their trips, or eliminate the trip entirely, a
modal
shift of 25 percent was entered into the
129
calculation.
The
25 percent is also based on the results of a logit model
employed in deriving the travel times.
e. Core ALS with Bus Improvement
revenues
The
to
be generated
from
this
scheme
are
essentially the same as for the core ALS.
f. Toll Increase for Tunnel Gates
The
revenues resulting from the toll increase can
be
calculated based on the daily revenues presently collected at
The daily revenues collected at tunnel
tunnel gates.
gates
are presented in Table V-6.
Table V-6
Daily Revenues Collected at Tunnels
1st namsan
3638
3rd Namsan
5804
Geomwha
3660
Source:
Tunnel division in Transportation Bureau of Seoul, 1983.
In
the
which
order
to calcalate revenues to be collected
morning peak hours,
represents
it is necessary to apply
a
during
factor
a ratio of the trips moring peak hours
130
to
the trips during the entire day.
A ratio of 0.4 was used for
this purpose.
g. Toll Charge at Bridges
Average daily traffic volumes were used to calculate the
annual
revenue
The
revenues.
to
ratio
used for
morning
that for the entire day is 0.4.
traffic volumes for four bridges shown in Table
peak
Based
V-3,
hour
on
the
annual
revenues turned out to be 29,193 million Won.
Table V-7
Revenues Resulting from Five ARZs
ARZs
Annual revenues (mil. Won)
Parking Meters
285
Core ALS
76013
Core ALS with Bus Improvement
76013
Toll Increase for Tunnel Gates
4782
Toll Charge for Bridges
Table
V-7
alternatives.
24283
summarizes revenues resulting from five
The
magnitude of revenues appears to be
greater than capital and maintenance costs.
131
ARZ
far
Pr I "
.
T-1-11
I -
I.WI - I-
time
-
I'll,
-
the
In order
by imposing some restricitions.
insights concerning likely changes in
develop
;..
reduce
to
schemes proposed are designed
ARZ
The
.--
; I 17
Travel Time Saving for Auto, Taxi and Bus
(5)
(4) and
travel
'
to
time,
travel
one needs to predict travel flows in an equilibrican state in
transportation system.
the
The supply and demand functions
following
should provide a basis for predicting flows in the
flow
A
way.
system
and
V,
For
travelers.
activity
occur,
Lo,
pattern F,
system
level
the
a
of
L,
service
experienced
particular transportation system
A,
Fo = F(T,A),
determined
consists of the volume using
the
flow pattern which
is the volume Vo
will
and level of
T
the
by
and
actually
service
as the equilbrium solution to the supply and
demand relations:
L = S(T, V)
(Vo,
Lo)
V = D(A, L)
In other words,
T
and
the
activity
by specifying the transportation system
system A the
transportation can be defined.
supply
and
demand
for
The use of supply and demand
representation to reach a prediction of travel flow is called
132
.
an "equilibrium -solution".
In
view
of
the
equilibrium
Package 2.1 (TTP 2.1) was
Teaching
Transportation
of
prediciton
flow,
used
to
predict the equilibrium flow which leads to the estimation of
the
change
in
by
developed
TTP 2.1,
travel time.
and
Tsunokwan
a
Manheim
computer
pakage
(1981),
models
representation
and
demand using origin-destination flows and a choice model
for
supply
transportation
modal split.
an
through
network
Once the supply demand relations are
solution is approximated and travel
equilibrium
defined,
impacts
assessed on the basis of those solutions.
The particular demand model used in this analysis is
the
multinomial
decisions
of
logit type,
calibrated on observed
travel
corridor
individuals living in the Jongro
of
in
Seoul by maximum likelihood estimation (Jaimu Won, 1980).
The
general
form
of multinomial logit
model
expressed as follows:
V (x1 , st
e
P(i:A)
=
----------------
E.eVj (x
j A
,s
P(i:A) = probability that traveler "t" will select
alternative "i" from the set of available
alternatives A .
133
can
be
x
s
t
V (x ,
i
= a vector of attributes of alternatives "i"
(e.g., level of service variables)
= a vector of socioeconomic characteristics
describing traveler "t" (e.g., income,
occupation, etc.)
s ) = the utility of consumer "t" associated
twith the selection of alternative "i".
The
parameters of the utility functions
The independent variables
of the models travel choices mad'e.
out-of-vehicle-travel-time
The models are
(OVTT) and out-of-pocket-travel-cost (OPTC).
used
estimated
The dependent variables
using observation of actual choices.
are invehicle-travel-time (IVTT),
are
to predict the magnitudes of changes in travel behavior
(relative to present conditions) resulting from a
particular
change in independent variables from present levels.
Table
V-8
shows
the parameters of . the
logit
model
for the Jongro corridor.
Out-of-vehicle-travel-
time appears to be more onerous than
in-vehicle-travel-time.
calibrated
Travelers
living
around Jongro corridor are also much
more
sensitive to out-of-pocket-travel-cost
Table V-8
Mode
1
2
3
Logit Parameters for Jongro Corridor
Constant
0.00000
2.75000
1.83000
IVTT
-0.02600
-0.02600
-0.02600
OVTT/Dist.
-0.05700
-0.05700
-0.05700
OPTC/Income
-0.04200
-0.04200
-0.04200
Note:
IVTT: In-Vehicle-Travel-Time (includes roadway travel time
(zone-to-zone) for auto mode) or transit running time)).
134
OVTT: Out-of-Vehicle-Travel-Time (includes terminal time at
origin and destination for both auto and transit).
Out-of-Pocket-Travel-Cost (includes parking cost at
OPTC:
destination for auto mode and fare for transit mode).
Mode 1: Automobile
Mode 2: Bus
Mode 3: Taxi
network charateristics of the Jongro corridor such as
paths and volume-delay curves.
lines,
the
Link characteristics
and link
cost.
links over which the line is operating
headway.
line
links,
volume-delay curve for in-vehicle
out-of-vehicle time,
and
define
lanes,
distance,
include
time
using
transportation supply was modeled in TTP 2.1
The
Line *data
and
Path data are input by destination by
the
mode,
and they specify travel paths in terms of link-line pairs and
path costs.
An
incremental
assignment technique is
travel flow prediction in TTP 2.1.
for
used
for
the
The volume to be assigned
a particular iteration is divided according to the modal
split,
and
assigned to links on the specified minimum
for each mode.
The network serivce level is updated, and the
next iteration is considered.
TTP 2.1 appears
path
in
The overall procedure built in
Figure V-6.
135
Figure
V-6
Procedure built in TTP 2.1
Base case
Test of a
Policy
* LOS:
Level-of-service
136
using
Seven alternative ARZ policies have been analyzed
The results are presented in terms of the reduction in
TTP.
time for bus,
travel
taxi and auto.
are
Only three modes
examined in each case because of TTP's built-in computational
The
limitation.
on
representative of the Jongro
data
therefore,
sense,
strict
.Further,
cosidering travel charateristics
when
ring area,
generally
represent
Therefore,
1980).
availability,
parking
congestion,
the
the characteristics of the Jongro corridor
parts of the CBD (Jai Mu Won,
travel
CBD.
within
been shown to correspond to large western and
CBD.
a
are applicable only to the Jongro
in its broad charateristics to the rest of the
similar
have
in
and
corridor,
However, the Jongro corridor has been shown to be
corridor.
outer
estimated savings in travel time are based
southern
The levels of traffic
and
transit
uniform patterns throughout the
service
entire
the underlying assumption here is that the
characteristics
of Jongro corridor
are
similar
to
those of the entire CBD.
Table
ARZs
become
V-9,
shows the results of the travel times after
effective.
Annual time savings
presented
in
Table 5-10 are derived from simulated individual travel times
during the peak hours in Table V-9.
137
Table
V-9
Simulated Individual Travel Times During
Peak
Hours After ARZ Alternatives Using TTP2.1 Flow
Prediction Models
Out-of-Vehicle-Travel Times
Modes
Base
Bus
10
a
b
c
d
10
10
11
12
(Minutes)
f
e
9
10
11
Taxi
15
13
13
12
10
12
12
13
Auto
8
3
6
6
5
5
7
7
In-vehicle-Travel Times
Modes
Base
(Minutes)
a
b
c
d
e
f
g
Bus
38
38
37
34
34
34
35
35
Taxi
28
28
26
26
22
23
26
25
Auto
24
22
20
20
18
19
20
22
ARZ Alternatives:
(a) On-Street parking Fee Increase,
(b) Parking meters in Core, (c) Public Transit priority
Signals,
(d) Core Area License Scheme,
(e) Core ALS
with
Bus
Improvement,
(f) Toll Increase
for
Tunnel
Gates, (g) Toll Charge at Bridge Entrances.
Base:These are the travel times when no action is taken.
The simulated travel times derived from the logit models
are generally consistent with a priori expectation. The logit
models
changes
imbeded in TTP.2 turned out to be
in values of OVTT and IVTT.
138
sensitive
For example,
to
the
OVTT for
auto
decreases from 8 minutes to 3 minutes if alternative
parking
(On-Street
Fee Increase) were
to
be
1
implemented.
Area License Schemes apperar to be most effective in reducing
both OVTT nad IVTT for all modes.
Table
Three
for
Annual Time Savings
Simulated
Modes Using TTP2.1 Flow Prediction Models
5-10
Out-of-Vehicle-Travel Times
(Minutes)
Modes
a
b
c
d
e
Bus
0
0
+315
+315
-315
0
+315
Taxi
-630
-630
0
-1575
-945
-945
-630
Auto
-1575
-630
0
-945
-945
-315
-315
In-Vehicle-Travel-Time
Modes
a
Bus
f
g
(Minutes)
b
c
d
e
f
g
0
-315
-1260
-1260
-1260
-945
-945
Taxi
0
-630
+630
-1890
-1575
-630
-945
Auto
-630
-1260
+630
-1890
-1575
-1260
-630
* 315 working days per year are assumed.
increase in travel time after
"t" means
Note:
"-" means decrease in travel
effective.
become
effective.
become
after ARZs
Table V-10 was constructed from the Table
working
days
ARZs
time
V-9.
Annual
(315 days) are multiplied by teh differences
139
in
individual
travel
times between base case and each
proposed alternative.
indicate
the
of
Travel times listed under the
the
"Base"
current travel times taken from the origin
to
the destination when no action is taken.
(6) Reduction in Pollution
So
far,
we have only considered only those
objectives
for which monetary values and travel times can be objectively
and accurately assigned.
For the last three criteria, there
are no quantified data from which to derive the values.
For
this
the
reason,
one
has
to resort to the
decision
maker or expert.
measured
through
such
jridgement
of
The level of pollution can
measures as HC
and
CO,
but
be
these
measures may be very difficult or costly to obtain.
Given
easily
the fact that the citeria to be followed are
quantifiable,
weighting
were
method.
critically
approximate
Ackoff.
it
necessary
After the concepts of
examined,
weighting
The
seems
it
method
was
to
weighting
decided
advanced
employ
by
to
not
some
methods
use
an
Churchman
and
Churchman and Ackoff technique (CAT) has
been
described in numerous texts, and successful applications have
been
reported
in
the
field
140
of
operation
research
and
management
science
(Turban
and
Metersky,
1971;
Stimson,
1969).
The CAT consists of five steps to be carried out by
the
decision makers or chosen panel of experts (henceforth called
"decision makers"):
(1) ranking the objectives by importance level.
Let ol represent the outcome that is judged to be
the most important, o2 is the next, o3 the next,
and o4 the last ;
(2) Tentatively assign the value 100 to the most
valued outcome and assign values that initially
seem to reflect their relative values to the
For example, the deciiosn maker might
others.
assign 100,80,50 and 30 to ol,o2,o3 and o4 respectively.
Call these tentative values vl,v2,v3, and
These are to be considered as
v4 respectively.
first estimates of the true values of Vl,V2,V3, and
V4;
ol versus
(3) now make the following comparison.
if the decision maker had
(o2 and o3 and o4) i.e.,
the choice of obtaining ol or rhe combination of
o2,o3, and o4, which he would select? Suppose he
asserts that ol is preferable. Then the value of vl
ahould be adjusted so that V1 > v2 +v3 + v4;
(4) now compare o2 versus o3 and o4. Suppose o3 and
o4 are preferred. The further adjustment of the
values is necessary.;
be
may
It
are completed.
evaluations
(5)
convenient, however, to normalize these values by
dividing each by Tv.
and
In
order
to
provide
to establish a degree of relative
a
test of
weighting to follow,
government
The
public
consistency
in
the
numerical
the same public officials in Seoul city
were interviewed to make a series of
officials
importance
in
the
Urban
Department were asked to "role play",
141
Planning
comparison.
and
Police
assuming the point
of
view
of
thought
they
ratings
most
were
to
asked
on
ratings
to
give
provide
of
The public
the
the
a 100 value to the
officials
of
ARZ
They
were
effects
on the reduction of the pollution.
alternatives
requested
provide
to
The initial ratings were
as displayed in Table V-ll.
obtained
order
representative
Bureau's position.
Transportation
were
in
Transportation Bureau,
the
most
ARZ
effective
alternative and 0 value to the worst effective one.
Table
V-ll
Officials'
Public
Reduction
Reduction in
Pollution
Level
a
b
c
Values
23
36
51
Evaluation
d
e
f
100
75
44
on
Pollution
g
Alternatives:
(a) off-street Parking Fee Increase,
(c) Public Transit
(b) Extensive Parking Metera,
Priority Signals, (d) Core Area License Scheme, (e)
(f) Toll
Area License with Bus Improvement,
Core
Increase for Tunnel Gates, (g) Toll Charge at Bridges.
A
series
members
in
the
of comparisons were conducted with the
manner described
in
Table
to do comparisons because
V-12.
most
panel
It
was
public
extremely
difficult
officials
interviewed were not cooperative after a couple of
meetings.
The
evaluations
were tentatively
follows:
142
completed
as
vl
v2
v3
v4
v5
v6
v7
100
75
68
51
44
36
21
weeks
later,
Two
the
same officials
provide second set of ratings.
were
asked
to
The results of second ratings
showed that the consensus was in favor of v2 plus v-7 over
vl
100.
A
although
the
original
assessment was
96
versus
necessary adjustment was made such that the ratings coincided
with the majority decision.
75
to
80.
v2 was arbitarily increased from
Since panel members decided that v3 is
than the sum of v4 and v7, v3
greater
was increased from 68 to 74.
143
The CAT for the Evaluation of Pollution Reduction
Table V-12
Comparisons
Yes
vl>v2+v3
vl>v2+v4
vl>v2+v5
vl>v2+v6
vl>v2+v7
vl>v2+v3+v4
vl>v2+v3+v5
vl>v2+v3+v6
v2>v3+v4
v2>v3+v5
v2>v3+v6
v2>v3+v7
v2>v3+v4+v5
v2>v3+v4+v6
v3>v4+v5
v3>v4+v6
v3>v4+v7
v3>v4+v5+v6
v4>v5+v7
v5>v6+v7
Decision
No
Equal
2
2
4
7
5
0
0
0
1
3
1
6
0
0
0
5
12
0
2
10
14
13
19
9
9
16
16
16
15
12
13
10
16
16
15
11
4
16
12
5
0
1
2
0
2
0
0
0
0
1.
2
0
0
0
1
0
0
0
2
1
The next decision,
ratings.
Therefore,
described
previously,
ratings.
The
The
Majority Decision
(c:consitency
i: inconsistency)
c
c
c
c
i
c
c
c
c
c
c
c
c
c
c
c
i
c
c
c
the- third, confirmed with the second
final ratings,
are
based on the procedure
the same as the
final ratings were followed by
normalization
the summation of
second
set
normalization.
was done by dividing each value of v
i
v.
]
vl
v2
v3
v4
v5
0.246
0.197
0.182
0.126
0.108
144
v7
0.089
of
0.052
by
(7) Institutional Preference
feasibility
The
upon
depend
institutional
The
factors.
largely
will
alternative
an ARZ
of
willingness
and
alternatives,
ability of government to carry out certain ARZ
in general, are likely to be important factors in determining
implementability.
depend
also
The
feasibility
institutional
and
established
this,
In view of
1982).
and objectives (Jaimu Won,
goals
control
on the ability of the agencies to
ARZ alternatives in ways which meet the
operate
will
of implementation
policy
preference or attitude toward each ARZ
should somehow be reflected in the evaluation process.
ratings on instituitonal
As in the pollution criterion,
with
preferences
were
regard to ARZ alternatives
elicited
members
gave
ratings based on their own perspectives and experience.
The
the
from
panel
The
same public officials.
initial ratings were obtained as follows:
Institutional
Preference
Based
tests
were
susequent
a
b
c
d
e
f
g
86
78
62
49
42
100
35
on the initial ratings,
performed.
tests
After
the following consistency
adjustment
were conducted using the
Table V-13 shows the results of consistency
145
of
values,
adjusted
tests.
the
values.
Table V-13
Institutional Preference
Comparisons
Decision
No
Equal
Yes
---
--------------------------------------------------------------
vl>v2+v3
vl>v2+v4
vl>v2+v5
vl>v2+v6
vl>v2+v7
vl>v2+v3+v4
vl>v2+v3+v5
vl>v2+v3+v6
v2>v3+v4
v2>v3+v5
v2>v3+v6
v2>v3+v7
v2>v3+v4+v5
v2>v3+v4+v6
v3>v4+v5
v3>v4+v6
v3>v4+v7
v4>v5+v6
v4>v5+v7
v4>v6+v7
*
c:
Majority Decision
0
1
2
5
3
0
0
0
0
2
0
2
0
0
2
4
6
2
6
4
-
16
14
14
9
11
16
16
16
16
14
15
12
16
16
14
12
10
14
10
11
Consistent with initial
0
1
0
1
2
0
0
0
0
0
1
2
0
0
0
0
0
0
0
1
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
ratings
-------------------------------------------------
It
out
is
to
interesting to note that
be
institutional
viewed
of
comparisons
consistent
with
preferences
rated by public officials
the
initial
ratings.
preferences may change
factors such as pressures
the normalization
The
over time due
or education.
conducted.
to a host
The
follows:
vl
v2
v3
v4
v5
v6
v7
0.221
0.190
0.173
0.137
0.108
0.193
0.077
146
turned
should be
as opinions at a time when the survey was
Institutional
of other
all
result
(8)
Enforceability
the
same
obtained
ratings
initial
The
scheme.
ARZ
each
to
Police
enferceability
were more knowledgeable about the
respect
the
on
to
belonging
officials
Public
The
policies.
ratings
used to provide
were
members
panel
Department
are as follows:
Enforceability
a
b
c
d
e
f
g
91
69
61
47
32
100
56
occured
second ratings showed that inconsistency
The
for
transportation
success of ARZ types of
enforceability.
with
in
enforcement is one of the most critical factors
The
v3 versus v6 and v7 where v3 was preferred to the sum of
adjusted
members
With this
Thus v3 was increased from 69 to 77.
v6 and v7.
value,
voted
V5
majority.
the third ratings revealed that
the
in favor of v5 and v6 over vl by a
therefore,
was,
11
from 56 to
raised
panel
5
to
A
59.
consensus was reached at the fourth stage.
vl
100
V
Normalized
Values
0.215
Having
v2
91
0.196
v4
v3
62
77
0.165
0.133
v5
v6
59
45
31
0.127
0.097
0.067
derived the ARZ preference weights for
147
v
v7
each
of
the
three unquantifiable criteria,
weights
for
weights
calculate
twelve
it is also necessary
criteria
to
The
themselves.
for the twelve criteria are of particular importance
for conducting empirical analyses in the next chapter.
Since
there
are twelve criteria to
be
weighted,
the
method employed in deriving weights for the three criteria is
not particularly suitable.
Churchman and Achoff suggest more
relevant alternative procedure for case where there are
than
seven criteria involved.
more
The alternative procedure is
described as follows:
(i) Rank the entire set of criteria in terms of
preference without assigning quantitative values.
(ii) Selecet at random one criteria from the set.
Then, by random
Let cl represent this criteria'.
assignment subdivide the remaining set of criteria
into groups of no more tham five, and preferably
groups of approximately equal size. Each
into
outcome other than cl should be included in one and
only one group.
(iii) Add cl to each group and assign the value of
1.00 to it (vl = 1.00)
(iv) Use the method employed in deriving the rankings for unquantifiable criteria to obtain unstandardized values for the criteria in the groups
formed in step 3 of this procedure.
(v) Compare the rankings obtained from steps 2
through 4 of this procedure with those obtained by
the method employed in deriving the weights for
unquantifiable criteria. If the rank orders differ,
reconsider ranking and if necessary proceed again
from setps 2 to 4 of this procedure.
(vi) Once consistent results are obtained, normalize the values obtained in step 5 of this procedure by dividing the value assigned to each criteria by the sum of the values assigned to all the
criteria.
The
procedure
just
described can be
148
applied
to
the
derivation
of
weights
for twelve
of the criteria by public officials
ranking
The
criteria.
was
initial
requested.
An average of all the values for each criteria was calculated
for the sixteen public officials and these ratings become the
The initial rankings are shown
starting point of adjustment.
as:
Ranking
Criteria
7
6
5
12
11
2
108
1
9
3
4
cl
c2
c3
c4
c5
c6
c7
c8
c9
c10
cll
cl2
c2
is selected at random and the remaining criteria are
assigned at random to four groups as follows:
(c)
(d)
c6
cl
c3
c12
c9
cll
c1o
c4
c8
c5
c2
c2
c2
c2
(a)
(b)
c7
Using
weights
for
the same method as one employed in
deriving
the three unquantifiable criteria (as shown
Table V-li), The following initial results were obtained.
149
the
in
(a)
v7 =
(b)
.30
(d)
(c)
v6 = 2.50
vl =
vl2= 1.75
v9 = 3.50
vll= 1.75
vlO=
v4 =
v8 =
v5 =
v2 = 1.00
.00
v2 = 1.00
.75
v2 = 1.00
v3 = 1.10
.50
.25
.65
v2 = 1.00
The procedure described in Table V-li was used to
check
the consistencies with the help of the same public officials.
v7 > v12 + v4
(a)
v7 > v12 + v2
v7 > v4 + v2
v12> v7 + v2
V12> v4 + v2
V12> v7 + v4 + v2
V6
v9 + v8
v6
v8 + v2
V6
v9 + v2
V9
v6 + v8
V8
v6 + v2
Inconsistency
(b)
vi > vil
+ v5
vi > vil
+ v2
Vi > v5
+ v2
vll> v5
+ v2
(c)
.. 0 .
