Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A multiple criteria decision process : the case study of rapid transit in Greater Vancouver Oyhenart, Gregory John Peter 1996

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1996-0330.pdf [ 10.81MB ]
Metadata
JSON: 831-1.0099062.json
JSON-LD: 831-1.0099062-ld.json
RDF/XML (Pretty): 831-1.0099062-rdf.xml
RDF/JSON: 831-1.0099062-rdf.json
Turtle: 831-1.0099062-turtle.txt
N-Triples: 831-1.0099062-rdf-ntriples.txt
Original Record: 831-1.0099062-source.json
Full Text
831-1.0099062-fulltext.txt
Citation
831-1.0099062.ris

Full Text

A MULTIPLE CRITERIA DECISION PROCESS: THE CASE STUDY OF RAPID TRANSIT I N GREATER VANCOUVER by GREGORY JOHN PETER OYHENART B.A., Simon Fraser University, 1992 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS (PLANNING) in THE FACULTY OF GRADUATE STUDIES School of Community and Regional Planning We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May 1996 © Gregory John Peter Oyhenart, 1996 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. department of G^^mo/li'ftj OJJAA f^loyici P / k n f t ^ c v The University of British Columbia I Vancouver, Canada Date DE-6 (2/88) Abstract One of the most fundamental issues in the field of planning revolves around evaluation within the decision making process. What determines how governments, or government agencies, allocate public resources? What social, economic, environmental, and political considerations are taken into account? Who is involved in this process? The purpose of this thesis is to examine evaluation methodology using the case study of the decision on the next phase of rapid transit within the Greater Vancouver area. The thesis begins by reviewing the major evaluation methodologies available for the decision: Cost-Benefit Analysis, Planning Balance Sheet, Goals Achievement Matrix, Multiple Criteria Analysis, Multiple Accounts, and the Delphi Process. An analysis of each model's strengths, weaknesses, and historical applications shows that no one is adequate for our decision. The emphasis of evaluation is typically to produce an answer for decision makers. The purpose of the literature review is to see what opportunities exist within these various models to include a process for decision making. The emphasis is not on identifying the best route, but how to decide which would be the best route. The focus is on what factors should be accounted for, and who should be involved in the process. Rapid Transit options have been reviewed and evaluated by two government agencies, the Greater Vancouver Regional District (GVRD) and BC Transit, with conflicting results. The GVRD's evaluation used a quasi-Multiple Accounts, and the BC Transit review was a combination of Multiple Accounts and Multiple Criteria Analysis. The two studies, like the theoretical models they are based on, suffered practical shortcomings. Prominent in both was a lack of public participation in the7 final decision-making process. After reviewing the various methodologies and practical problems with the case study, the thesis offers a model based on simple Multiple Criteria Analysis, Multiple Accounts, and the Delphi Process. The hybrid model is sufficiently comprehensive to account for all of the relevant economic, social, and environmental factors and sufficiently robust to include public input to the decision making process. The aim of the thesis is not to radically change the scope of evaluation methodology, but to set it in a broader socioeconomic context. iv TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vi List of Figures vii Acknowledgment viii Chapter 1 The Role of Evaluation in Planning 1 Thesis Outline 3 Chapter 2 Cost-Benefit Analysis 7 Strengths 12 Weaknesses 13 Conclusion 15 Planning Balance Sheet 16 Strengths 20 Weaknesses 20 Conclusion 21 Goals Achievement Matrix 22 Strengths 26 Weaknesses 26 Conclusion 27 Multiple Criteria Decision Making Models 28 Strengths 35 Weaknesses 36 Conclusion 37 Other Methods - Multiple Accounts Evaluation Guidelines 37 Strengths 43 Weaknesses 43 Conclusion 44 Other Methods - Delphi Process 44 Strengths 46 Weaknesses 46 Conclusion 47 Conclusion: Evaluating the Methodologies 47 Conclusion 52 V Chapter 3 Context 56 The Evaluation Studies of the GVRD and BC Transit 59 GVRD: Transport 2021: Goals, Objectives and Criteria 60 BC Transit: Multiple Account Evaluation of Rapid 64 Transit Options in Greater Vancouver The Current Situation 69 Conclusion 70 Chapter 4 Information 73 Stakeholders 75 Forum 77 An MCA Process for Evaluation the Decision on Rapid Transit in 79 the Lower Mainland Step 1: Identify Stakeholders Groups 79 Step 2: Identify a Set of Decision Options 82 Step 3: Structure Stakeholder Value 84 Step 4: Build Common Value Structures 87 Step 5: Build Quantitative Value Models for each 89 Stakeholder Group Step 6: Use the Value Models for Communication, 93 Conflict Diagnosis, and Resolution Conclusion 94 Chapter 5 Strengths, Weaknesses and Practical Shortcomings of the 97 Combined MCA Process Strengths 97 Weaknesses 99 Practical Shortcomings 100 Items for Future Consideration 101 Conclusion 103 Bibliography 105 Appendix A 109 Appendix B 142 vi LIST OF TABLES Table 1 Planning Balance Sheet Summary: Ipswich Expansion Schemes 18 Table 2 Goals Achievement Matrix 25 Table 3 Weights for the Bicycle Example 32 Table 4 Single-Dimension Utilities of Bikes 34 Table 5 Computation of Aggregate Utilities 34 Table 6 Multiple Accounts Summary 42 Table 7 Greg's Weights of Methodologies 51 Table 8 Comparison of GVRD and BC Transit Evaluations 68 Table 9 Summary of Decision Options 83 Table 10 Hypothetical Weights for Rapid Transit Objectives 88 Table 11 Final Computation of Aggregate Utilities 92 vii LIST OF FIGURES Figure 1 Value Structure for the Bicycle Example 32 Figure 2 Value Structure for the Rapid Transit Decision 86 ACKNOWLEDGMENT A thank you to some key people without whose help this effort would have been impossible: Professor Craig Davis, who was unwavering in his support, enthusiasm, and editing. This thesis would not have been possible without his help. Tricia Kobayashi, who showed remarkable patience and understanding trying to help me make this document what it was. Misa Izuhara and Leanne Martinson, my two best friends from SCARP who will be remembered long after this thesis. Thanks. Colleen Lowe, who has been my partner in this educational odyssey every step of the way. We should all be so fortunate to have a Colleen in our corner. Thank you. 1 CHAPTER 1 One of the most fundamental and intriguing issues in the field of planning revolves around evaluation within the decision-making process. Every year, governments at all levels make public decisions on providing resources for transportation, housing, the environment, and even social welfare. The key question is how, or on what basis, are the decisions made to allocate resources to, and within, these various activities? For example, how does the provincial government decide whether or not it will spend $1 billion dollars on a rapid transit system for the Greater Vancouver area? More specifically, how will they decide on what type of technology to use, and where to locate the system? How does the government decide what factors should be taken into account when making this decision? Finally, who makes the decision on which parties or groups are to be consulted, or included, in this decision making process? The purpose of this thesis is to examine such questions in the context of planning for the decision on the next phase of rapid transit in Greater Vancouver. The Role of Evaluation in Planning The definition and role of evaluation are relatively straightforward. A succinct yet thorough definition provided by McAllister (1980:3) outlines the activities associated with this function. "Evaluation - obtaining, organizing and weighing information on the consequences, or impacts, of alternatives." As McAllister (1980) further noted, the evaluation function actually pervades most of the activities in the planning process. The simplified steps which he outlines provides an adequate understanding of this role. The five step planning process is as follows: 2 1. Identify the problem to be addressed 2. Design alternative solutions to the problem 3. Evaluate the alternatives 4. Decide on the action to be taken through the appropriate political process and implement it 5. Monitor the results Evaluation, whether formal or informal, is inherent throughout this process. Identifying the problem involves evaluation because a decision has to be made as to whose interest is being served, or where scarce resources have to allocated. Designing alternative solutions requires value judgements as to alternatives, or designs of different technologies. In our case study the choice, for practical purposes, is between Light Rail Transit (LRT) or Rapid Bus technology. Finally, monitoring the results requires judgement as to success of an action, or plan, for the purposes of providing feedback for further changes. From a transportation planning perspective, the model described by, Roads and Transportation Association of Canada provides another example of the hierarchical, yet interrelated aspects of planning. The guide breaks the planning function into six steps. Step 1 is to delineate the goals and objectives of the issue. Step 2 is to gather the data. Step 3 is to analyze the policy implications, which in this case are how the transportation policy options for roads, public transport and terminals will impact on land use policy options such as location and intensity of housing and employment. Step 4 is the analysis of option performance, through the use of modeling techniques, of the land use and transportation plans. Step 5 is the evaluation process. Step 6 is the implementation of the plan. The model 3 only differs only slightly from that described by McAllister, yet shows how the steps in the process can be related to the specific aspects of the project being undertaken (1977:109-110). McAllister (1980) divided the evaluation phase into two parts: analysis and synthesis. The analysis phase is the objective phase of the evaluation that defines and estimates the various impacts of the intended action. The synthesis phase attempts to bring these various impacts together, as a whole, so that an integrated view - or judgement - can be formed on whether or not the action should be taken. The synthesis stage is the more contentious of the two because of its subjectivity. Impacts are relatively discernible for accuracy because they are usually based on some objective fact. However, the weighting or consideration that these impacts should have on the final decision are based more on opinion or value. The dilemma is that people, decision makers, may reach different conclusions from the same set of facts about the impacts because their values differ. To further complicate the matter, the evaluation process offers the opportunity to use a number of different tools - or models - from which to draw upon. Evaluation methodologies range from the informal, personal, judgemental type of evaluation, to the more formal, mathematically-based methods, such as Cost-Benefit Analysis (CBA). The type of methodology chosen, as will be seen in Chapter 2, can have a tremendous impact on the decision, and what considerations, or whose considerations, are factored into the decision. Thesis Outline Chapter 2 of the thesis will examine the different methodologies and models available to planners and decision makers. The list includes Cost-Benefit Analysis (CBA), Planning Balance Sheet (PBS), Goals Achievements Matrix (GAM), Multiple Criteria Analysis (MCA), Multiple Accounts Evaluation, and the Delphi Process. This is not a complete list of models 4 available. Indeed, Multiple Criteria Analysis alone encompasses a range of different models, each with its own unique characteristics. These approaches will, however, provide us with a broad enough understanding of the complexities of the evaluation process. As well as examining each of the models theoretical underpinnings, strengths, and weaknesses, we will also examine each of their practical applications and appropriateness for our case study. Chapter 3 will examine the context for this case study, and review two evaluation documents prepared by the Greater Vancouver Regional District (GVRD) and BC Transit. The decision to be made is on which of three routes a rapid transit line should be constructed. The three routes under consideration are: the Broadway-Lougheed corridor, the Coquitlam-New Westminster corridor, or the Richmond-Vancouver corridor. These three options are geographically distinct, and the provision of rapid transit to any one of them would likely have local economic, environmental, and social impacts. The two types of models utilized for the evaluation of options provides a valuable overview as to how the evaluation and decision-making process works, and how the theoretical evaluation models fare when used in practice. The decision on the next phase of rapid transit has been a contentious one for all of the stakeholders involved. The different conclusions reached by the use of these models, and their respective agencies, has not curtailed the debate on the issue. Chapter 4 is the synthesis of the work of the two previous chapters. After reviewing the various models available for the decision, and reviewing how the decision has been conducted, we have the opportunity to suggest an alternative. At the outset of this introductory chapter, we reviewed how evaluation is assessing the impacts of a decision, and then having to subjectively decide which impacts to take into consideration, and how these impacts should be considered. The model that is described in Chapter 4 is an attempt to deal 5 with the problems of exclusion, of people and information, in the evaluation process. More specifically, there is an attempt to include in the process all of the relevant information -economic, environmental, social, and political - and all of the relevant stakeholders necessary for consensus decision-making. By making the process more comprehensive, and less exclusionary, there exists a opportunity for political decision-makers to have new insight as to how their constituents evaluate the issue, and what factors they consider important. Chapter 5 will analyze the strengths and weaknesses of the new model which was developed, and outline how such a process and model could be made practical. The aim of the thesis is not to offer a radically new solution to the problem of evaluation. The answers to overcoming the obstacles to effective and inclusive evaluation lie within the existing evaluation models, or some combination of them. 6 CHAPTER 2 In the landmark book in the field of economics and ecology, For The Common Good, authors Herman Daly and John Cobb (1989:138) take an interesting look at the nature of how the discipline of economics has transformed itself over time. Oikonomia, from which the word "economics" is derived, relates to the management of the household (or community) so as to increase its value to all of its members in the long run. Chrematistics, on the other hand, is that branch of political economy that relates to the manipulation of property and wealth so as to maximize short run monetary exchange value to the owner of the resource. Economics, despite its historical beginnings in the field of moral philosophy and obvious concern with oikonomia, has today firmly entrenched itself in the fields of mathematics and the sciences and thus seems more concerned with chrematistics. This shift from "soft" economics to "hard" economics has, according to Daly and Cobb (1989:19), served greatly to undermine the role and credibility that modern economics plays in our lives: "our intention is not that economic theory begin over again, but that it be reconstructed on the basis of a paradigm that both clarifies the excellence of its past work, and sets it in a larger context." The paradigm shift would include less concern with chrematistics and more with oikonomia. The larger context would include more concern for the community, social, and ecological aspects of economic functioning. Daly and Cobb's remarks provide an interesting and relevant starting point from which to investigate the field of evaluation methodology from a planning perspective. As mentioned previously, there exists a wide range of tools available to planners to evaluate transportation as well as other decisions. While it would be beyond the scope of this analysis to review all of 7 the methodologies in detail, the subset that has been chosen accurately displays the problem which Daly and Cobb discussed. For purposes of clarity, all of these models should be classified as economic evaluation models even though only CBA and PBS make explicit attempts to convey all (or most) of their information in monetary terms. The social and ecological well-being of the community, in our case the people who reside in the GVRD's current boundaries, must be factored into the economic or "oikonomic" analysis, whether or not this is expressed in dollar terms. By starting with the assumption that evaluation can be more than a purely chrematistic deductive mathematical exercise, the planner has an opportunity to expand the scope of what is meant by economic evaluation. Cost-Benefit Analysis The history of CBA can be traced back to 1844 when it was used to measure the utility of public works in France, althoughthe more contemporary notion of CBA was first utilized in the U.S. in 1936. The Flood Control Act of 1936 required that in order for Federal assistance to be offered in flood control measures, economic benefits must be shown to exceed economic costs (Hill 1973). In essence, CBA was to achieve two goals. First of all, it was to indicate economic efficiency. In comparing two or more alternatives, the one which provided the efficient use of the capital resources available was the "best choice". Furthermore, the assumption was that economic efficiency would also ensure that the second goal was met - the maximization of economic welfare. In order to better understand the goals of CBA, it is necessary to examine the main component parts. 8 Pareto Efficiency The concept in economics of Pareto efficiency provides the theoretical foundation of CBA. The Pareto criterion for efficiency dictates that the marginal benefit realized from using a good will be equal to the marginal cost of producing that good. For public goods, such as rapid transit, the sum of marginal benefits for all consumers should be equal to the marginal costs of providing that good. From a CBA perspective, we would judge a decision as a move towards Pareto efficiency if the well-being of one individual were improved without diminishing the welfare of anyone else. Once further welfare improvement cannot be obtained, then the decision is Pareto efficient. It is important to note that this efficient allocation, and thus the maximization of social welfare, looks only at the aggregate costs and benefits: it makes no distinction, in terms of equity, between the individual impacts of the costs and benefits. (Smith 1986) Measurement of Costs and Benefits While the Pareto criterion provides the theoretical basis for meeting the goal of efficiency, the economic concept of utility helps the economist (or planner) understand what the decision holds for both individual and aggregate economic welfare. CBA makes the assumption that social welfare (SW) is equal to the product of the utility (U) that an individual derives from consuming a good or service and the weight (w) of relative importance attached to that individual. The aggregate social welfare is expressed as a weighted sum of the utilities of all individuals. Expressed mathematically: SW=wlUl + w2U2 + ... +wnUn. Evaluation using CBA attempts to measure the changes in social welfare as a result of changes in the different utilities (McAllister 1980). For example, measuring a positive change in consumer surplus would denote, for purposes of CBA, a positive change in social welfare, or a benefit. 9 Costs There are two types of costs which need to be accounted for in any type of project: direct and indirect. Direct costs, the more easily derived of the two, are those that are used to provide the project's output. For instance, the direct costs associated with providing rapid transit would be the total outlay of capital needed to provide the final product. It would include, for instance, the track, the stations, and the train cars. These are also referred to as project resource costs (Schofield 1987). Indirect costs relating to any project, including rapid transit, are also important to measure. Two aspects which are invariably impacted with any such project are the social and environmental areas. Social dislocation costs would not only relate to the direct monetary compensation which may have to be paid, but also to the values people place on items such as loss of neighbourhood, visual intrusion, and increased inconvenience. Negative environmental impacts constitute a further cost for such elements as increased noise levels, increased pollution, and loss of open space for habitat or recreation. These are only a sample of the possible indirect costs that need to be accounted for in evaluating any decision; many more could result from specific local circumstances (Schofield 1987). Benefits The measurement of benefits of a project can also be categorized into two groups: direct user benefits and indirect, non^ user benefits. Relating to a rapid transit project, user benefits are calculated on vehicle operating costs savings, travel time savings, accident costs savings, and comfort and convenience. Non-user benefits are calculated, for example, on the net time, money and risk savings of those people who continue to drive their cars in less 10 congested traffic (Schofield 1987). Once the cost and benefit categories have been determined, the CBA analyst then has the task of assigning values to them. Willingness To Pay/Accept Although there are many methods that a CBA practitioner has at his or her disposal to determine values, the most common and convenient method to use is contingent valuation (CV) measurement of a person's willingness to pay (WTP), or willingness to accept (WTA), a particular condition. As there are usually no market price indicators for many of the costs and benefits associated with a change, CBA must try to ascertain the values that those affected place on the costs and benefits. Through the use of surveys or interviews, respondents are asked to indicate a price, usually expressed in dollar terms, as to what they would be WTP or WTA for a change in, for instance, a visual amenity. If, for example, a park were to be provided in a residential area, CBA would question the residents, using CV, to find out what would be their maximum WTP to have the park. The sum of the responses, or the average of the survey responses multiplied by the number of total residents, would be tabulated to provide a dollar value benefit that those people put on having the park provided. Conversely, one may need to calculate a cost for having to remove the park from the neighbourhood. Respondents would then be asked how much financial compensation they would be WTA to lose their park.1 One of the major assumptions of WTP and WTA which should be noted here is that the individual's responses under these hypothetical circumstances correspond exactly to what they would be if the choice was actually purchasable under market conditions (Smith 1986). Although CV is perhaps the most common way for a CBA to determine the direct and indirect costs and benefits, other methods do exist. In the field of transportation planning, 11 other methods that have been used are travel time savings and accident costs savings. For travel time savings, benefits can be determined by finding the amount of working time saved by the reduction in travel time. If, for instance, a person saves 20 minutes on a trip per day and his or her hourly rate is $18.00 per hour, then that trip savings is assumed to be a benefit of $6.00. Improvements in transportation infrastructure, which have historically been new roads, are usually expected to result in a reduction in accident rates. As such, there should also be a cost savings in terms of property damage, injury, and even death as a result of the new roadway. These cost savings are seen as benefits of the project.2 As it would be impossible to tabulate these costs and savings precisely, the CBA practitioner would be expected to estimate their potential effects. A further method, and one not related specifically to transportation, is hedonic or shadow pricing (Davis 1990). Hedonic pricing assumes that a commodity that does not have a market price can be "priced" by examining its various attributes which do have market prices.3 By examining the parts and establishing a price for them, the analyst may then use these component prices to estimate the price of the whole commodity. As with CV, the benefits are expressed in dollar terms. The Role of Time and Uncertainty Once the CBA practitioner has defined and calculated the costs and