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