..
150
.
Inconsistency
vll> vl
+ v2
v2 > vl
+ vll
v3 > v10 + v2
+ v2
vl0> v3
(d)
v2 > v3 + v10
A series of comparisons were offered to the public offiTwo
cials.
inconsisencies
the summation
than
also
suggested
in
that
of v6 and v8 in group
v9 should be decreased
They
(b).
from
3.50
least
ten officals agreed that vll in group (c) should be at
same
increased
expressed.
as
from
the
.50
to
Of the sixteen offcials interviewed,
somewhere around 2.80.
the
and
(b)
group
Two-third of public officials felt v9 to be
(c)respectively.
less
occured
summation of vi
to
.75 and no
and
v2.
further
vl
was
objection
was
Thus
No consistencies occured at fourth interview.
It
was unnecessary to conduct further interview for this reason.
The final values obtained were normalized in Table V-15.
151
Normalized Weights for 12 Criteria
Table V-14
Criteria
It
Normalized Weights
cl
.056
c2
.073
c3
.080
c4
.000
CS
.019
c6
.182
c7
.021
c8
.056
c9
.202
c1O
.047
cll
.136
c12
.128
should
be noted that it was extremely
difficult
to
articu late weights of the public officials because they were
usuall
busy
with
routine
work and
were
not
willing
to
participate in the thinking process for deriving the rankings
They have little understanding about
the
rrequirements of the technique or little appreciation of
the
comparisons.
and
benefits which whould accrue from the interview.
The
that
Churchman-Ackoff technique employed has the advantage
the decisions required of the panel members are
straightforward and simple.
rather
On the other hand, the technique
152
be
must
from the fact that large number of comparisons
suffers
the
unless
made
the
Furthermore,
change over time.
number
of
small.
are
alternatives
may
values obtained at the final decision
Therefore, this tequnique should be viewed
as an approximate and tentative weighting method.
4. Conclusion
This
chapter has examined the relevant ARZ alternatives
District
and criteria within the context of Central Business
The ARZ alternatives and criteria developed
of Seoul, Korea.
in
this chapter will be used for an empirical analysis to be
conducted in next chapter.
central
The
determine
of this
first
part
of the chapter was
description of ARZ alternatives
locations.
proposed
and
discussed
third
of
the
quantifiable
travel
demand
criteria,
the
have
models.
devoted
chapter
was
to
the
153
types
chapter
criteria.
The
with
the
concerned
using
remaining
accounting
and
unquantifiable
Churchman-Achoff method was used
the weights.
the
Those criteria which can
been examined
For the
to
rationale,
The second part of
objectives and corresponding
the
part
including
quantification of criteria weights.
be
been
has
chapter
relevant criteria and ARZs in the context of Seoul
The
CBD.
objective
to
derive
EMPIRICAL ANALYSES OF THREE MULTICRITERIA METHODS
CHAPTER VI
1. Introduction
the ARZ alternatives and criteria advanced in
With
previous
chapter,
discuss
the
Concordance
the
underlying
Analysis,
to
objective of this chapter is
theoretical
of
structure
the
Goal-Achievement Matrix and
and
2)
to
conduct
the
1)
the
the
empirical
the
Compromise
Solution;
analyses.
The ARZ alternatives and criteria examined in the
chapter
previous
will provide a vehicle for conducting
the
foundation
for
analysis.
empirical
discussion
of
This will provide the
limitations and advantages of
three
methods
which is the subject of the next chapter.
2. Concordance Analysis (CA)
a. Theory
This
by
Roy
section examines a concordance analysis
(1968)
and subsequently modified by Van
154
developed
Delft
and
Nijkamp
(1977),
and Guigou
(1971).
method is used for selecting one alternative from a set
This
of
Nijkamp and Vos
(1977),
a weight which
assigned
is
criterion
relevant
with respect to
alternatives
Each
criteria.
corresponds
its
to
relative importance.
analysis
concordance
The
denoted by a ,
sets,
alternative
number
starts with a
i =
1,
...
n.
of
n
These
i
are evaluated against a set of criteria,
alternatives
=
k
expressed
as
represents the outcome
a typical element v
where
ki
th
criteria can
outcomes of
The
alternative.
i
of
c ,
c ,
2
1
follows:
c .
m
...
V =
effect
An
. . . . .
V1l
.
.
.
vln
.
.
.
v
matrix can be
.
v
.
.
.
.
.
mn
ml
the
c
expressed
any
in
quantifiable
reasonable
unit
of
measurement.
and non-quantifiable criteria could be
in this matrix.
155
be
Both
entered
step is to provide a weight corresponding
next
The
The weight is determined by the preference
each criterion m.
The weight vector can be
of the decision makers.
structure
to
represented as follows:
W
=
where
,w
.......
[w ,
k
m
1
k = 1
..
each
w
.
.
.
.
.
reflects
k
attached to criterion m.
.
. m
the
weight
These weights can be represented on
interval scale,
a ratio scale,
preference
relative
or even cardinal scale.
In
this way, the method offers considerable flexibility in using
non-quantitative
information.
-
information
as
well
as
uncertain
It is assumed that the weights are normalized,
A weghted effect matrix can be
1.
k
its
with
v
multiplying each element
by
obtained
ki
corresponding relative weight wk.
i.e.,
that
The
W
next step is to make a pairwise comparison
between
The preference relationship between
competing alternatives.
th
th
sets of n alternatives can be established for
and j
the h
th
th
j
and
h
the
The relationship between
each criterion.
alternatives
for
a given criterion can
follows:
156
be
represented
as
h preferred to
j
h and
The
j
; h >
j;
j
equivalent ; h =
associated with pair
CS
set
concordance
j)
(h,
is
hj
defined as follows:
CS
(k
=
v
>
)
v
kj
ki
hj
One subset consists of all criteria k for which alternative h
is
preferred to
to
i
alternative
j.
The symbol
alterntive
j,
> denotes a
preference
of
m
is
as far as criterion
concerned.
concordance set is expressed in terms
The
values
to
respect
with
alternatives
and
of
absolute
The
criteria.
relative value of the concordance set is measured by means of
an index.
the
The concordance index is then equal to the sum of
preference
weights defined above.
which ranges from 0 to 1,
all
weights.
Thus
To obtain an
the index is divided by the sum of
the concordance index for any
alternative sets, h and
j
index
pair
can be expressed as follows:
157
of
Wk
k 6 Chj
CIij
=
-------------
m
Wk
It is assumed that weight vector w are normalized, i.e.,
wk=1
k=l
Then, the concordance index is generally expressed as follows:
CIhj
=
Wk
(k
C
hj
This is the sum of weights with respect to those criteria for
158
outranks alternatives
a
h
measure of the degree to which the
alternattive
which
providing
a
a
thus
,
preference
over a
are in agreement with the dominance of a
h
The following relationships hold in concordance analysis.
weights
(i) if
(ii)
if
The
Chj = 1,
there is a complete dominance
of all criteria of h with
respect to j.
Chj = 0,
there is no doxfiance for any
criterion.
indices derived can be expressed in matrix form
of
j
= 1,
respect
to
CI with with n x n.
...,
n,
includes
Each entry CI
the
,
h = 1,..., n;
hj
concordance index with
pairwise comcparison of a with a
j
h
159
r
..
,..CI ,.....
12I
13
12
. . . . . . . . . . .
CI
..
C
iln
.
0
CI
2n
21
CI=
CI
,
n2 .
nl
next
The
discordance
n,n-1
CI
.
.
.
.
CI
step is to examine the
between
the various
relative
alternative
of
degree
This
pairs.
h
measures the degree to which outcomes of alternatives a
worse
than those of a
.
As was the case with concordance
a discordance index will be established.
analysis,
to establish the discordance index,
to all k criteria can be defined.
scale
of
This is done by defining a
100 is assigned to each criterion (the choice of
and
any
other
In order
an interval scale common
such that a certain number of points out of a
arbitrary,
are
maximum
100
number will work equally
is
well).
The choice of the number of points to assign to each criteria
depends
on the level of importance the decision makers
attach to the range between the best and worst levels
hj
is expressed
The discordance set DS
each criterion.
to
follows:
160
wish
of
as
hj
DS
symbol
The
kj
kh
=
(k
v
<
v
)
The
< denotes "not preferred to".
discordance
index can be expressed as follows:
kj
kh
v
DIhj = max
( ktDhj)
-
v
--------------------kmax
DI
the
where
expression
kh
v
kj
v
-
the
denotes
absolute
j
i
between
differenc
the effects of alternative a
and
following relationship holds in discordance anaysis.
(i) if dhj = 1, there is maximum discordance
between two alternatives.
(ii)
if
dhj = 0,
there is minimum discordance
between two alternatives.
161
a
.
The
A discordance matrix can thus be established as follows:
ln
12
.
DI
.DI
.
.
.
2n
21
.
.
DI
DI
DI
defining
a
and
only
criterion.
.
.
.
.
DI
is
to
eliminate
(1968) and Delft
elimination
critical
value
can
for
less
(1977) have
proceed.
both
favourable
shown
two
by
Firstly,
concordance
and
the average value of all
h
j
if
a
outranks
a
alternative
one may statethat
hj
hj
with respect to each
< d
> c and d
if c
discord ance
indices) ,
Roy
.
n,n-1
DI
step
which
by
.
..
next
alternatives.
methods
.
n2
nl
The
.
.
indices (in general,
These relatoinship can be expressed as follows:
162
hj
> c
c
hj
>~
d
Alternatively, one may construct a zero-one concordance
hj
dominance matrix G with elements g , which can be expressed
as follows:
hj
= 1,
if
hj
= 0,
g
if
g
Similarly,
hj
c
hj
c
a zero-one dis/c ordance dominance matrix Y with
hj
can be constructed as follows:
elements h
hj
= 1, if
h
hj
= 0, if
h
If
the
second
hj
d
hj
d
method
described
163
above
used,
it
is
to calculate the aggregate dominance matrix G and
hj
of T
The elements t
discordance dominance matrix Y.
necessary
satisfy the folowing conditions:
hj
= 1, if
t
hj
hj
g
1 and y
=
= 1
hj
t
=
h
then alternative a is favoured over alternative
hj
= 1,
If t
a
0, otherwise
The final
both for concordance and discordance criteria.
strategy
selection
nondominated solution,
find
to
is
alternative
hj
kh
=
0,
k = 1,
with
a
namely, an alternative that satisfies
the following conditions:
t
a
...
,
n; k /
164
h; k
7
j
Then, based on the relationships examined, the conditions can
also be expressed as:
hj
kh
< c < c
c
kh
hj
d
< ~c< d
b. Implementation of Concordance Analysis
As was stated in the theory section, an effect vector or
impact
matrix
can be calculated based on information
which
alternatives.
The
outcome of an effect vector is presented in Table VI-1.
The
reflects
criteria
all
relevant outcomes of
all
are measured in such a way that a high value of
impact is preferred to a low value,
an
except criteria 1 and 2,
for which a low value is preferred to a high value.
Table
out
the
VI-1 includes the information necessary to
concordance
anaysis.
165
The
implementation
carry
of
a
analysis requi'res transformation of
concordance
matrix into dimensionless units.
the
impact
The normalization procedure
gives
rise to the normalized impact matrix as shown in Table
VI-2.
Each criterion is divided by the maximum value of that
criterion.
Maximum value of 160 and 1268.5 were used for the
criteria -
capital and maintenence costs.
For the criteria
10, 11, 12 shown on Table VI-1, the normalized values already
obtained in chapter V were used.
Having
established a normalized impact
it
is
asdescribed earlier to specify weights in order to
necesary,
The need to specify a weight
perform a concordance anaysis.
vector
matrix,
is
considered
one
of
the
major
drawbacks
concordance anaysis because of the difficulties and
of
inherent
arbitariness in choosing weights.
A set of weights,
from
denoted by row vector w, was obtained
the sixteen public officials discussed in the
chapter.
This.set is presented as follows:
1
w = ( .056
2
.073
4
3
.080
.000
8
5
6
7
.019
.182
.021
10
11
12
.047
.136
.128
9
.202
.056
166
)
previous
Using the normalized impact matrix and a set of weights, a
concordance matrix can be calculated.
needed
at this point.
Table
VI-3.
These
Concordance pairs are
Concordance pairs are
calculated
are
for
presented
pair
each
in
of
means that preference relationships
th
are
sets of n alternative sets
j
This
alternatives.
th
h
between
and
established for each criterion.
Impact Matrix before Normalization
Table VI-1
7
1
2
3
4
5
1.Capital
cost
0
80
65
48
41.5
2.Operat.
cost
0
78
190.5
268.5
14.4
4782 24283
-315
-315
3.Revenue
4.0 VTT for
auto
5.QVTT for
taxi
6.OVTT for
156
0
-1575
285
-630
0
0
76013
-945
76013
-945
-630
-630
0
-1575
-945
315
315
-315
0
0
bus
7.IVTT for
auto
8.IVTT for
6
4
160
7.2
-945
0
-630
315
-630
-1260
630
-1890
-1575
-1260
-630
0
-630
630
-1890
-1575
-630
-945
0
-315
-1260
-1260
-1260
-945
-945
taxi
9.IVTT for
bus
10.Reduction
.052
.089
.126
.246
.197
.108
.182
pollution
ll.Institu.
.190
.173
.137
.108
.093
.221
.077
preference
12.Enforce-
.196
.165
.133
.127
.067
.215
.097
ability
.5
non-
Source:
This
quantifiable
table
incluaes 9 quantiriaole ana
with respect to each of
impacts
alternatives.
Capital costs operating costs and revenues are
167
the
7
based on table V-4,V-5 and V-7 respectively. The criteria
relating to travel time come from table 5-10. Criteria 10,11
and 12 are based on the Churchman and Ackoff technique in the
chapter 5.
* ARZ Alternarives: 1. Off-Street Parking Fee Increase
2. Parking Meters in Core Area
3. Transit Priority Signals
4. Core Area License Scheme (ALS)
5. Core ALS with Bus Improvement
6. Toll Increase for Tunnel Gates
7. Toll Charge at Bridges
Table VI-2
Normalized Impact Matrix
al
a2
a3
a4
a5
a6
cl
.000
.500
.406
.300
.259
.025
1.000
c2
.000
.291
.581
.709
1.000
.054
.027
c3
.000
.000
.000
1.000
1.000
.063
.319
c4
1.000
.400
.000
.600
.600
.200
.200
c5
.400
.400
.000
1.000
.600
.600
.400
c6
.000
.000
-1.00
-1.00
1.000
.000
-1.00
c7
.333
.667
-. 333
1.000
.833
.667
.333
c8
.000
.333
-. 333
1.000
.833
.333
.500
c9
.000
.250
1.000
1.000
1.000
.750
.750
c1o
.052
.089
.126
.246
.197
.108
.182
cll
c12
.190
.196
.173
.165
.137
.133
.108
.127
.093
.076
.221
.215
.077
.097
in
table VI-1
Source:
Each
max imum
value of that criterion.
criterion
168
is
a7
divided
by
the
Table VI-3
Cl2
C13
C14
C15
=
=
=
=
C16 =
C17
C21
C23
C24
C25
=
=
=
=
=
C26 =
Concordance
Pairs
(sets)
(cl,c2,c4,c5,c6,cllc12)
(clc2,c4,c5,c6,c7,c8,cllc12)
(cl,c2,c4,c6,cllc12)
(cl,c2,c4,cll,c12)
(cl,c2,c4,c6)
(cl,c2,c4,c5,c6,c7,cll,c12)
(c7,c8,c9,clO)
(c2,c4,c5,c6,c7,c8,cllc12)
(c2,c6,cll,c12)
(c2,c6,cll,c12)
(c4,c6,c7,c8)
C27 = (cl,c4,c5,c6,c7.cll.c12)
C31 =
C32 =
(c9,clO)
(cl,c9,clO)
C34 = (c2,c9,cll,c12)
C35 = (c2,c9,cll,c12)
C36 =
C37
C41
C42
C43
C45
C46
C47
C51
C52
C53
=
=
=
=
=
=
=
=
=
=
C54 =
C56
C57
C61
C62
C63
C64
C65
C67
C71
C72
C73
=
=
=
=
=
=
=
=
=
=
=
C74 =
C75 =
(c9,clO)
(clc9,cll,c12)
(c3,c5,c7,c8,c9,clO)
(cl,c3,c4,c5,c7,c8,c9,clO)
(cl,c3,c4,c5,c6,c7,c8,clO)
(c2,c3,c4,c5,c7,c8,c9,clOcllcl2)
(c3,c4,c5,c7,c8,c9,clO)
(cl,c3,c4,c5,c7,c8,c9,clOcllcl2)
(c3,c5,c6,c7,c8,c9,clO)
(clc3,c4,c5,c7,c8,c9,clO)
(cl,c3,c4,c5,c6,c7,c8,clO)
(cl,c6)
(c3,c4,c5,c6,c7,c8,c9,clO)
(cl,c3,c4,c5,c6,c7,c8,c9,cl)
(c3,c5,c7,c8,c9,clO,cll,c12)
(cl,c2,c3,c5,c9,clOcllc12)
(clc2,c3,c4,c5,c6,c7,c8,cllcl2 )
(cl,c2,c6,cll,c12)
(cl,c2,cll,c12)
(cl,c4,c5,c6,c7,c9,cll,c12)
(c3,c8,c9,clO)
(c2,c3,c8,c9,clO)
(c2,c3,c4,c5,c6,c7,c8,clO)
(c2,c6)
(c2,cll,c12)
C76 = (c2,c3,c8,clO)
(c2
is
One subset, called the concordance pair(set),
Note:
preferred
composed of all criteria for which alternative h is
over alternative
j.
169
concordance matrix.
will be applied to
pairs
concordance
These
a
This matrix, denoted by CI, is presented
concordance index CI 16 can be
For example,
in Table VI-4.
derive
obtained through the concordance pair of cl6(clc2,c4,c6)
in
Table VI-3.
c16 =
(cl,c2,c4,c6)
CI 16 = (.056,
.594
1
.671
.575
.393
.311
.797
.615
.519
.519
.259
.542
.539
.539
.249
.522
.762
.425
.745
.607
.663
2
.326
3
.249
.305
4
.425
.481
.461
5
.628
.481
.461
.238
6
.689
.741
.751
.575
.393
7
.385
.458
.478
.255
.410
In
.182) = .311
Concordance Matrix
Table VI-4
CI=
.000.
.073,
addition
to
a concordance
170
.744
.256
matrix,
a
discordance
matrix
should
concordance
pairs.
be calculated as well.
matrix,
it
As in
the
is necessary to derive
Discordance pairs are shown in Table VI-5.
171
case
of
discordance
Discordance Pairs
Table VI-5
D12
D13
D14
D15
D16
D17
D21
D23
D24
D25
D26
D27
D31
D32
D34
D35
D36
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
(sets)
(c7,c8,c9,clO)
(c9,clO)
(c3,c5,c7,c8)
(c3,c5,c7,c8,c9,clO)
(c3,c5,c7,c8,c9,cl0)
(c3,c8,c9,clO)
(cl,c2,c4,c5,c6,cll,c12)
(cl,c9,clO)
(cl,c3,c4,c5,c7,c8,c9,clO)
(cl,c3,c4,c5,c7,c8,c9,cl0)
(cl,c2,c3,c5,c9,cl0,cll,c12)
(c2,c3,c8,c9,cl)
(cl,c2,c4,c5,c6,c7,c8,cll,c12)
(c2,c4,c5,c6,c7,c8,cll,c12)
(cl,c3,c4,c5,c6,c7,c8,clO)
(cl,c3,c4,c5,c6,c7,c8,cl0)
(cl,c2,c3,c4,c5,c6,c7,c8,cll,c12)
D37 = (c2,c3,c4,c5,c6,c7,c8,clO)
D41
D42
D43
D45
D46
D47
D51
D52
D53
D54
D56
D57
D61
D62
D63
D64
D65
D67
D71
D72
D73
D74
D75
D76
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
(cl,c2,c4,c6,cll,c12)
(c2,c6,cll,c12)
(c2,c9,cll,cl2)
(cl,c6)
(cl,c2,c6,cll,c12)
(c2,c6)
(cl,c2,c4,cll,c12)
(c2,c6,cll,c12)
(c2,c9,cll,c12)
(c2,c3,c4,c5,c7,c8,c9,clOcll,c12)
(cl,c2,cll,cl2)
(c2,cll,c12)
(clc2,c4,c6)
(c4,c6,c7,c8)
(c9,clO)
(c3,c4,c5,c7,c8,c9,cl)
(c3,c4,c5,c6,c7,c8,c9,clO)
(c2,c3,c8,clO)
(cl,c2,c4,c5,c6,c7,cll,c12)
(cl,c4,c5,c6,c7,cll,c12)
(clc9,cll,cl2)
(clc3,c4,c5,c7,c8 ,c9,clOcll,c12)
(cl,c3,c4,c5,c6,c7,c8,c9,clO)
(clc4,c5,c6,c7,c9,cll,c12)
One subset, called the discordance pair(set), includes
Note:
preferred over
criteria for which alternative h is not
all
alternative
j.
172
Instead
in
Table
of using the normalized impact matrix presented
VI-2,
a weighted impact matrix can
be
used
for
calculating the set of discordance indices.
This is done by
mn
multiplying each element v
in normalized impact matrix with
m
its corresponding relative weight w
Table VI-6
al
Weighted Normalized Matrix
a2
a3
a4
a6
aS
a7
cl
.000
.028
.023
.017
.015
.001
.056
c2
.000
.021
.042
.052
.073
.004
.002
c3
.000
.000
.000
.080
.080
.007
.034
c4
.000
.000
.000
.000
.000
.000
.000
c5
.008
.008
.000
.019
.011
.011
.008
c6
.182
.182
.000
.000
.364
.182
.000
c7
.007
.014
-. 007
.021
.017
.014
.007
c8
.000
.019
-. 019
.056
.047
.019
.028
c9
.000
.051
.202
.202
.202
.152
.152
c1o
.002
.004
.006
.012
.009
.005
.009
.024
.021
.019
.017
.015
.016
.013
.009
.030
.028
.010
.012
cll .026
c12 .025
Table
VI-7
predominantly
contains
discordance
values above .500.