benefits associated with a project, the final task is to adjust the costs and benefits to a net present value (NPV). This is done because a large public project, such as rapid transit, will have streams of costs and benefits which will occur far into the future. Consequently, CBA must obtain the present values of those future effects so that a comparison can be made on an equal and rational basis. More important for our purpose is to highlight two important assumptions CBA makes in dealing with the future. First of all, it assumes that societal preferences, as obtained in WTP 12 and WTA measures, do not change over time. The values expressed today are treated as a constant for future generations. Secondly, and perhaps more importantly, the CBA practitioner must establish a discount rate which will be the rate by which the future costs and benefits will be discounted. Generally, analysis is conducted with several rates so that the sensitivity of the results to different calculations can be compared. Sensitivity comparison allows the CBA analyst the opportunity to compare impacts under different sets of assumptions. If the outcome does not change despite the different discount rate which is applied, the project is not sensitive to the assumption, and the evaluation is not significantly dependent on the particular rate being used.4 CBA Strengths and Weaknesses In reviewing the literature on CBA, there is definitely no shortage of opinion on the strengths and weaknesses of the technique as an evaluation model. Indeed, the misgivings of CBA range from general ethical concerns to more abstract theoretical microeconomic shortcomings. For our purpose, however, we will focus on the more general concerns which are found in the majority of literature discussing the subject.5 Strengths • CBA provides an opportunity for analysis from a social viewpoint. • CBA offers a systematic approach to decision making that is derived from a consistent and long established body of economic principles. The use of NPV and sensitivity analysis using different discount rates are but two examples of how the 13 technique utilizes standard and accepted measures in dealing with future economic uncertainty. • CBA, by using CV surveys, attempts to include the values of all those affected and not just a select few. • The methodology is explicit in the assumptions and figures that it uses and can convey its results in easily understandable terms. As such, it offers opportunity for debate and challenge once the analysis has been conducted. • The model is relatively adaptable, and is open to modification depending on the circumstances. Perhaps the biggest strength of the model is that it has been used in such a wide variety of public sector decisions, ranging from water supply, transportation, land use, health, education, research and development, and defence. Such an extensive body of literature, on obviously diverse topics, serves as a valuable tool for future CBA practitioners.6 Weaknesses7 • Deficient and often unreliable data can serve to undermine the credibility of the results. Moreover, as the model is data demanding, it is relatively expensive to conduct. • Secondary benefits are often included in the analysis, which results in a double counting - or over estimation - of the benefits of the project. As such, including secondary benefits tends to favour the project being undertaken. • The treatment of equity considerations has traditionally been poor. Who bears the costs, and who benefits from a project are not necessarily explicitly stated. The 14 Pareto criterion is only concerned with moving towards efficient allocation and aggregate social welfare. • The measurement of costs and benefits of non-market commodities, such as environmental and social impacts, rests on a tenuous theoretical basis. Not only are the accuracy of WTP and WTA values often somewhat suspect in their own right, they also require that the respondents have the ability to pay for the values they are expressing. As such, the responses of lower socio-economic groups, who may be affected by a project and have strong feelings against it - or perhaps towards it - are excluded from the process. In general, there has been some reservation to the pracitce of monetizing impacts, especially those regarding human lives. • The assumption that contemporary social values will not differ greatly from future social values is an important shortcoming. Few would have predicted the overwhelming growth in the concern of general attitudes towards the environment, especially in British Columbia, just twenty years ago. The inability to deal with future uncertainty calls into question the validity of the values expressed today being constant for future generations. • The discount rate which is used to bring future benefits and costs to the present day is a source of great contention, and is easily subjected to the forces of politics. The scope and magnitude of these weaknesses serve to question whether or not CBA is always the best evaluation tool to use. Despite its historically widespread use, the inherent weaknesses of the model should be a caution to those who wish to use it in the future. While 15 it may be possible to conceptually modify the model to handle some of these shortcomings, in practice it proves to be difficult, expensive, and time consuming. Conclusion CBA provides decision makers with a valuable evaluation tool provided that its practitioners understand the limitations of the model. These limitations, according to Smith, (1986:31) "do not imply the equivalent of a 'council of paralysis.' They identify the importance of (1) quantifying the degree of uncertainty in each set of benefit-cost estimates ... and (2) implementing a program of research that systematically evaluates the consequence of pragmatic judgements ..." Smith, like other evaluation specialists, believes the future of CBA lies in the refinements of its techniques, like CV and hedonic pricing, and dealing with intangibles.8 Two areas in which CBA would seem to be most appropriate are: (1) For public projects which are primarily financial in nature, and where the social and environmental externalities would not differ significantly from what currently exists. The S.T.O.L. Airport Project would seem to be the type of project for which a CBA is well utilized under this criteria (Davis 1990:115-117). (2) For projects which will have an easily recognizable and local impact on a relatively small group of people. Such a situation would allow CBA to deal effectively with the problem of value elicitation, and incorporate relatively more people into the process than would otherwise be included.9 For the purposes of our case study, and in light of its practical weaknesses, CBA would prove an awkward and difficult tool to utilize. 16 Planning Balance Sheet In his review of CBA as a tool for regional science evaluation, economist John Schofield (1991:319) recounted a humorous anecdote for why PBS and GAM (to be discussed next) are better tools than CBA, which is sometimes referred to as 'a horse and rabbit stew.' Great care is taken to measure the precise monetary terms of the small portion, in this case the rabbit, and then the great amount of intangibles, in this case the horse, is thrown into the pot, thus "leaving the analysis with a rather dubious flavour." The analogy of the problems of a horse and rabbit stew is an accurate introduction as to why the PBS was developed. PBS was the brainchild of planner Nathaniel Lichfield, and was developed in response to two of the major shortcomings of CBA. First of all, PBS attempts to deal directly with the problem of identifying the incidences of costs and benefits - essentially who is bearing the burdens of the decisions and who is reaping the benefits. Secondly, PBS is interested in trying to emphasize those aspects which are not easily computed into dollar values, and thus de-emphasize somewhat those costs and benefits which usually received the bulk of consideration (Davis 1990). As Lichfield (1970:156) states: "... planning balance sheet groups the community into various homogeneous sectors distinguished by the kinds of operations they wish to perform. It then evaluates and compares the alternatives from the point of view of the advantages (benefits) and disadvantages (costs) accruing to every sector from each alternative, to see which would provide the maximum net advantage (benefit).... It follows that as well as those benefits and costs which are measurable in money terms, there are others that are only measurable in some other unit (time, physical) and others that are not measurable. Thus, the balance sheet cannot, and does not, aim to provide a conclusion in terms ... measured by money values... Its value lies in exposing the implications of each set of proposals to the whole community and to the various groups within that community... The purpose of the approach is the selection of a plan which, on the information available, is likely to best serve the total interests of the community." 17 As PBS is primarily an adaptation of CBA, the basic economic principles and notions of efficiency, consumer surpluses, and social welfare are unchanged. Indeed, the planner using PBS would undertake to gather the same information he or she would to undertake for a comprehensive CBA. What is added, however, is the presentation of where, or to which groups, the various costs and benefits are accruing. The final results are to be laid out in sets of detailed accounts on the balance sheet, and decision makers are then called upon to weight the relative importance of the impacts, both monetary and non-monetary, as well as how those impacts affect the different groups (Schofield 1987; Lichfield 1975). The case of the Ipswich town development study conducted by Lichfield (and reviewed by Schofield) is an excellent example of a typical PBS analysis. There are three basic steps in PBS evaluation: (1) Identification of the different community groups which could be impacted. These are often disaggregated into producer and consumer groups. (2) Classification of the various impacts on the different groups. The impacts are stated in monetary terms where applicable, or else denoted by other measures (time impacts, physical impacts) or simply noted as an intangible. At this point in the evaluation, the plan should have some stated objectives or goals which the plan is presumably to working towards. The impacts stated in the various terms could alternatively be designated by a "plus" or "minus" sign to show both the order and direction of magnitude towards the objective. (3) The final step is to then compile a summary comparison of the various schemes with one another, and in light of the various sectors and groups.10 The following is an example of how the impacts are presented to decision makers. 18 Table 1 Planning Balance Sheet Summary: Ipswich Expansion Schemes Sector Schemes I II III IV V Producers/Operators Development Agency 0 -4 -4 -5 -7 Current Landowners Displaced: In Urban and Village Areas 0 0 0 0 0 Agricultural Landowners and Farmers 0 +3 +1 +2 +4 Not Displaced: In Urban and Village Areas 0 -1 -1 -1 -2 Farmers 0 +2 +2 +1 +2 Local Authorities and Ratepayers 0 -4 -2 -1 -3 Producers/Operators Overall Reduction 0 -4 -4 -4 -5 Consumers The Public in the Expanded Town In Town and District Centres:: Commercial Occupiers 0 -4 -4 -7 -7 In Residential Areas: Remaining or New Occupiers 0 -4 -4 -4 -5 In Principal Industrial Areas: Industrialist Workers 0 0 0 0 0 Users of Regional Open Space and Countryside 0 -1 -1 -1 -1 On Principal Communications System Vehicle Users: Internal Traffic 0 -2 -6 -4 -3 External Traffic 0 +1 0 -1 +2 Through Traffic 0 +1 -1 0 0 Public Transport External 0 0 0 0 0 Pedestrians 0 0 0 0 0 Current Occupiers Displaced: In Urban and Village Areas 0 -1 -1 -1 -2 Agricultural Occupiers Not Certain Not Displaced: In Urban and Village Areas 0 -1 -1 -1 -2 Agricultural Occupiers 0 +2 +2 +1 +3 Consumers and Overall Reduction 0 -9 -16 -18 -15 Producers/Operators and Consumers Overall Reduction 0 -13 -20 -22 -20 Source: Schofield (1987:151) 19 Table 1 represents the final product of a PBS analysis. The table, and the numbers, represent the relative merits of five different land expansion schemes from the viewpoint of two different groups of townspeople: Producers/Operators and Consumers. Within each group, there are various sub-groups delineated by economic function and geographical location. The scores for each scheme and each group reflect the magnitude and impact (benefit denoted by +, detriment denoted by -) of the particular scheme on the group. For instance, for current landowners who are displaced from agricultural areas (farmers), all but scheme I have a positive impact, but the impact of scheme V is the best one of the group, and is four times more positive than scheme III. Overall, the land expansion schemes appear to have a negative impact on both the Producer/Operator groups as well as the Consumer groups. Scheme I is completely neutral to all groups and is likely a status quo scenario. If a scheme were to be adopted, scheme II would be the obvious choice as is appears to be the least negative of all the options, assuming scheme I is being ruled out because the status quo is not a political option. The mechanics of the model are relatively straightforward. The scores attributed to each scheme would represent the monetary and non-monetary impacts of that scheme on the group. As PBS is an extension of CBA, any monetary impacts would likely be presented in NPV terms. However, this PBS model is more concerned with representing the distributional impacts than with presenting the timing of the impacts or dealing with future uncertainty. This may not be indicative of all PBS models, but it is clear that this PBS model is less concerned with a single deductive answer incorporating all of the factors that a CBA would. Although we have only highlighted the results of what a typical third step in a PBS process would look like, it is important because it gets to the central theme of what Lichfield 20 was trying to show with the PBS - the advantages and disadvantages of different schemes to different groups without a single, monetary based deductive answer. To be sure, these impacts were assessed for both their monetary and non-monetary implications. The evolution from CBA, however, is that the monetary impacts are now no longer the primary objective of the process of evaluation. Finally, Lichfield's model does offer the decision makers the opportunity to compare the impacts on the different groups, and all the groups as a whole. PBS Strengths and Weaknesses Strengths As far as overcoming the two principal shortcomings of CBA mentioned previously, the PBS makes a somewhat successful attempt. It offers decision makers the opportunity to compare the impacts of all the schemes, on all the different groups, in a fairly straightforward manner. Both the efficiency and equity considerations are undertaken simultaneously. Furthermore, the results, such as with the Ipswich example, are relatively easy to interpret. By beginning its search for groups potentially affected by a proposed plan, it would help to identify which groups are going to receive the bulk of the benefits (advantages) or the bulk of the costs (disadvantages), including the externalities of a scheme. Finally, there appears to be a concerted effort to incorporate community goals or objectives into the plan.11 Weaknesses Despite its success in overcoming some of the problems of CBA, PBS has its own share of shortcomings as an evaluation methodology. As PBS attempts to include even more information than the data-demanding CBA, it has a practical limitation of being costly and expensive to perform in terms of both time and money. The different groups are generally 21 treated as equals in the model. However, in looking at the Ipswich town model presented above, the larger numbers of consumer groups over producer groups tends to skew the model as their negative impacts are two to almost four times higher, but they are treated in the end as an aggregate. The model makes no allowances for the groups weighting their own impacts, even though it would be relatively easy to accomplish. A final weakness is that the assessment of impacts between steps two and three are left to the discretion of the decision makers (or as Lichfield refers to them, decision takers). However, there is no explicit consideration given as to how the evaluators weight of the impacts in their own right, and also for how they affect the various groups. Furthermore, the assumption that various groups have homogeneous interests, and are easily categorized, is a weak one. Once again, the sheer quantity of information needed for PBS analysis makes it a complicated process.12 Conclusion Although PBS is a useful evaluation tool, and a meaningful adaptation of CBA, it has been little utilized in urban planning. One of the reasons it has failed to entice planners is because it is quite a demanding model to use. Furthermore, the fact that it deviates quite significantly from its neoclassical welfare economics roots in CBA, by departing from the sole reliance on monetary evaluation, means that economists have little use for it.13 For the purpose of our case study, PBS would seem to offer little more than CBA. However, the attempts at defining impacts to different groups, and the ranking of those impacts by decision makers will prove important to our task. As such, PBS at least offered the notion that this should be done. Further models may help to show how it could be done. 22 Goals Achievement Matrix The Goals Achievement Matrix (GAM) was developed in the late 1960's by Morris Hill and was intended to evaluate transportation proposals. In fact, GAM was developed following PBS, to deal with some of the shortcomings that Hill saw with Lichfield's model, specifically the exclusion of community objectives or goals in the process. Hill believed that various proposals or alternatives should be evaluated on the basis of whether or not they are moving towards meeting community goals (in which case they are benefits) or whether they are moving away from them (costs). Like PBS, GAM attempts to incorporate the impacts of a proposal or project on the different groups in the community, but on the basis of what the community's objectives are.14 The explanation by Hill (Lichfield 1975:96-97) of what GAM is hoping to achieve is useful: "By determining how various objectives will be affected by proposed plans, the goals-achievement matrix can determine the extent to which certain specific standards are being met. Is the transportation plan likely to meet the minimum accessibility requirements and minimum standards of comfort and convenience? Are levels of air pollution and noise likely to exceed specified standards? Is the fatal accident rate within prescribed acceptable limits? These are the types of questions that the goals-achievement matrix is designed to answer." Like CBA, GAM attempts to discern whether the benefits of an alternative will outweigh the costs, remembering, however, the specific definitions of the terms benefit and cost stated earlier. For each community goal defined, a cost-benefit account is prepared. Like PBS, tangible costs and benefits can be expressed in monetary terms where applicable. Furthermore, where the tangible costs and benefits cannot be converted into monetary terms, they can be expressed in some other quantitative fashion in terms derived from the stated goal (e.g. decrease of x parts per billion of airborne pollutants). Finally, intangible items are accounted for. Once the costs and benefits for each of the three subgroups are determined by the analyst, the decision-maker(s) is then ready to weigh alternative courses of action against 23 one another. The weighting in GAM analysis pertains both to the community's perceived value for all of the goals, and the relative valuation of the incidence of costs and benefits for the specific goals.15 Like CBA, GAM makes assumptions about two important aspects of evaluation: uncertainty and time. Hill (1973) suggests that uncertainty could be mitigated by using conservative estimates of impacts, and looking at possible ranges of future impacts rather than trying to predict one outcome. In terms of valuing the impacts of costs and benefits which will accrue at different times, Hill believes one way to treat the problem is to show what impacts will happen in the different periods, and leave the comparison to the decision makers. If a discount rate is necessary, he feels that the market rate is not the best one to use. The discount rate should reflect social and political objectives as well as pure market objectives. The presentation of the results of GAM would take the form, obviously, of a matrix of alternatives by objectives or goals. The presentation of results for comparison purposes could conceivably be done on four separate scales: (1) a nominal scale which classifies numbers and entities; (2) an ordinal scale which simply ranks entities; (3) an interval scale which provides equal intervals between entities and indicates the distance of entities from some arbitrary origin; (4) a ratio scale which provides equal intervals between entities and indicates the distance from some non-arbitrary origin. Although Hill suggests utilizing the highest order scale which can be employed, he conceded that the ordinal ranking system was the easiest to use, and the one subject to least criticism.16 Once the relative weightings have been assigned, the final task is to compare the results. To this end, GAM has several possibilities. The simplest method is to present the results in different accounts without attempting to synthesize the alternatives and objectives, thus leaving the weighting to the decision makers. A second way is to combine the 24 alternatives and objectives to form a weighted index for each alternative based on its ordinal rankings. This however, requires assumptions about the interrelationship between the measurement of an individual objective and that objective's measurement of achievement. In essence, to treat the answer as a simple product may not be totally accurate. Another more complicated method is to utilize a transformation function to aggregate different units into a single unit of measurement. Hill (1973) feels that this is conceptually promising, but difficult to put into practice. The following is a simple GAM model utilizing measurement by ordinal scale. 25 Table 2 Goals Achievement Matrix Wt. Weighted Accessibility Wt. Weighted Community Disruption Community Group a Group b 2 3 1 Plan A +6 -2 +4 PlanB -6 _±2 -4 1 3 2 Plan A -3 0 -3 PlanB 0 -2 -2 Weighted Index Method Weighted Index of Goal Achievement Plan A = 1 Plan B = -6 Plan A is thus preferable to Plan B Source: Hill (1973:39) With the example above, Plans A and B are being compared by two community groups, a and b, on the basis of two criteria: accessibility and community disruption. For the community as a whole, accessibility carries twice as much weight as disruption, 2 to 1. For each of the two criteria, the two community groups attached weights as to how the plans meet the criteria. As well, each group gives its own weighting for the importance of each criteria. Accessibility is obviously more important to group a than b (300%; 3 to 1), and disruption is only marginally more important for group a than b (50%; 3 to 2). For group a, Plan A increases accessibility considerably (3x2 = +6), but also causes disruption (3x1= -3). The answer is positive if it moves towards meeting the community goal, or negative if it is moving away from meeting that goal. Plan B, on the other hand, fares poorly in terms of accessibility (3x2 = -6). Plan B is neutral in terms of disruption (3x0 = 0). If the plan neither moves towards the goal nor away from it, the weighted score must be assigned a 0 score. For group b, Plan A offers poor accessibility, but is neutral in terms of disruption. Plan B is 26 considerably better in terms of accessibility, but has negative impacts in terms of community disruption. When the weighted indices of each groups measurement of each plan is tallied, Plan A is preferable to Plan B. The final score is calculated by adding the scores for each plan in terms of the two criteria. Plan A = +4 + (-3) = 1. Plan B = (-4) + (-2) = -6. Overall, therefore, Plan A with a positive score of 1 is moving towards meeting the community's goals. Plan B, with a score of -6, is obviously moving away from meeting the overall goals. GAM Strengths and Weaknesses Strengths As with one of the major strengths of PBS, GAM does a credible job of attempting to show the incidence of cost and benefits of different groups affected by a proposal or alternative. Furthermore, by the use of one or more scales, GAM makes a conscious effort to rank or weight the costs and benefits. It makes an explicit assumption that not all costs and benefits can be treated equally. An obvious strength of the model is its attempt to incorporate the values of the community into the decision process, although the model as expressed by Hill does not say how this is to be accomplished. Finally, the model offers consistency of application because results can be expressed as a function of simple, or complex, mathematics. The example above could easily include a third group or alternative plan, and arrive at answer consistent with the principles and logic of the original example.17 Weaknesses One of the limitations of the model which was pointed out by Lichfield is that GAM is useful only when evaluating decisions in a single sector, and that planning issues, and costs and benefits, typically affect more than just one sector of society. However, this is somewhat 27 like criticizing a Volkswagen because it is not a Porsche. Lichfield ignores the fact that the model was intended primarily for single sector analysis. Hill would likely agree that GAM is not necessarily suitable for multi-sectoral decisions, such as building highways or schools. Indeed, others have taken Lichfield's view that the model is better at representing information than providing decision makers with answers. As a limitation, it is a minor one. One of the significant weaknesses of the model is that, although it attempts to incorporate the values of the community, it is not explicit on how this should be accomplished. Indeed, Hill's model begins with the assumption that the values are already in place, and the GAM practitioner proceeds from there. A second weakness of the model is how it treats time and uncertainty. In this sense, it really does not progress beyond the problem as it occurs with CBA. Expecting the decision makers to use conservative estimates or low discount rates does not really solve the problem. In fact, one could easily argue that Hill's model leaves too much to the discretion of decision makers in terms of weightings and rankings. Finally, the mathematical rigor of the model tends to be easily compromised when implemented in practice. The simple math utilized in the example provided here may not necessarily capture the complexity of a different problem or decision (Hill 1973; Lichfield 1975; McAllister 1980; Schofield 1987; van Delft 1977). Conclusion The interesting departure that GAM has made from its predecessors, and a point that is a major criticism by Lichfield and others, is that GAM seems more appropriate for displaying information than pointing to a decision for the betterment of social welfare. Although this is true, this does not undermine the credibility of the model. An evaluation model need not necessarily give an answer, or point the decision makers in the direction of 28 economic efficiency or improvement of social welfare. As McAllister (1980:169) notes, "One might take the position that an evaluation method based on a sophisticated but erroneous theory of social welfare is more dangerous than one without a theory at all." It is easily argued that the notions of what action would constitute the best efficiency or social welfare are really a matter of opinion and perspective, and that no results from the model would be totally defensible or dependable. If this is the case, then at least GAM offers some insight to decision makers about how to structure the decisions, and present the information needed for these decisions. The importance of GAM is that it shows that community objectives or goals should be factored into the decision, even if these goals are competing and conflicting. To improve upon the model would be to find a way to structure the decisions with more input from the public, and to define and present society's conflicting objectives. Multiple Criteria Decision Making Models Hill's GAM model was one of the earlier iterations of what could technically be described as multiple criteria decision making models. Whereas both PBS and GAM are easily categorized because each was primarily the work of one person, these latter models are more difficult because they have many proponents and many differences. The purpose of this section is to review some of the background of what these models are trying to achieve beyond what GAM has achieved. Like CBA, PBS, and even the more complex forms of GAM, there are economic, mathematical, and even computer programming aspects of each model that are beyond the scope of this study. The purposes of this section are to discuss these models in general terms, and to show how they have been utilized in practice. According to Edwards and von Winterfeldt (1987:142-143), multi-attribute utility theory (MAUT), one of the main forms of Multiple Criteria Analysis (MCA), involves a six 29 step process. MAUT " is a method for quantifying and analyzing complex individual preferences among decision alternatives that vary on multiple conflicting objectives." The six steps are: 1. Identify stakeholder groups; 2. Identify a set of decision options; 3. Structure stakeholder values; 4. Build a common value structure; 5. Build quantitative value models for each stakeholder group; 6. Use the value models for communication, conflict diagnosis, or conflict resolution.18 Identifying stakeholder groups is the first step in the MCA process. This step actually goes beyond Lichfield's PBS, because it is not only designed to evaluate the impacts on these groups, but also to elicit the values of the different groups in order to assist the process. MCA assumes from the outset that any type of decision will have differing impacts, be they economic, social, or environmental. The need to identify and include as many possible stakeholders is an important step in ensuring the decisions eventually reached with the help of the model are equitable and legitimate. As a rule of practice, it is important to involve stakeholders in the process as early as possible, because their input may help to guide the data collection needed. Essentially, the model at this point is not fixed, and depending on the number of, and differences among, the stakeholder groups, the model may take on a variety of forms (Keeney 1982; Keeney 1988). Identifying a set of decision options builds on the concepts of GAM, where decision makers are faced with deciding amongst a set of alternatives in light of various community objectives. In MCA, once the stakeholder groups have been identified, this task involves 30 eliciting both the objectives or goals which a group deems important, and also what relevant alternatives it views as fulfilling those stated objectives. While this may seem like a relatively simple task, it is perhaps the most difficult for the decision makers to deal with. Aside from the practical problems that will be discussed later, it is important for decision makers to be able to deal with the vast quantity of information from the different groups, and be able to categorize it so that no objectives or values are excluded. At this point, the different stakeholder groups should be working with roughly similar objectives, (e.g. economic, social, environmental, and economic development) and alternatives, (e.g. option A, B, or C.) A final note about objectives is that they should be geared towards ends, not means. For instance, an "end" environmental objective may be to lower the level of air pollution. A "means" objective may be to have fewer cars on the road. In this sense, having fewer cars on the road is only important insofar as it helps to lower the levels of pollution; it is the means to the end.19 Structuring stakeholder values is the key to the process of MCA. This is the part of the process that best resembles the value elicitation process in CBA, where people are asked to express their values for the objectives and alternatives at hand. The stakeholder groups, either as individuals or through a spokesperson, work with the MCA practitioner in putting values on the items which have been delineated. For a simple problem, the values for objectives may be expressed simply as an ordinal ranking, e.g. 1 through 5. A more complex method would attempt not only to identify the ordinal rankings, but also their relative position to one another. For instance, 1 may be four times as important as 2, which is twice as important as 3, while 4 and 5 are only half as important. An even more complex method, as described by Keeney and Raiffa (1976:96-97), would attempt to analyze the range and sensitivity of the value tradeoffs that are being made by the stakeholders. Through a series of questions, respondents would be asked how much of x they would be willing to trade for.y, 31 given a series of possible numerical values. With all of these methods, the MCA analyst works with the stakeholders to elicit and establish their values for the objectives and alternatives. The major difference from other methods, most noticeably CBA, is that the values are not necessarily tied to monetary values.20 The initial three steps as outlined are the integral steps in MCA. The final three steps essentially involve integrating the stakeholder groups responses and values into a common structure. How this is done, however, is dependent on the type of decision which needs to be reached, and how many, and how different, the stakeholder groups are. Indeed, there are numerous examples of MCA studies, both simple and complex, which show how the process differs once the values are obtained.21 To elaborate how MAUT would work for a practical application, Edwards (1981) used the example of how to choose which of three bicycles to purchase. Step 1 was to identify the stakeholder groups, in this case Edwards and his son Page. Step 2 was to identify a set of decision options. The objectives (or criteria) which were decided upon were performance and comfort. Performance related to the number of gears, assuming more gears relates to better performance, and weight of the bike, assuming that a lighter bike translates into better performance. The criterion of comfort related to how comfortable the seat was, and whether the size of the bike was appropriate. Skipping ahead quickly to Step 4, Edwards produced the following Value Tree. 32 Figure 1 Value Structure for the Bicycle Example Aggregate Value A B Performance Comfort AA AB BA BB # of gears weight seat comfort size Once the common value tree had been produced, Edwards was then able to structure the stakeholder values into the model. The following table shows how the values were structured. Table 3 Weights for the Bicycle Example Ward Page Common Normalized Final Normalized Final Normalized Final Weight Weight Weight Weight Weight Weight (A) Performance .4 .8 .6 (B) Comfort .6 .2 .4 (AA) # of Gears .5 .20 .6 .48 .55 .33 (AB) Weight .5 .20 .4 .32 .45 .27 (BA) Seat Comfort .8 .48 .25 .05 .575 .265 (BB) Size .2 .12 .75 .15 .425 .135 Using Ward Edward's value information, it can be seen that comfort is given marginally greater consideration than performance (A = .6, B = .4). In relation to the smaller component parts relating to performance, the number of gears and weight of the bike are 33 weighted equally (AA = .5, AB = .5). Seat comfort, however, is over four times as important than size (BA = .8, BB = .2). Once the normalized weights for each of the sub-components has been elicited, the final weight is calculated by multiplying the given weight by the normalized weight provided for the main criteria under which it is listed. For Ward Edwards, the value of seat comfort will have the greatest impact on his decision (B x BA, .6 x .8 = .48) Seat comfort will account for 48% of the decision that Ward makes on the bike. For Page, however, the number of gears and weight are very important (.48 and .32 respectively), while seat cornfort is basically insignificant (.05 or 5% weight for the decision). The table includes a compromise weighting for the two stakeholders which represents their compromise, possibly even consensus, value structure. Under this category, the normalized weights are simply the average of each of the two participants normalized weights. The final compromise weights are once again the product of the normalized weights (AA final = AA normalized x A, .33 = .55 x .6). Now that the stakeholder values have been obtained for the criteria, the participants decided on the alternatives, which in this case is the option of which bike to purchase. Three bikes were chosen, and a score was given to each as to how it performed in each of the four sub-component criteria. This score represents the utility values and is subjective. A score of 0 means that it performs the worst for that element, and a score of 100 mean it performs the best. Although not stated, the scores were presumably obtained by compromise between the two stakeholders. 34 Table 4 Single - Dimension Utilities of Bikes Gears Weight Seat Comfort Size BikeX 30 10 100 100 Bike Y 100 80 40 100 BikeZ 100 100 10 100 Aside from size, which was not a determining factor because both stakeholders require the same size of bike, the other three values are linear functions of objective measures. In terms of gears, Bike X is a three speed bike, while Y and Z are both high performance ten speeds. In terms of weight, Bike X is the heaviest, and poorest performing, while Bike Y is marginally heavier than Bike Z. In terms of seat comfort, Bike X is by far the best of the three. And although size was not a factor in this study, it could have been in another case and was thus included. The final part of the exercise is to compute the aggregate utility scores for each stakeholder and the alternative choices they have. The following table is how the MAUT example worked for Edwards and his son. Table 5 Computation of Aggregate Utilities BikeX Bike Y BikeZ Dimensions of Value W. wt. P. wt C. wt. W. wt. P. wt. C. wt. W. wt. P. wt. C. wt. AA (gears) 6.00 14.40 9.90 20.00 48.00 33.00 20.00 0.28 33.00 AB (weight) 2.00 3.20 2.70 16.00 25.60 23.60 20.00 0.32 27.00 BA (seat) 48.00 5.00 26.50 19.20 2.00 10.60 4.80 0.50 2.65 BB (size) 12.00 15.00 13.50 12.00 15.00 13.50 12.00 15.00 13.50 SUM 68.00 37.60 52.60 67.20 90.60 80.70 56.80 95.50 76.15 35 Quite simply, the final score for each alternative is the weight that each stakeholder gave to that criterion, multiplied by the utility measure of how that alternative achieved on the criterion. For Edwards, the total utility of Bike X is (30 x .2) + (10 x .2) + (100 x .48) + (100 x .12) = 68. This calculation is done for each of the stakeholders on each of the alternative Bikes, including the computation for their common value. From the table, it can be seen that Bike X fits best according to Edward's value structure, but Bike Z is best for Page. Bike Y, however, offers the best alternative based on a compromise weighting. At this stage, Edwards has demonstrated Step 6, essentially providing a table of values for communicating results and assisting in the decision process.22 MAUT Strengths and Weaknesses Strengths The most obvious strength of the model that it is inclusive. The use of the public as stakeholders in the process ensures relevant and important input into the process. As well, not as much discretion is left to the decision maker or practitioner as with other methods. As such, MAUT is a very defensible methodology. The values determined by the process reflect social (including environmental), political and economic priorities. Even if these values are competing and conflicting, the process offers the opportunity for debate, compromise, and ultimately consensus decision making involving all affected parties. The decisions reached may not be "optimal", but they will likely "satisfy" most of the parties involved in the process. Finally, MAUT is both systematic and adaptable. The methodology is dynamic in that it not only offers shape to the process, but it allows for flexibility once the process has begun. It is not uncommon for new alternatives to be generated once the process has begun. As 36 mentioned, the model offers varying degrees of complexity from simple mathematics to creating more complex utility functions.23 Weaknesses One caveat of using the methodology is the potential problems in developing the range of alternatives. Often, alternatives can be brought to the process which are unreasonable. The process can be paralyzed by having to deal with too many alternatives (or even criteria). This is a fundamental weakness inherent to almost any public process, but can be overcome through compromise and reasonableness on the part of the participants. However, it does pose a potential weakness. As M A U T necessarily involves people in the decision process, it can also be rather time consuming, and perhaps expensive to conduct. A more theoretical critique, especially for simpler M A U T models, is that simple averaging, multiplying, and ranking of weights as demonstrated in our model may not be totally accurate. Obtaining the scale and range of individual values may alleviate this concern. Further to this point, obtaining values for items without specific reference to units (such as dollars) can be considered meaningless. Like G A M , M A U T does not deal specifically with the elements of monetary evaluation or discounting impacts. The model does not explicitly deal with the timing of impacts or with future uncertainty by using a discount rate, or expressing terms in NPV. Implicitly, however, the values expressed in the model by stakeholders could incorporate future uncertainty. In our bike example, Edwards may have placed a high premium on seat comfort because he realized that he would own the bike a long time, and knew that an uncomfortable seat may not bother him in the near future, but in the long run it would make riding less enjoyable. Placing a high value on it now would ensure his needs 37 would be met much later on (van Delft and Nijkamp 1977; Gregory and Keeney 1992; Keeney 1982). Conclusion Multiple criteria decision making, or MAUT, is an extremely useful evaluation tool. Despite its weaknesses as a model, it offers the opportunity to evaluate public decisions with genuine input from the public. The use of stakeholder groups provides a workable and meaningful opportunity for decision practitioners to work with the public. The model has the capability to explicitly consider social, environmental, political, and economic objectives. As well, the model assists the process greatly by being structured in such a way as to deal with the essence of decision making - value trade-offs. MAUT does not require agreement or acceptance of its participants before the process begins, but attempts to build this during the process. Indeed, the outcome a MAUT model is perhaps best seen as a communications tool. Once the values have been structured and the alternatives evaluated, then the real decision making can take place by way of further trade-offs, compromise, and ultimately consensus. A decision reached by way of MAUT may not be Pareto efficient or optimal, but it likely will be better understood and accepted by both the participants and the public than one reached by other methods. Other Methods - Multiple Accounts Evaluation Guidelines and Delphi In 1992, the British Columbia Provincial Crown Corporations Secretariat developed a Multiple Accounts Evaluation Guideline which is to be used by Crown Corporations such as BC Transit. Although the model is fairly similar to PBS, it does have some specific requirements that are worth examining. The model has many shortcomings, but could be a 38 useful start as the basis for developing a model to evaluate the rapid transit decision. The Delphi method is a group decision making technique made notable by the RAND corporation after World War II. Although Delphi is best known as a technique for predicting future outcomes, it has some implications for the type of decision that we will be evaluating and is therefore quite useful. Multiple Accounts Evaluation Guidelines The rationale for developing this type of structure for governmental decisions in BC was based on elements which are factors in almost all decisions. Of primary consideration is the fact that Crown Corporation plans and projects are not undertaken in isolation; their decisions have impacts on society at large, as well as on other governmental agencies. Second, there is the understanding that any public project is not end in itself, but simply a means to addressing society's needs. In our case, a transportation project would serve the Lower Mainland's need for a better way of transporting people. A third, and perhaps the most fundamental objective, is the understanding that given the diverse nature of interests, objectives, and alternatives associated with any one project, one single measure will not adequately reflect or summarize what needs to be decided. Multiple Accounts will at least try to factor more elements into the decision and evaluation process than is normally done. Finally, constructing decisions in this type of framework will help deal with the uncertainty of making decisions for the future. The understanding of the Crown Corporate Secretariat (1993:1-4) in developing these guidelines is that "effective planning requires the explicit recognition and focus on higher level goals and strategies. Effective pursuit of provincial objectives requires integrated planning across interrelated organizations both in the development and evaluation of alternatives." 39 Multiple Accounts Evaluation is composed of five different categories: Financial Performance, Customer Service, Environment, Economic Development, and Social (Impact). Financial Performance This account is primarily concerned with the performance of alternatives plans or projects in terms of the impacts on revenues and expenditures. The performance is to be calculated from the perspective of both the Crown Corporation undertaking the project and the rest of the Government. Furthermore, the model is flexible enough to include more local concerns, such as with a municipality. After revenues and expenditures are calculated over the length of the project, the model utilizes a discount rate for base case analysis. Essentially, this category is a discounted cash flow analysis of the project. Many of the assumptions for financial performance are the same as for CBA. First order effects are the only ones calculated; secondary benefits, or stemming effects, are ignored, Furthermore, interest costs and depreciation are not factored into the calculation of expenditures. As well, the model calls for assumptions on the part of the analyst when it comes to items like economic growth, population forecasts, and future prices. The model suggests that sensitivity analysis be conducted using a variety of possible forecasts (1993:10-12). Customer Service Customer service is concerned with assessing the costs and benefits resulting from the changes in quantity and quality of the conditions when the project is provided. The evaluation would take into consideration the impacts on both groups that are directly impacted, and those that are not. As well, the model is flexible enough to include analysis of the impacts on different geographical or social groups. Utilizing the CBA concept of consumers' willingness 40 to pay, values are obtained and calculated on a base case analysis by using an appropriate discount rate. Such values would be obtained by way of CV surveys, as mentioned earlier. Where it is not possible to obtain a monetary valuation of the customer service implications, the model allows for non-monetary measures of quality of performance and numbers of customers affected. Such information is not directly comparable to financial performance, but it can be taken into consideration by the decision makers. Unfortunately, the model is not very specific as to how to display the non-monetary impacts (1993:12-14). Environment This account is not intended to replace the more detailed environmental impact assessments usually associated with planning decisions. Instead, it attempts to categorize any major biophysical changes or impacts resulting from a new project. Items such as pollutant emissions, resource use and depletion, and impacts to wildlife and habitat can all be calculated and factored into the evaluation, and the information can be used to assess trade-offs which have to be considered at all levels of government. As with customer service, both monetary and non-monetary costs can be calculated. Quantitative considerations are treated identically as with customer service, and in accordance with CBA principles. Qualitative assessments can be made as to the severity of impacts, and how changes fit into the context of broader policy goals and interests (1993:15-17). Economic Development The focus of this account is not on the short term direct and indirect jobs created by the project or plan. Rather, it attempts to assess how the project will contribute to sustained economic activity. The difference is between measuring "gross impacts" of the direct and 41 indirect aspects of the project, and the "net benefits" to the economy that will accrue later because the project was undertaken. A monetary net benefit can be developed by estimating the increase in income generated by the project to persons who are unemployed or underemployed. The yearly benefits would be compared from both a regional and provincial perspective, and would be calculated using NPV. Non-monetary economic developments would be categorized by statements of significance about the impacts on the project. Again, two or more perspectives can be incorporated (1993:17-18). Social This account is not meant to provide a detailed social impact assessment of the project, although a formal assessment can be incorporated. It is meant to document the major community or distributional impacts and the trade-offs that are entailed. This is the one category in which the model makes no attempt to monetize the impacts. A summary statement of the nature of the effects would be included (1993:19). The Multiple Accounts Summary would look as follows: 42 Table 6 Account Alternative A B C Financial Crown Corporation NPV Rest of Government NPV Local Government NPV Customer Service Consumer Surplus NPV or Summary Statement Environmental External Cost NPV or Summary Statement Economic Development Incremental Income NPV or Summary Statement Social Summary Statement The summary matrix would provide decision makers with information as to the advantages and disadvantages of the project or plan, as well as the trade-offs that would be entailed. The final component of multiple account analysis which would be incorporated, if necessary, is to deal with risk and uncertainty. The model suggests using sensitivity analysis (as in CBA) or scenario analysis. In the latter, sets of assumptions about a project component are altered to analyze the combined effects of any changes. An example would be to alternate between relatively more optimistic and pessimistic outlooks of the environmental aspect of a project to see how it would affect the other categories, as well as the project as a whole (1993:21-22). Multiple Accounts Strengths and Weaknesses As the model is localized to British Columbia and is relatively new, little has been written as to the merits of the model. However, it is clear that Multiple Accounts contains 43 many elements similar to those of models and methodologies that have already been reviewed. As such, it contains many of the same fundamental strengths and weaknesses. Strengths The most notable strength of the model is that it attempts to incorporate a broad range of considerations into a decision. As well, the fact that it analyzes the financial aspect of a plan from the view of the three levels of government is unique and helpful. It is often the case that those who are paying are not the ones benefiting. The Columbia River Treaty is an example of the Province of BC reaping the benefits, while the residents of the Kootenay region were paying a large environmental and social cost. Clearly, the attempts to utilize the economic principles of CBA helps the model to be systematic and consistent. Finally, the model is relatively straightforward and easy to understand in terms of the information it requires and how it is presented. It also allows for monetary and non-monetary information to be included. Weaknesses Like CBA, PBS, and GAM, the model seems to place much of the responsibility for evaluation on the practitioner. There is no explicit mechanism for involving stakeholders in the process, as the model relies mainly on CV surveys to elicit values. A further shortcoming relating to the previous one, is that the model offers no explanation as to how to evaluate the trade-offs involved between the five major components. Like PBS, the decision maker is left to "weight" the considerations for the decision. As such, using a Multiple Accounts model will likely mean that the decisions will be more open to criticism and debate. Without a stakeholder process, consensus is difficult to achieve. 