173
indices,
showing
Discordance Matrix
Table VI-7
1
2
.336
1
3
4
5
6
7
1.00
1.00
1.00
.752
.752
.748
.800
1.00
1.00
1.00
.500
.688
.500
.627
1.00
.938
.685
1.00
.973
2
3
.500
.575
.507
4
.712
.500
.250
5
1.00
.712
.425
.421
6
.055
.000
.248
.913
.913
7
1.00
.875
.589
.696
1.00
.982
are
derived
DI
Note:
Discordance
indeces(DI)
.338
discordance pairs presented in table VI-5.
174
from
the
Then zero-
where both inequalities have to hold.
<d,
>c
hj
3
h
and d
c
if
if and only
outranks a
that alternative a
can
one
d of concordance and discordance respectively,
state
of
elimination
By setting critical values c
favourable alternatives.
less
and
step to be undertaken is the
next
The
hj
one
concordance
are
constructed.
anc
discordance
and discordance dominance matrices G and
concordance
The threshold values for the
others.
values some of the alternatives are preferred to all
Aggregation
concordance
(T) was obtained by inter'secting the
and discordance matrix Y.
matrix G
these
given
indices are assumed such that
Y
Using the elements of T,
the undominated alternatives can be identified.
threshold
The
and
discordance
values
of
"satisfying"
value for the concordance index of
and d are,
c
effect,
in
Now,
manner.
one
arbitary,
can
alternatives by checking each row od T.
alternatives
contain
one
element
The
used.
were
index of .300
equal
identify
threshold
set
in
a
dominated
In Table VI-8,
to
.700
unity.
two
The
alternatives corresponding to those rows should be eliminated
because a value of 1 indicates that the alternatives are less
favourable.
The remaining undominated alternatives are the
desired output.
175
Table VI-8
0
Determination of G, H and T by .700 and .300
0
0
0
0
G=
0
0
1
0
0
0
0
0
T=
0
0
0
0
110
0
0
0
0
0
1
0
0
0
0
0
1
0
0 1
0 0
1
0
0
0
0
0
00
1 00
0
0
0
0
0
0
0
d
0
0
1
_ .0
0
0 U
1 0
0
0
0 0
0 0
00
0
0
0
0
0
0
The
too
0
0
0
0
0
0
00
H=
0
0
1
0
01
00
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0 0
0
0
0
0
remaining undominated alternatives
were,
however,
large to provide any meaningful basis for selecting
desired alternative.
decrease
c
Thus it was deemed necesary to
or increase d.
decrease c to .500 and increase d to .500.
presented
in Table VI-9.
either
After examining the indices
both concordance and discordance matrices,
Of seven ARZ
176
the
in
it was decided to
The results
alternatives,
are
four
alternatives
contained
alternatives
were
to
equal
elements
1.
The
implying that the
eliminated,
four
remaining
three alternatives warrent closer scrutiny and evaluation.
Table VI-9
Determination of G, H and T by .500 and .500
11 1 0 0
1
0
0
0
1
1
[Q
0
0
0
0
1
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
_
0
0
0
0
0
0
0
0
1
0
0
0
1
1
0
0
11
11
00 d
0
0
0
0
1
1
19
0
1
0
1
i
1
1
0 0
0
0
0
0
0
1
0
0
0
1
177
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
1
0
the use of varired values of c and
Through
to greatly reduce the number of alternarives.
possible
c and d was varied by restricting
of
value
it
d,
the
only to current undominated subset of alternatives.
parameters
of .300 for the concordance matrix and
the discordance matrix were used,
was
The
operations
When new
.700
for
only two alternatives were
The two undominated alternatives were: the Core
undominated.
Area License Scheme and the Core Area License Scheme with Bus
Improvement.
Determination of G, H and T by .200 and .700
Table VI-10
1
T
G=
1
1
1
1
1 1
1 11
1
1
0
1
1
1
1
1
1
11
1
0
T
1
1
1
1
i T
1 1T
111
1 1
T
0
0
0
1
0
1
T
1
0
0
T
l1
T
0
0
0
0
1T
1 00
1 1
0
0
0
0
0
0
0
1
0
0
0
1
1
T
H=
0
0
1
1
1
0
1
178
1
0
0
T
0
0
1
0
0
0
1
1
110
0 0 0
0 1 1
1
0
0
0 0 1
0 0
T
1
0 0
0
0
index
discordance
the
and
index
values for
threshold
When
alternatives.
of
concordance
the
As shown in
altertnative
(Core
5
Area
License
IVTT for the bus mode,
The
weights.
high
both of
final
the
other reason for
seems to be that the ALS with Bus Improvement
outcome
been
have
which
not
require any municipal expenditures associated
bus
improvement.
a
reducing
expectation as the ALS is very effective in
assigned
Bus
with
Scheme
outcome is generally consistent with
This
Improvement).
and
to
the sole undominated alternative turned out
Table VI-10,
priori
.700
and
.200
one alternative was undominated.
respectively,
OVTT
ARZ
schemes appear to be more favourable than other
license
be
area
findings based on this analysis sugggest that
The
would
the
with
The bus improvement costs are to be borne
While the capital and maintenance
by private bus companies.
costs required for instituting ALS schemes seem rather
high,
remaining criteria are more than enough to offset
these
the
Despite the low weight scores on the
cost-related criteria.
preference
institutional
and enforceability,
the two
ALSs
outranked the other ARZ alternatives on criteria relating
be
should
examined
is
information
framework
and
revenue
that
are
generally
important.
considered
It
pollution
time,
travel
to
emphasized that the
intended
as
to
provide
concordance
analysis
and
explicit
empirical
the basis for decision making and
that
imbeded in the concordance analysis is helpful
179
the
in
. .1.-W
structuring
the
decision
making
process.
the
However,
results of the analysis should be critically discussed rather
Viewed in this way,
than simply adopted.
it may be wise to
use the results of the analysis as a tool to narrow the range
of
ARZ alternatives.
best
ARZ
will
Ultimately the final selection
through
be accomplished
of
negotiation
a
among
implementing agencies, planning agencies and interest groups.
are
There
to
ways
several
a
conduct
sensitivity
analysis.
The sensitivity analysis is performed to test the
robustness
of
the
in
weights,
(2) variations in critical
(1) varia-
values,
be established which. stresses criteria other than
heavily in the original anlysis.
weighted
maker
wishes
criteria,
weights
weight
and
interactions
1,
priority
do
The
have an impact
same
with
to
the
finalcial
given
criteria.
on
of
180
high
Therefore,
the
selecting
criteria.
decision makers are required
various sensitivity analyses.
those
decision
is true for variations in
-changes in the number
(3)
If the
2 and 3 should be
to other remaining
relative
changes
assign high
criteria
The
alternative.
values
to
to
For example, a scenario
increase in the number of criteria.
can
respect
with
The several ways are:
changes in some parameters.
tions
analysis
concordance
best
critical
Numerous
to
conduct
3. Goal-Achievement Matrix (GAM)
a.
Description of Theory
The GAM method of evaluation was advanced by Morris Hill
1967;
(Hill
Miller 1980; Hill and
evaluating public projects (Hill 1971;
1972;
Tzamir
disaggregate
1974).
and Alterman;
Hill
treatment
benefit
of cost and
for
used
Hill 1973) and has been
1968;
Hill
The
allows
GAM
accounting
by
objectives, and by incidence of goal achievement throughE 2 the
relevant population.
The
method retains qualitative data in the accounts and
explores simple alternative weightings for the objectives and
differential impacts.
their
realism
full
can
The method's
contribution
the
of
be briefly summarized as a recognition
to
subjectivity of any formal weighting process and of the
concomitant
disaggregate
provided
need
to
find a balanced
information.
The
means
decision
of
presenting
may
makers
be
figures
with both a variety of alternative summary
and the full set of underlying disaggregated accounts.
The
inherent
GAM
in
traditional
because
essentially attempts to
cost-benefit
analysis.
overcome
Hill
limitations
criticizes
cost-benefit evaluation of transportation
non-economic
objectives
181
are
neglected.
He
the
plans
also
criticizes
multi-criteria
methodologies
which
attempt
to
relate all categories of impacts to a common scale.
Thus one of the limitations of the development of balance sheet is that it does not
always appear to rec gnize that benefits and
costs have only ins mental value, that they
have meaning only in relation to a welldefined objective. A criterion of maximizing
net bemefits in the abstract is meaningless.
Whereas benefits can be computed referring
to different planning objectives, the benefits and costs are not necessarily additive
or comparable (Hill, 1973).
proposes a methodology in which a
Hill
benefit
analysis
cost-
separate
would be constructed for -each
goal,
each
goal weighted, and then the degree of achievement of weighted
goals compared for each alternative.
The
first
step
in
Hill's
is
methodology
the
determination of goals, which are defined as "an end to which
a
is
course of action
planned
directed".
Goals
include
increased accessibility, reduction of displacement, separated
pedestrian
and vehicular traffic,
and minimized traffic congestion.
to goal selection,
approaches
with
minimized air
Hill suggests a number of
among which are
elected officials and community groups,
sampling,
pollution,
consultation
public opinion
and examination of previous allocations of
Hill
investment.
believes that in practice,
public
the number of
important goals will be relatively small.
Once
goals are selected,
alternatives
the impact of
on the goals are examined.
182
transportation
Up to this
point,
Hill's
Goals-Achievement Matrix methodology is similar to
standard
environmental impact statement.
are identified
a
Important
issues
(goals) and the future impacts of each
alter-
native are estimated.
But
Hill
consequences
make
of
these
Hill
time
to
content
alternative actions,
consequences
presentation
driving
not
is
states
report
the
his methodology
must
merely
comparable.
that
Early
benefits
such
in
as
his
reduced
cannot be directly added to the costs
of
air
solution to this problem of comparability is
to
pollution.
Hill's
aggregate similar impacts within each goal category and
For example,
compare the results for different alternatives.
the
air quality impacts of highway alignment A at
locations
air
with a
is
to
total
B.
One
could
and
the
loser
impacts
of highway
alignment
better alternative with a "+"
the
A more rigorous approach, which Hill describes,
"-".
convert
numerical
different
would be added together and compared to the
quality
indicated
then
scale
the
aggregated air
quality
impacts
which is referred to as its degree of
to
a
goal
achievement.
Alternatives
specific goals.
can
now
be
compared
with
respect
"A" may be superior to "B" with
to
respect to
the
goals of air quality and residential displacement,
but
"B"
may be superior to "A" with respect to travel
and
accident rates.
time
If alternative selection is to be made on a
183
rational
basis,
the dissimilar categories must be compared.
It would be naive to assume that all goals are equally valued
by the community.
Hill recommends using similar
and consultation techniques to weight the goals
survey
respect
each other.
Since "the public interest"
to
weighting
the
of
we
alternatives,
with
does
a bodily form with the ability to answer our
take
as
Therefore,
not
questions
must
somehow
aggregate the goal preferences of all those consulted.
Hill
of
does not describe how to weight the importance
consulted party other than to say it is difficult,
each
and
that "relative valuation of objectives by the community ought
to be approached from a particular theory of government.
Once
each alternative has its goal
achievement
score,
and each goal is properly weighted, then,
"the combined weight of the objectives and their
incidence is assigned to the measures of achievement of the objectives.
The weighted indices of
goals-achievement are then summed and the preferred
plan among those alternatives compared is that with
the largest index (Hill, 1973)."
By consulting the public and decision makers as to their
favorite
the
goals
alternative
chosen
without
and determining how they order those
which best serves the
public
any political interference.
goals,
is
interest
It is
a
very
rational process.
The
GAM
reaches
out
to the
diversity of evaluation criteria,
184
community
to
embrace
a
as opposed to the economic
focus
cost-benefit
of
apples
to oranges,
It does not try
analysis.
to
add
If one were
at least not directly.
to
evaluate a set of transportation alternatives in a
community
was a clear consensus as to the most
important
where
there
goals and their order of importance,
surely
would
this methodology
determine the transportation alternative that was
in
This methodology might also work well
the "pulic interest".
if
it were possible to rationally determine,
of
consensus,
in the absence
the proper weighting of objectives or of
the
relative importance of affected parties.
transportation
persons
may
consensus
Such
are
exist
improvements
or
in
the
in rural
of
areas
where
But Morris
directly affected.
small
case
Hill's
few
Goals-
Achievement Matrix would have serious deficiencies if applied
to
urban
settings
where pluralism is more
prevalent
than
consensus.
In order to properly weigh the goals,
it
as Hill suggests,
is necessary to determine either a shared
community-wide
valuation of goals or calculate an average goal weighting
all
affected parties.
decisions
objectives.
whole
to
determine
This
displays
individual.
a
of
Hill suggests looking at past public
how
a
community
would
weight
technique assumes that the community as a
constant set of preferences
as
does
an
But, if we assume that past decisions are merely
the outcomes of battles fought out by different parties in
a
pluralist system, the past may not always be a good indicator
185
of future political choice.
the same winners.
the
The battles may not always have
Furthermore, people's values determined in
may not remain constant when applied
abstract
to
real
life choices.
An alternative technique that Hill suggests is to survey
affected
parties and then somehow average their preferences.
But how much weight is to be given to the preferences of each
group?
Should
all groups be given the same weight
differences in size,
gains and losses?
or income,
despite
or differences in potential
Any weighting process must make decisions
This
as to the relative importance of competing groups.
is
deep a political choice as one can make and should not be
as
hidden under the guise of Eational planning.
If
rather
the
the
weighting
process is indeed
than a rational one,
politicians
a
political
why not leave those choices
Hill thinks that
and public?
rational planner's duty to make these choices.
"The simplest strategy is to present the
goalsdecision-maker with the entire
achievement account without attempting to
synthesize the extent of goals-achievement.
This approach is obviously least demanding
of the planner and most demanding of the
decision-maker. The onus would then be on
the decision-maker to trade off the extent
of achievement of the set of objectives
with the weights of those objectives (Hill
1973)."
186
it
is
one
to
the
Implementation of the GAM
b.
for each goal
Following the logic of the GAM described,
distinguish
a
to
each group a goal-achievement account is established
and
goal
between impacts that represent a regression from
(Hill's
represent
definition
of
costs)
a progression toward a goal
and
impacts
that
(Hill's definition
of
benefits). A/discussed in the theory section, the first step
in
the GAM is to determine goals and estimate impact
based
on these goals.
VI-1
concordance
in
requirement.
matrix
Table
The impact matrix presented in Table
analysis is
also
for
suitable
this
In order to facilitate'the comparison of impact
accounts,
VI-ll
matrix
the
accounts should first be
normalized.
shows the normalized impact matrix in
which
ratio of each account to the highest account for criteria
a
is
used.
The
next step is to derive weights for goals
in terms of criteria.
analysis
Since
the
expressed
The set of weights used in condordance
was directly applied in calculating overall scores.
set of weights employed in
included small numbers,
concordance
analysis
each weight is multiplied by 1000 as
shown in Table VI-12.
Each normalized account is then multiplied by the weight
for each alternative,
and product are summed to generate
187
an
overall,
weighted,
normalized account for each alternative.
The outcome of this procedure is presented in Table VI-13.
The
next
step is to aggregate similar
impacts
within
each goal category and then compare the results for different
Benefit categories include criteria 3 to
ARZ
alternatives.
12.
Cost categories contain criteria 1 and 2.
The summation
sign at the bottom of a column in Table VI-14 indicates
that
all the impacts for the corresponding goal are quantified and
therefore
can be totaled.
and costs are thus obtained.
The differences between benefits
These differences provide
framework for camparing each alternative policy.
188
the
Execution Procedure for the GAM
Goal:
Objectives
Alternatives
Goal
Achievement
Acount
Normalized
yGoal-Achievemment
Account
Weights
for Goals
Overall
GoalAchievement
Score
189
Impact Matrix
Normalized
Table VI-ll1
a4
a5
a6
a7
300
.259
.025
1.000
.709
1.000
.054
.027
.000
1.000
1.000
.063
.319
.400
.000
.600
.600
.200
.200
.400
.400
.000
1.000
.600
.600
.400
c6
.000
.000
-1.00
1.000
.000
-1.00
c7
. 333
.667
-.333
1.000
.833
. 667
. 333
c8
.000
.333
-. 333
1.000
.833
.333
.500
c9
.000
.250
1.000
1.000
1.000
.750
.750
c10
.211
.362
.512
1.000
.801
.439
.740
cll
.860
.783
.620
.489
.421
1 .000
.348
c12
.912
.767
.619
.619
.312
1 .000
al
a2
a3
cl
.000
.500
.406
.
c2
.000
.291
.581
c3
.000
.000
c4
1.000
c5
-1.00
Source: This Table is derived from Table VI-1
190
(Impact
.451
Matrix).
Table VI-12
Weights for Criteria
Criteria
Weights (Add up to 1000)
1
56
2
73
3
80
4
0
5
19
6
182
7
21
8
56
9
202
10
47
11
136
12
128
Source: Each weight in Table VI-15 is simply multiplied by 1000.
191
Table VI-13
Goal-Achievement Scores
al
a2
a3
a4
a5
cl
.0
28.0
22.74
16.80
14.50
1.40
c2
.0
21.24
42.41
51.76
73.0
3.94
1.97
c3
.0
.0
.0
80
5.04
25.52
c4
.0
.0
.0
c5
7.6
7.6
.0
c6
182.0
182.0
.0
c7
7.0
80
.0
19.0
.0
.0
a6
.0
11.4
11.4
364.0
182.0
a7
56.0
.0
7.6
.0
14.01
-7.0
21.0
17.50
14.01
7.0
56.0
46.65
18.65
28.0
c8
.0
18.65
-18.65
c9
.0
50.50
202.0
c1O
9.92
17.01
24.06
47.0
37.65
20.63
34.78
cl1 116.96
106.49
84.32
66.50
57.26
136.0
47.33
c12 116.74
98.18
79.23
79.23
39.94
128.0
57.73
202.0
202.0
151.5
151.5
Source: This Table is constructed by multiplying each value
by
the
Matrix)
(Normalized Imapct
VI-ll
Table
in
corresponding weight of criteria in Table VI-12.
192
Table
VI-13
al
Goal-Achievement
Matrix
a2
a3
a4
a5
494.44
363.96
570.73
856.4
49.24
65.15
68.56
a6
a7
cl
c2
cll
c12
Benefits
Costs
440.22
0
440.22
445.2
298.81
502.7
87.50
768.9
667.23
359.46
5.34
57.97
661.89
301.49
Source: This Table is derived from Table VI-13. Benefits and
and aggregate costs are
costs
are summed seperately,
subtracted from aggregate costs.
193
The
Area License Scheme with Bus
Improvement
receives
the highest overall weighted scores (696.9), followed closely
by
As far as the
the Toll Increase for Tunnel Gates.
ARZ alternative is concerned,
desired
analysis
Bus
of
elimination
the results obtained
an
alternatives
through
a
in the CA turned out to be
similar
to
favourable
goal-achievement
obtained by subtracting cost-related
those
benefit-related goal-achievement scores in
from
scores
both
by
less
comparison
of
the underlying structure
While
differs substantially,
pairwise
concordance
and the goal-achievement matrix select the ALS with
Improvement.
methods
both the
most
the
Thus it may be argued that the undominated solution in
GAM.
the concordance. analysis is the same as the alternative
that
receives the highest score in the goal-achievement.
The
between
Toll
comparison
on
the second and
the
third
ranking
these two methods reveals that the core ALS and
for
Increase
respectively
in
Tunnel
Gates
place
second
concordance analysis
the
while
and
the
the
third
Toll
Increase for Tunnel Gates and Core ALS receive the second and
the
third ranking in the GAM.
received
The reason why the core
ALS
the third ranking in the GAM compared to the second
ranking in the CA is that the Core ALS in the GAM has smaller
values
for
criteria
benefit
611 and
12
than
the
Toll
Increase for Tunnel Gate for the same criteria.
It
directly
appears
that
reflected
the weighted
in the GAM.
194
scores
themselves
Recalling that
the
are
goal-
the
achievement scores are derived from directly multiplying
impact scores by the weights,
normalized
too
of
multiplication
does
The
unpreditable.
not
the
GAM
as
all
the
the concordance
analysis
concordance
impact scores are converted into the
normalized
direct
with
associated
impact scores by the weights in
in
exist
problem
this result is not
indices and the discordance indices.
In
view
analyses,
it
appealing
and
difficult
to
wider
seems
contacts
gained
obvious that the GAM
is
the
analyst's
is heavier in the
between
burden
Concordance
theoretically
but it is
apply to transportation projects
While
empirical
two
from
computationally efficient,
areas.
computation
experience
of
which
in
rather
cover
terms
Analysis,
close
the analyst and the decision maker can
established.
195
of
be
4. Compromise Solution (CS)
a.
Theory
The
Compromise Solution was developed by Zeleny (1980).
The CompromiseSolution
of
the
family
Compromise
maker
can be considered as being a
of multicriteria
Solution
evaluation
methods.
is an attempt to provide
a basis for selecting the
best
member
the
decision
This
alternatives.
and is
method is used in a situation involving multicriteria
for
designed
deriving
compromise
The
between
solutions
the
analyst and the decision maker.
The method retains both qualitative and qu
and
explores weights for various criteria.
Solution
maker
is
s hould
judgments
indicates
the
analyst
gives
the
The Compro mise
that
with
analyst
itative data
the
means
of
of
membership
achieving
decision
membership
functions
compromise.
functions are determined by interactions
and
decision maker).
revealed
information
On
the
by the decision-makers,
to
the
value
several
regard to preference weights and
(be choice
membership
preferences
provide
with
functions
the
value-oriented approach in
k
decision-makers
basis
simply
Relevant
between
of
the
the
analysts
as
to
the
feasibility range of outcomes and about reasonable compromise
196
7
solutions.
conflicting
(1974) suggests that because of the
Zeleny
nature and noncommensurability of multicriteria a concept
of
solution,
is
solution,
compromise
probably
more
useful
reliable
construction
rather
than
optional
for an
analysis.
By
a ulility function
of
arguing
that
trade-off
or
function is often too complex or unrealistic to be practical,
he
viewed the Compromise Solution as an effort to
help
makers to reduce the set of nondominated
decision
the
solutions
by eliminating "obviously bad" solutions.
The
Compromise
Solution starts with a set
of
initial
feasible alternatives, denoted by
=
,
(a
n
2
1
A
a
a
,........,
These alternatives are evaluated against a set of n criteria.
th
The k
alternative can be expressed as
k
k
a
(a
=
1
k
,
k
,.......,
a
2
a
n
k
=
,...,m
k
Each
a
represents
the level of criterion i
i
197
attained
by
alternative
where
k,
i=l,...,
n;
k=l,...,
m.