44 Conclusion To be fair, Multiple Accounts Evaluation Guidelines makes modest claims as to what it can achieve. The information it contains is meant to assist in the decision making process, not to replace it. For our purposes, the model is very important. As BC Transit is a Crown Corporation, this is the model that will form the basis for its decision process. The key task is to adapt the model to overcome its inherent weakness of passive public input. Delphi Process As mentioned earlier, the Delphi process is a decision making technique that was developed and modified after WW II. Delphi gained notoriety as technique for predicting the outcome of future events on the principle that "two heads are better than one" (Dalkey 1972). From various experiments which it carried out, the RAND corporation concluded that a Delphi group decision making technique enabled decisions to be up to 45% more accurate than if the decision had been made alone. Although this applies to judgement on matters of fact, such as the probability of a future event occurring, the conclusion was that on judgement of matters of values - what we are more interested in - that decisions would likely be more satisfactory. As we are interested in incorporating stakeholder values into our process (or model), the implications of Delphi are worth examining. The Delphi process has three major features: anonymity, controlled feedback, and statistical group response. Anonymity is achieved by the use of questionnaires, and is intended to reduce the effect that dominant individuals may have on the process. Controlled feedback is a way of introducing results or information into the process in such a way that it could be used in formulating future decisions. Typically, the first round of decisions are made 45 on limited information. When more information is introduced, decisions will likely change. The new information, and the results of the previous round, will be factored into the next round of decision making. The process is controlled so that no one participant has more information than the others. The final element is the use of a statistical group response to ensure that the responses of each individual are factored into the final group response, and to reduce pressure on group conformity. Under Delphi, no one person controls the decision (1972:19-20). A classic Delphi (1972:21-22) experiment which helps to explain the process is of the type where respondents were asked to answer a series of factual almanac type questions, such as "How many suicides were reported in the U.S. in 1967?" Because this type is easily verifiable for correct answers, it was used. A group of participants is presented with some background information, and answers are obtained. The group is then provided with feedback, which may include more information or simply the results of their first round responses, such as the upper and lower quartiles. Following this, the group then attempts to answer the questions again. In a Delphi process, the number of rounds can vary depending on the task. In general the outcome can be characterized by the following traits. The initial round produces a wide variety of answers. With feedback and iteration, the distribution of responses for a group would become smaller. Finally, the group response (defined as the median of all responses) becomes more accurate. Again, even though this is more true for factual responses, one would suspect intuitively that the same process would hold true for value elicitation. At the very least, the final group response would at least be more reflective of the groups values as a whole. 46 Delphi Process Strengths and Weaknesses Strengths One of the major strengths of the Delphi Process is that it allows a forum for participants and individual stakeholders to have equal input. What is being said will become more important than who is saying it. The Delphi Process is also quite adaptable, and can be useful in a number of contexts. Clearly, it can be of great use for value elicitation. However, it may also be helpful in dealing with the future uncertainty of planning projects by being able to assemble a sample of expert opinion on a topic. Finally, Delphi is a useful technique to enable decision makers to incorporate the values of whole groups, and not just individuals, in the decision making process.24 Weaknesses Despite its seemingly scientific basis, Delphi is still just a method for obtaining opinion. One would assume that the use of experts in a structured and controlled process would lead to more enlightened opinions and better decisions, but such is not always the case. The Delphi Process is an abstraction from the real world, and does not offer the type of relatively objective measures that are comforting when dealing with real world problems. It was summed up best by Pill (1971:62) when he wrote, "It is in the questions of intuitive judgement, the marshaling of subconscious processes, dredging half-formed ideas from the group memory, that Delphi is useful and as such, one cannot judge it on the same basis as a concrete measurement."25 47 Conclusion It is easy to see where the Delphi process could be merged with MCA (and even Multiple Accounts) to enhance the stakeholder process for decisions. Indeed, one of the drawbacks of a stakeholder roundtable meeting may be the dominance of an individual or group. A Delphi process could help to ensure that stakeholder views and values are given equitable treatment. Furthermore, to ensure the anonymity of responses as well as the quick tabulation of statistical responses tools such as a computerized Group Decision Support System (GDSS), or an electronic meeting system (EMS) means that utilizing a Delphi Process is quite viable.26 Conclusion: Evaluating the Methodologies We have now reviewed the most prominent evaluation methodologies that Planners have at their disposal. The question then becomes which one to use for our decision Each method has strengths and weaknesses, both in the theoretical and practical senses, which will determine its usefulness. The one which will be of most use for our type of decision may not be useful in a different context. Our task now is to evaluate the evaluation methodologies. McAllister (1980: 72-73) provided a list of criteria for assessing the evaluation methodologies. None of these is intended to be an absolute directive, but an acceptable method must fall within these guidelines. The criteria are as follows: 1. Systematic. The method should be systematic so that the results are reproducible. 2. Simple. The method should not be so complicated that only a few people are able to use it or understand the results. 3. Quick. People using the method should be able to generate answers within a reasonable length of time. 48 4. Inexpensive. The method should be capable of providing useful information on a reasonable evaluation budget. 5. Comprehensive. The method should be capable of taking into account all of the factors relevant to the decision it is intended to aid. For the most part, all of the methodologies we have discussed would fulfill these criteria to some degree. However, not all of them are suitable for the type of decision we are investigating. In light of these criteria, and our type of decision, which methodology is most useful? C B A Although CBA is clearly the most systematic of our methodologies because of its grounding in the field of economics, it fares rather poorly by the other criteria. Aside from the CBA practitioner, others involved in the decision process, such as the provincial and local politicians - as well as the public - would have little comprehension of the basis for the recommendations or decisions. While it may be possible to include all of the financial, social, and environmental factors into the decision on rapid transit, it would clearly be a large undertaking in terms of time and money to conduct consumer or public value surveys on these factors. This would be further complicated by the fact that those in the areas which may benefit from the project are likely to express greater values than those who will not benefit. Are these values truly representative, or are they being expressed so the project will be undertaken? In general, CBA would focus on providing an answer (or recommendation) which would indicate what would be the most efficient use of the resources on narrow economic grounds. The decision would provide an optimal economic solution. However, 49 given the politically contentious nature of the problem, the optimal solution may not be the most socially acceptable. P B S As mentioned earlier, PBS shares many of the same practical shortcomings as C B A because like its predecessor, it is grounded firmly in the field of economics. In addition, it attempts to be even more comprehensive by way of integrating community impacts, however it is a demanding model to use in terms of time and money. The PBS model would help in the decision on rapid transit by attempting to focus on the different impacts on the three areas. Once these have been identified, however, the model leaves it to the decision makers to identify and decide the trade-offs. The model does not offer the opportunity for any public involvement in the process other than at the value elicitation stage. Essentially, the model progresses little from C B A , and offers little more in the way of a viable alternative. GAM G A M could prove a useful tool to evaluate the decision on rapid transit. As mentioned, Hill developed the model to evaluate transportation proposals. The system of rankings and weightings is easily understood, and can be conducted by the practitioner with more ease than the more economic-based value elicitation measures. The model is practical for our purpose because it focuses more on displaying the information involved in the decision without attempting to prescribe a "right" answer. However, the model still leaves much discretion to the decision makers with little input from the public. The decision on rapid transit in the G V R D , if it is to be defensible, must have a better way of incorporating the public in the process. 50 Multiple Criteria Of all the methodologies, I believe the MCA model provides the most effective evaluation methodology for our purpose. The models are not necessarily as systematic as CBA or PBS, but they can be constructed relatively easily and cheaply. Indeed, the one strength of the model is that it is adaptable, and can be made as complicated and sophisticated, or as simple, as the problem requires. Although our problem seems to be complicated given the scope and magnitude of the decision, the evaluation can easily be done along the lines of the simple MAUT analysis done by Edwards. Indeed, all of the technical information has been gathered. The primary task of the model would be to display the information in such a way that will allow those involved in the decision process the opportunity to assess the trade-offs involved between the various financial, service, social, and environmental criteria. To come full circle, we may arrive at a decision that is not necessarily optimal - in the economic sense -but will be satisfactory and acceptable to all involved. Multiple Accounts and Delphi In accordance with their mandate, the Crown Corporations Secretariat for BC has undertaken a Multiple Accounts evaluation of the rapid transit decision. This report will form part of the basis for the next chapter. As mentioned previously, the decision will be made by the provincial government, because it has the responsibility to do so. The Multiple Accounts guidelines offer a viable method for evaluating the decision. It fulfills most of the criteria for being a useful evaluation tool. Moreover, what it lacks in terms of active public input can be overcome by modifying it with a relatively simple MCA model, and incorporating some of the elements of the Delphi process. 51 The following table is an example of the type MCA evaluation we have already seen. Although the values are somewhat subjective (being my own values), it provides an outlook as to how such a format can be constructed. By incorporating a Delphi process, different values may be obtained if more than myself were doing the evaluation. The values expressed in the tables represent two aspects. The five objectives (or criteria) defined by McAllister have been given factor weights as to how they would impact on the final decision; each weight is representative as a percentage. The six evaluation methodologies (or alternatives) have been given a score out of 100 as to how they meet the requirements of that criteria. Table 7 Greg's Weights of Methodologies Systematic Simple Quick Inexpensive Comprehensive Normalized Weight .15 .25 .10 .10 .40 Methodology Score Norm Score Score Norm Score Score Norm Score Score Norm Score Score Norm Score Final Weighted Score CBA 40 13.5 20 5 20 2 40 4 40 16 40.5 PBS 50 7.5 20 5 20 2 30 3 60 24 41.5 GAM 60 9 50 12.5 45 4.5 65 6.5 60 24 56.5 MCA 60 9 40 10 50 5 60 6 90 36 66 Mult Acc 50 7.5 45 11.25 55 5.5 65 6.5 65 26 56.75 Delphi 50 7.5 75 18.75 75 7.5 70 7 30 12 52.75 From the table, it is clear that MCA fares the best of all of the alternatives, given my assigned scores. It is clearly not as simple to understand - especially in its more complex forms such as concordance analysis - nor as quick to provide answers. However, it is by far the most comprehensive of the methodologies, which is important when trying to evaluate a very public transportation project which has a variety of impacts. 52 Conclusion At the beginning of this chapter, we outlined the difference between oikonomia and chrematistics and what they mean in the context of modern economics. Further to this, we have examined a series of models, which in light of these two interpretations, are all economic evaluation models. The type of evaluation that is usually considered standard procedure for the issue we are trying to decide has been primarily chrematistic. Indeed, CBA was the predecessor of all of the models. But what we have seen, in our reviews of PBS, GAM, and especially MCA is that hard economics forms only part of the analysis. Modern planning issues, such as the one around rapid transit, must take into account a multitude of factors, many of which are not easily quantifiable - especially in monetary terms. The paradigm shift that Daly and Cobb discussed, in the context of evaluation methodology, would see a greater utilization of MCA to help decide on these critical issues because it can incorporate the many conflicting and competing societal objectives inherent in decisions, and do so in a way that is inclusive, comprehensive, and responsible to all those involved in, and affected by, the decision making process. 1 McCallister (1980) makes the distinction that a WTP to secure a desired change is a benefit, while a WTP to avoid an undesired change is a cost. This latter case is roughly similar to asking respondents what they would be WTA in compensation for the undesired change taking place. 2 Schofield (1987); Gwilliam (1972). The type of saving associated can be relegated to the objectives of "accessibility" and "increased mobility" according to Gwilliam (1972), and are primarily user benefits. However, a study done in Toronto, and reviewed by Gwilliam, points out that there are pitfalls to using this as a single measure. The "savings" may be higher in some areas as opposed to others. This aggregating effect makes it difficult to use this as an evaluative measure, especially when used in isolation from other factors. 3 Davis (1990) cited a study conducted in three U.S. cities which compared house prices against air pollution, and showed that house prices were negatively correlated to air pollution. The change in house prices (and rent) are a way of shadow pricing the effects of the air pollution. 53 4 Smith (1986). As discussed in Davis (1990), the Treasury Board of Canada recommends that a 10% discount rate be used, with sensitivity analysis conducted at 5% and 15%. Essentially, any evaluative analysis is sensitive to its inputs. The systematic measurements of sensitivity allow for the impacts of the inputs to be monitored, should the inputs change. For a review of the calculation of Net Present Value and Discounting, see Davis (1990:119-122). 5 Davis (1990); McAllister (1980); Schofield (1987). In surveying the literature, there is general agreement over the strengths and weaknesses of CBA. Rather than duplicate information by reviewing each author separately, their remarks have been grouped together. For an example of a simple and easy to understand CBA study, see the case of the STOL Air Transport study in Davis (1990:115-117). 6 Lichfield et al. (1975). For a fairly detailed bibliography of applications of CBA, McAllister (1980: 144-147) provides 43 citations of CBA's done on a variety of general and environmental projects. 7 The general weaknesses of CBA methodology were taken from the following sources: Davis (1990); McAllister (1980); Schofield (1987); van Delft and Nijkamp (1977). An excellent review of the problems of Contingent Valuation, both in general and the specific case of environmental decisions, can be found in the article by Gregory et al, (1993). Some points raised by the authors are that CV questions are often complex and require the respondent to deal with more information than they can actually "digest" in a satisfactory manner. (1993:14) The authors contend that non-market values do not already exists for environmental goods, but are constructed during the elicitation process. The authors generally feel that standard CV elicitation does not do an adequate job of helping to construct these values. (1993:20) One of the major shortcomings of C V is the difference between peoples WTP money to receive a "benefit", and peoples WTA compensation for having a "cost" foisted upon them. In Pearce and Turner (1990), a survey of eight different studies into the differences regarding WTP and WTA values shows that in all eight cases, people require more to accept a cost than they are will ing to pay to obtain it. The differences range from just under twice as much to over seventeen times as much. (1990:156-157). An interesting article which attempts to overcome the weaknesses of value elicitation in hypothetical circumstances is by Brookshire and Coursey (1987). The authors found that using market like elicitation procedures, such as bidding at an auction, and making it an iterative feedback oriented process resulted in respondents refining and modifying their values. Having a simulated market, the authors felt, disciplined and sharpened the responses. On the subject of long run certainty and the application of discount rates, the article by Haveman (1977), argues for a premium to be added to the expected value of future damages, and a discount be subtracted from the expected value of future benefits to balance out the risk of activities to future generations. He argues against the Arrow-Lind theorem which states that with an increasing number of individuals participating in an activity (or being affected by a project), the cost of uncertainty to any one person diminishes proportionately. At the limit, any cost vanishes. Haveman argues however that the numbers of people required to "diminish" the costs of future uncertainty are so great that the theory has practical limitations (1977:369-370). Kelman's (1981) ethical critique of CBA is valuable because it gets to the heart of the problem about social costing in general. Kelman gives a spirited argument that the utilitarian assumption that benefits must outweigh costs in order to have a project is not a moral judgement. Furthermore, he argues that the very act of fixing a price for something tends to decrease the value of the good being priced. He argued that putting a dollar price on ones values of clean air or water is ethically wrong. He contends that modest efforts to assess benefits and costs are justified, but he warns against the dangers of government and their agencies extending the practice. An article by Anthony Boardman et al. (1993), looks at the effects of politicization of a CBA relating to the Coquihalla Highway Project. The article reviews the perceptions of three types of bureaucratic roles on the evaluation of costs and benefits of the highway project. The three roles are: Analysts, Guardians, and Spenders. Analysts are primarily concerned with allocative efficiency and tend to perform CBA the proper way. Guardians tend to perform revenue-expenditure analysis. They are primarily concerned with keeping the dollars flowing out to a minimum and the dollars flowing in to a maximum. They ignore any environmental or social externalities associated with the project. Guardians are usually found in the budget or accounting offices of government agencies. Spenders, who come from service departments, tend to view expenditures not as costs but as benefits to constituents, such as in the form of jobs. In reality, they are 54 performing economic impact analysis and not CBA. Furthermore, they tend to favour all projects over the "do nothing" option. (Such is the case with Glenn Clark and BC 21) The authors point out that Spenders tend to underestimate the costs of a project, and the case of the Coquihalla Highway, which was pushed forward by Spenders, resulted in "pathological underestimation." The lesson for the authors was that correct and accurate CBA is difficult to conduct, and that it is quite easily manipulated depending on ones bureaucratic role and tendency. Further to this point, the CBA of the STOL Airport project mentioned earlier in Davis (1990) was approved even though two CBA studies showed that costs exceeded potential benefits. One of the reasons cited, which relates to governmental roles, is that the larger the share of non-budgetary costs (i.e. environmental or social externalities) in total economic costs, the more likely it is for a project with a negative NPV to be recommended. The reason would appear to be that despite their inclusion, externalities are not given "real" weight in the decision. This is a problem much like the double counting of benefits. 8 Schofield (1987) points to areas of future improvements to CBA specifically in the urban and regional planning context. 9 Schofield (1987) discusses the problem of project implementation as it relates to partial equilibrium. He contends that disregarded interdependencies in the economy, as a result of implementation, could invalidate the piecemeal recommendations associated with a limited CBA. "It is for this reason that the method is more safely applied ... to small scale projects than to larger-scale endeavours..." 1 0 Schofield (1987). Table 12.2 and 12.3, which Schofield provides are excellent examples of how the impacts are both categorized and rated for the decision makers. 1 1 Hi l l (1973); Lichfield (1970); McAllister (1980); van Delft and Nijkamp (1977). Lichfield (1970), in evaluating eight types of methodologies, compares them against a list of ten criteria which would account for a "good" evaluation methodology. Not surprisingly, PBS fares the best by fulfil l ing most of the criterion fully and some partially. However, the list is a bit self-serving to PBS, and should not overstate the usefulness of the method. 1 2 McAllister (1980); Schofield (1987); van Delft and Nijkamp (1977). Hi l l (1973:19-20) had an important critique of PBS which was rectified in later models by Lichfield. Essentially, Hil l 's critique was that PBS did not make attempts to explicitly include community objectives. According to Hi l l , PBS "... does not always appear to recognize that benefits and costs have only instrumental value, they have meaning only in relation to a well-defined objective." As such, Hi l l believed that PBS could have the tendency to include costs and benefits which the community would find irrelevant. Hi l l , p. 19-20. For a quick but comprehensive overview of PBS, along with a comprehensive bibliography, see McAllister (1980:170). 1 3 Schofield (1987); Schofield (1991). Schofield indicated that many feel that PBS (and GAM) are really in the same genre as CBA. Furthermore, he feels that the departure from economic principles undermines the credibility of the models. The author would disagree with the latter assessment. 1 4 Davis (1990); Hi l l (1973); Lichfield et al.(1975); Schofield (1987). The review of GAM covered in these pages provide a comprehensive enough overviews of the methodology. The review in Lichfield et al. is essentially a response to Hil l 's earlier work, and an attempt to compare which is the more useful method. 1 5 Hi l l (1973). The overview of GAM by McAllister (1980:159-165), like with PBS, is brief but thorough. See also the bibliography in McAllister (1980:170-171). 1 6 For classification of the four different scales, see Hi l l (1973:33). Comments on the ordinal scale, Hi l l (1973:38). 1 7 McAllister (1980). These are general strengths of the model as observed by the author, and in light of how it enhances the efforts on CBA and PBS. Most of the literature other than McAllister's and Hil l 's (1973) tends to focus on the weaknesses of the model. 1 8 Edwards and von Winterfeldt, (1987): In an earlier article by Edwards (1981), there was actually a 10 step process outlined, however steps 7 through 9 dealt primarily with the statistical treatment of the data, and step 10 was to "decide." 1 9 Edwards and von Winterfeldt (1987); Gregory and Keeney (1994); Keeney (1988). Gregory and Keeney (1994) point out that including both means and ends objectives could result in a double counting problem. 2 0 Nijkamp et al (1990);. Keeney and Raiffa (1976). 55 In actuality, there are many more ways to elicit values than the tradeoff, rating, and ranking methods discussed. Nijkamp et al. (1990:45-48) outline other methods such as: verbal statements on weights relating to a five point scale, i.e. outstanding, good, satisfactory, fair, poor; paired comparisons; formulating scenarios where stakeholders would give values on different hypothetical situations. Most of these methods, like the three main ones, are direct estimation methods. 2 1 A good example of another form of MCA is Concordance Analysis. As explained by Nijkamp et al. (1990:69-70), the idea is that a ranking of alternatives is arrived at by a pairwise comparison of alternatives. When an alternative performs better than (or equal to) another alternative, based on a set of criteria, the resulting pairwise alternative is said to be in concordance. When an alternative performs worse than (or equal to) another alternative based on a set of criteria, that set is said to be in discordance. The indices scores for the two types of sets are calculated, and the "best" option is the one with the highest values of concordance indices. The theory has actually been used in a transit study for Singapore, see Massam (1978). The study attempted to rate 6 alternative plans on 8 different criteria. The results of this study were presented in both a ranking format, and a mapping format. The latter method is capable of mapping in two or three dimensions, so that plans which may end up being relatively similar, depending on the criteria, can be viewed together. 2 2 Edwards (1981). Al l tables were copied directly from the article. Another excellent example of a simple MAUT procedure was done by Einhorn and McCoach (1978). The exercise in this case was to try and predict those NBA players who would be on the All-Star team in the 1974-75 season based on their previous years performance. 2 3 McDaniels (1994); Delft and Nijkamp (1977). An excellent example of a more complicated MAUT process was prepared by McDaniels (1992) for the British Columbia Commission on Resources and Environment (CORE). Gregory and Keeney (1994) point out that many evaluation or decision problems start out as dichotomies when there are actually a range of alternatives between these most often focused upon choices. 2 4 Tell's (1978:490) conclusion for his study was that Delphi was effective to ensure that preferences expressed were for the group as a whole and not just the individual representing the group. 2 5 Pill (1971). The article contains 40 citations of Delphi experiments. 2 6 The EMS system at the British Columbia Institute of Technology (BCIT) is but one example of the type of computer assisted decision evaluation tools available. Some of the significant features of EMS is anonymous and simultaneous input, clear documentation, control of tools, and equal voice. Ideas may be brainstormed, organized and evaluated The meeting environment is very amenable to a stakeholder roundtable process. Results of votes and information is easily and quickly presented in statistical form i f necessary. EMS, like Delphi, is not intended to replace meetings, but to enhance their input and output. The information obtained on BCIT's EMS system was from Matt Baxter, Project Manager, Group Decision Support Systems, BCIT. 3700 Willingdon Avenue, Burnaby BC. Canada V5G 3H2. 56 CHAPTER 3 The evaluation, and subsequent decision, on the next phase of rapid transit routes has been reviewed by two separate entities. The first agency to do so was actually a joint effort of the Ministry of Transportation and Highways, BC Transit, and the Greater Vancouver Regional District (GVRD), which in 1993 produced both a Long-Range and Medium Range transportation plan for the Lower Mainland area. The second agency to review rapid transit was BC Transit. The latter review was published in 1995 as part of a ten-year comprehensive transportation development plan for the province. For the purposes of this thesis, these two efforts provide an interesting juxtaposition of ideas and philosophies of evaluation, and are very helpful in highlighting the shortcomings of the decision process thus far carried out. The aims of this chapter are threefold. First of all, we will undertake a brief review of the political and bureaucratic context for the rapid transit decision. Secondly, we will review both the GVRD and partners' and BC Transit's summary evaluation documents to discern both their theoretical underpinnings as well as their practical assumptions, and to see what impacts these had on the agencies' final recommendations. Finally, we will review what decisions have been made and what, if any, problems have occurred with the decision making and evaluation processes. Context The provision of rapid transit in the Lower Mainland has been a topic of discussion for a number of years. Ever since SkyTrain was completed in the mid 1980's, there has been much speculation as to where the next distinct phase of rapid transit would be built. The one route which was prominent in the public eye was the transit line from Richmond to 57 Vancouver. Indeed, most of the research of that line had been narrowed to which of two routes through Vancouver, Cambie or Arbutus, that the line would take. Ultimately, however, no action was taken on the decision. Since the latter 1980's there have been some significant changes in the Greater Vancouver landscape with implications for the future of rapid transit. Certainly, the biggest change is the population growth of the Greater Vancouver region, and the concern over future continuation of this growth. In the 30-year period from 1961 to 1991, the populations of Richmond and Coquitlam, two areas now vying for more transit, almost tripled in size (GVRD 1993a:6).1 However, it was the growth predicted to take place over the next thirty years, to 2021, that was the cause of concern. At the risk of oversimplifying the problem, the concern was that the type of settlement pattern, if allowed to continue at current trends, would be further urban sprawl to the outlying areas of the Fraser Valley. The fear on the part of planners, and especially the GVRD, was that this type of ecologically unsustainable growth would not only consume valuable farmlands for housing, but would also bring with it a serious transportation problem in terms of more road space needed for more cars. As part of their Livable Region Strategic Plan, the GVRD had proposed that growth and development be more concentrated to areas such as the Northeast Sector, Burnaby, North Surrey/Delta, and to some extent Vancouver(1993a:26).2 With such development, the provision of transportation solutions other than additional roads would be more economically and environmentally feasible. There now existed serious competition for future transit resources. The second change was a political one. The provincial government of BC had for the better part of 40 years been formed by the Social Credit Party. It was well known that the west side of Vancouver and Richmond, the areas that would benefit from rapid transit, were areas in which the Social Credit Party were politically strong. In 1991, the New Democratic 58 Party (NDP) was elected to form the next government, and their political base of support was very strong in areas such as East Vancouver, Burnaby, Coquitlam, and New Westminster. It would likely not bode well for the NDP going in to the next provincial to bypass its political base of support. As columnist Jamie Lamb (1994) bluntly pointed out: "The route that benefits the greatest group of MLAs belonging to the government that is proposing to spend $1 billion on transit is the route that you're going to see built. The Broadway-Lougheed line (which is the one the Government eventually did endorse) would run through five Vancouver ridings, all of them currently represented by NDP MLAs including premier Mike Harcourt and Transit Minister Glen Clark. The decision clearly had other merits. However, the relationship and influence of politics to this type of decision is difficult to ignore. One of the more contentious aspects of this decision, perhaps the most contentious aspect, revolves around the issue of who will decide the next phase of rapid transit. Clearly this is an area ripe for debate. The various municipalities, through their representation in the GVRD, had hoped that their forum would be the one to decide the issue. Indeed, the transportation plan that they proposed, Transport 2021. was a thorough and comprehensive study of all the transportation needs throughout the GVRD region, going well beyond the scope of rapid transit. The problem, however, is that the GVRD has no regional land use or transportation planning powers. Despite its efforts, the GVRD is merely an advisory board with no mandate to implement plans. The actual mandate belongs to the cabinet of the provincial government and the Crown Corporation responsible for such decisions: BC Transit. Although the BC Transit Ten Year Development Plan would have obviously taken into account the needs of the various municipalities, these local governments did not have the type of input into the decision that they would have had with the GVRD. This situation poses some important questions about the legitimacy of the decision making process on local issues 59 when there is seemingly no avenue in which to allow meaningful participation of the various municipalities, or for the people they represent (Simpson 1994).3 It is important to understand the context of the decision making process on rapid transit before reviewing the following two evaluation documents. Although both efforts are based on rational subjective data, it is important to understand that each has its own political agenda. The politics of the two organizations, the GVRD and BC Transit, do not necessarily compromise the integrity of their respective works. However, the emphasis that each places on certain aspects of the decision, such as the trade-off between cost-effectiveness and lands-use shaping, is born out of a need to fulfill a particular political role. The Evaluation Studies of the GVRD and BC Transit The purpose of this section is to review the two evaluation studies of the GVRD and BC Transit. For ease of reference, both summary documents are contained in Appendices and will be referenced as such. The review begins by juxtaposing the studies with the evaluation methodologies discussed in Chapter 2. Secondly, we will review the strengths and weaknesses of each study, paying particular attention to the underlying assumptions. Ultimately, the assumptions that are made as to what are the important, or critical, elements of the decision have a pronounced effect on the outcome of the decision. Finally, comments will be offered as to how each of the studies has, at all, helped in the decision making process for rapid transit. In particular, do these models offer the comprehensive evaluation needed for the rapid transit decision? 60 GVRD: Transport 2021: Goals, Objectives, and Criteria The Transport 2021 plan prepared by the GVRD and partners in late 1993 was released in two parts. In September 1993, the GVRD published its Long-Range Plan (1993c) which outlined four broad policy levers to shape future transportation development. The policies were intended to control land use, adjust service transport levels, apply transport demand management to change travel behavior, and, finally, supply additional transport capacity.4 In October of 1993, the GVRD produced its Medium-Range Plan (1993d), which delineated its proposals over the first 15 years of the plan to 2006. Relevant to our purposes, more specific information was included in this report as to the proposal of the three possible routes for further rapid transit expansion. From an evaluation perspective, the fundamental elements of the plan were the subject of a working paper produced in May 1993, titled Goals. Objectives and Criteria for Developing a Long Range Transportation Plan for Greater Vancouver (1993b). These latter two documents will form the basis of our analysis here. Methodology The introductory section of the Goals. Objectives and Criteria document provides a very clear understanding of what the GVRD and partners (1993b: 3) hoped to achieve with the evaluation methodology: "Within the exercise of planning, one of the most important component tasks is the development of a value structure, to guide both the generation of alternative plans and the evaluation and selection of a final recommendation. This value structure is typically conceived as a three-level hierarchy, consisting of goals, objectives, and measures of effectiveness. It defines what one wishes to achieve, while the alternative policies or plans define how one might achieve it. This distinction is most critical." The methodology proposed is very similar to MCA, and somewhat related to GAM and Multiple Accounts. Certainly, the emphasis on creating value structures, and using those 61 values to generate alternatives for the plan is an idea consistent with those expressed for MCA. The inclusion of measures of effectiveness, or how the alternative choices are meeting the goals and objectives, is an idea that is common to all three. As can be seen from Figure 1 (Appendix A:5), seven broad transportation goals were derived from a set of five general regional development goals. The seven transportation goals are as follows: Livability, Economic Development, Land Development Interaction, Social Equity, System Performance, System Management and Implementation, and Public Accountability. It should be noted that all of the transportation goals are related to at least one or more of the regional development goals. The regional development goals are: Maintaining a Healthy Environment, Conserving our Land Resources, Serving a Changing Population, Maintaining the Region's Economic Health, and Managing our Region. (Appendix A: 5) Table 1 (Appendix A: 7-10) contains the basic structure of the evaluation methodology. As can be seen, each goal explicitly states what it hopes is achieved. Furthermore, each goal contains a subset of more specific objectives relating to the overall broad goal. Finally, a list of criteria is provided to measure the effectiveness of how the goals and objectives are being met, and this would ultimately be used to test the effectiveness of any of the alternative rapid transit options. As can be seen, many of the criteria are quantifiable, especially those relating to the goals of Livability, Economic Development, and System Performance. Beyond the generation of this Table, however, there is no specific mention as to the mechanics or procedures of how it is to be used. Presumably, once the quantifiable data are gathered for each of the objectives, they are put into the model, and the decision makers are left to make judgements and trade-offs, as they do with Multiple Accounts. 62 Strengths Clearly, as an evaluation model, the GVRD and partners proposal has the strength of comprehensiveness. The scope of objectives is impressive in that it covers not only the traditional technical chrematistic aspects one would expect, such as System Performance and Economic Development, but also the more oikonomic categories of Livability and Social Equity, with strong human and natural environmental components. As well, there is no shortage of quantifiable data demanded by the model in terms of criteria for measures of effectiveness. While some of the measures are somewhat vague, such as the "access index by mode to activity centres" found in some of the categories (Appendix A: 7 -9 ) , most ask for specific and relevant measures. Overall, the model provided is very easy to understand, as each of the relationships of the criteria to the transportation goals and objectives, (and even the regional development goals) is clearly defined. Weaknesses The model, however, suffers from some fundamental shortcomings that ultimately limit its effectiveness. The primary weakness of the model is that once the data are gathered and placed within the model, there is no specific mention as to what is to be done with the model, much like Multiple Accounts. There is no mechanism to deal with the ranking of the importance of the goals and objectives to one another. Surely each of these goals is valuable in its own right, but the model implicitly assumes that they are all equal, which can hardly be the case. Furthermore, the model was meant to evaluate all transportation proposals, so it is somewhat general. However, when it was adapted to evaluate the provision of Intermediate Capacity Transit Systems, or ICTS (of which light rapid transit is one option), the GVRD and 63 partners abandoned many of the model's comprehensive goals and objectives. From the table on Transit Corridor Performance (Appendix A: 40), it is obvious that the concern for evaluation purposes was focused on System Performance and Land Development Interaction. Although this is a fault of the GVRD and partners and not the model, the fact that the model is easily manipulated and changed suggests that it lacks the type of rigorous framework and methodology which would not allow it to be so easily compromised. As the Medium-Range report summarized, if transport efficiency is the goal, then Richmond-Vancouver is the line to build. If, however, shaping of future land use patterns is important, then the Coquitlam-New Westminster line is the one to build. All of the other social and environmental goals and objectives are conspicuously absent (1993d:42-43).5 A final and very noticeable shortcoming of the evaluation model is that it completely ignores the impact of finances on the evaluation. The model's goals and objectives focus on the benefits and secondary costs in monetary terms, particularly in the Economic Development category. However, there is no evidence presented in either document that relates to the specific costs of the ICTS system, nor how to deal with discounting the values obtained. Conclusion The Goals, Objectives and Criteria model presented by the GVRD and partners does not offer the type of evaluation methodology which would be rigorous enough to use in this case study. The format of value structure around the seven goals would be useful, but only if the goals were more specific to the case study at hand. As we saw, only two of the seven goals were presented in the summary evaluation of the issue. However, if the format were to be looked at as a Multiple Accounts format, which it essentially is, it could be adapted using the processes of MCA. If there was a procedure for ranking the objectives, and incorporating 64 meaningful value trade-offs through a stakeholder process, then perhaps the broader goals would not be dispensed with so easily. Finally, the exclusion of any serious financial calculation from the evaluation seriously undermines the credibility of the process, and any decisions reached by that process. BC Transit: Multiple Account Evaluation of Rapid Transit Options in Greater Vancouver The Multiple Accounts Evaluation study was prepared by BC Transit in May of 1995 as part of their Ten Year Development Plan. As the report stated in its section on Policy Objectives (Appendix B:2-3), the evaluation was done with reference to the objectives of the provincial government, the GVRD, and BC Transit. From the provincial perspective, rapid transit would provide "a more strategic approach to new investments ... focusing on the movement of people and goods as opposed to vehicles..." (1995:2). The strategy of the GVRD has already been outlined in the previous document. The policy objective of rapid transit for BC Transit is "to provide a fast, reliable, and cost-effective transit service; to promote regional land-use goals; and to reduce automobile use and resulting environmental impacts" (1995:3). The three objectives of BC Transit do not appear to differ from those of the GVRD. Methodology The theoretical model for this evaluation was the Multiple Accounts Guidelines discussed in Chapter 2. Three of the accounts remain unchanged: Financial, Customer Service, and Environmental. Two of the accounts are new. The Urban Development account would provide a qualitative assessment as to how each of the three corridor options would 65 support the regional land-use goals of the Livable Region Strategic plan discussed earlier. The scale used would be low, medium, or high. The Systems Operations account would provide a qualitative assessment as to how flexible the new system is to adjust to different ridership capacities (this is because the evaluation takes into account technology other than fixed rail), and how easily it is integrated into the existing BC Transit system (1995:9). For each corridor, assessments of costs and benefits were done for each of the five accounts. Further, two additional measurements were included. Using NPV calculations as with CBA, the net social cost of each option was derived from the discounted sum of the costs and benefits over the life of the project. A further measure, and one which proved integral to BC Transit's final analysis, was the development of a Lifecycle cost per Boarding. This measure is designed to reveal the amount that each person using the service, over the lifespan of that service, would have to be charged in order to recover the total cost of the project. Lifecycle cost per Boarding is calculated as the ratio of the present value of the total financial costs of the option over the discounted sum, or present value, of the future boardings on that option (1995:15).6 The use of the Lifecycle cost per Boarding is an intriguing measurement, and will be discussed in further detail below. The purpose of this analysis on the methodology is not to criticize, per se, the data that has been gathered or referenced.7 That would be the task of a more technical study taking issue with the specific results of the various reports. The BC Transit document is essentially a summary evaluation product, and our concern is more with the evaluation process. As such, we will review the strengths and weaknesses of the report more on the basis of how it performs as a process without challenging too many of the conclusions it makes as to specific results. 