Now
the
multicriteria decision problem can be expressed as:
k
Max a
which
subject to
a
A
represents the vector function
maximization
Max stands for maximization over feasible set A.
solved
by
assessing the utility functions of
problem.
This can be
the
decision
of
decision
makers such as
k
U(A) =
U( a
k
,
an
assessment
of
utility functions
makers is often a difficult task.
functions
is
)
n
2
1
However,
k
a
,........
a
The assessment of utility
also beyond the scope of
present
study
(for
utility function assessments, see keeney and Raiffa, 1976).
The
built
underlying logic behind the Compromise Solution
upon
the notion that among all achievable scores
is
for
th
any i
criterion,
there must be at least one extreme
value that is preferred to all others.
be denoted by a
.
The ideal point
This can be stated as:
i
198
ideal
can
k
*
ai
Max ai
=
i = 1, 2,....., n
*
If there is a
*
A such that
(a
*
*
)
a
,.....,
=
A
i
i
n
Thus it is not necessary to
all alternatives are "ideal".
,
bother with decision problems.
Such an ideal case is hardly
attainable
in real decision-making practices.
therefore,
to
Our goal is,
find a solution which would be "as
close
as
possible" to the "ideal".
k
into membership
i
k
is
functions.
The membership function, denoted by z ,
i
th
criterion into the interval
designed to map the scores of i
The
(0,
next
1)
(Zadeh,
interpreted
as
step is to transform all a
1974).
The
membership
the degree of closeness to the
*
a
ideal
can
be
point
*
.
The degree of closeness to a
following properties:
*
1.
funciton
If ai
is a maximum, then
k
a
k
i
-------
z
i
*
a
i
199
,
ideal point,
has the
If a * is a maximum, then
i
2.
a
zk =
k
i
a
*
is
*
a
If
3.
I
is a feasible goal value,
k
preferred to all a
,
a
for example,
k
smaller and larger than
a
i
then
,
ii
k
a-
a
1
k
*
i
z
i
2
*
a
4.
zero
*
ai
the most distant feasible score is to be labeled
If
by
*
regardless
of
its actual closeness
define:
a
=
i*
min
a
ik
200
to
a
,
we
can
k
and write z
as
i
k
-a
a
i
k
------------
=
z
i
*
- a
a
i*
i
k
where a
is
i*
anchor value in the case of
min
i
k
of
can be varied because
i
learning and other dynamic factors.
value
The
situations,
of
z
changed
The change
is possible through the following two operations:
1. Concentration:
a
k
----
z
k
0 < a<1
(z)
ii
2.
Dilation:
k
kca
z
(z
i
The
i
first
operation reduces the
201
degree
of
closeness
k
and relatively more
less for higher values of z
i
k
Similarly, the operation of
z .
lower values of
relativey
for
i
dilation has the opposite effect than that of concentration.
Then we can construct the following table.
Degree of closeness with respect to i and k
Table VI-15
k (alternatives)
al...
i (criteria)
1
1
l
.
.
.
.
.
am
..............
.
.
.zlm
Zl
Zn
znl.................znm
n
The
next step is
to derive
(z
e
i
of
the
th
i
increasingly
data.
be
criterion.
used
The
entropy
),
the entropy measure
i
measure
in the analysis of business and
being
is
economic
It is simply another measure of dispersion which
related to the moments of the probability function.
202
can
The
is
larger
criterion,
the
intensity
entropy or contrast
The entropy,
is taken to derive the weights
in this case,
criterion's importance. Criterion's importance g
relative
its
of
measure
the
the greater is the amount of decision information
in and transmitted by the criterion.
contained
of
th
i
and
,
is
the
a
i
given
decision
criterion's
contrast
a
importance in
situation.
In table VI-14, we observe
m
z
-
i
Z=zk
*..,n
i
k =i
i
1,
=
which indicates the strength of vector z
i
1
z
(z
m
............
i
i
i
).
z
th
Then
the
entropy measure of the
i
intensity is
m
k
k
z
z
e(z
)
-kZ
ln
k = 1
z
203
z
where k>O
and
k
0 < z
<1
i
e (z )
0
i
Tota-l entropy of Z can be introduced as
n
e(z )
E=T
i=1
i
larger e(z ) is,
the less information is transmitted by
i
th
the i
criterion.
Since weights of a criterion's importance'
are
i
reversely related to e(z ), 1 - e(z ) would be used.
The
i
=
g
i
-----
n -
[1
E
-
i
e(z )}
i
204
where n is the number of criteria.
and
that by reducing Z we could shift the ideal point
change zk as well as e(z ),
E and the
9's.
i
i
Note
thus
Ultimately
such
changes
get
reflected in
a
new
set
of
gs.
Now it is necessary to have information on a weight
vector, wi = (wl,.....wm) which represents the subjective
i
i
th
assessment of importance of i
criterion.
The assessment
can be expressed as
importance, g , as:
one possible hypothesis of a criterion's
i
g .w
g
i
ii
or after normalization:
g .w
i
i
i = 1,.......,n
g
i
n
g
i=l
The
minimizing
matrics,
next
step
.
w
ii
is to derive
Lp( g,k) for p = 1,
distance
2,
compromise
.
membership functions,
205
solutions
A family
of
is used as a
by
Lpwide
k
range
*
as
expressed
The Lp-matrics can be
and z .
z
of
in
to a
of geometric measures of closeness of any a
k
terms
*
follows:
n
k
p
Lp(g ,k) =
(1-z
g
1/p
p
I
i=1
the distance parameter 1 _.
represents
p
where
Therefore Lp(g
,
p
<__
-
k) evaluates the distance between the ideal
*
and the actual vector of degrees of closeness
k
The power 1/p can be
an alternative z .
z
alternative
induced
by
for 1
disregarded
affected.
.
since the solutions would not
p
be
n
g
Setting
i=1
1,
=
k) with respect
Lp (g
i
p values can be stated as:
varying
k
Z
n
1
Ll (g, k)
g
-
=1
iI
k
2
L2(g, k)
r
g
=1
(
1/2
2
l-Z
I
ii
k
L
(g,
k)
=
max
g
(l-Z
i
206
)
i
I
to
can be observed that as "p" goes up,
It
more and
more
weight as given to the largest distance ( Distance means that
ik
the set of alternatives are mapped through z 's into a
Ultimately,
space).
distance
the
distance
completely
for p = 1 and p = 2, there is at least one
k
distance for any a with regard to a certain criteAs
dominates.
maximum
rion i.
The choice of p then reflects the strengh of our concern
about
making such maximal deviation from the ideal as
as posible (Zeleny,
simply
Lp-matrix
1974).
reflects
small
Put differently, the choice of a
the
means
of
a
achieving
Therefore, that alternative which is the closest
compromise.
to the ideal can be considered "compromise alternative".
The
alternative can be obtained by minimizing Lp( g
compromise
k) with respect to p values.
b.Implementation of the Compromise Solution
The
theory
of the Compromise Solutions just
described
was applied for seven alternatives and twelve criteria.
ideal point for each criterion with respect to an
alter native
For example,
The
individual
was obtained which is presented in Table VI-15.
3*
Since a low value
(ideal point) is 68411.
a
207
is preferred to a high value for criteria 1 and 2,
the ideal
point for this criteria is, of course, the lowest value.
individual
Each
degree
of
closeness
component
was
to the points,
transformed
for
the
12
into
the
criteria
*
respectively, a
1890,
of
ARZ
-1260,
0, 0, 68411, -1575, -1575, -315, -1890, 0.245, 0.211,
alternatives
0.215.
A were mapped
In other words, the set
ik
's into a
z
through
distance space as shown in Table 6-16. For criteria 1 and 2,
*
ik i*
ik
was a minimum.
was used because a
= a /a
z
208
Impact Matrix for ARZ Alternatives and Criteria
Table VI-16
al
a2
a3
a4
cl
0
80
65
48
c2
0
78
156
c3
0
285
a5
a6
a7
4
160
41.5
190.5
268.5
14.4
0
76013
76013
4782
24283
7.2
c4
-1575
-630
0
-945
-945
-315
-315
c5
-630
-630
0
-1575
-945
-945
-630
315
315
-315
0
315
0
c6
0
c7
-630
-1260
630
-1890
-1575
-1260
-630
c8
0
-630
630
-1890
-1575
-630
-945
c9
0
-315
-1260
-1260
-1260
-945
-945
c1O
.052
.089
.126
.246
.197
.108
.182
cll
.190
.173
.137
.108
.093
.221
.077
c12
.196
.165
.133
.127
.067
.215
.097
1. Off-street Parking Fee Increase;
*ARZ Alternatives:
2. Parking Meters in Core; 3. Transit Priority Signals;
Core ALS; 5. Core ALS with Bus Improvement; 6. Toll
4.
Increase for Tunnels; 7. Toll Charge at Bridges.
2. Maintenance Cost; 3.
1. Capital Cost;
*Criteria:
Revenue; 4. OVTT for Auto; 5. OVTT for Taxi; .6. OVTT for
8. IVTT for Taxi; 9. IVTT for
IVTT for Auto;
7.
Bus;
Instituitonal
11.
Pollution;
in
Reduction
10.
Bus;
Preference;
12. Enforceability.
209
Table VI-17
Transformation of Alternatives and Criteria
al
a2
a3
a4
a5
a6
a7
cl
1 .00013
.00015 .00021 .00024 .0025
c2
1 .00013
.00006
c3
0
c4
j
c5
2
.0042
0
.5000
0
2.5000 2.5000
0
c6
0
0 -1.000
.00005 .00004 .0007
(D
1 .667
O
0
.0721
Zi
.00006
1.0033
.0014
1.0024
.4267
2.5030
1.667 5.000
5.000
16.8340
1.667 1.667
2.500
11.8340
0
-1.000
-2.0000
-1 .000
c7
3.0000 1.5000 -3.000
1
1.200 1.500
3.000
8.2000
c8
0 3.0000 -3.000
1
1.200 3.000
2.000
7.2000
c9
0 4.0000
1.333
1.333
9.6660
0
1
(
c1O
.2110
.3620
.512
1
.801
.439
.740
4.0650
cll
.8600
.7830
.620
.489
.421
1
.348
4.5210
c12
.9120
.7670
.619
.591
.312
1
.451
4.6520
The
next step is to derive a normalization table
the information in Table 6-16.
criterion
was
added up to 1.
using
Individual components in each
implying that the set of
alternatives was mapped into unit interval (0, 1).
ARZ
Table VI-
17 includes the information necessary for deriving an entropy
or
a measure of contrast intensity.
210
As was stated
in
the
section,
theory
the entropy measure is required in order to
derive a weight of each criterion's importance.
Since there are seven ARZ alternatives,
max
and k were
e
obtained as follows:
max
e
=
k
l/e
n 7
1.946
max
Using
max
e
,
the
=
.5139
entropy measure can
be
example, e(z3) was obtained as:
e(z3)
= -(.514)[.002(ln.002)
+ .4(ln.4)
+.4(ln.4) + .029(ln.029)
.171 (ln.171)
=
.592
211
+
calculated.
For
Normalization
Table VI-18
al
a3
a2
a4
a7
a6
a5
Zi
cl
.997
0
0
0
0
.003
c2
.998
0
0
0
0
.001
.001
1
c3
0
0
1
.002
0
.400
.400
.029
.171
1
c4
.059
.149
0
.099
.297
.297
.297
1
c5
.211
.211
0
.085
.140
.140
.211
1
.500
.500
-. 500
0
.500
1
-. 366
-. 417
.122
.139
.146
.167
.183
417
.366
.278
1
0
.183
.417
0
.414
.103
.103
.103
.138
.138
1
c6
0
0
.366
c7
c8
c9
.
1
c1O
.052
.089
.126
.246
.197
.108
.182
1
cll
.190
.173
.137
.108
.093
.221
.077
1
c12
.196
.. 165
.133
.127
.067
.215
.097
1
Then E, the sum of all e(z ) was calculated.
a weight of each criterion's importance. Then
i
values of g 's stand for relative contrast intensities
to derive8
the
"E" was used
measuring
,
the
intrinsic average information transmitted
by
to
be
importance
i
assigned were then calculated using8 i and w
each
criterion.
The
final
weights of
212
i
Table VI-19
Values for g , w
i
i
i
and g
wi
9i
1
.230
.056
.159
2
.231
.073
.207
3
.095
.080
.098
4
.038
.000
.000
5
.103
.019
.024
6
.150
.182
.329
7
.050
.021
.012
8
.032
.056
.024
9
.040
.202
.098
10
.012
.047
.012
11
.007
.136
.012
12
.012
.128
.024
i
As
analysis
to w
was
,
the set of weights used
in the
also used as the set of weights
of public officials interviewed.
i
i
i
contains the values for g , w and g
preferences
213
concordance
reflects
Table
the
VI-19
ik
i*
Table VI-20
Values for (z
a3
a2
al
-z
a5
a4
a6
a7
1
cl
0
1
1
1
1
.997
c2
0
0
0
0
0
.999
.999
c3
1.
.996
0
0
.928
.573
c4
0
-2.5
1
-1.5
1
0
c5
-. 667
-. 667
-4
-. 667
-. 667
-4
-1.5
c6
1
1
0
0
0
1
0
c7
-2
-. 5
4
0
-.2
-. 5
-2
c8
1
-2
4
0
-. 2
-2
-1
c9
1
-3
0
0
0
-. 333
.333
.561
.260
.199
c1O
.789
.638
.488
cll
.140
.217
.380
.511
.579
0
.652
c12
.088
.233
.381
.409
.688
0
.549
0
i
next step is to sibstitute the values of g 's into the
p
distance membership functions, L - matrics. Then the compromise
The
solutions
includes
can
be
obtained for p
1,
2,
i*
.
Table
VI-20
ki
which would
- z
values of deviations, z
ki
*i
i
-z ) in conjuction
a basis for calculating g (z
the
provide
p
with L ( , k).
214
*i
(z
i
Table VI-21
al
0
c2
0
.158979
.158976
.158967
a5
a6
a7
.158962
.158960
.158990
0
0
.207000
.207000
.098000
.098000
0
0
.091000
.056000
0
0
0
0
0
0
-. 036 -.036000
.024000
0
.329
c7
a4
0
c4
c6
-z
0
.098
c5
a3
a2
cl
c3
Values forg
ki
0 -.016000 -.016000 -.036000
0
0
.329000
0
-. 024 -.006000
.048000
0 -.002000 -.006000 -.024000
0 -.005000 -.048000 -.024000
0
.329000
c8
.024
-.048000
.096000
c9
.098
-.294000
0
0
c1O
.009
.008000
.006000
0
.002000
'.007000
.003000
cll
.002
.003000
.005000
.006000
.007000
0
.008000
c12
.002
.006000
.009000
.010000
.017000
0
.013000
.502
.219000
.445000
.175000 .162000
.329
.329000
.158976
.158967
Max
The
minimum
0 -.033333 -.033000
.158962
.329000
results are presented in Table VI-21 in
values
The results reveal
are circled.
fifth alternative,
.353000
.690000
.207000
which
the
that
the
Area License Scheme with Bus Improvement,
has the smallest value,
.162, representing the closest value
215
to
ideal point with respect to p = 1.
the
.158962,
alternative turned out to be the closest value,
the ideal point with respect to p = max.
identify
show
the
concerned.
point
in
Compromise
Solution
Thus
values as far as 3 decimal
4 and
points
calculation was performed up to 6
order to derive a ranking.
Solution
seem
to
imply
The
that
may not be effective in providing a
the preferred alternatives.
216
to
It was difficult to
the closest value because alternatives 3,
identical
ARZ
same
The
results
the
5
are
decimal
of
the
Compromise
ranking
among
i
*i
Table VI-22 Values for g
a2
a3
cl
0 .02500
.02500
c2
0
al
c3
.01000
c4
0
.01000
0
0
c5
.00025
.00025
c6
.10800
.10800
c7
0
0
c8
.00100
.00400
c9
.01000
.09000
clO
0
0
ki 2
- z
(z
a4
a5
a6
a7
2i
.02500
.02500
.12500
.02500
.02500
0
.01000
0
.00100
0
0
0 .04300
.04300
0
0 .00900, .00300
.01000
0
0
.00000
0 .00044
0
.00044
0
.00025
.00100
0
0
0 .10800
0
.10800
0
0
0
0
.00000
.01600
0 .00004
0
.00400
.00100
.00100
.00100
.01000
0
0
0 .00100
0
0
0
0
.00000
0
0
0 .00030
.00000
.00100
0
cll
0
0
0
0
0
c12 .00002 .00005 .00010 .00020 .00050
.12927 .23730 .28940 .02520 .02598 .15644 .07355
When
the
alternative,
value
of
p
2
used,
the
fourth
Core Area License Scheme was the closest to the
ideal point as shown in Table VI-22.
Scheme)
was
The ALS (Area
License
alternatives have the property of being "as close as
possible" to the ideal solution.
The reason seems to be that
the ALS alternatives received relatively higher weights
217
from
the
groups for what is
making
decision
for
IVTT
bus trips seem far more effective
under
than
schemes
be
to
For example, the criteria such as OVTT
"important criteria."
and
considered
ALS
under
the
public
government
incurs
other alternatives because
officials provided high weights on these criteria.
other
In
while
words,
city
the
in
its return expressed
burdens in the ALS case,
financial
"revenue" seem very high, perhaps inordinately so in the case
of the ALS with Bus Improvement.
the ALS schemes
In total,
retain their superior advantage on the travel times for
bus.
Also it would seem that the clear advantage of the ALS scheme
is
too
not
instituiton'al
eroded by
heavily
low
received
weights
preference and enforceabiltity criteria.
for
The
oucome confirms that ALS alternatives are clearly superior to
the other alternatives.
could facilitate
These nonextreme undominated solutions
the
decision making
could
legal,
the
alter the process of deriving
the decision maker after careful
compromise
of course, be
The final decision making should,
solutions.
by
undominated
these
in that different policy scenarios for an impact
solutions."
made
However,
should be interpreted as "temporary and short-term
solutions
matrix
process.
consideration
institutional and social implications of
of
recommended
nondominated solutions.
The
solutions
ARZ
alternatives
selected through
the
are similar to those chosen under the
218
compromise
concordance
analysis.
were
Two
ALS alternatives
before the sole undominated
selected
fifth)
(the fourth and the
alternative
was
chosen in the concordance analysis.
The reason why the results from the compromise solutions
are similar to those in the concordance analysis seems to
that
two
methods
a
similar
theoretical
of effects and weights which
identification
to
has
identification
of the nonextreme
be
structure:
generally leads
undominated
solutions
within which the best-compromise solutions lie.
5. Conclusion
This
chapter has considered theories of three
methods.
Empirical analyses within the context of Seoul have been made
the
following
procedures
of
each
Several
method..
conclusions can be drawn from the above empirical analyses:
(1) The three methods have been shown to be implementable
in
practical circumstances;
(2)
The applicability of the methodologies to transportation
shown to be
projects
have
analyses
demonstrate
possible.
In
the practicality and
fact,
empirical
utility
of
the
three methods for such TSM type of projects as ARZ policies;
(3)
The
successfully
been
multicriteria,
incorporated
observed,
only
the
As
has
relevant
for
into the methodologies.
GAM seems
219
were
intangibles,
including
to
be
presenting a number of criteria and alternatives to a variety
of decision making groups.
Compromise
The
Solution
Concordance
The Concordance Analysis and the
similar.
appear to be methodologically
Analysis and the Compromise
Solution
are
particularly appropriate as elimination procedure so that the
ultimate
selection
is reduced to a very limited
remaining alternatives.
number
An evaluation of the three
of
methods
on the basis of empirical analyses is the subject of the next
chapter
in which methodologies are evaluated in a number
respects.
220
of
Evaluation of the Three Methods
Chapter VII
I.
Introduction
VI was devoted to the discussion of the
Chapter
and
empirical
with respect to each of the three
The purpose of this chapter is to describe criteria
methods.
can
that
analyses
theories
be
used to evaluate the
strengths
relative
and
limitations of the three methods and to present an assessment
of the three methods in terms of such criteria.
The comparithree
son reveals that the overall results obtained from the
are
methods
quite
computational burdens,
similar
but
that
they
differ
in
in the degree of interaction with the
decision maker, and in many other regards.
Synopsis of the Three Methods
II.
As
the
previous
chapter
discussed
the
underlying
theoretical structure of the three methods in greater detail,
it
seems
appropriate
to
summarize
these
assessing their strengths and limitations.
221
methods
before
The objective of
the
synopsis
is
to give the reader an averall
picture
of
their structure (see Figure VII-1).
Table VII-1
Major Components of the Three Methods
CS
GAM
CA
Alternatives
and
Criteria
Alternatives
and Criteria
Alternatives
and Criteria
Impact
Matrix
-- We igh t s
i---Concordance
&DiscorSets
Goal-Achievement Matrix
Impact
Matrix
.:
Aggregate
Costs&Benifts
Transformaof
tion
Matrix
Concordance&
Discordance
Entropy,
Measures
Weight
Threshold
Values
Weight of
Criterion's
Importance
Concordance&
Discordance
Aggregate
Matrices
Functions
Undominated
Solutions
Compromise
Solutions
Overall
Goal-Achieve
ment Scores
AS
shown
in Figure VII-1,
inputs or
222
initial
information
by
required
the
(transportation
and
analysis,
methods.
chosen
methods
are
alternatives
plans such as ARZ alternatives) and criteria
These are the essential inputs for
measures).
(performance
the
three
After
criteria
inputs are the same for
the
impacts are predicted on
and the
alternatives,
three
all
the basis of
weights
the
be
should
assigned to all criteria.
maker
decision
These
criterion.
(or
Analysis,
Concordance
decision
making
groups)
weights can provide a basis for
comparisons
pairwise
of
competing
alternatives
that
remains
alternatives
is
for
the
each
deriving
in
aggregate impacts in the GAM and
weight of criterion's importance in the Compromise
All
of
weights represent the relative preferences
These
to
choose
the
best
the
the
Solution.
(undominated)
by threshold values in the Concordance Analysis
and by membership functions in the Compromise Solution.
The
GAM does not need these types of values to arrive at the best
alternatives
decision
as
making
its
underlying
logic
assumes
that
the
in
groups' preferences are fully reflected
the aggregate impact matrix.