66 Strengths The primary strength of the Multiple Accounts Evaluation prepared by BC Transit lies in the presentation of its data. As with CBA, the use of a discount rate, and discounting costs and benefits to a NPV figure offers some consistency of application to the process. The Financial accounts for each option (Appendix B:Table V- l 16;Table V-5 21;Table V-9 25) do an excellent job of representing the financial implications to the evaluators. The analysis is sufficiently comprehensive to include both the costs of each of the systems, as well as the revenues those systems would generate in terms of increased fares and taxes. Similarly, the amalgamation of the Customer Service and Environmental accounts to give a picture of the quantified social benefits of the rapid transit options also provides evaluators with a comprehensive understanding of the financial and social implications of the alternatives (Appendix B:Table V-2 18;Table V-6 22;Table V-10 26). The presentation of all of the results, and especially the final comparison (Table V-l3 30), are also relatively easy to understand. According to BC Transit's analysis, the Broadway-Lougheed corridor clearly fares the best. Weaknesses The primary weakness of this Multiple Accounts Evaluation is one that is derived from its primary strength. Although it does an excellent job in terms of financial and economic calculation, this type of analysis tends to be less inclusive than it needs to be. In looking at the final analysis (Table V-13 30), the decision comes down to three factors: Contribution to Regional Land-use Goals, Transit Ridership, and Cost-Effectiveness. Originally, Transit Ridership was at first contained as a subset of the Financial account, but in this final analysis it 67 appears to have a full one-third weighting in the final decision. The final decision is clearly tied to the issue of Cost-Effectiveness using the Life-cycle Cost per Boarding. This measurement, however, favours the assumption that the goal of transit, especially rapid transit, is to make the system as least costly as possible. It completely ignores the issue of why transit is a requirement in the first place: to deter further urban sprawl in the face of rapid population growth. The problem with the study and the recommendations generated by it is that there was no debate. A better evaluation process would have taken the data gathered, which are formidable, and then have a stakeholder process with elected officials, transportation experts, and the public to debate and decide on what the priorities should be. The Multiple Accounts Evaluation provided an answer to a question that in many people's minds has yet to be formulated. Is the provision of rapid transit to serve current needs and be economically feasible in the short run? Or, is rapid transit best used as a land shaping mechanism to deal with future growth and urban sprawl? The public priority is unknown.8 Conclusion From a technical standpoint, and re-iterating the discussion in Chapter 2, the Multiple Accounts Evaluation provides a usable framework for evaluation. The type of analysis offered by the study is definitely needed to do a thorough and comprehensive evaluation. The problem, however, is still one of process. The analysis was, for the most part, completed in isolation. In true chrematistic fashion, the data were gathered, the numbers were "crunched" and an answer was provided. The answer provided may or may not be the most appropriate solution - it is contingent on the question asked. If the question is what is the most cost-effective way of spending resources on rapid transit in the three corridors, then it may be that the answer provided by BC Transit is correct. However, if the question is which of the three 68 corridors would do the best job of shaping land-use and lessening urban sprawl to the outlying areas - even at the risk of less potential ridership - then the BC Transit solution may not be the most appropriate. To offer a comparison of the two models, it is useful at this time to re-visit McAllister's criteria for assessing evaluation methodologies at the end of Chapter 2. We will ignore the criteria of whether the model was Quick to generate an answer, or whether the process was Inexpensive, due to a lack of information. It would be fair to surmise that neither report fared well in these categories. However, given that this project is worth over $1 billion, it is difficult to say what a reasonable evaluation budget and time-frame should be. Most of the strengths and weaknesses will be discussed in the context of whether the model was Systematic, Comprehensive, and Simple. The following table represents a comparison of the two evaluation models, and to what extent they are strong or weak for that attribute. Table 8 Comparison of GVRD and BC Transit Evaluations Methodology Attributes GVRD . Quasi-Multiple Accounts BC Transit Multiple Accounts Comprehensive Strength Strength/Weakness Systematic Weakness Strength Simple Strength/Weakness Strength The GVRD and partners' model is comprehensive to include economic, social, and environmental objectives. As well, the model is relatively simple in terms of reasonableness of data demands and ease of understanding. However, the model is so simple that it lacks even 69 the most basic financial analysis, and offers no systematic process for meaningful stakeholder value structuring and trade-offs amongst the alternatives.9 To be fair, at least the GVRD model was an outcome of a lengthy and substantial public process, and there were later attempts to incorporate some financial analysis. The BC Transit model fares comparatively well in all three attributes. It is sufficiently comprehensive to include all the relevant objectives. As well, the use of financial measures offers some consistency of application. Finally, the model is relatively easy to understand in terms of its application and results. The comprehensive nature of the model was easily compromised, however, with the final results and recommendations based primarily on the financial aspects of the model. More specifically, the almost sole reliance on the Life-cycle Cost per Boarding ratio indicates a bias towards financial performance. As with the GVRD and partners'model, the BC Transit model offers no opportunity for a stakeholder process, and it lacked any of the public consultation that occurred in developing the plan outlined by the GVRD. The Current Situation The decision on where the next phase of rapid transit will be provided has, in fact, been a contentious one all along. As late as December 1994, the GVRD had decided to recommend to the provincial government that the Coquitlam-New Westminster route should be the first one built. This decision, however, did not bode well with the City of Richmond. As Councillor Corisande Percival-Smith argued, "We feel we've been cheated ... [T]he whole focus of Richmond's city centre ... was predicated on having transit through the area. It's all very well talking about choosing our future. But if you can't even recognize what's already here, that's very short sighted" (Strachan 1994). By April of 1995, however, the GVRD 70 backed off on their earlier endorsement until all of the municipalities had the opportunity to review the BC Transit documents being prepared. By May of 1995, it had become clear that neither Burnaby nor Vancouver was going to endorse the GVRD recommendations of the Coquitlam-New Westminster route (Bula 1995; Simpson 1995a) By June of 1995, the GVRD was going to propose two lines, not including Richmond. The lack of financial resources on the part of the Province, however, meant that this was likely not to be a viable option. Although a formal proposal was never made, the mayors of Coquitlam and Port Coquitlam both charged the GVRD with weakening the interests of the region by proposing the two plans. Coquitlam Mayor Lou Sekora was further angered by what he saw as the provincial government engaging in politics by naming Burnaby Councillor Derek Corrigan as Chair of BC Transit. Corrigan and the Burnaby Council were both supporters of the Broadway-Lougheed route. "Its a conspiracy of the worst kind," Sekora was quoted as saying (Simpson 1995b). The provincial government, on the basis of the BC Transit report and the political situation discussed previously, eventually decided to endorse the Broadway-Lougheed line for rapid transit. Conclusion There are many questions over the way in which the decision was handled. How did two separate agencies review the same issue, yet propose recommendations that are markedly different? In the absence of a legitimate regional governance structure, what process could be implemented to deal with divisive decisions such as this one over rapid transit? Finally, if a reasonable answer, well-grounded in economic principles, was provided for the problem, why does there continue to be such debate? 71 Although there may be no straightforward answers to the above questions, they do raise some important points as to the role of the planner in this process. As I have heard professor and transportation planner Setty Pendakur say many times, the process is the product that the planner must provide. The major fault with the two evaluation studies reviewed is that they offer little structure or opportunity for achieving a process by which to effectively evaluate the decision. The fundamental goals and objectives were made explicit, but there was no attempt made at attributing values and weights to the different aspects of the evaluation, be they financial, environmental, or social. Furthermore, there was no attempt to include the public's input. The public was viewed merely as the recipients and end users of the product. I would argue that with a more structured and inclusive process, the public would be able to participate and contribute meaningful input into the decision; thereby providing input that could ultimately prove valuable to the political decision makers. The task, therefore, becomes one of showing how the inclusion of a more process oriented evaluation methodology would have improved this evaluation. 1 The following are the growth rates for 1961-1991 for both Richmond and Coquitlam, as well as some of the other Municipalities involved in the rapid transit Decision: Vancouver 23%, Burnaby/New West 51%, North East Sector (including Coquitlam) 218%, North Shore 76%, and Richmond/Delta 272%. 2 As a comparison of current trends to the Livable Regions strategy for growth rates, the GVRD proposed the following growth rates Current versus Proposed: Vancouver 26%, 36%; Burnaby/New West 26%, 81%; NE Sector 73%, 146%; North Surrey/Delta 105%, 164%. In comparison, areas such as Richmond, and the whole of the Fraser Valley, which would be expected to take the growth, would see much more limited future development. Richmond 79%, 47%; North Fraser Valley 138%, 87%; South Fraser Valley 155%,. 27%. 3 The question of who decides was brought up in the debate when it became obvious that the Provincial Government was not going to endorse the GVRD proposal. The GVRD Chairperson is, ironically enough, the Mayor of Richmond Greg Halsey-Brandt. He was still under the assumption (back in November of 1994) that the regional government would decide have the final decision on the route. The Provincial Opposition Leader for the Liberal Party, Gordon Campbell, concurred with Halsey-Brandt in feeling that the local governments and GVRD should have the autonomy to make the decision. 4 The Plan takes a general look at the future transportation problems assuming current trends in land development and transportation investment. Overall, it calls for $10 billion in transport investments over a 30 72 year period covering everything from new road construction, construction of future ICTS routes, and new bridges. 5 In the conclusion to the report, the GVRD make two recommendations. I f there is sufficient money to build two lines, then they suggest that Richmond-Vancouver be given top priority, then Coquitlam-New Westminster second. If, however, there is only to be money provided for one line, then it should be Coquitlam-New Westminster because of its ability to shape land use. In both cases, the Broadway-Lougheed corridor was to be built last. It is somewhat ironic that the BC Transit study, which wi l l be discussed later, chose this as the most desirable route. 6 The discount rate used is 8%, and the lifespan of the technology is assumed to be 47 years. 7 Although the EMME/2 transportation demand model which was employed in determining potential and future ridership rates is well established, the environmental values which have been calculated (notes 17 and 18) may be open to scrutiny as they are estimates generated by research in progress. As of yet, there seem to be very few widely accepted measures for quantifiable environmental impacts. Overall, however, the report does a credible job of providing reliable figures. 8 Not surprisingly, the GVRD was critical of the merits of the BC Transit study. In real dollars, the Broadway-Lougheed line is anywhere from $3-19 million more expensive than the Coquitlam-New Westminster line. As well, they take issue with the ridership projections of BC Transit. The GVRD contends that it is not appropriate to present data in terms of ridership levels and the cost per ride when there are differences in the two routes trip length. They contend that by using a cost per passenger per kilometer traveled would be a better measurement. Doing this, says the GVRD, the two routes come out virtually equal. The GVRD, however, did not produce any calculations (Griffin 1995). In my opinion, The Broadway-Lougheed line, which is one of the highest usage transit lines at the present time using buses, would only be shifting riders who are already using transit to a more enhanced form of transit. The Coquitlam-New Westminster line, however, would be creating mostly new transit users. From the data supplied by BC Transit, the cost of Rapid Bus would be $82 million to carry 17.3 million passengers on the Broadway-Lougheed route. However, the cost of rapid transit would be $562 million to carry 21 million passengers. Based on this, they would be spending almost 7 times the money to get an increase of 20% in passenger ridership. On the Coquitlam-New West route, the cost of Rapid Bus would be $62 million to carry 4.3 million passengers, rapid transit would cost $559 million to carry 12.2 million passengers. They would be spending 9 times the money, but they would be increasing ridership by 284% ! The criticism of Rapid Bus is that it wil l not shape land use as wil l Rapid Transit. However, when focusing on what is being achieved by spending the dollars, the Lougheed-Broadway line may not make as much economic sense as BC Transit may like to admit. For figures, refer to Appendix B, and Munro (1995a). 9 In April of 1995, the GVRD did in fact release a staff report on the three rapid transit corridors. The report detailed cursory financial information. However, it contained nothing as detailed as BC Transit's report. 73 CHAPTER 4 The purpose of this chapter is to provide an example of how the decision and evaluation of the next phase of rapid transit could have been structured, utilizing a more process oriented approach. Thus far, we have identified and delineated a number of alternative approaches to evaluating a decision such as this. Furthermore, we have outlined how the evaluation has been structured by two separate agencies, and examined the merits and shortcomings of each approach. The emphasis of this section will not be to offer a radical overhaul of what has already been done. Indeed, much of what has been produced and accomplished can be synthesized into the process I advocate. As such, the two documents that were discussed in the previous chapter will continue to form the basis of further discussion here. This chapter will examine the three distinct elements that are required for evaluating the decision: information, stakeholders, and a forum for decision making. Information One of the most obvious difficulties in evaluating transportation decisions is the volume of information which needs to be collected. First, there is the information that needs to be gathered on corridor selection. The corridors that are decided upon have to be evaluated for technical and financial efficiency. As mentioned earlier, most of the debate about the past proposal of a Richmond-Vancouver rapid transit line centered on whether the line should be set in the Arbutus or Cambie corridor. With the three routes being discussed, there are financial, environmental, and social costs associated with each route. Identifying these costs (and potential benefits) in technical terms is an integral part of comprehensive evaluation. The Medium-Range and Long-Range Transportation plans prepared by the 74 GVRD and partners were derived from the efforts of no fewer than seventeen background reports covering a wide range of financial, environmental, and social topics.1 As well, the Multiple Accounts Evaluation study prepared by BC Transit incorporated the technical work of many background studies relating to engineering issues and environmental and financial costs.2 The second type of information which needs to be investigated and explored is the type of general transportation issues around transportation demand. Both of the studies we have reviewed have used the GVRD's EMME/2 computerized model of transportation demand to forecast the patterns and growth of travel demand. This type of modeling, however, is only one type of analysis that can be conducted. Another very important type of modeling examines consumer choice. Demand for travel, and for alternative types of travel, can be viewed as choices in which consumers are looking to maximize their utility. The type of travel chosen from amongst a set of alternatives, such as the private automobile, the bus, or rapid transit, will be a function of a consumer's desire to utilize the service he or she believes will be most beneficial to them. The utility of any mode of travel is related to the time, cost, and convenience to the consumer. Acting in a rational manner, the choice of travel will be a function of the consumers desire to maximize utility. To relate this to our specific case study, will providing buses or rapid transit to consumers in each of the three corridor options be enough to affect consumer choice, that is, take people away from the automobile? There are many general theories within transportation planning relating to consumer choice for mode of travel. However, for this case study, pertinent consumer choice information is missing (Meyer and Miller 1984).3 The information which has so far been gathered is quite comprehensive. The information tabled in BC Transit's Multiple Account Framework would provide most of the 75 information needed for the discussion and evaluation by the different stakeholders. One drawback to the way the problem has been evaluated with respect to information is the timing of the information. For example, it is difficult to make value trade-offs regarding the environment without having the detailed route information to do so. Any further information for our case study which needs to be included will be discussed later. Stakeholders The concepts of stakeholders and stakeholder participation are very prevalent in planning. It is important to understand that being a stakeholder implies more than just participating in a process. For example, an individual may fill out a CV survey on how he or she "values" the environment. Further, he or she may be asked to provide monetary values as to how much money he or she would be willing to pay for less air pollution. The respondents are participants primarily for the purposes of information gathering. However he or she also has a vested interest - or stake - in ensuring that their natural environment is the way they want it to be. When an individual is a stakeholder, his or her role becomes much more involved. Not only will stakeholders provide values, monetary or other, but they may advocate for those values with other stakeholders, who may or may not share those same values. If the values of different stakeholders are conflicting and competing - which is often the case - then a stakeholder process, in which participants have the opportunity to discuss the issue, will allow for the exchange, debate, and even clarification of these values. The different values expressed do not simply cancel each other out, but they allow for the potential development of new values, and an opportunity for reaching a decision that is based on consensus. 76 For the purposes of our case study, we can identify a number of stakeholders. Certainly, the local politicians and delegates from each of the participating cities: Vancouver, Richmond, Burnaby, New Westminster, Coquitlam, and Port Coquitlam would all be considered as stakeholders. So too would the local politicians of those areas not involved in the decision. For instance, the decision to use transit to shape land-use patterns in the Northeast Sector has consequences for the areas of the Fraser Valley. Surrey, Delta, White Rock, Langley, and Chilliwack all have an interest in how this project proceeds. Failure to shape land-use patterns may result in further growth in each of the respective areas. Depending on local circumstances, this growth may or may not be desired. As we have seen from a review of documents, there are definitely competing and conflicting objectives amongst the various municipalities. Another set of stakeholders is the provincial government and BC Transit. Through BC Transit, the province has the responsibility and obligation to deliver these transit services. Their concerns will relate not only to doing what is in their best interest politically, but also to what is in the best interest of all of the taxpayers of BC. Furthermore, they are expected to account for the environmental and social implications of their decisions. As such, within this stakeholder group, there is likely to be a subset of stakeholder groups with competing and conflicting priorities about the decision. One group which will definitely be encountered - as with the GVRD - is the transportation planners and engineers. Traditionally, decisions such as the ones we are evaluating have been left to the experts. However, are the values expressed by these experts during the decision making process representative of the population as a whole, or even of the political members which they serve? We simply do not know. 77 A final set of stakeholders is the public. It is very clear today that in public planning matters such as this, people want - in fact expect - input into the process. Although elected representatives are ideally supposed to represent the interest of their constituents, it is obvious that in a pluralistic society such as ours, there are many different constituents with a wide variety of interests. Relating specifically to transportation matters, for example, there has been recent debate on both sides of the issue on whether or not private automobiles should bear a greater financial cost for their environmental and social externalities.4 The public stakeholder group would likely be the most diverse, in terms of values, of the three. If we were to try and restructure the evaluation process for this decision on rapid transit, we would have to find some meaningful way of incorporating the values of these different stakeholder groups into the process. We know at the outset that the values expressed are likely to be competing and conflicting. Therefore, the process must allow not only for the clarification of values for the objectives, but also for the opportunity for debate, compromise, and ultimately consensus on not only the actual objectives, be they social, environmental, or financial, but also on the trade-offs involved between these objectives. Forum It is recommended that the forum for structuring this stakeholder process will be through the use of the Delphi Process discussed in Chapter 2. The three major features of Delphi - anonymity, controlled feedback, and statistical group response - would prove to be key elements in a stakeholder process. Once again, anonymity would help to ensure that the discussion of values would not be dominated by any one individual. In other words, what is being said is focused upon, who is saying it is not. Anonymity would have even greater 78 impact when the stakeholder process is expanded to include non-homogeneous groups, i.e. politicians, transportation experts, and the public, all involved in one stakeholder meeting. The controlled feedback aspect will arguably be less important in this exercise than in a typical Delphi Process. Delphi has been used traditionally to allow for more accurate group decisions as to matters of fact. In an exercise of value clarification, the beliefs, opinions, biases, and knowledge that form individuals' values in the first place are not ignored. In fact, most of the information which the participants will be working with is already well known, such as the BC Transit Multiple Accounts Evaluation report. The controlled feedback in this case will come from the MCA facilitator who will control and guide the discussion on value trade-offs within the objectives, and how each of the alternatives is meeting the individual objectives. The statistical response of the group (or individual representing a group) is important to the type of MCA stakeholder process which we are constructing. While we are attempting to quantify values for the different objectives, these values will only have a basis in statistics. For our case study, we will be using relatively simple statistics. However, as with the discussion of some of the more complex forms of MCA, such as Concordance Analysis which was discussed in Chapter 2, the statistical response mechanism would prove very useful. If at all possible, I would advocate utilizing a computer based group decision software model, such as the GDSS model at BCIT.5 Although this process could work without such technology, having the use of such a system would save time in calculating responses, as well it would better insure the anonymity of responses. Finally, the role of the facilitator would be made easier as the computerized process would allow for less bias and coaching of responses. 79 An MCA Process for Evaluating the Decision on Rapid Transit in the Lower Mainland The following is a suggestion as to how to structure a process to decide and evaluate the next phase and form of rapid transit. Because this is not an actual research experiment being proposed, discussion of time frames, budgets, and resources will not be a major focus of discussion. Nevertheless, in the concluding remarks, efforts will be made to show why this type of process is feasible both in terms of time and money. A critical assumption here is that the basis for discussion of objectives and alternatives is limited to the documents found in Appendix A and B. Although it would be possible to gather more information, such as the background feasibility and environmental studies, the information contained in the reports already discussed forms a solid basis for useful discussion. A caveat which needs to be noted is the fact that the process as described here will only deal with three major stakeholder groups. Although there are likely to be many more involved individualized groups involved in the decision, such as environmental groups or automobile advocates, the emphasis here is to show how three distinct groups would evaluate the decision, and if there would be fundamental differences with the results they obtained. The significance of the potential differences will be discussed in Step 1. Step 1: Identify Stakeholder Groups There are three proposed major stakeholder groups involved in evaluating the decision. The first group would be comprised of the political figures (or their delegates) involved in making the decisions at both the local and provincial level. Locally, we would expect representatives from each of the twenty Greater Vancouver municipalities represented in the GVRD. Further to this, we would expect representation from the provincial level to 80 come from the Chair of BC Transit, the Ministers responsible for transit and finance, as well as representatives of the official opposition in the same capacities. In total this group would represent potentially 20-25 people. The reason that this group of individuals is being singled out as a stakeholder groups is not so much that they are of a common mind on the issue. Indeed, they would likely have very diverse opinions. However, their commonality lies in the fact that they are the group with the ultimate responsibility of making the decision, especially those members from the provincial government. Their common agenda is to ensure that the decision is financially, socially, and environmentally sound, but also that it does not do harm to them politically. The second proposed stakeholder group to be convened is the transportation experts. Within this group, one is likely to find mostly transportation planners and engineers. The groups would likely be made up of delegates from the various municipalities and those who work in the appropriate government Ministries. Although with this group, as with all three, there is the potential for engaging in regional biases, this group would be expected to evaluate the decision based solely on their professional judgement and experience. Although many of the participants will likely have ties to the various political entities involved, there is an opportunity to engage others from the field of academia, as well as those from outside the geographical area. As such, it is strongly suggested that some of this group be comprised from the various fields of economics and business, environmental studies, and social planning. The various universities and post-secondary institutions throughout the province would have a plethora of individuals eager to participate in such a process. Again, this group could easily be comprised of twenty people. The final stakeholder group would be comprised of citizens. Most of the people involved will have some prior knowledge of the issue, as well as some regional biases. 81 However, it would be important to draw participants from a wide geographical area, as well as diverse socio-economic backgrounds. It does not make sense to fill the group with people who are all transit users any more than it does to have an entire group of people who always drive their cars. This citizen stakeholder group would be expected to evaluate the decision from more than what is in their own personal interests. They would be asked to consider all of the financial, social, and environmental details from the perspective of what is in the best interest of the Greater Vancouver community. With this stakeholder group, it would be easy to include hundreds of people. One way around this may be to have two or three public stakeholder groups selected in a random draw or lottery. The group evaluations could be taken together to compare to the other two groups. At this point it is worthwhile explaining why there is the focus of the three groups. An MCA analysis, unlike CBA, does not rely on finding the "optimal" solution. Much of the documentation we have reviewed has included all of what one would find within a CBA study of the problem, especially the BC Transit report. MCA, on the other hand, is more concerned with the values which people ascribe to the different objectives, and how the objectives are met. We have identified three main groups for the purposes of examining the differences of the values. This can serve two purposes. Primarily, since all of the groups have a stake in the decision - political, professional, and personal - it would be worthwhile to see if their differing roles impact on of how their values are expressed. In essence, do the three different groups think alike, or are the values they derive unique to their own group? As well, we can even attempt to measure the group's cohesiveness of response to examine the extent to which their values are divergent. Secondly, for the benefit of the politicians and experts in charge of making the final decision, such an exercise in value clarification may yield more insight into how other stakeholders view the decision being made, and to see if their own values are 82 representative of the people they are representing. If there are significant value differences between the public and the politicians and experts, it would seem to lend credence to the argument that a more structured and inclusive public process on planning decisions such as this are warranted. More will be discussed later as to how these values, and their differences, can be measured. Step 2: Identify a Set of Decision Options The decision options for this case study have been set. From Appendix A and B, we can see that the GVRD and partners outlined seven basic objectives: Livability, Economic Development, Land Development Interaction, Social Equity, System Performance, System Management and Implementation, and Public Accountability, and BC Transit outlined five objectives: Financial, Customer Service, Environmental, Urban Development, and Systems Operations. Both reports covered most of the major categories that we would expect to deal with. However, the GVRD and partners' final recommendations were mainly based on the trade-off between System Performance and Land Development Interaction. Likewise, the BC Transit study finalized their recommendation on the Life-Cycle Boarding Cost, which was mainly derived from the Financial Performance and Customer Service accounts. The first step here would be to present the whole list of decision options to each of the stakeholder groups for their input to see if they would want any changes. The Multiple Accounts Guidelines, in their theoretical and practical form, would prove a workable basis from which to set the decision options. As well, since the provincial government and BC Transit ultimately have the final say in the decision, the use of their model would also be somewhat expected. This is not to say that other objectives could not be added, but most of the aspects of this decision that need to be addressed can be so done through the Multiple 83 Accounts Guidelines. The four accounts, or objectives, which we will work with are: Financial, Customer Service, Environmental, and Urban Development (Land Development Interaction in the GVRD and partners study). I would advocate deleting the System Operations account on the basis that it only measures the flexibility of the chosen system to adapt to future needs. (Appendix B: 15) When looking at the summary for the three corridor options in comparing Rapid Bus to the various forms of rapid transit, the bus obviously fares much better in terms of future adaptability than would a fixed rail transit scheme by virtue of its technology. BC Transit never really factored this into their final recommendations, and the fact that it is more flexible would seem to carry little weight in the final analysis. There is no significant benefit to including this category in the model, and there is nothing lost by excluding it. For the purposes of simplifying the decision, it will not be included (Appendix B:30). The decision options can be summarized in the following alternatives by objectives matrix: Table 9 Summary of Decision Options Financial Customer Service Environmental Urban Development Broadway - Lougheed LRT Rapid Bus Coquitlam - New Westminster LRT 84 Financial Customer Service Environmental Urban Development Rapid Bus Richmond Vancouver LRT Rapid Bus Stakeholder involvement in developing decision options is a key component to having an effective MCA process. If this were being done before the GVRD or BC Transit studies had been produced, then the various stakeholder groups may have arrived at a different set of decision options. This would be something for governmental decision makers to consider the next time they are dealing with such an issue. However, for practical purposes, these would most likely be the set of decision options the stakeholders involved in this process would have to consider. Step 3: Structure Stakeholder Values Once the decision options have been formalized, the task now turns to structuring the stakeholder values. As outlined in Chapter 2 , there are a number of ways in which this can be accomplished. The method that is being described is similar to that of Ward Edwards' simple MAUT framework. For the purposes of simplification, I will describe the process for one stakeholder group, but the process would be carried out with each group separately. Step A would be to present the four objectives (in the form of accounts) to the group, with all of the background information provided in the reports plus any further information. 85 For instance, under Financial, the group would be advised of the costs of the rapid transit options in general terms, as well as discussing how that money would be raised, e.g., increased taxes, types of user fees. It is important that discussion at this point be generated about the provincial government's ability to provide resources for the project.6 The Customer Service issues would be discussed around potential ridership numbers, and travel time savings expected. The Environmental aspects would be discussed as to the possibilities of reduced air pollution - specific and general - as well as the "softer" environmental issues such as potential neighborhood disruption and ambient noise levels. Finally, the discussion would center on the ability of rapid transit (in either form) to shape land-use patterns, and to relate this back to the ideas expressed in the Livable Regions strategy, and the potential impact of future urban growth. Although this discussion is in relatively general terms, it is an important starting point. What is communicated to the stakeholders is that these are four objectives which the decision on rapid transit is to serve. Each is important in its own right, but the final weighting of each objective is to be considered against the other three. At many levels, these objectives are conflicting and competing. For instance, the need to create a better environment in terms of less pollution needs to be considered against the financial costs of providing this enhanced environment. As well, the need to maximize customer service and have high ridership levels is competing with the fact that rapid transit may be built in an area that does not yet have a large existing ridership base, but building in such an area would have the greatest potential to positively impact on future minimization of land-use. The following diagram represents the relationship of the secondary objectives (or accounts) to the primary goals. As mentioned previously, the primary trade-off is between the 86 ability to serve current transit needs and the ability to shape future land-use. The four secondary objectives are derived from the primary goals as follows: Figure 2 Rapid Transit Serve Current Transit Needs A A Financial Performance A B Customer Service B Shape Future Land-Use B A BB Environmental Urban Impacts Development In serving current needs, the financial and customer service performance of an alternative choice would be important. In shaping future land-use, the consideration of the environmental impacts and urban development implications would be paramount. Ultimately, however, it is clear that the financial and customer service objectives would also have to be considered in shaping future land-use, and that the environmental and urban development objectives would be considered under serving current needs. After this initial discussion, the stakeholders would provide a ranking and weighting of the primary goals and secondary objectives for the final analysis. They would be asked to give a score to each, with the total adding up to 100. From this response, two calculations would be made. The first would be the average response of the group to each objective. This would form the group response. The second would be to calculate the standard deviation (SD) of responses to each objective, and communicate that to the group. This latter measure is important because it is a good indicator of the group's cohesiveness of values. If the SD is 87 large, then the group and the facilitator would know that there is a wide range of opinion. This would be an opportunity for further discussion and clarification of issues. A second round of scoring would allow stakeholders the opportunity to reconstruct their values if, indeed, any have changed. After this second round, or any subsequent round, the average and SD of responses would be finalized. Once again, the average of the group would form, that group's response. The SD of responses is likely to change after a round of discussion and debate, and would indicate to the facilitator the cohesiveness of the groups response. A small SD of response would indicate that the group is of similar opinion as to how they value the various objectives. A large SD would indicate that there is less cohesiveness in that group. The SD would be valuable to the facilitator because it could identify problems with the objectives being too general or vague. If the SD were large, perhaps disaggregating the information in the objectives and creating more specific objectives would allow the group to form different, and perhaps more cohesive, responses. Nevertheless, simply reporting the SD, without changing any of the models structure, would still provide the decision makers with valuable information as to stakeholder values. Step 4: Build Common Value Structures At this point in the process, it is time to construct a common value structure similar to Table 3 of Chapter 2. As with the Edwards example, there are to be six components in calculating the four final objective weights of each group. Initially, a group would decide, on the basis of a total score of 100, how much weight should be given to rapid transit: (A) Serving Current Transit Needs, or (B) Shaping Future Land-Use. For example, the group could decide that A should be given a weight of 65 and B 35. The use of the 100 point scale facilitates the use of percentages, therefore, serving current transit needs would influence their 88 decision by almost two to one. For each of the four secondary objectives, the group would present their score, out of 100, with the total of the four objectives adding up to 100. The hypothetical distribution may look as follows: AA Financial - 35, AB Customer Service - 25, BA Environmental - 15, BB Urban Development - 25. These six normalized weights would then be used to calculate the four secondary objectives final weights by multiplying as follows: AA x A = final weight Financial, AB x A = final weight Customer Service, BA x B = final weight Environmental, and BB x B = final weight Urban Development. Using the hypothetical values constructed above, the table would look as follows: Table 10 Hypothetical Weights for Rapid Transit Objectives Group 1 Normalized Final Weight Weight (A) Current Needs .65 (B) Shape Land-Use .35 (AA) Financial .35 .2275 (AB) Customer Service .25 .1625 (BA) Environmental .15 .0525 (BB) Urban Development .25 .0875 From the final weights in this hypothetical example, we can see that the overwhelming bias towards serving current transit needs means that the final weights for the Financial and Customer Service accounts will be large, and will thus influence the decision considerably. The Financial weighting is almost three times that of Urban Development, and four times that 89 of Environmental. As shown, this is the table for only one of the groups. The table for the other two groups would be constructed in the same fashion. Although we have stated an interest in finding out the differences between the three groups, there is also the opportunity to derive a common score at the end. The compromise score would simply compute the average weights, in each of the primary goals and secondary objectives, for the three groups. This common weighting would be used, as will be shown later, to highlight the alternative which is best suited to all three groups. Although in Edwards bike example the common score worked as a solution to picking the bike which suited both he and his son, it would not likely work for our case study. It would be naive to think that the decision on such a project would be based on a compromise of the three groups - especially if the compromise choice was different from the choice of the politician group who are the ultimate decision makers. Nevertheless, including a common score, and ultimately choice, could prove useful when discussing the differences between the groups. Step 5: Build Quantitative Value Models for each Stakeholder Group At this point in the process, each of the groups will have discussed, debated, and clarified their values with respect to the primary goals and secondary objectives of this rapid transit evaluation. It is important to note, however, that this has been done in the absence of any discussion as to the alternative choices which are available. As discussed earlier, and modeled in Table 8 of this Chapter, there are three alternative route choices, and within each route choice there is the option of two types technology: LRT and Rapid Bus. In actuality, the inclusion of Rapid Bus as a choice is primarily for academic purposes. As discussed, this technology was given virtually no consideration by either the GVRD and partners or BC Transit. However, given that evaluating Rapid Bus could have significant implications with 90 respect to all four secondary objectives, I would contend that such an exercise as this should include it as a choice. By the same token, consideration could have been used to include Advanced Light rapid transit (ALRT) on the three corridors had more information been obtained. The cost of ALRT, however, excluded it from even the type of cursory examination that Rapid Bus was subjected to. As with Table 4 from the Edwards example discussed in Chapter 2, the next step is to evaluate each of the alternative options with respect to how it performs under each of the four secondary objectives. For example, our hypothetical group would be presented with information as to the cost and workings of each technology, and how they would meet the objectives with respect to the four accounts. Such information could easily be compiled in the form of notes presented to each participant in the stakeholder groups, and followed up with a question and discussion period to clarify any issues of information which is uncertain. To be sure, there would be an opportunity for biased information to be presented as the GVRD and BC Transit reports do have different interpretations of costs and impacts. Nevertheless, the facilitator could present both sides, and leave the stakeholders to decide which information is the most credible. With this information in place, the stakeholders would then give each alternative a score, once again out of 100, as to how that alternative measured up to the needs of the objective. A score of 100 would mean that the stakeholder believed the alternative to meet all of the criteria for that objective, with a score of 0 meaning that none of the criteria were met. This score represents the utility value of that alternative for the stakeholder. The one difference from the previous type of ranking done on the common value structure is that now each score given is independent of the next. For each alternative choice, such as LRT in the Broadway-Lougheed corridor, it may achieve a 100 in each of the four accounts. The total of 91 its score does not necessarily add to a total of 100. The following is an example of a partially constructed table, looking at only one corridor. Broadway-Lougheed LRT Rapid Bus Financial 60 90 Customer Service 95 45 Environmental 80 60 Urban Development 95 55 The utility values for each of the two alternatives are represented in this table. The values indicate that the stakeholder (or stakeholders if this was the average of the group response) felt that from a Financial perspective, Rapid Bus had significantly more utility than did LRT. However, with respect to the other three objectives, LRT had greater utility. Even though these utility scores are bound to be somewhat subjective, I would advocate to keep them separate for each of the three groups. These scores do represent relatively individual values, and it would be consistent with monitoring the decision of the three groups for differences. The final part of the exercise is to compute the aggregate utility scores for each stakeholder (or stakeholder group) with respect to the alternatives available. This is done by multiplying the final value weight for a given objective, e.g. AA Financial, by the route and mode utility score for that objective. Using the hypothetical values already obtained, we know that the Financial objective (or account) weighting is .2275. Further, we know that for the Broadway-Lougheed corridor, the LRT had a utility score of 60, while Rapid Bus had 90. The aggregate Financial