The
foregoing discussion has been necessarily brief,
but
it indicates the basic stages of the three methods, namely:
(1)
the prediction of the impacts and the generation of
the
alternatives;
(2) the formulation of the weights of the decision makers;
(3)
the
derivation of the best
223
alternatives
by
threshold
values in the Concordance Anaysis and membership functions in
the Compromise Solution.
With
the background of the major components of the
three
methods, a critical assessment of each is made in the section
following.
The
purpose
of this assessment is
to
clarify
relative strengths and weaknesses of the three methods.
3. Evaluation of the GAM, CA and CS
the
Evaluating
problem.
conauct
three methods is itself
the evaluation.
The intent here is not to
for,
methods
counterbalancing
examined
weaknesses.
evaluating
into
have
Although
the
circumstances
the
methods
and
the
circumstances.
a.
multicriteria
A number of criteria can be established in order to
which is the best method,
enter
a
as will become apparent, the
needs
and
strengths
thus
are likely to vary
decision makers,
evaluation
of
several
methods
in
criteria
for
and
for
particular
criteria
virtually
must
all
The criteria to be considered are as follows:
Computational burden
b. Incorporation of transportation effects
c. Degree of interaction with the decision maker
d.
suggest
Ease of testing for sensitivity
e. Real-world applicability
224
f. Applicability
to
Transport Planning
it
is necessary to
in
Environment
Seoul
At
the
outset,
the
wherein
present
show
unfavorably.
importance.
table
above-
The purpose of the table is to give
overall look at the results of the assessment.
to
the
three methods are assessed against
mentioned criteria.
used
the
whether
each
method
scores
an
A ranking is
favorably
or
Undoubtedly, the criteria are not all of equal
Table
VII-1
shows that each method
has
both
advantages and disadvantages; no method is strongly favorable
without
some
drawbacks.
The
results
obtained
from
assessment are discussed in the section to be followed.
225
the
Assessment of the methods against the criteria
Table VII-1
Methods
Cri-teria
CA
GAM
CS
a. Computational Burden
(1) Information requirement
(2) Computational efficiency
-
+
+
(3) Possibility of
sensitivity analysis
++
-
++
++
+++
++
(4) Inclusion of qualitative
information
b. Degree of Interaction with
the Decision Maker
(1) Understandability of
++
methods
(2) Encouragement of
public participation
+++
(3) Interaction requirement
++
-
++
++
+
++
++
-
+
+
-
-
+
-
++
-
c. Real-World Applicability
(1) Applicability to ARZ type
(2) Applicability to largescale transport project
-
(3) Incorporation of
Uncertainty
(4) Linkage to planning
process
(5) Linkage to decision+
making process
d. Applicability to transport
environment in Seoul
(1) Adaptabality to scarce
+
+
information
(3) Applicability to rapidly
++
changing environment
Note: +++ very favorable
++ favorable
+ moderately favorable
-
less favorable
226
-
+
a.
Burden
Computational
The
of
amount
the
various
be
must
aspects
that the
fact
decision
information
as
such
the
varies,
conducted
This is attributed
parameters and iteration.
requirements,
the
that
computation
upon
depending
to
burden may be hard to evaluate as
computational
makers'
for
requirements
necessary
information
vary a great deal from one method
to
another.
Also,
different
alternative
be
for
established
many
testing each method's
weights and effects.
will
be
can
to
sensitivity
the
Thus rather than attempting to evaluate
the computational
all these aspects in the present analysis,
burden
ways
examined with reference
to
the
empirical
analyses performed in the previous chapter.
Information Requirement and Computational Efficiency
the three methods none of them is clearly
Of
The
GAM,
however,
superior.
appears to be easier to use because
computational burden is rather light compared to the
Analysis
dance
However,
the
(CA)
and
advantage
the
of
the
Compromise
method's
approach
may be offset by the quantity of
must
assimilated by the analyst.
be
the
Concor-
Solution
(CS).
straightforward
information
To apply this
that
method
well, considerable resources are required for data collection
227
and for adequate public participation.
of
analysis
an
While
sectors
incidence
various
("incidence sectors" refer to the affected groups, classified
by
income,
area,
land use and so forth) to be affected
has
projects
transpotation
not
this
in
undertaken
been
by
the GAM requires, in its original concept, an impact
thesis,
evaluation of these various incidence sectors. In particular,
the
incidence
the
amount of empirical work on
sizable
and costs may be overly time-consuming unless those
benefits
impacts are large.
The success of incidence anlaysis depends
In this
upon the appropriate choice of sectors for analysis.
some
neglected,
to
consideration of sectors to
be
implementation of the GAM in our case
sense,
comprehensi'e
extent,
affected by transportation plans.
that
would be used to
GAM
the
focusing
of
prejudice
incidence sectors
trivial
on
Therefore, a danger exists
results
the
more
ignoring
and
by
important ones.
The
Concordance
Analysis has a
burden than the GAM.
the
Concordance
assimilation
computational
Furthermore, the results obtained from
Analysis need still further analysis to
burden
The
useful.
heavier
of
information
on the analysts and the
and
collection
information-processing
burden on the decision makers are clear disadvantages of
Concordance
Analysis.
implying
that
the
Recall that the Concordance Analysis
begins with determining the weights to develop a full
matrix,
be
prior computation must be
228
impact
performed
with a technique such as the Churchman-Ackoff
sensitivity
information
objective
analysis
to
performed
increases
method.
the
be absorbed just as it improves
scope of the study (except in the
Each
amount
multi-
the
happy
of
instance
where all results point the same way).
The
computational burden of the Compromise Solution
similar to that of the Concordance Analysis.
for developing non-dominated sets involve
tation.
order
The algorithms
substantial compu-
An entropy calculation is quite time-consuming.
to select the best alternative,
through
membership
require iterations.
analyst's
functions
should
compromise
be
In
solutions
which
undertaken
The Compromise Solution also limits the
by ruling out more intuitive approaches
scope
is
to
It is hard to imagine a consultant content
the information.
to be merely a mixer of paints, as it were.
Possibility of Sensitivity Tests
Analysis
Concordance
sensitivity tests.
is
superior
for
conducting
The involvement of the decision maker in
the
sensitivity tests in the CA is much easier than that
the
GAM
as the Concordance Analysis aims at single
of
few
or
decision makers.
Sensitivity
the
requirement
provides
tests seem to be a difficult task because
of various public
inputs.
The
GAM
results at the end of what is likely to be a
229
of
only
very-
time-consuming process,
since
unfortunate,
seems
This
fixed.
be
of alternatives which are assumed to
design
the
and it cannot readily feed back into
through completion
the
of
procedure the analyst would inescapabily have a great deal of
By this
specialized knowledge to offer about the situation.
decision-making
the
in
point
are
modifications
no
longer
process,
and
welcome,
with the way in which impacts
familiarlity
those
however,
the
special
distributed
are
would be in a sense wasted.
The
Compromise
In
criteria.
reached
Compromise
Solution,
the
decision
by an iterative process concentrating on
identification
likely
the
to
of
of
Solution is attractive in terms
"bad" solution.
Sensitivity
this
is
sequential
tests
be burdensome unless the computation is aided
are
by
computer programs.
Inclusion of Qualitative Information
Qualitative effects can essentially be incorporated into
the structure of the three methods.
conducive to the inclusion of
particularly
effects.
statements
alternative
Corcordance Analysis is
the
qualitative
This is because Concordance Analysis only requires
as
to
over
the superiority and
inferiority
another according to a certain
regardless of the specific scale used to measure
230
of
one
criterion,
performance
an interval or ratio
However,
according to this criterion.
to be superior measurement scale for the Concordance
appears
deriving
among alternatives in
ranking
a
particularly
scale is
interval
An
Analysis.
useful
for
the
Concordance
to
incorporate
Analysis.
The
weighting
formal
preserve
and
accomodate
to
regard
process with
measurement
Any
recognized.
information
is
essentially
be used in the GAM although Hill
that
qualitative
Also full subjectivity of
in a meaningful way.
information
any
to
ability
means
a
as
The chief advantage of the GAM is
information.
qualitative
its
best
functions
GAM
qualitative
scale
can
(1973) suggests
and
a ratio scale is preferrable to interval
interval
scale is preferable to ordinal.
The
qualitative
information
ratio
Solution
Compromise
scale
information
as
long
as
the
is expressed in terms of measurable
seems
to
be the most
accomodate
easily
can
useful
Compromise Solution because noncommensurable
qualitative
units.
A
for
the
scale
(dollars, travel
times, etc.) must be transformed into the degree of closeness
to the ideal point.
231
c. Degree of Interaction with the Decision Maker
None
all methods can be viewed as decision
human decision making;
aids
for
the three methods is intended to substitute
of
into
that synthesize the contents of the impact matrix
more essential and meaningful information.
Understandability of Methods
complexity
Technical
not
with non-experts.
communication
the
because
drawback
would
Analysis
Concordance
make
decision maker or
the
difficult for
may
Concordance
layman
seem to
be
Analysis
comprehend.
to
conducive
a
This seems to be
Concordance Analysis is
to
major
designed
to
replicate the thinking process of human mind.
The GAM,
the
on the other hand,
understood by
and
analyst
can easily be constructed by
the
decision
maker.
simple
distinct adavantage of the GAM is that the concept is
and
One
underlying process of goal-achievement is understandable
for non-experts.
insistance
maker.
on a bottom-line recommendation to
While
alternative,
Another advantage of the GAM is the lack of
the
results
fo
the
GAM
the
reveal
decision
the
best
the GAM does not lead to a definitive statement
of the best alternative.
232
The
similar
theoretical structure of the Compromise Solution
drawbacks to those of Concordance Analysis in
has
terms
Membership functions may
of communicating with non-experts.
be difficult to understand on the part of the decision maker.
Encouragment of Public Participation
The
not
Concordance Analysis and the Compromise Solution
for encouraging public
appropriate
these methods,
limited
to
participation.
are
In
public or administrative participation seems
an
after-the-fact-analysis
review
process,
although it could'be argued that public values are
partially
accounted for in the impact matrix.
The
GAM
appears to be the best way to
participation.
The GAM is designed to evaluate consequences
across a broad range of community objectives.
The GAM could
conceivably lend itself to participatory reviews,
goals
In
are preserved,
developing
public
encourage
and weights are simple to
goals,
elected officials,
the
analyst needs
to
since
all
interprete.
consult
with
community groups, public opinion sampling
and examination of previous transportation investment.
Interaction
Maker
Requirement
between
the Analyst
and
Decision
For evaluating the three methods with reference to
233
this
criteria,
stages
maker.
it
seems
necessary
to stress
particularly
of interactions between the analyst and the
In contrast to the CA and CS,
the
decision
the decision maker in
the GAM is the community to be affected by the transportation
plans.
234
Figure
VII-2 Stages of Interactions Between the Analyst
the Decision Maker
and
Concordance Analysis
*objectives & alternatives(l
*impact matrix(2)
Weights & Threshold values(3)-1
DECISION MAKERI
undominated solutions(4)
*compromise undominated
solutions (5)
GAM
*objectives & alternatives(l)
*impact matrix(2)
*weights(3)
g COMUNITY
goal-achievement
)
COMPROMISE SOLUTION
*objectives & alternatives(l
*impact matrix(2)
ideal point & entropy(3)
*weight of importance(4)
membership functions(5
.*relevant functions(6)-*compromise solution(7
Note:
Elements
marked
by "*"
are preceded
between the analyst and the decision maker.
235
by
interaction
From
the differences
The
interaction
decision
requirement
the
The methods differ in
In
between
Concordance
undominated
the
analyst
there
the
an
for active involvement of the decision maker
in
and
Analysis,
and
is
many stages of the procedure,
objectives
three
Concordance Analysis is attractive in terms
maker.
opportunity
the
among
of value judgments required for deriving
solutions.
of
VII-2,
are generally apparent.
methods
degree
Figure
including specification of the
criteria and determination of
the
weights.
Close interaction between the analyst and the decision
maker
is expected at an early stage in the process.
The
In
GAM is a weak method with respect to this criteria.
contrast
philosophy
of the GAM
rather
than
maker.
The
interest
to the
Concordance
Analysis,
groups,
underlying
is the achievement of specified goals
the provision of information
GAM
the
to
the
decision
also is designed to cater to a variety
making
it difficult for
the
analyst
of
to
interact with multiple decision making groups.
The
Compromise Solution can be regarded as a
interactive
active
method.
The
Compromise Solution
practical
requires
an
role on the part of the decision maker throughout the
analytical
process.
Instead of a rather
passive
position
after helping to formulate a set of weights as in the case of
the GAM, the decision maker is deeply involved in the process
of analysis.
236
e. Real' World Applicability
Applicabiltity to ARZ Types of Transport Projects
inherent strength of the Concor'dance Anlysis and the
An
that
they
potential
a
have
this
In
to a large number of alternatives.
applicability
the CA and the CS are useful for ARZ types of actions
sense,
a
since
is
Solution
Compromise
variety
ARZ
of
types
similar but alternative
of
schemes often exists within the urban transportation context.
The
be
regarded
as
quickly,
of
types
ARZ
flexible
being
because these methods can be put
changeable
the
can
which
actions
rather
the CS are also suitable for
and
CA
and
operation
into
and are designed to be a short-term aid for
for
decision maker rather than a guiding framework
the
transportation process as a whole.
One weakness inherent in the CA and the CS is that these
methods
are
not
decision makers.
effective in handling a
of
number
large
There are wider variety of decision makers
involved in decision making environment surrounding ARZ types
of
actions.
These
decision makers' perspectives
are
not
coherent and each decision maker has his own preference.
The
GAM
appears
transportation projects.
to be weak for handling
ARZ
types
of
The incidence sectors in terms of
237
different interest groups
gains and losses between
important for ARZ types of actions in the context
seem to be
of
of schemes.
such
for
worthwhile
transportation
plans.
and
effort
be
service-oriented
GAM is also of little
utiltity
are
minimal.
projects
effects of transportation
the
process
making
and weights may not
small-scale
The
not
general
siginificant
Also the
necessary for formulating objectives
when
in
are
sectors
into the planning and decision
types
ARZ
incidence
The
Seoul.
incorporated
for
do not
This is often the case of ARZ schemes. situation prevails.
Applicability to Capital-Intensive Transportation Projects
In
respect
to
transportation
generally
considerable
of
The effects
this criteria.
intensive
applying
schemes,
and the Compromise Solution
Analysis
Concordance
with
ARZ
contrast to their usefulness for
projects
and complex.
to
be
are
weak
capital-
considered
There is a danger
these methods that these effects may be
the
are
in
aggregated
in a crude way.
Capital-intensive
projects usually require huge capital
time
and
span
investment
,longer
analysis.
Capital-intensive
cost-benefit
projects may be
type
successfully
handled by cost-benefit type of analysis in cases where
238
of
such
criteria
evaluation
as
discount
The CA and the CS can
opportunity are considered important.
be
capital-
of
a means to narrow down the number
as
used
and
price
shadow
rate,
intensive projects when the above-situation exists. After the
of alternatives are narrowed,
number
analysis
elaborate cost-benefit
of
will be employed for a systematic evaluation
a
small number of alternatives.
making
While it is difficult to generalize the decision
surrounding
environment
projects,
capital-intensive
are
fewer number of agencies and decision makers
relatively
projects.
in making decisions on capital-intensive
involved
This environment is conducive to the underiying structure
of
the CA and the CS.
appears
GAM
The
and
practically
incorporating
in
projects.
few
because
basis
for
all relevant sectors that must be
taken
into
the planning process for such
of
transportation
would
capital-intensive
The GAM has a relative advantage for cases
alternatives
philosophy
for
a
capital-intensive
GAM
methods
other
provides
It
comprehensiveness.
its
account
than
relevant
the effects of capital-intensive projects
analyzing
of
more
conceptually
The GAM is
intensive transportation projects.
capital-
the
to be satisfactory for
the
are
available
projects.
GAM is
as
often
In this sense,
closely
matched
the
case
239
with
the underlying
with
on-going
practice of capital-intensive projects.
function best when sufficient time
where
is
The
available
Major disadvantages occur
for formulating goals and weights.
the
by
affected
be
to
area
the
when
capital-intensive
projects become large.
Incorportion of Uncertainty
In
Analysis and Compromise
Concordance
problem
of
methods
suggest performing sensitivity analyses for
itself
dynamic
dealing
sensitivity
that
however,
seems,
It
uncertainty.
with
These
stressed.
superficially
uncertainty is
the
Solution,
cannot effectively handle the uncertainty because
requirements.
underlying
the
Since
computational
complex
and
effects
of
nature
of
philosophy
these
of
methods is the provision of information to the decision maker
it appears difficult to incorporate
on the "one-shot" basis,
uncertainty in a systematic manner.
uncertainty
capable
projects
of
is
uncertainty
fact
The
uncertainty.
with
dealing
conducive
can
capital-intensive
be
that
the
GAM
capital-intensive
to dealing
generally
transportation
240
with
to
handle
GAM
inherent
difficulties
the
recognizes
explicitly
The
methods.
other
than
better
the GAM may enable it
of
adaptability
The
projects.
in
potentially
is
transportation
uncertainty
considered
also
important
In
the
since
in
GAM,
project
consequences
statements
may be
change)
environmental
or
increase
(e.g.,
in travel
accompanied
by
time
and
probablistic
they may be expressed by a range of
possible
outcomes.
Linkage to Planning Process
Figure VII-3
Stages of Planning and Decision Making Process
'Problems
Def inition
Planning
Process
Technological,
- Legal,
(---Administrative,
-Financial
Cons iderat ion
Generation of
Objectives
Genera tion o f
HAlternatives
Selection of
Alternatives
Evaluation of
Alternatives
Bargaining and
L
Negotiations
Selection of-
DecisionMaking
Process
]AlterativeInter-agency and
Inter-agency conflicts
and Coordination
241
linkage to the overall planning process is weak
The
Analysis is oriented toward helping the decision
Concordance
Also this method is aiming
maker formulate preferences.
Analysis
Concordance
of
stage
the
in
employed
be
of alternatives" in the overall planning process.
"selection
In
can
at
VII-3,
As shown in Figure
few decision makers.
or
single
of
structure
underlying
The
Analysis.
Concordance
in
not
the Concordance Analysis would probably
this sense,
occupy more than a fraction of the larger planning process if
it were applied in the real world.
for
The philosophical basis of the GAM is more suitable
the
The analyst doing the GAM is
overall planning process.
the
expected not only to formulate community objectives with
of
help
planners
or
evalaute selected alternatives.
the
from
first
inkling of
also
but
decision making groups
to
A continuity of the analsyt
design
project
final
through
evaluation is, therefore, essential.
The
Compromise Solution is not attractive with
to
the planning linkage criterion.
is
a
method
of searching out the
respect
The Compromise Solution
best
solutions
through
maker.
interaction between the analyst and the decision
It
can be regarded as a decision-oriented rather than a planningoriented method.
alternatives
and
The Compromise Solution begins by examining
impacts
previously
designed.
Compromise
Solution
which are assumed
Therefore,
that
it
is
have
implicit
the analyst enters
242
to
the
in
been
the
planning
process at a relatively late point, particularly the stage of
"evaluation of alternatives".
Linkage to Decision Making Process
The Concordance Analysis is satisfactory with respect to
makers.
As shown Figure VII-3,
consists
of
procedure
and
the
because
selection
alternative
implementation must be heavily influenced
and
making
decision
different
many
the decision making process
If there are multiple decision makers, this
inferior
is
decision
both the selection of the best alternative
implementation.
method
few
if there is only a single or a
criteria
this
real-world
in
groups
by
situation.
The GAM is considerably more realistic about the decision
Even in
making process than are the other two methods.
setting
goal
process,
it
tries
to
structures of the diverse decision making
the
suggests
GAM
reflect
groups.
consultation with relevant
the
the
value
Although
agencies
and
review of planning documents for setting the goals, intra and
inter-agency
nature
of
relationship
stemming from
the
transportation projects could not be
intersectoral
taken
into
account.
The Compromise Solution is not sensitive to the decision
243
.
. .
making
process ,
implementation
address
the
including the alternative
process.
There
and
selection
is in fact no
attempt
issues surounding the decision making
to
process.
Legal, administrative and financial considerations are hardly
incorporated into the analytical process.
limitation
planning
because
This is a
the environment in which
transportation
takes place is charactrerized by the
legal, administrative and financial constraints.
244
major
existence
of
f. Applicability to Transportation Environment in Seoul
in
There
is no common pattern in terms of the
environment
which
transportation
in
developing
and
orientations
philosophical
Distinct
countries.
operate
systems
to
traditions of government may lead each developing country
differ
in
its
evaluation
draw
and decision making.
of
handling
of objectives and alternatives and
recognition
problems,
transportation
interpretation of
Thus it would be futile to
conclusions about the issues surrounding this
criteria
solely from the experience of Seoul.
Before assessing the three methods in terms of the Seoul
characteristics
and
perception
planning.
This
necessary
seems
it
experience,
of
is
environment
the transportation planning
of public
transportation
toward
officials
based on the author's
the
describe
to
experience
with
public officials interviewed in the Seoul city government.
While
understand
articulating
participate
the
public
officials
Difficulties
for
the importance of the ranking technique used
their
in
the
they
preferences,
thinking process.
obstacle for conducting the analysis.
perception
general
in
interviewed
of
ARZ
arise
types
of
reluctant
This is
the
to
major
Problems lie in their
transportation
for a variety of
245
are
reasons
projects.
ranging
from
to
ones
transportation-related
methodology-related
ones.
They may be summarized as follows:
Transporation -Related Problems
1.
schemes
from
improvement
with
Failure to derive needs for ARZ types of
comprehensive
program
of
transportation
designated priority for action.
2. Failure to appreciate potential importance of ARZ
types
of transportation projects.
3.
Failure to relate ARZ types of
projects
to
capital-
intensive transportation projects in an integrated manner.
Methodology-Related
1.
Problems
Failure to understand the linkage between
outcomes
of
ranking technique and decision making process.
2. Failure
to incorporate multiobjective or
multicriteria
into evaluation process.
3.
Unwillingness to actively participate in
process.
246
the
analysis
Formulation of Weights
the
Churchman
heavy
its
is
(CAT)
technique
Arkoff
on the value judgements assigned to criteria by the
reliance
avoided since the CAT implicitly
be
encourages
the
criteria
and
to come up with preliminary "screened
analyst
judgements
value
analyst's
The
interviewer.
or
analyst
cannot
and
as
such
essential problem with an interview method
An
alernatives.
problem is particularly accute in several agencies
This
in Seoul where interviews were conducted.
notion
They have no clear
This lack
context of the central business district in Seoul.
of
the
about the objectives for urban transportation in
a clear notion makes the analysis
significant
himself
in
educate
the
amount
of
work
this situation.
very
A
cumbersome.
analyst
the
is required
for
The analyst
should
only
not
decision making groups in the logic behind
the
potential
CAT but also conduct the preliminary evaluation on
criteria and alternatives.
groups
making
CAT becomes more difficult for the decision
The
to
follow
criteria increases.
consistency
checks
painstaking
tracing
as the number of consistency
checks
The public officials find a process
a
of
difficult undertaking
relationship.
The
requiring
CAT
is
and
of
the
also
hampered by the fact that even after its completion, there is
247
no acceptable means of assessing the validity and accuracy of
the weights derived by the decision maker.
that,
Experience with these public officials also reveals
only the financial
among some public officials interviewed,
and
costs
(returns) are of interest and
revenues
environmental
pollution and enforcement
externalities
degree,
some
To
problem.
generally considered as
are
the
not
one
of
several criteria in transportation-related agencies in Seoul,
but in most cases these are given low priority.
Applicability in the Case of Scarce Information
situation
for
Analysis works reasonably well
Concordance
The
in which scarce and vague information is
a
avaiable
if
the decision maker is willing to provide value judgements
on
several parameters.
The burden of the analyst
significantly reduced in this case.
in the absense of full
provide
still
the
for
be
While a conclusion made
information is less reliable,
results
can
drawing
it can
about
inferences
desirable alternatives.
The GAM is rather weak in terms of applicability in
case
of
vague
theoretically
rating
the
may
be
possible to derive weights through ranking
or
process
of
and scarce information.
techniques,
a
difficulty
248
lies in
While
the
it
articulating the common denominator (weights) for the various
interest groups.
of
scarce and vague
the
situation
Compromise Solution is favorable for the
The
case
of
can
Solution
situation
Concordance
the
used
be
in
information for the reason discussed
as a "screening
this
if
purpose"
are
alternatives
After the number of
prevails.
Compromise
The
Analysis.
narrowed, effort should be made to collect more accurate data
for more detailed analysis
Applicability to a Rapidly Changing Environment
in
place
economic
takes
environment in which transportation planning
The
characteristics,
preferences.
A
demand,
socio-
in
Seoul is characterized by rapid change
objectives and
people's
method capable of incorporating one or more
The
iterations can be potentially useful for this situation.
method
be that of constant
should
feedback
and
iteration
with objectives being redefined and modified.
The
Concordance
requirements
between
the
iteration
of
best
Analysis functions
and
feedback.
under
Close
these
contact
analyst and the decision maker in this case
is
important because of the requirement of a constant feedback.
The
GAM
is
an
inferior
249
method
for
incorporating
In GAM, the decision makers would be
iteration and feedback.
figures
with both a variety of alternative summary
provided
and the full set of disaggregate accounts behind them.
the
iteration.
incorporating an
for
difficult
to make
tends
requirement
underlying
This
inherently
GAM
The
GAM
also
suffers from a lack of a dynamic dimension: it is essentially
it is difficult to deal with the change in
Thus
"one-shot".
on-going transportation systems.
Compromise- Solution
The
works
performing iteration and feedback.
of
ranking
experiment
priorities.
The
and revise
situation
also
on
past
plans.
The
his own judgmental weights based
using
would,
therefore,
be the case
his
may
analyst
actions and documents related to transportation
ideal
of
In addition to specifying
limitations or enforcement problems,
ordinal
terms
in
the decision maker can add constraints, such as
new weights,
budget
well
where
the
compare
the
decision making unit is a single decision maker.
4. Conclusion
The
goal
of
this
chapter has
250
been
to
The three
limitations and usefulnesses of the three methods.
methods
The
have been evaluated from a number of respects.
usefulnesses
has highlighted the limitations and
assessment
of each.
The
both
None of the three methods
favorable without some drawbacks.
is
methods
contents
making.
All
synthesize
the
to substitute for human decision
intended
can be viewed as decision aids that
of impact matrix into more esential and
brief
A
information.
summary
strongly
is
no method
and disadvantages;
advantages
has
method
results of assessment reveal that each
of
meaningful
results
asessment
as
follows.
Comparision reveals that the underlying structure of the
Concordance
The
Solution.
Compromise
the
is similar to that of
Analysis
Concordance Analysis and Compromise Solution
methods.
are decision-oriented rather than planning-oriented
The
philosophy of the GAM is the
underlying
effects
incidence sectors
various
for
objectives
articulated
while
the
prediction
of
reference
to
with
internal
logic
the
of
Concordance Analysis and Compromise Solution is the provision
of information to single or a few decision makers.
technical
The
and
Analysis
methods
difficult
comprehend
analyst
the
while
and
complexity inherent in
Concordance
the
for
the decision
maker
and
layman
the GAM can easily be constructed
understood
by
the
decision
these
make
Compromise Solution seems to
maker.
by
to
the
The
Concordance Analysis and Compromise Solution have, however, a
251
interaction
potential usefulness as a tool to facilitate the
the
between
make
and the decision maker and to
analyst
a
preliminary cut of a large number of alternatives.
Concordance
The
single
hand,
other
on the
The GAM is,
decision makers.
few
are
Solution
Compromise
for the transportation situations involving
suitable
or
and
Analysis
useful for the case where a variety of decision making groups
sectors
exists as it requires the public inputs of incidence
when formulating goals, objectives and weights.
Concordance
The
for ARZ types of transportation projects because of
superior
there
are
ARZ
numerous
types
alternatives.
-Since
transportation
projects that can be feasible in given
contexts as examained in Chapter IV,
of
number
large
a
to
applicability
potential
their
are
Solution
Compromise
and
Analysis
of
urban
two methods are
these
The GAM functions best
particularly useful for ARZ schemes.
for capital-intensive and large-scale transportation projects
comprehensiveness.
its
to
due
conceptually
practically
and
GAM
The
more
relevant
appears
for
to
capital-
intensive projects.
The
Concordance Analysis and Compromise Solution are
particularly adaptable to the overall planning process
the
GAM
is
considerably more realistic about
the
not
while
overall
planning process.
None
although
of
the
the
methods
GAM attempts
is superior
to
252
reflect
for
the
decision
value
making
structure
of
a
large
However, none of the
number of
decision
making
groups.
three methods deals explicitly with
the
environment in which major transportation decisions are made.
This
is
a
major
environment
of
characterized
by
limitation
major
a
number
decision
because
projects
transportation
of
legal,
making
is
administrative
and
of the thesis will be presented in
the
financial constraints.
The
next
conclusion
chapter
summarizing
the
thesis
and
its
potential
contributions and suggesting directions for further research.
253
SUMMARY AND CONCLUSION
CHAPTER VIII
The purpose of this chapter is to summarize and conclude
the
objectives of the thesis.
in
the
context
of
In view of the experience
research,
this
major
two
are based on the
conclusions
The
thesis.
gained
useful
potentially
directions for further research are suggested.
(1) to discuss
The objective of the thesis was twohold:
restricted
auto
behind
rationale
zone policies in general and the
restricting
auto
traffic
in
central
in
cities
developed countries and developing countries as well;
(2) to
the
three
assess
usefullnesses
the
multicriteria
limitations
and
of
evaluation methods through the application
to
auto restricted zone policies in Seoul.
1. Thesis Summary
The
experience
first
of
multicriteria
chapter briefly introduced the
ARZ schemes.
concept
and
It continued on to discuss the
problems inherent in a transportation
254
context
evaluation
and to review the literature on the multicriteria
also
chapter
The first
transportation field.
in
methods
discussed the objectives and organization for thesis.
Chapter
in
policies
II
restrction
began with a discussion of auto
Europe, Asia
examination of cases of ARZ application in U.S.,
This
chapter
with
experience
ARZ
of
The
ARZ
chapter
policies
in
and
both
Also issues associated
ARZ application were addressed from
relevant
examples.
continued with a comparison of objectives
and developing countries.
dimension.
other
taxamony of these and
a
characteristics
underlying
developed
and
cursory
a
restricting automobile traffic in central cities,
(Singapore),
behind
rationale
philosophical
the
general:
intercultural
the
progressed with an exploration 'of
and
techniques
in
developed
and
developing countries as well.
Chapter III examined the evaluation methods with respect
to
transportation
discussed
along
Conventional approaches
projects.
with
their
drawbacks
and
were
usefulnesses.
Posing the problems associated with these approaches directed
the
attention to alternative methods.
This was followed by
the
review of multicriteria evaluation
mewthodologies
with
reference to transportation plans or projects.
Chapter
IV
described
the
Central Business District and transportation
shares
of
Seoul
systems.
Mode
characteristics
and socio-economic characteristics of travellers were
also discussed.
255
Chapter
Quantifiable criteria
ARZ alternatives.
VI was devoted to the implementation
Chapter
the
goal-achievement
the
compromise
solution.
with a discussion of theory followed by a
method began
Each
matrix and
three
of
the concordance analysis,
evalaution methods:
multicriteria
in
qualitative weights.
of
terms
were expressed
and unquantifiable criteria were expressed
measurable units,
in
and
V centered around the discussion of criteria
discussion of the implementation.
Chapter VII evaluated the three multicriteria evalaution
methods
in terms of several criteria.
for
essential
assessing
the
The criteria seen as
methods
the
include:
computational burden, degree of interaction with the decision
maker,
real-world
applicability
and
applicability
to
developing countries as explained by the case of Seoul.
Chapter
conclusions
thesis.
VIII summarized and concluded the thesis.
were
based on the two major objectives
of
The
the
In view of the experience gained from this research,
potentially
useful
directions for further research
suggested.
256
can
be
2. Thesis Conclusions
Conclusion Related to ARZ Experience
objective
The main conclusion with respect to the first
of
the
restricted
thesis is that the rationale behind auto
zone policies generally varies within the U.S. and Europe and
Accordingly, the
between developed and developing countries.
was
experience
based
on
Most American ARZ
revitalization.
economic
ARZs in European cities.
difffer
according
traffic
congestion
Examining
to
in
While the objectives
cities and
types
ARZ,
of
the central cities
the
is
a
key
ARZs
reducing
principal
countries,
to
and
each
has
essentially similar
distinct
approaches
and
transportation
problems.
the major reasons for
the differences seem to be because each region has
themselves.
of
Europe
ARZ experience of the U.S.,
Grossly summarizing these experiences,
perceptions
The
developing countries.
objective in
solutions
another,
objective is generally stressed in the concept
environmental
developing
to
city
upon individual circumstances.
depending
of
ARZ schemes vary from one
of
objectives
different
and attitudes toward the transportation problems
Decision making is,
factor is that
therefore, different.
each of the cities has
257
different
The
goals
which greatly affect the ARZ objectives and types.
the
Also,
hard
useful in accomplishing particular objectives is rather
to
determine.
most
cases,
specific
types
of
ARZ
relationship
A direct
difficult
seems
schemes
exists
matrix
no obvious
ARZ schemes and objectives.
between
with
In
to
objec-
because of other forces operating on these
establish
be
to which certain ARZ schemes may
degree
tives.
Conclusion Related to Evaluation of Three
The
conclusion with respect to the second objective
is that the three methods
thesis
the
Methods
can
for ARZ types of transportation
implemented
multicriteria
evaluation
methods
have,
despite certain drawbacks,
tools
for assisting
be
successfully
The
projects.
employed in
of
this
thesis
considerable potential
as
the decision maker and others concerned
with the transportation decision making process.
Their major
advantages over traditional transportation evaluation methods
is
that
and
costs
they allow the consideration of a
benefits,
wider
and are oriented towards the
range
of
decision
makers.
The empirical analysis also
258
illustrates the
usefulness
of
including qualitative criteria into the evaluation framethem to
transforming
without
work
a
dimension.
monetary
While the three methods have certain limitations that prevent
complete confidence from being placed in their results, it is
methods
three
the
the positive aspects of
that
contended
outweigh those disadvantages.
methods can be viewed as decision aids that synthe-
All
essential
size the contents of the impact matrices into more
and meaningful information.
incorporating
of
capable
plans.
transportation
A
The three methods are in general
the
imbeded
multicriteria
assessment
the
brief summary of
in
results is presented as follows.
to
easiest
The Goal-Achievement Matrix (GAM) appears to be
(1)
use
because
the Concordance Analysis
to
compared
rather
is
the computational burden
(CA)
light
Compromise
and
Solution (CS) .
(2)
As
general
in
are
the three methods
examined in Chapter VI,
capable of encompassing a variety of primary effects
form
resulting
the
of
implementation
transportation
projects.
(3) The technical complexity inherent in the CA and CS
to
make
seems
and
these methods difficult for the decision maker
layman to comprehend while the GAM can easily be
constructed
by the analyst and understood by the decision maker.
(4) The CA is not relevant for
tion
while
encourage
the
public
GAM
encouraging public participa-
appears to be
participation.
259
the
approach
best
Because
of
the
to
time-
consuming
and
technical nature of the CS,
be
may
its use
restricted to the analyst and key decision maker.
(5) The CA and CS are decision-oriented rather than planningoriented approachs.
The underlying philosophy of the GAM is
with
the prediction of effects for various incidence sectors
reference
to articulated objectives while the internal logic
information
imbeded in the CA and the CS is the provision of
to a single -or to a few decision makers.
(6)
sensitivity
CA and the CS are superior for testing
The
to
because these methods are designed to provide information
a single or few decision makers
include
to
needs
a
variety
The GAM is weak in that it
.
of
incidence
weights.
reformulating goals and objectives and respecifying
(7)
CA
The
was
found
to be superior
for
when
sectors
ARZ
of
types
transportation projects due to its potential for assessing
large
number of alternatives.
a
Since there are numerous ARZ
types that can be feasible in given urban contexts, the CA is
particularly useful for ARZ schemes.
of
handling
necessary
range
of
to
ARZ
schemes
because
The GAM is weak in its
the
effort
significant
formulate objectives and weights from
incidence sectors may not be worthwhile
a
wide
for
such
small-scale and service-oriented transportation actions.
The
CS has a potential usefulness as a tool for screening a large
number
of transportation alternatives such as ARZ
types
of
actions.
(8) The CA and CS appear to be inferior for capital-intensive
260
transportation
large-scale
and
conceptually
is
The
GAM
for
capital-
relevant
more
practically
and
projects.
intensive projects.
the
The linkage to overall plannging process is weak in
(9)
the CS while the philosophical basis of the
and
CA
is
GAM
more suitable for the overall planning process.
(10)
While none of the the methods can effectively deal with
decision making process,
the
realistic
the GAM is
more
considerally
transporta-
about the decision making process for
tion projects than are the CA and the CS.
The Churchman-Ackoff Technique (CAT) was found to be
(11)
difficult
preferences
public
to
approach
decision-making-group
articulating
The
in transportation-related agencies in Seoul.
officials
interviewed were not generally willing
participate in a repetitive thinking process.
a
process
of
to
They appear to
have no clear notion about about objectives and ARZ types
actions in the context of Seoul.
a
of
The public officials found
consistency checks
undertaking
difficult
a
requiring the painstaking tracing of relationship.
(12) The CA and the CS work reasonably well for the situation
where scarce and vague information is available while the GAM
seems rather weak for this case.
(13)
and
The CA and the CS function best as a means of iteration
feedback.
transportation
This
is a considerable
planning environment
261
in
advantage
Seoul
in
for
which
the
rapid
characteristics
and
In view of the experience gained in the context of
this
in
change
demand,
socio-economic
4. Consideration for Furture Research
research,
several
suggested.
First,
objectives
and
areas
there
for
furture
is a need to explore
further
cities
relevant ARZ policies in central
both developed and developing countries.
can
research
Information on
be
the
in
ARZ
developed an-d developing countries should
be
searched in order to compare the objectives and relevance
of
experience
ARZ
in
policies.
Further
research
should also
exploring the decision making and institutional
This
directed
be
environment.
will lead to identifying the motives and reasons behind
restricting automobile traffic in a given urban area.
Second, further research is required to establish a more
explicit framework for identifying, measuring and valuing the
effects
research
A
related
secondary
impacts
from the proposed transportation plans.
project would be to incorporate
into the impact matrices.
Third, The Churchman-Ackoff technique employed to assess
262
the
appears
structure of decision making groups
preference
not to be highly reliable.
The assessment procedure is timeSome public officials
consuming and requires repetitiveness.
interviewed exhibit aversion to the thinking process required
the
be
to
groups
or
would
analyses for different interest
similar
conduct
Related research
procedure.
assessment
or modify
Research is required to improve
by the technique.
experts which can lead to important additional insights.
multicriteria
evalaution
be
validity of the three methods could further
implementation
Further
these
of
in.
methods
tested.
different
contexts would provide a test of whether
transportation
obtained from new applications are
results
methods
require extensive and continued development so that
employed
the
three
the
Fourth,
the
methodologically
similar to those found in this study.
Fifth,
the
accurately
patterns
trips
further
to
delineate
city,
including
should be done
research
travel behavior
in
Seoul
of demand for both individual trips and
The logit parameters used
the entire city.
for
aggregated
this thesis represent the travel behavior of one corridor
the
central city.
for
in
Thus attempt should be made to calibrate
the logit model on the basis of the entire city.
Sixth,
the
Concordance
Analysis and
the
Compromise
Solution
are methodologically designed to deal with a single
decision
maker
environment
or a few
decision
makers.
However,
the
made
are
in which transportation decisions are
263
by
characterized
Clearly,
a
making
exists for incorporating the
need
the
decision
of
variety
groups.
of
values
these decision making groups into the analytical process.
the uncertainty concerning the estimates of
Seventh,
criteria such as cost,
be
represented
explicitly
the uncertainty can
Generally,
should further be analyzed.
a
by
etc.,
travel time savings,
revenue,
a
probabiltiy
distribution
function.
This representation ensures more confidence in the
estimates
of criteria as criteria
tribution
of
values.
values
treated
of being
instead
dis-
are described as a
as
single
Along, this line of research, the way to incorporate
a probability distribuiton function into the framework of the
three methods examined should be explored.
5. Extensions of the Research
In
addition to the potential research areas
the
discussed,
use of interactive
the
techniques
analyst
methods can be
First, the three methods
dimensions.
with
three
and
would
facilitate
extended
in
The
interaction
interactive
between
the decision maker and help the decision
264
several
could be supplemented
techniques.
the
previously
the
maker
articulate weights.
of
Through interactive technique, the logic
can
decisions is constantly tested and value preferences
be explicitly and clearly incorporated.
the
Second,
three methods could be supplemented
by
a
tradeoff analysis. The tradeoffs among different criteria are
important.
Little
previous
reseach has been done
in
this
particular area.
The
final
suggested
extension to the
would be to develop computer packages for
present
research
the three methods.
The various sensitivity analyses and weight articulations can
be effectively aided by the computer programs.
265
REFERENCES
Its Effects on
"Boston's Downtown Crossing:
Algmin, Jane,
Downtown Retailing," TRANSIT JOURNAL, Vol. 6, No. 2, Spring,
1980.
and
Decision-Making
"Quantification,
R.,
Bisset,
Kingdom,"
United
in the
Impact Assessment
Environmental
JOURNAL OF ENVIRONMENTAL MANAGEMENT, 7, 43-58, 1978.
Systematics
Cambridge
MODELLING, 1976.
I.
Cochrane, J.
DECISION MAKING,
1973.
INTERNAL
Inc.,
CBD
TRAVEL
DEMAND
MULTIPLE CRITERIA
and Milan Zeleny Eds.,
University of South Calorina Press, S.C.,
Cohon, J.L. and D. H. Marks, "Multiobjective Screening Models
WATER RESOURCE RESEARCH,
and Water Resource Investment,"
9 (4) , 826-836, 1973.
MULTIOBJECTIVE
J.L.,
Cohon,
Academic Press, New York, 1978.
PROGRAMMING
AND
PLANNING,
C. ReVelle, J. Current, T. Eagles, R. Eberhart
Cohon, J.L.,
Facility
"Appicaiton of a Multiobjective
and R. Church,
Location Model to Power Plant Sitting in a Six-State Region
COMPUTERS AND OPERATION RESEARCH, 7(1-2), 107of the U.S.,
123, 1980.
"Multi-criteria Ranking of
L. Duckstein,
and
L.
David,
WATER
Systems,"
Resources
Long-Range Water
Alternative
RESOURCES BULLETIN, 12(4), 731-754, 1976.
De Neufville, R. and R. L. Keeney, "Multiattribute Preference
Evaluation,
Systems
Transportation
for
Analysis
1973.
63-76,
7,
TRANSPORTATION RESEARCH,
Delaware Valley Regional Planning Commision, AUTO RESTRICTED
ZONES IN THE DELAWARE VALLEY REGION; AN EVALUATION OF TRENTON
COMMONS AND CHESTNUT STREET TRANSIT WAY, August,
266
1977.
"Weighting Multiple Criteria,"
R.T.,
Eckenrode,
SCIENCE, Vol. 12, No. 3, pp. 180-192, 1965.
Fishburn, P.,
Wiley, 1964.
DECISION
AND VALUE THEORY,
New
Estimating Additive
"Methods of
P.,
Fishburn,
1967.
435-453,
13(7),
MANAGEMENT SCIENCE,
MANAGEMENT
York,
John
Utilities,
in
Use,"
Reduce Vehicle
to
"Actions
J.,
R.
Fisher,
Research
Transportation
MANAGEMENT,
SYSTEMS
TRNASPORTATION
Board Special report 172, 10-12.
"Comparision of Multiobjective
and Others,
Friez, Terry
Evaluation,"
in Transport Project
Methods
Optimization
Research
Transportation
751,
TRNASPORTATION RESEARCH RECORD
Board, 1980.
L. Duckstein and M. Fogel, "Multiobjective
A.,
Goicoechea,
RESOURCES
Management," WATER
Watershed
in
Programming
RESEARCH 12(6), 1085-1092, 1976.
Hansen and L. Duckstein, MULTIOBJECTIVE
D.
Goicoechea, A.,
DECISION ANALYSIS WITH ENGINEERING AND BUSINESS APPLICATION,
Jone Wiley & Sons, New York, 1982.
A. and G. Fauth, "Downtown Auto Restraint
Gomez-Ibanez, J.
JOURNAL OF
Boston,"
for
Benefits
and
Costs
Policies:
TRANSPORT ECONOMICS AND POLICY, Vol. XIV no. 2, MAy, 1980.
Heaton, C. and J. Goodman, "Automobile Restrictive Measures
Findingd and
Some Recent
Business District:
in Central
Transportation
747,
RECORD
TRANSPORTATION RESEARCH
Views,"
Research Board, 1980.
of Transportation
"A Method for the Evaluation
M.,
Hill,
Highway Research Board,
Plans," HIGHWAY RESEARCH RECORD 180,
1967.
Evaluation of
in
"A Goal-Achievement Matrix
M.,
Hill,
Alternative Plans," JOURNAL OF AMERICAN INSTITUTE OF PLANNERS
34, 1968.
and Masud, A. S., MULTIPLE OBJECTIVE DECISION
L.
C.
Hwang,
METHODS AND APPLICATIONS, Lecture Notes in Economics
MAKING:
and Mathematical Systems, Springer, Berlin, 1979.
Keeney R. L. and H. Raifa, DECISION WITH MULTIPLE OBJECTIVES,
Wiley, London, U. K., 1976.
267
Keeney, R. L. and G. A. Robilliard, "Assessing and Evaluating
Imapcts at Nuclear Power Plant Sites," JOURNAL
Environmental
OF ECONOMICS AND MANAGEMENT 4, 153-166, 1977.
Kettle, P. and M. Whibread, "An Ordinal Method of Evaluation:
A Comment," URBAN STUDIES X, 95-99, 1973.
STREETS FOR PEDESTRIANS AND
and R. Edminster,
D.
Koffman,
AN EVLAUATION OF THREE TRANSIT MALLS IN THE UNITED
TRANSIT;
and Associates, Report UMTA-MA-06-0049-79-1,
STATES, Crain
1979.
Impact,"
"Evlaution of Highway
and M. Wohl,
S.
A.
Lang,
1960.
Washington,
268,
BULLETIN
BOARD
RESEARCH
HIGHWAY
COST-BEMEFIT ANALYSIS:
ed.,
R.,
Layard,
London, Penguin, 1972.
SELECTED READING,
INCOME DISTRIBUTIONAL EFFECTS OF ROAD PRICING,
Kulash, D.,
Urban Institute, 1974.
Lichfield, N., "Evalaution Methodology of Urbaan and Regional
Plans: A review," REGIONAL STUDIES 4, Pergamon Press, 1970.
P. Kettle, M. Whitbread, EVALAUTION IN THE
N.,
Licfield,
PLANNING PROCESS, Pergamon Press, Vol..10, 1975.
W. Pecknold and C. Kern, BOSTON AUTO RESTRICTED
Loudon, W.,
Systematics
EVALUAITON PLAN, Cambridge
ZONE DEMONSTRATION:
Inc., Report UMTA-MA-06-0049-79-4, March 1979.
Manheim,
ANALYSIS,
OF
FUNDAMENTALS
M.,
The MIT Press, 1979.
TRANSPPORTATION
"The Art of Traffic Management," URBAN
Midgley, P.,
Vol. 4, No. 2, Feb., 1980.
D. M.,
McAllister,
MIT Press, 1980.
SYSTEMS
EDGE,
EVALUATION IN ENVIRONMENTAL PLANNING, The
the Goals
Location Analysis Using
"Project
D.,
Miller,
PLANNING
AMERICAN
Evalaution,"
of
Method
Achievement
ASSOCIATION JOURNAL 46(2), 195-208, 1980.
"Operational Mathematical Programming Model for
F.,
Muller,
the
Relation to
Economic Activities in
Planning of
the
7, pp.
Environment," SOCIO-ECONOMIC PLANNING STUDIES, Vol.
123-138, 1973.
268
C., D. Pearce and J. Stanley, "Criteria for Evaluating
Nash,
Project Evalaution Techniques," JOURNAL OF AMERICAN INSTITUTE
OFR4 PLANNERS 41(2), pp.83-89, 1975.
SPATIAL
MULTIDIMENSIONAL
Nijkamp, P.,
ANALYSIS, John Wiley, New York, 1979.
DATA
AND
DECISION
OECD, EVALAUITON OF TRAFFIC POLICIES FOR THE IMPROVEMENT
THE URBAN ENVIRONMENT, Paris, 1975.
OFR
K., "Car-Free Zones and Traffic Restraints: Tools
C.
Orski,
RECORD,
HIGHWAY RESEARCH
Environmental Management,"
of
No.406, 1972.
Peat, Marwick, Mitchell and Co. / RTKL Associates, A MANNUAL
FOR PLANNING FOR PEDESTRIAN FACILITIES, Prepared for Federal
Highway Administration, June, 1974.
S. Rosenbloom, PEAK PERIOD TRAFFIC CONGESTION:
Remak, R. and
C.,
D.
Washington
PROGRAMS,
CURRENT
FOR
OPTIONS
Highway
Cooperative
National
Record,
Transportation Research
Research Programs, Report No. 169, 1976.
OFR4
S. Rosenbloom, IMPLEMENTING PAKAGES
R. and
Remak,
WITH
DEALING
FOR
STRATEGIES
CONGESTION-REDUCING TECHNIQUES:
PROGRAMS,
COOPERATIVE
OF
PROBLEMS
INSTITUTIONAL
National Cooperative Highway
Transportation Research Record,
Programs, Report No. 205, 1979.
NATIONAL URBANIZATION POLICIES IN DEVELOPING
B.,
Renaud,
COUNTRIES, World Bank Staff Working Paper No. 347, July 1979.
Rietveld, P., MULTIPLE OBJECTIVE DECISION MODELS AND REGIONAL
PLANNING, North Holland, Amsterdam, 1980.
Environmental
"A Regional
Spofford,
and W.
C.
Russell,
JOURNAL OF
An Assessment,"
Management Model:
Quality
ENVIRONMENTAL ECONOMICS AND MANAGEMENT 4, pp.80-110, 1977.
T. J. Atherton and W. Pecknold, "A Preliminary
L.,
Sherman,
Restricted
Assessment of the Transportation Imapcts of Area
FORUM 27,
RESEARCH
TRANSPORTATION
Zones in the U.S. Cities,"
1976.
"A Compararive Methodology for Assessing
Sondheim, M. W.,
Impact," JOURNAL OF ENVIRONMENTAL MANAGEMENT
Environmental
6, PP 27-42, 1978.
269
of Performance Using
"Comparative Evalaution
H.,
Terry,
9, No. 3, pp.
Multiple Criteria," MANAGEMENT SCIENCE, Vol.
431-442, 1963.
GREAT CITIES AND THEIR
J. M.,
Thomson,
Victor Gollancz Ltd., 1977.
TRAFFIC,
METHODS OF TRAFIC LIMITATION
M.,
Thomson, J.
AREAS, Working Paper No. OECD, 1972.
IN
London,
URBAN
Tokunaka, K., STUDY OF CONGESTION PRICING : THE CASE OFR4 LOS
ANGELES, Unpublished Doctoral Thesis, Urban Planning Program,
UCLA, 1981.
RELIEVING TRAFFIC
Holland,
E. P.
and
L.
P.
Watson,
THE SINGAPORE AREA LICENSE SCHEMS, World Bank
CONGESTION:
Staff Working Paper No. 281, June 1978.
AUTO
DOWNTOWN CROSSING:
Loudon,
and W.
G.
Weisbrod,
UMTAInc.,
Systematics
Cambridge
BOSTON,
IN
ZONE
RESTRICTED
MA-06-0049-82-3, 1982.
Undominated Solutions
P. and M. Zeleny, "The Set of All
Yu,
Methods," JOURNAL
Simplex
in Linear Cases and A Multicriteria
1975.
49(430),
APPLICATIONS
OF MATHEMATICAL ANALYSIS AND
P., "Linear Multiparametic Programming by Multi-Criteria
Yu,
Simplex Method," MANAGEMENT SCIENCE 23(2), 1976.
MULTIPLE CRITERIA DECISION
M.,
Zeleny,
Springer-Verlag, New York, pp.207-233, 1976.
MAKING,
MULTIPLE CRITERIA DECISION MAKING,
Zeleny, M.,
Book Co., New York, 1982.
270
Kyoto,
McGraw-Hill
Appendix II-1 Charateristics of Widely-Used ARZ Schemes
a. Traffic Cell
This
1)Concept.
and encourage public transit usage.
drivers
auto
scheme is designed to discourage
cells
Several
are created within the city center, and borders between cells
public
and
pedestrians
of:
arriving before peak periods
to another mode,
switching
to
auto drivers are then faced with a choice
The
transit.
traffic except
to
closed
are
or
parking away from the cell.
2)Mechanism.
is
center
and
the
a traffic cell at
held
by
or
transportation
the
city
towns
usually within the authority of cities and
generally
department
Establishing
traffic
Appropriations for
or its functional equivalent.
the purchase of signs and necessary items, labor for creating
cells
and installation of signs and
may be from a city's general funds,
department's
funding
structures.
determining
streets
own resources,
may
A
the number of
be
enforcement
additional
or may be covered by the
daparmental
depending upon the
procedure
would
cells to be
closed until the core
271
be
required
for
created.
Pedestrian
is
effectvely
area
traffic
The
connections.
few
traffic cells with
separate
into
divided
through
or
cells can be supplemented by
the
creation of one-way street and street closing at key points.
of
example,
For
that can be created in the city center.
cells
number
Traffic cells may vary in the
3)Variations.
the number of cells to be created could be by size of area or
by
specify
may
Restrictions
number of main streets.
the
where new pedestrian streets can and cannot be constructed. A
The
terminates.
service
traffic
should
scheme
cell
transit
rapid
be applied to the area where
also
may
cell
be
designed so that undesirable traffic spillover does not occur
boundary
around
roads or into
if necessary, to absorb the traffic to be
should be created,
the
in
held
Parking
areas.
surrounding
ring road or surrounding
area
following
the
creation of the cell system.
4)Possible Changes in Commuter Travel
*diversion to public transit, if available
*diversion to shared rides
*continued
arrive
at
occupant vehicle
single
cell
use,
but
tendency
a
area earlires in order to secure
to
scarce
space - possible shift or extension of peak period
*dropoff by family member
*possibly
parking
outside cell area and taking
transit
or
road
or
walking to walk or shop
5)Side Effect.
*possible
increase
of
traffic
927)
in
the
boundary
surrounding area
*possible increase in vehicle kilometers traveled
6)Factors Which May Influence Success.
*the stringency of enforcement
*availability of public transit
is
street system in the city center
the
*whether
able
to
serve a significant amount of automobile traffic
*the intrinsic attractiveness of the area
b. Pedestrian Street to Area
This scheme is designed to give pedestrians
1)Concept.
of
the city center priority over automobile
pedestrian
usually
is
street
created
in
traffic..
This
area
where
an
conflicts between pedestrians and auto vehicles are high. The
specified
street is closed off for all traffic.
This can be
applied to the CBD or to suburban shopping malls.
Most local governments have the
2)Mechanism.
to
The
establish pedestrian streets within their
decision
reasonable,
scheme
may
about
which
necessary,
affect
the
and
streets
authority
jurisdictions.
to. restrict
must
be
non-discriminatory because this
business
activities.
Funds
for
construction and enforcement may be appropriated from general
funds,
but
in other cases,
the department in charge may be
able to utilize its own funds for such a scheme.
Enforcement
of
traffic restrictions is the responsibility of
group
the
enforce all other traffic scheme - usually the
which
police
department.
The
3)Variation.
pedestrian volume,
business activity.
type of land use,
heavily used by pedestrians.
other streets.
streets
Transit and paratransit may
is
their
for vehicular traffic.
its
are
be
in
connections
that
particular
public
Relevant roadways may
benefit
surface
pedestrians.
be
The
pedestrian
street,
area and classification into
single-
dimensioning of the width of
density,
are
from
demarcation
cannot be blocked off completely because
into
meaningful
clear
This applies
service vehicles must have access.
turned
which
What seems important for areas which
pedestrians
for
they
where
In others. the pedestrian street
of the pedestrian area should not be detrimental
arrangement
purely
and characteristics of
depending upon the characteristics of the area. The
allowed,
to
volume,
traffic
applied only to streets in the core area
be
vary
may
may be applied to an entire
This scheme
city center in some cities.
may
area,
the
to
according
significantly
street or area
pedestrian
mixed purposes depend on the size of
purpose
and
center,
the volume of pedestrian traffic and
buildings.
4)Possible Changes in Travel.
*diversion to public transit, if available
274
the
city
the density of
*dropoff by family member
*diversion to shared rides
*possibly parking outside pedestrian area and walking to work
or shops
5)Side Effect.
*increase
or
decrease
of
business
activities:
potential
impact on area development unclear
*possible increase in non-wark trips
*may
encourage
center:
this
pedestrian
some people or business to
leave
the
applies to those groups who cannot afford
area
locations
which are expensive
because
city
the
of
their scarcity or their exclusivity.
6)Factors Which May Influence Success.
*stringency of enforcement
*avaiablity of transit
*high pedestrian volume
*existence
of
significant
amount
of
conflict
between
automobiles and pedestrians
*intrinsic attractiveness of the area
*whether
or
not
city
center
is
tightly
structured
walkable with a good mix of land uses and activities
275
and
c. Area License Scheme
1)Concept.
commuting
are
who
to
required
center.
make
people
This scheme is directed toward the
to the city
center.
Auto
are
drivers
city
purchase a lincense in order to enter the
to
This scheme is intended to force those commuters
choice between paying a lecense fee and taking
another
mode to work or arriving at work after peak period.
2)Mechanism
traffic
Because
regulation scheme,
an
Area
license
is
scheme
a
same
it usually falls under the
sort of authority as described under the cell system. Whether
area license system is legal or not depends on the
an
situation
(legal structure and law).
local
Legal challenge to the
constitutionality of area license schemes may be possible
in
certain area. Permits are typically issued to auto drivers of
the city in which the scheme is to be undertaken. A procedure
would be established for determining the license structure to
be charged and the time of day to be affected. Fringe parking
lots
and
boundary
roads outside the
area
license
scheme
(ALS)area are essential before implementing these scheme.
3)Variation.
The
license
fee struture
that
can
be
charged to auto drivers may be varied. For example, the level
of the license could be some fixed fee,
it could be a higher
fee for peak hours and a lower fee for off-peak hours, or any
of
a variety of schemes.
This scheme may also be varied
276
by
the target groups to be charged. For example, car with 3 or 4
passengers could be entitled to enter the city center without
paying
the license.
Restrictions may also specify where new
boundary roads or parking lots can and cannot be constructed.
Restrictions may be applied to entire metropolitan
well as the CBD;
area,
regulations for all jurisdictions should be
so that undesirable spillover does not occur
coordinated
as
in
one city or district.
4)Possible Changes in Commuter Travel.
*diversion to public transit, if available
*diversion to shared rides
*dropoff by family member
*possibly
parking outside restricted area and taking transit
or walking to work
*increase in vehicle kilometers traveled particularly in
the
boundary road
*distribution of overall pattern of transportation system
5)Side Effect.
*possible
decrease in peak period travel if sufficient
auto
drivers wait untill the off-peak period
*change
in land use - change in home and firm locations (may
be longer term effect)
*possible fee increase when demand becomes greater than
capacity (supply)
*potential impact on urban development - unclear
*imbalance in the use of road system
277
road
.......
.......
6)Factors Which may Influence Success.
*stringency of enforcement
*availability of public transit alternatives
*strong will and power of government and effective management
*availability of boundary roads and fringe parking lots
*intrinsic attractiveness of the area.
d. Pedestrian Shopping Mall
It is the taffic-free shopping streets
1)Concept.
of
the inner cities and suburban centers combined with malls and
The idea was developed when the concept of
shopping centers.
the
renovating
city
scheme
This
center was advanced.
is
intended to protect pedestrians from vehicular traffic in the
general
Auto traffic is
usually the city center.
shopping area,
allowed to enter into this area.
not
is
apllied to the area where the
unable
to
serve a significant amount of auto
intent
of
this scheme is often to enhance the
this
Usually
system
street
scheme
in
traffic.
vitality
is
The
of
economy in the city center.
pedestrian
A
2)Mechanism.
shopping
mall
may
be
accomplished through urban development policies such as urban
renewal
or
conceivable
accomplished
a
downtown
that
in
through
project.
It
is
this
scheme
may
be
to
existing
revitalization
some
cases,
modifications
278
zoning
local governments have the
Most
regulations.
shopping
pedestrian
create
malls (exceptions
to
authority
include
may
streets which are privately owned, or owned by other state or
regional
agencies).
crossing
the
streets
The reduction in the number of
pedestrian area may be necessary in
to
order
separate vehicles and pedesrians. Rerouting crosstown traffic
may also be preceded so that major streets do not cut through
pedestrian areas.
This scheme may be applied to the
3)Variations.
entire
city
center ot it may be applied only to few streets in
core
area
traffic.
which are currently used by pedestrians and
This
scheme
may
be
designed
to
auto
eliminate
automobile or to separate vehicular and pedestrian
The
the
movement.
pedestrian streets may be created in conjunction with
revised
traffic
pattern or it may be established
the
with
a
the
current road network in the city center.
4)Possible Changes in Travel.
*diversion to public transit
*diversion to shared rides
*dropoff by family member
*possibly parking outside pedestrian shopping mall or walking
to work or shops
5)Side Effects.
*may cause a serious loss of activity in the city center
*banning
cars from one street may have repercussions on
279
all
downtown traffic in some circumstances
*possible increase in non-work trip
*surface
parking
parking
near
battlegrounds
lots which provide most of the
pedestrian
of
shopping
malls
will
the fight for space between
off-street
become
shoppers
a
and
downtown employees
6)Factors Which May Influence Success.
*stringency of enforcement
*availability of public transit
*high pedestrian volume
*existence of boundary or adjacent streets which are
capable
or absorbing some diverted traffic
*intrinsic attractiveness of the area
*a good mix of land uses and activities to maintain forces of
econimic viability in the city center.
280
Detailed Assessment of ARZ Evaluation Cases
Appendix III-1
a.Cost-Benefit Analysis
(i) Jose Gomez-Ibanez and Gary Fauth (1980)
The decision problem that is addressed
Problem
Decision
exceed
in this study is the issue of whether or not benefits
employed
in
principal benefit estimated is the savings
in
travel time resulting from less congestion on the streets
in
The
Boston.
forms
two
when
costs
of auto restraint
are
The principal. cost estimated is
and around the central area.
the increase in travel time, expense, and/or inconvenience to
auto users who change to transit or to another route.
Parking
Alternatives
of various
charges
amounts
for
different times of the day are tested. Area license scheme to
both local streets and expressways were compared.
Two other
Boston's
commercial
policies
- a
district
and toll increases on the three facilities
in
small auto-free zone
already
charging tolls are analysed less throughly.
Analysis
demand
and
forecasting
behavior
calibrated
employed
travel
models to predict the changes in
travel
This
Findings
and congestion levels.
from
study
The logit
other metropolitan areas
281
models
were
already
used.
The
results
average traffic speeds
increased
significantly
every
Almost
highways.
calculating
Results show that area restraint schemes
costs and benefits.
area
for
these medels provide the basis
of
type
of
on
central
period
peak
auto
restraint measure produces positive net transport benefits.
and
In
other
in
calibrated
model
over-reliance
addition,
transportation studies,
savings
cost
problem
raises
of
an
cities
they
constraints,
few
of
is placed on the
logit
Like
other
in
U.S.
this study was directed toward
the
users.
The
(time) likely to accrue to
how
highly
are
from formally analyzing more than a
prevented
them.
investgate
Because of resource and data
ambitious.
were
authors sought to
the
effects
alternatives
The number and diversity of
Assessment
road
time
to place a monetary value to
issues commonly discussed in
the
savings
evaluation
of
transportation plans. A subjective judgement entered into the
process of evaluating the value of time.
to externalities,
made
While reference was
there was no attempt to estimate the
benefits and costs associated with them.
Another concern is that time is an important variable in
evaluation.
Effects or benefits and costs of ARZ schemes may
attempt
to
distinguish these effects and costs and benefits in terms
of
change
over
time.
In this study there was no
short-rum and long-run time frames.
reasoning,
consideration
Along a similar line of
of uncertainty does not enter into
282
the evaluation.
and
Given the fact that the system itself (human
it
Hence,
only on
analysis
provides
the
this
above,
objective
does not incorporate a multiple
study
with
associated
As was already implied
plans.
transportation
is
it
flux,
of
state
uncertainty
consider
to
important
in a
be
may
transportation)
framework.
time
transport
savings as a benefit and time increase as a cost and does not
address other objectives such as environment, safety, economy
and so on.
(ii) J.M. Thomson (1967)
The decision problem is to ascertain
Decision Problem
whether or not two auto restraint proposals are beneficial in
the
losses
administration,
restrained
from
The
London.
central
The
cars.
to
costs
diverted
include
estimated
cars
and
losses
benefits estimated include
benefits
the
higher speeds and benefits to traffic generated by
license system.
The author evaluates two schemes.
One is a
daily licensing scheme whereby all cars and cycles moving
the area between 7 a.m.
license
valid
parking
tax
for
to
in
and 8 a.m. are required to display a
that day only.
is
The other scheme
a
to
the
This study is limited to
the
scheme whereby an hourly .tax is
added
parking charge at all parking sites in the area.
Analysis
and Findings
283
traffic
of
assessment
The
charges.
of
resulting from different levels
benefits
economic
associated
and
effects
demand functions of different classes of traffic are examined
volumes and speeds, are then calculated.
fee or parking tax are derived.
(1980)
Fauth
employed
Thomson
methodological
its
whereas
equations for
In this sense,
prediction
the
work
high
activities.
benefits of an ARZ in these two cities turned out to
significant.
be
In the case of London, little advantage is seen
in the proposal for a parking tax.
the
these
relatively
density and financial and commerical
population
The
where
London and Boston,
applied are characterized by
were
schemes
areas,
Both
Thomson.
of
Fauth's
Gomez-Ibanez and
seem to have been heavily inspired by the earlier
work
of
in
and Fauth use logit models
demand
behavior.
travel
Benefits of licence
The work by Gomez-Ibanez and
appears quite similar
Gomez-Ibanez
approach.
the traffic
The supply functions i.e.
elasticities.
using
Thomson concludes that if
licence fees were used for the benefit of
motorists
it
be theoretically possible for all classes to be better
would
In
off.
terms
of
implementability,
no
administrative
consideration was discussed in this study.
Assessment
area
of
London.
restraint
From the administrative
point,
covers 7 square miles in the
Thomson's
center
of
This appears to be large area when compared with 2.5
square mile in Boston in Gomez-Ibanez and Fauth's work.
284
The
question which arises in this context is that
other
The demand functions are rather
elasticities were estimated.
them
demand functions need more variablells in
The
naive.
how
of
than just straight cost and taxation.
discussed in the context of Gomez-Ibanez and Fauth's
As
study,
this study is also centered around the evaluation
of
no mention
is
made
Suprisingly,
costs and benefits.
transport
method
the
in
The
savings.
this
are
Intangibles
savings.
externalities were not included in the
such,
time
of
magnitude of benefit of the alternatives
particularly sensitive to travel time
and
value
used to calculate the
evaluation.
study is purely based on the single objective
As
-
maximization of transport benefits over costs.
(iii) Damian Kulash (1974)
Decision
or
whether
progressive
fashion.
above
Problem
and
Alternatives
The
issue
not the effects of introducing road pricing
or regressive is explored in
rather
of
are
systematic
Several values of time saving are used to test the
decision problem.
Three metropolitan areas
(Boston,
San Fransisco, and Washington) are used for this analysis.
Using 1970 work trip patterns
Analysis and Findings
in Boston,
San Fransisco and Washington,
distributional
consequences
of
285
roadway
Kulash studied the
pricing.
As
he
as
less the value of
charges
user
road
the
are
savings
time
Kulash) depends on time savings and road user
larger
defined
the regressivity of net costs (net cost are
argues,
charges.
The
road
user
time savings or smaller are the
because
the more regressive would be the net costs
charges,
Whether
of the greater trip demands by wealthier motorists.
or
and
his hypothetical combinations of roadway prices
not
by
resultant travel time savings are appropriate assumptions
in
three metopolitan areas is left on open question.
regressivity
travel
conclusion
The
Assessment
regarding
drawn
the
in three areas is somethat questionable as
characteristics
of
different from one another.
areas
three
are
significantly
In te.rms of net costs defined by
Kulash, road pricing in a city like Los Angeles would be very
because considerable time savings may be achieved
regressive
in
typical "bottle-neck" situation where traffic flow
rather
decrease
than
increase
response
in
a
of
appropriate.
A
"average"
income group does not seem to be
breakdown
by various income groups seems necessary in
to
make
a
convincing argument.
Also,
assumed to be same for all commuters.
286
demand
the choice
Along a similar line of argument,
increase.
to
would
trip
order
distance
is
(1981)
(iv) Kenich Tokunaka
Decision
Several
Problem
(1) whether or not overall benefits
addressed in this study:
of
congestion
travelers
costs;
exceed
pricing
would
or
winners
be
distributional effects?;
(2)
most
terms
in
losers
kind
what
of
of
(3) which toll scheme is better - a
flat toll per trip or a toll per vehicle=mile?
the
are
decision. problems
comprehensive cost-benefit analysis
This study is
that
address
various issues surrounding the congestion pricing.
systematically
toll
per
following
The
Alternatives
considered:
vehicle-mile;
rich traveler vs.
are
alternatives
(1) a flat toll per trip
vs.
a
(2) income distributional effect
-
(3) average trip length
-
poor traveler;
rich traveler vs. poor traveler.
Analysis
comparison
and
For illustrative
Findings
between
this studies and other
performed as follows.
287
purpose,
studies
can
a
be
Comparisons
Table III-A-1
u-------------------------------
studies
other
study
Tokunaka's
-
No considersation of
distribution
-
issue is
Distributional
through
addressed
clearly
systematic analysis
-
are of the
trips
All
length
same average trip
-
High (low) income travelers
make longer (shorter) trips
-
The average
is assumed
-
by income,
Time values differ
journey purposes, mode and
small and large time savings
-
cost
collection
Toll
zero
be
to
assumed
-
Toll collection
into account
cost is
-
Modal choice is fixed
regardless of toll
-
Modal choice is
of toll
a
-
No
time value
between
distinction
considered
constant toll
and constant
is
-
the
This
taken
function
difference
is
per trip
per mile
toll
-------------------------------------------------
The general
spite of the regressivity of
as
-
10
and
However,
to make
in
20 cent/trip
rich
off
better
an efficient
toll
road pricing,
Los Angeles)
Rather,
the poor worse off.
poor
study is
conclusion of this
would
as
follows:
In
low charges
(such
not always
make
low charges might make both the
in
(63
the poor person worse off.
288
a
highly
congested
cent per one way trip)
Bus
area.
seems
passengers would gain
if
greater
wide
demand for bus travel lead to more frequent
and
One thing the author neglects is
the
spread service.
of distribuitonal weight.
application
pricing
roadway
regressive
is
In
weight chosen.
distributional
which
The extent to
on
the
because
this
depend
also
addition,
is geared toward the distributional effects of roadway
study
pricing
other concerns
for motorists,
soley
in
terms
of
environmental and social effects are not treated.
(v) Peat, Marwick, Mitchell & co.
Decision problem and Alternatives
This study tries to
explore whether pedestrian schemes produce positive benefits.
What
factors are important in evaluating pedestrian schemes?
Several different types of alternative pedestrian schemes are
included in order to evaluate these alternatives.
and Findings
Analysis
the
principles
This study attempts
apply
to
of traditional cost-benefit analysis to
evaluation of pedestrian schemes.
the
As with most such efforts,
the difficulty lies in quantifying abstract benefits, such as
convenience,
variable
comfort
affecting
and perceived safety.
pedestrians
quantified, in terms of number and
the
greatest
attention
Impacts to abutters,
which
is
Safety is
most
the
readily
severity of accidents; so
is given to safety in
this
study.
motorists and others are also necessary
289
to be counted in the calcualations.
analysis
The application of cost-benefit
Assessment
The
here runs into the same problems as in previous studies.
major problems in this study are:
(1) the arbitrariness with
which benefits are assumed to be relevant, discount rates are
and
chosen,
decision
allow
upon
decided
are
rules
considerable bias to be incorporated into the process, for or
outcomes;
particular
against
incidence
the
(2)
and,
of
benefits or impacts tends to be largely ignored.
(i) Kiran Bhatt (1976)
Trent Bertrand
(ii)
(1978)
These two studies treat benefits and
Decision Problem
The decision problem
costs in a more general fashion.
is
addressed
in
Bhatt's study is the
combining
of
issue
that
(or area license) with expansion of services by high
charges
occupancy
transportation
modes.
The decision
study is to ascertain whether or
Bertrand's
problem
not
in
congestion
pricing produces potential benefit in the context of Bangkok.
He
when
also
examines the "second best" decision often
there
are
constraints
associated
with
required
congestion
pricing.
Alternatives
the
benefits
of
Using hypothetical data, Bhatt compares
road pricing scheme
290
with
the
do-nothing
In Bertrand's study, congestion pricing and the
alternative.
data.
second best alternative were compared with the Bangkok
Congestion pricing and benefits were analyzed less thoroughly
in Bhatt's study than Bertrand's study.
to
estimated
was
pricing
the
to
benefits
net
produce
congestion
study,
In Bhatt's
and Findings
Analysis
he calculated the toll revenues to be
12.5
percent greater than the loss of benefits to the users.
The
And
community.
be
then
can
revenue
transportation.
public
the
improve
to
used
Bertrand's work was rooted in the concept of
externality which is long recognized in economics literature.
Based
congestion
not
pricing
use-reduction
vehicle
increases
the use of buses and taxis.
the
on
covers
an
appears
traffic system of
metropolitan
extensive
be
to
examined.
The
Bertrand.
He
than
greater
far
- Thonburi
which
area
study
The
area.
the
was
Bertrand's work
Bangkok
urban
other
areas
raised
question of implementability was
on
also
but
in private cars
aggregate
based
impact
substantial
only has a
that
Bertrand argues
on the results of his analysis,
by
calls the attention to the fact that attempts
to raise registration fees on cars in Bangkok have met fierce
including
resistance,
Implied
mechanism
transport
here
of
that
is the notion
revenue
systems
do
policymakers.
treats to the lives of
be found
should
not
appropriate
lose
291
to
so
the
redistrition
that
extent
users
of
without
the
undermining
for reducing and
incentives
the
altering
structure of demands on transport systems.
Unlike
Assessment
a wide spectrum of issues surrounding the ARZ
covers
such
legal
the
as
on
the
and
financial
scheme
fiscal
the possible
the administrative difficulties,
implications,
impacts
aspects,
study
Bhatt's
studies,
other
activities and employment in the ARZ
area.
In
Bertrand study, the problem definition and the description of
the study area seem
criteeria
detailed
to be weak.
for evaluation were used
process of designing alternatives.
may
value
while
countries.
markedly differ from developed
and benefits.
guide
the
The treatment of value of
fail to include any indication of the
studies
costs
to
the context of Bangkok is not explicit
in
time
Neither objectives nor more
incidence
time
Both
of
These two studies are hampered by many
of the same difficulties of intangibles and externalites that
appear in systematic cost-benefit analysis.
(iii) Cambridge Systematic Inc.(1976)
Decision
Problem and Alternative
Four different
ARZ
schems (or policies) were evaluated using disaggregate demand
The four policies investigated were:
models.
Policy 1:
Policy
2:
Total all-day auto exclusion, no transportation
improvement.
Total all-day auto exclusion, fixed routes
bus system plus areawide transit improvements
within ARZ.
292
Policy
3:
Policy
4:
Total all-day exclusion, fixed route bus
system plus areawide transit improvement within
ARZ.
Peak-periods exclusion only, no transportation
improvements.
disaggregate
demand
models to predict changes in two categories of travel
demand
Analysis
This
study
employed
(1)
associated with the proposed schemes:
choice models;
- mode
which
regional work mode
(2) intra-CBD workers frequency - destination
choice and non-worker destination-mode choice models,
incorporated the walk mode.
The models used
demand changes associated with travel time and cost
predict
changes,
i.e., elasticity-based changes rather than absolute levels of
demand.
The
general
conclusion derived from this
lengthy
analysis can be summarized as follows:
Policy 1:
Policy 2:
Policy 3:
Policy 4:
Increasing the size of the ARZ without providing
other transportation improvements results in an
increase in the public transit mode share but in
an overall decrease in the number of trips
to downtown.
For work trips, the latter is the more important
result, as shoppers find suburban alternatives to
replace downtown trips.
For smaller-sized ARZs, providing an internal
distribution system results in little time saving
over policy 1 for transit users and former CBD
for larger ARZs, walk time savings
auto users;
associated with an internal bus system become
significant, and
the number of transit trips and overall shopping
trips increases.
This results in a significant shift to public
transit, including a large increase in ARZ-destined
shopping trips for smaller ARZs.
This policy is similar to policy 1 in its impact
upon work trips, since these occur primarily in
the peak period. The total number of auto shoppers
increases because of reduced competition from
workers for CBD parking spaces.
293
worker's
intra-CBD
mentioned,
harder to access,
demand
the several categories of
Among
Assessment
mode choice
much
is
model
because of a high percentage of walk-trips
and non-home-based trips included in intra-CBD trips, and the
most
limited data on pedestrian movements available in
very
size
The
cities.
survey in this study
is
results of this type of study would
The
small.
of
relatively
provide
basis for evaluating a more comprehensive benefit
useful
a
or
cost-effectiveness analysis.
b. Impact Analysis
contrast to cost-benefit analysis,
In'
impact analysis
has been largely used for demonstration and pilot syudies
in
the transportation field.
(i)
Peter and Watson (1978)
Decision Problem and Alternatives
The decision problem
is addressed in this study is whether the Area
that
Scheme
travel
has resulted in significant impacts on
(ALS)
behavior,
License
traffic
performance,
and public opinion.
environment,
business
activity,
the
Therefore, no alternatives
were examined.
and
Analysis
This study
Findings
carries
out
an
extensive program of data collection before the ALS went into
effect
and
a
follow-up
program once
294
the
scheme
was
in
households,
constituted
which
operatrion.
Car-owning
approximately
41 percent of all households in
Singapore
in
1975, were analyzed separately from non-car-owning households
No
attempt
was made to carry out a
of the scheme.
evaluation
the
interest.
their response to the ALS was of particular
since
following
traffic
pedestrian activity,
A
travel
economic
data,
With the before-and-after
behavior,
air pollution,
identified:
were
areas of potential impacts
performance,
attitudes.
comprehensive
business
activity,
and public opinions and
number of different methods were employed
in
order to examine the effects on these areas.
This study represents a comprehensive
Assessment
ex
post analysis, relying primarily on a number of measurements.
The
data collection efforts were concerned not only with the
of physical or behavioral changes but also
measurement
with
The success of the project was determined in
public opinion.
terms of the results of these measurement.
As with many
of
the ex post analysis, some estimates in this study were based
Figures derived from surveys of respondents,
on respondents.
however,
may
not be terribly reliable.
This kind of study
seems important because the results of the study can
a basis for cost-benefit analysis.
in
travel
times
For example, the changes
resulting from the ALS could
estimate the benefit of time savings.
295
provide
be
used
to
(ii)Delaware Valley Regional Planning Commission
In the case of
Problem and Alternatives
Decision
(1977)
the
Chestnut Street Transitway in Philadelpia, both merchants and
area employees were surrveyed in order to ascertain attitudes
its impact on shopper behavior,
about the Transitway itself,
purchasing patterns,
perception of the downtown area,
Similar surveys were
as well as sugestions for improvement.
by
conducted
Planning
Regional
The
Trenton Commons in New Jersey.
in
Commission,
Valley
Delaware
the
etc.,
stated
purpose of the surveys were "to evaluate the effectiveness of
ARZ as a transportation system management strategy to improve
mobility in activity centers in the Delaware Valley Region".
The questions asked in the
Analysis and Findings
for what
survey were designed to elicit such information as:
activities
walking,
been
do
transitway
people use the
eating lunch,
successful
in
riding transit);
achieving
certain
user
(shopping,
other
has the transitway
goals
(enhanced
commercial vitality, environmental improvement, eased traffic
conditions, encouraged use of transit, relaxed atmosphere for
pedestrians);
transitway
(remove buses,
more parking,
and
would
what
other);
disadvantages
improve
the
quality
allow autos after working
of
the
hours,
mode of transport to work; advantages
area
compared
and personal statistics
(income,
of downtown as a shopping
with other shopping areas;
296
of
perceived
changes
in
since the transitway opened;
changes
in
effect
on
business
activity
customer
accessibility
to
transitway
stores;
collected
conclusions drawn from observation of the
general
The
suggested ways of improving accessibility.
deliverics;
Street
Chestnut
that
were
data
location;
the transitway
in
tenure
length
size of business;
merchants included;
of
asked
The questions
auto ownership, occupation, etc.).
sex,
age,
pedestrian
atmosphere,
vitality,
and
been
downtown
strengthening
improving
successful
its
and
aeshetic
in
downtown
creating a relaxed
its major goals:
accomplshing
has
commercial
environmental
It has not, however, resulted in increased transit
quality.
use or reduced transit vehicle travel times.
This
Assessment
study can be considered one
are
conclusions
addressed.
achieved by the transitway,
directions
ealier
approach
transportation
terms
of
planners.
of
success
and
to
the
In
contrast
in which "success" was defined in terms
of
success
in
efficiency,
this
study defines
the perceptions of merchants,
Such
meaningful
the reasons behind success,
further improvement.
for
allows
study
drawn concerning the degree
be
to
The
the
Wider
ARZ.
most comprehensive surveys in the evaluation of
concerns
of
a
definition
seems
implementers of particular ARZ projects.
297
patrons
more
and
local
meaningful
to
(i) Transportation Research Record
(NCHRP 81)
In general, the aims
Decision Problem and Alternatives
of this study were to provide information for determining the
success
of ARZs for
failure
or
Nicollet
in Nicollet Transit Mall.
and traffic counts and surveys
Analysis
and
surveys
major
for
and
The method employed were two surveys in Boston
Transit Mall.
two
Boston
For the Boston ARZ
Finding
were conducted on the
downtown shoppers and workers.
evaluation,
behavior
of
the
The results of surveys are as
follows:
ARZ
The volume of pedestrians on streets within the
and
on
weekdays
by
5 percent overall
increased
pedestrian traffic increased by
weeknight
Saturdays:
17 percent.
Sales volumes increased between 5 and 10 percent.
Walking to ARZ increased from 48 to 55 percents of a-11
Auto use decreased form 11 to 6.5 percent of
trips.
total trips to the ARZ.
Transit
use to the ARZ by employees imcreased by 23
percent.
1.
2.
3.
4.
For
the Nicollet Transit Mall evaluation,
cordon
for the CBD indicated no decline in auto trips.
was
apparent
that
mall
the
volumes,
but
pedestrian
increase
in the sidewalk area.
showed no change.the
increased
circulation
was
total
of
No evidence
pedestrian
improved
Total pedestrian
by
an
accidents
Importantly, surveys after the opening of
mall showed a decline in shoppers to the CBD.
surveys
counts
merchants
have reported
that
the
However,
predominant
attitude was that the mall had been a good investment.
Assessment
This study appears to be quite general
298
in
A more disaggregate treatment
comprehensive.
not
were
success or failure of project.
the
failure
The success or
be oriented around merchant attitude
to
tends
by
judge
confidently
group seems necessary in order to
impact
surveys
The target groups of the several
relatively small.
was
A sample size
no systematic effort was attempted.
that
rather
than
various impact groups.
(ii)OECD - Gothenberg
A number of studies
Problem and Alternatives
Decision
were
(1979)
carried out in order to evaluate the effect of the
CBD
The decision problem was whether or not
the
traffic cells.
traffic
cell
terms
beneficial in
were
schemes
of
their
effects.
Analysis
and
permanant monitoring activities,
done
in
evaluations
Some of the
Finding
were
others were special studies
of
connection with the implementation
cells.
It
should be emphasized that the time of evaluation was not same
for all impact categories.
success
or
confidence.
on
failure
of
Thus it is difficult to judge the
the
scheme
with
any
of
This research also suffered from a lack of data
some impact areas before the implementation.
methods
degree
employed
range from
interviews
distributions to counts and surveys.
299
and
Evaluation
questionnaire
Perhaps one of the most
of this study related to
aspects
neglected
consideration
No
activities.
was
the
pedestrian
to
pedestrian
given
The effects of
movement before and after the implementation.
The most
various traffic cell schemes were mostly favorable.
success of the cell schemes was the fact that
remarkable
interest groups seem to be worse off.
no
The following positive
effects were stressed in the report:
1.
2.
3.
4.
5.
6.
accompanying
the
cell system and
the
Through
improvements to the environment, the municipalities have
given the CBD a distinctive image and indicated a future
path of development.
The environment for people living, walking and working
in the CBD has improved considerably, particularly
pollution from car traffic and
noise,
concerning
conflicts between cars and pedestrians.
Traffic accidents have been reduced by 36 per cent, on
the average, for the CBD plus the ring road.
The level of service for transit has improved slightly
improvement
'through improved regularity, although the
was not as good as expected.
Traffic has been rerouted to the ring road without
causing problems of any size.
The travel speed of cars has improved in the CBD and on
on the ring road.
the
contributed
to
results
initial
After
method employed in
post
ex
the
note
It is of particular interest to
Asessment
the
this
of evalutions
responsible
done
authorities of
significantly
study
plans of traffic cell
after
that
schemes
through
implementation.
Gothenberg
observed
increases in travel time and a negative effect on regularity,
particularly in the direction of peak hour traffic, plans and
actions
widening
were carried out in 1973 to remove this
the street in order to create physically
300
problem
by
separated
The
effort
opened
the
The line is
now
further
of relocating a tramway line.
possibility
walking
service to passengers due to shorter
better
giving
tramway.
the
for
lanes
distances between the new stops and the main destinations
in
the CBD.
c. Weighting (or Scoring) Technique
(i)Transportation Research Board (1978)
Problem
Decision
studies examined,
a
comprehensive
and
the objective of this study was to provide
for
methodology
social,
the
evaluating
facility
pedestrian
and economic benefits of
environmental
previous
Unlike
Alternatives
comparison of alternative pedestrian facilities
proposals. A
were performed.
This study essentially represents a extension
Analysis
of
the
cost-benefit
of benefits.
enumeration
study,
and
normalized
within
a
established
Each
of
the
for
on
focusing primarily
Thirty-six impact
this
in
identified
usually
approach,
of
1
scoring projects on each of
measures is given a weight
to
were
10,
the
according
were
scales,
rating
elaborate
range
measures
hte
measures.
to
its
importance as seen by community, agency, and other interests.
Because many of the variable are subjective in nature- (e.g.,
comfort,
attracitveness, noise), the calculation of benefits
301
is performed using an unitless scale of positive and negative
values
varialbe.
allows
scoring
Unitless
alternatives without the need
of
comparison
dollar
each
for
assinging
for
pedestrian
values to the many noneconomic impacts of
For
facilities.
scoring
the
example,
the
technique
is
illustrated by the noise impacts of motor vehicles.
In this
case both an appropriate formula and a corresponding
graphic
scale
The appropriate score
are given for value selection.
value is selected after a series of sound readings have- been
taken
and
averaged
for the facility
under
evaluation
in
Table.
Noise Level Scoring
Table III-A-2
90
db(A)
-10
65
52.5
40
db(A)
db(A)
db(A)
db(A)
-5
0
+5
+10
77.5
Any noise level over 90 db(A) scores -10
Any noise level under 40 db(A) scores +10
Total Noise Score = -10 + [(90 -
----
observed or estimated
noise level) x 0.4)]
--------------------------- ---------------------------
302
The
Assessment
of
may
analysis.
eocnomic
values
method
By
can
can
the
from
community
reflecting
of
the
provide adequate justification for projects
not
that are not easily
use
quantified,
defendable previously by economic analysis
it
method
scoring
the
many of many benefit measures usually excluded
conventional
needs
of
in this study is that it allows and encourages
employed
use
advantage
However,
alone.
weights
be faulted for the arbitrariness with which
derived.
be assigned and scores
the
Also,
themselves are dimensionless,
"overall"
offer
and
little
scores
by
sisight
into the merits of a particular scheme.
Finally, the
a
rationalized
presentation
evaluation
information
of
scheme
list of measures in
this
does
required
to
not
obviate
the
make judgements
fact
about
variables may be extremely difficult to obtain.
303
that
the
individual
Download