utility score for each of these alternatives would compute as follows: .2275 x 60 = 13.65 weighted score for LRT in Broadway-Lougheed corridor .2275 x 90 = 20.475 weighted score for Rapid Bus in Broadway-Lougheed corridor 92 With the hypothetical values we have obtained for this one group, we can actually compare the alternative choices of LRT and Rapid Bus by calculating the sum of their weighted utility scores. The following is a partial table of the scores for the Broadway-Lougheed corridor. B road way-Lougheed LRT Rapid Bus AA Financial .2275 x 60 13.650 .2275 x 90 20.475 AB Customer Service .1625 x95 14.625 .1625x45 6.500 B A Environmental .0525 x 80 4.200 .0525 x 60 3.150 BB Urban Development .0875 x 95 7.875 .0875 x55 4.813 SUM 40.35 34.94 Our sample group arrived at an aggregate utility score of 40.35 for LRT, and 34.94 for Rapid Bus. For each of the other two corridors, a similar table would be constructed, leaving each group with six separate utility scores evaluating the six alternative options. The following table represents the relationship of how the information could be presented for final discussion. Table 11 Final Computation of Aggregate Utilities BROADWAY-LOUGHEED Group 1 Wt. Group 2 Wt. Group 3 Wt. Common Wt. LRT R-Bus LRT R-Bus LRT R-Bus LRT R-Bus Dimensions of Values AA (Financial) 13.650 20.475 AB (Customer Service) 14.625 6.500 BA (Environmental) 4.200 3.150 BB (Urban Development) 7.875 4.813 SUM 40.350 34.940 93 R I C H M 0 N D - V A N C 0 U V 1 ER Group 1 Wt. Group 2 Wt. Group 3 Wt. Common Wt. L R T R-Bus L R T R-Bus L R T R-Bus L R T R-Bus Dimensions of Values A A (Financial) A B (Customer Service) B A (Environmental) BB (Urban Development) S U M COQUITLAM-NEW W E S T M E NSTER Group 1 Wt. Group 2 Wt. Group 3 Wt. Common Wt. L R T R-Bus L R T R-Bus L R T R-Bus L R T R-Bus Dimensions of Values A A (Financial) A B (Customer Service) B A (Environmental) BB (Urban Development) S U M Step 6: Use the Value Models for Communication. Conflict Diagnosis. Resolution It is critical to understand that Step 5 is not the end of the process. An answer has not been achieved. All that exists is a clear and defined expression of values on behalf of three distinct groups. At this point, a true stakeholder process would convene the different groups to have further dialogue and discussion, and perhaps reach some sort of consensus on the decision. For our case study, this is not likely to happen. As mentioned, the decision on rapid transit had been left to a relatively small group to decide, namely the cabinet of the provincial government. The decision of Broadway-Lougheed is a matter of public record. However, the intriguing question is how might have the decision changed had such an exercise as this been 94 conducted, and the values expressed by the public, or even those of transportation experts, been different? Would the Broadway-Lougheed corridor still have been chosen for an LRT line if the other group's values had been dramatically different than that of the politicians? Perhaps an alternative corridor, or alternative technology, would have been deemed more appropriate from a consumer utility point of view. In the absence of using such a model, we will never know if the decision could have been - or perhaps more importantly should have been - different. Conclusion The relative merits and the shortcomings of this model, as well as opportunities for further applications, will form the basis of discussion for Chapter 5. However, it would be more appropriate here to discuss perhaps the element that will usually have the greatest impact on this type of public decision - the political context. As was discussed in Chapter 3, the decision to provide an LRT line on the Broadway-Lougheed corridor will be to many people nothing more than political patronage. However, if the previous provincial government were still in power, and the Richmond-Vancouver corridor were selected, one would doubt that the accusations and charges would be any different. A fundamental question is, will politics continue to influence this type of decision making? The simple answer is that it likely will continue. The type of MCA process discussed here, however, would lend some much needed credibility to the decision making process. Primarily, it would provide a meaningful opportunity for consensus decision making on difficult and contentious public issues. Furthermore, it would provide the political decision makers with invaluable information as to public values on these types of issues. With such a process, future decisions maybe no 95 less contentious, but they are likely to be more sensitive to the public, and certainly more defensible for decision makers. 1 GVRD (1993d:6) Of the seventeen papers, we have already reviewed "Goals, Objectives, and Criteria", and seven of the papers relate to the demand for transpiration, two relate to the supply of transportation, two discuss the costs of transportation, three discuss the land use implications, and the final two cover economic development perspectives and the implications of new transportation technologies, such as LRT and ALRT. 2 See Appendix B. The following footnotes contain the information on the background studies. P.4, Note 2.: three engineering and feasibility studies relating to the three corridor options. P. 10, Note 10: the use of the EMME/2 transportation trip forecast model. P. 12, Note 13: report on revenue sources to fund transit. P. 13, Notes 17 and 18: discuss the quantification and costing of environmental factors. 3 Meyer and Miller (1984:257) Chapter 7 provides a very detailed analysis of Demand Analysis for transportation. Most of the models for forecasting demand are grounded firmly in economics and statistics. The reason for focusing on consumer choice, however, stems from the fact that I am advocating a more public oriented process. Surely, many of the assumptions that the decision makers at BC Transit and the GVRD have taken these factors into account. However, the fact that there has been not noticeable attempt to gauge consumer opinion in each of the three corridors is a major shortcoming. Even the use of CV surveys would be helpful. For my model, having a separate public stakeholder group wil l be an attempt to overcome this shortcoming by having their values reflected directly in the process. 4 This is a very controversial topic, but it is one that is getting a fair bit of coverage. The rationale is that private automobile drivers are heavily subsidized by the taxpayers. The infrastructure which they utilize, such as roads and even traffic enforcement activities, are paid for by everyone but used only by them. The concept of road pricing is one whereby the users would be expected to pay for using these facilities. This would be achieved through higher fuel taxes, bridge tolls, and increased parking fees. These transport demand measures, which are expected to shape and change travel behaviors, are a central part of the GVRD's Long-Range Transportation Plan (1993c:22y Speaking recently at the National Sustainable Transportation Conference in Vancouver, transportation expert Alan Thein Durning discussed this concept in detail. He advocates that the cost of driving should reflect all of the associated environmental externalities, and be based on a variable instead of fixed scale, as with insurance premiums. He said that in the US, drivers pay about $0.10 for maintenance and fuel to drive one mile, while the actual cost is $1.00 i f everything from traffic police to accident costs are factored in. In the same article, it was pointed out that the GVRD calculates the subsidy to Lower Mainland motorists at $2.6 billion per year. Bonn (1995) Within BC, another public advocacy group has formed to lobby the government against the idea of further penalizing motorists with higher taxes and user fees. The group is headed by former Social Credit M L A Russ Fraser. 5 The GDSS model at BCIT would prove very useful to this type of experiment. It can accommodate over 20 people at a time, each with their own terminal, in a roundtable format. The room size is small enough so that discussion without the computers is possible and even recommended. The system allows for responses to be input verbatim into the computer, and for lists to be compiled and prioritized. Voting takes place though the computer, and the groups response can be tabulated in seconds. As well, the statistical package available offers the opportunity to assess group responses with simple or complex calculations. 6 Another issue which has been raised, and which could even be discussed and debated, is the possibility of private sector financing. This issue was raised by the GVRD as a way helping to get the systems built. Regional Manager Ben Marr argued that the cost of providing sewers and infrastructure for sprawling growth would cost up to $4 billion dollars more than to upgrade the transportation system. This savings could be utilized towards the rapid transit projects. The one flaw in this argument, however, is that the rapid transit would have to be paid for in real dollars, and these are relatively scarce. Others, such as BC Transit chair 96 Corrigan and McGill University Transportation Planning Professor Ron Rice have argued that the notion of private sector financing wil l likely not work because transit is subsidized. Other cities have attempted to entice investors with other concessions, such as free land, but these have not proven to be very successful. Munro (1995b) 97 CHAPTER 5 At the outset of Chapter 2, we reviewed the terms oikonomia and chrematistics and what impact they had on the field of economics and evaluation methodology. Further, we have reviewed a number of different evaluation methodologies, and discovered that they range from the almost purely chrematistic CBA, to the mostly oikonomic Multiple Accounts. These distinctions must be tempered, however, by the fact that any evaluation methodology contains elements of both. The combined MCA-Multiple Accounts-Delphi process I described in the last chapter offers, I believe, the best opportunity for an evaluation methodology to strike this balance between hard and soft economics in the field of evaluation. The combined approach can incorporate the traditional elements of monetary evaluation and discounting into the analysis, but it is used to construct stakeholder values as opposed to providing an answer or solution. The focus on construction of values around a wide variety of interests - economic, social, environmental, and even political - insures that the evaluation and decision process has become more comprehensive and less exclusionary. This broader scope, and mandate, will help to make the process more credible with decision makers, and with the public for whom these decisions are made. Strengths, Weaknesses and Practical Shortcomings of the Combined MCA Process Strengths The fundamental strength of the model lies in its comprehensiveness. The model is adaptable to deal with any number and type of decision options and stakeholder groups. Drawing on the two documents of the GVRD and partners and BC Transit, our model was able to incorporate elements of both and come up with four secondary objectives relevant to 98 both reports. However, in the absence of preconceived decision options, the model would have been able to generate options through the input of stakeholders. Indeed, in its purest form, MCA calls for stakeholders to set the decision options themselves at the start of the process. The second component of this comprehensiveness lies in the use of stakeholder groups. For this case study, we convened three distinct groups to evaluate the issue. However, the model is sufficiently flexible to include almost any number of different stakeholder groups, and to include their input in the final communication. Another strength of the model is that it is relatively systematic. With respect to the financial and monetary calculation of costs and benefits which are incorporated into the model's accounts, the use of discount rates and the calculation of benefits to NPV means that the model is capable of reproducing results. However, Edwards's six step approach to evaluation also means that the process is capable of being reproduced. Obviously, the flexibility inherent in any stakeholder process means that decision options and values can change from one case study to the next. But the process that the stakeholders will go through - from identifying and aligning themselves, to identifying a set of decision options, to structuring their own and then common values, and building quantitative models to communicate the results - is the same. Once the steps in the process are understood, using the model is relatively straightforward. The final strength of the model lies in its simplicity. Everybody involved with the process, from the evaluation facilitator to the individual stakeholders, should be able to understand the concepts expressed in this model. The values are constructed and communicated using only the most basic statistical concepts of the average and standard deviations of responses. The calculation of the SD of responses, as mentioned previously, is mainly for the benefit of the facilitator and the final decision makers. The simplicity of 99 structuring and communicating the stakeholder values, indeed the accessible nature of the process, means that the values constructed may be less open to questions about their authenticity. A more complicated process, as some MCA processes are, could lead to skepticism about the results. Weaknesses Despite the simplicity of the model as a practical strength, it is also a principal theoretical weakness. By constructing values in such a simple manner, and using the average of the group response, many would argue that those values are not completely accurate or reflective of the stakeholders' beliefs. The works of Keeney and Raiffa suggest, for example, constructing utility curves for each of the objectives, and establishing a range by which trade-offs take place. For our case study, this type of approach would have been less concerned with establishing that the Financial account had a greater weight than the Urban Development account. The focus would have been on questioning stakeholders as to how much of one account they would have been willing to give up to obtain more of the other. For instance, how much Customer Service would they be willing to trade for a more enhanced Environment. They would use this information not only to establish a ranking, but also range of results as to the sensitivity of the trade-off. Although this is a rather crude comparison, it provides a good example as to how much more complex MCA analysis can be. The decision on whether or not a more technical approach is going to be more appropriate or effective for this evaluation question needs to consider what would be sacrificed in terms of time, money, and confusion, by the more complicated process. For this case study, I believe that the simple model, in view of its strengths, offers a valid method for constructing stakeholder values around the question of rapid transit. 100 A second weakness of this model is that it is not as easy to use as other models. The BC Transit Multiple Accounts evaluation is typical of what would have been a final step. The data were gathered, the numbers and impacts were analyzed, and a report offering solutions was published. This is the norm for evaluation. Our case study model is obviously more time consuming because it goes beyond this step. As well, depending on the numbers of people and groups involved in a stakeholder process, the process of value clarification can take a long time. Groups have to be convened, meeting times set, and resources - such as the GDSS computer facility at BCIT - allocated. All of these have a detrimental impact on the time it takes to complete such an evaluation. However, the benefits derived from this extra time must be weighed against the costs of doing so. I believe that the extra time and resources needed for the MCA process, especially for a billion dollar project, is warranted. Practical Shortcomings Along with the weaknesses of the model, there are a number of practical shortcomings which need to be addressed for our case study. The previous chapter discussed convening three distinct stakeholder groups. For two of the groups this would be relatively straight forward to decide on who was to be involved. For the group of transportation experts, there is no shortage of people who could be involved. Each municipality in the GVRD would have at least one person they could designate. As well, rovincial Ministries of Transportation, Finance, Environment, and Social Services would have a number of people who would want representation in this group. From the field of academia, the various post-secondary institutions would have no shortage of interested parties wanting to have input. For reasons of effectiveness, a group should be limited to no more than twenty persons.1 One question is, who is to be left out? One way of overcoming this is to convene more than one of these 101 groups. However, when dealing with the third group, the public, this becomes rather difficult. If five hundred people are interested in participating, there is no structure in place to decide who is to be selected. For cost and time reasons, it would be problematic to convene twenty-five separate "public" groups, especially if the results obtained from one group to the next are not markedly different. A second practical shortcoming of the model as applied to this case study is that there exists no mandate for this type of participation in the decision making or evaluation process. For our case study on rapid transit, the mandate for the decision rested with the Provincial Government and BC Transit, with the help of the Crown Corporations Secretariat, who conducted the Multiple Accounts Evaluation. The problem of deciding who else may participate in such a process as this is an academic one. The answer, at this time, is no one. Items for Future Consideration Stemming from these weaknesses and practical shortcomings, there are a number of items for future consideration if the model is to be used. As the Crown Corporations Secretariat is used to evaluate projects or issues from the basis of Multiple Accounts, it may be possible to extend that mandate to include an MCA evaluation of the problem as well. As demonstrated, the information obtained by such a process would prove invaluable to the political decision makers as to public values on sensitive issues such as transportation, housing, and the environment. In fact, a great deal of work done with MCA (and referenced in Chapter 2) has been done around the issue of evaluating environmental decision. There is no shortage of local and provincial issues which would benefit by this process. Once the mandate is in place, there has to be a mechanism for involving the public in the decision process. Any local planning meeting, especially one around a contentious topic, 102 will turn up a number of interested parties who have a stake in the decision being made. Information needs to be obtained and preserved on interested parties so they may be contacted and used for their input. Certainly, agencies such as the Crown Corporation Secretariat or the GVRD could use this information to convene stakeholder processes on issues they need, or want, to evaluate. If the lists of interested persons is long enough, then selection could be made by random sample to choose who will participate. With the investment in facilities, such as a GDSS system, and staffing resources, an Evaluation Department could be organized to deal with these issues on an ongoing basis. If a mandate were not given to include the public in the process, this Department could prove useful from the perspective of an information gathering department. They would be able to elicit public values on issues, such as rapid transit, and present this information to the decision makers. The input may not ultimately change the decision, but it is doubtful that the decision makers would ignore the results. Another item for future consideration relates to some of the assumption that this model makes as to conducting a comprehensive and inclusive public process. It woud be useful, from the perspective of conducting further research, to do a thorough analysis of the history of transportation decision making in BC, and review the decisions on Skytrain and the Coquihalla Highway. What models or processes did they follow? Further, it would be useful to contact key participants from the GVRD, BC Transit, and the Crown Corporations Secretariat, to ascertain their feelings on how the transit decision they were involved with would have differed with such a stakeholder process at the end. Would such a process have provided a basis for consensus? Or rather, would the inclusion of the public, to such a large degree, complicate and paralyze the decision? Would the political decision makers, such as the provincial cabinet, be interested in such a process? Would the values obtained provide them with direction or distraction? Finally, what is the influence of politics, vis a vis all of the 103 other considerations, on this type of decision? These are a sample of the types of areas for future direction of research. Conclusion This thesis began by trying to examine the different evaluation methologies with a broader concept of economics than the chrematistic, technical type of which we have become accustomed. Daly and Cobb spoke of trying to shift economics into a new paradigm that would not sacrifice any of the traditional elements, but would incorporate some of the "softer" aspects which need to be addressed in our current lives. Our concerns for social and environmental well being do not necessarily fit well with the traditional type of models we use to evaluate the type of public decisions we face, such as with the environment, housing, or transportation. We cannot simply dispense Cost Benefit Analysis, but we have an opportunity, even an obligation, to try and expand our scope of what economic evaluation means. We have seen that this is possible. The evaluation efforts of the GVRD and BC Transit represent a strong foundation to the type of comprehensive evaluation that needs to take place on such projects. It may have been unfair to characterize their fundamental flaw as being a lack of process. Indeed, the GVRD report was the outcome of a substantial public process. However, the final results and recommendations did not seem to fit with the models from which they were derived. In the final analysis, both of the model's structures were compromised, and the decisions came down to very few factors. In the case of the BC Transit study, the Lifecycle Cost per Boarding seemed to carry the most weight, and it was not even an objective. A successful decision process would strive to ensure that the decision would flow from all of the primary and secondary objectives. 104 We have outlined many theoretical and practical obstructions to conducting a successful evaluation. The combined model that was described offers only a partial solution to overcoming these problems. One can argue that such a simple process does not adequately capture the values that stakeholders have, or the value trade-offs that are made. This may be true, but at least this is a start. If the process is too simple, then further efforts can be made to adapt and modify it, as are being done with CBA. Although there are obstacles and challenges that need to be addressed in trying to conduct a more inclusive evaluation process, these barriers need to be set against the potential benefits of what could be achieved. If we can move decisions from being contentious to consensus, then the efforts are well worth it. 1 Dalkey (1972:19) There area few reasons for limiting the size of the group to approximately twenty persons. The first, as they found with the Delphi experiments at RAND, was that there were diminishing returns to accuracy by adding more people beyond this limit. From a practical sense, the resources needed for our project, such as the GDSS, could easily accommodate this many irt one sitting. A larger group would require, for instance, a lecture hall as opposed to a classroom. Finally, to control discussion amongst twenty persons would be a challenge enough. Having a group larger than this would obviously impact on the amount of time needed for a stakeholder group meeting. 105 Bibliography Boardman, Anthony, Aidan Vining, and W.G. Waters II. "Costs and Benefits Through Bureaucratic Lenses: Example of a Highway Project." Journal of Policy Analysis and Management Vol. 12 No. 3 (1993): 532-555. Bohn, Glenn. "Make Drivers Pay Higher Taxes, Transportation Expert Says." The Vancouver Sun. 1 November 1995. Brookshire, David S., and Don L. Coursey. "Measuring the Value of a Public Good: An Empirical Comparison of Elicitation Procedures." The American Economic Review 77.4 (September 1987): 554-566. Bula, Frances. "Vancouver Rejects New Transit Route." The Vancouver Sun. 17 May 1995. Butters, Gerard, John Calfee, and Pauline Ippolito. "Reply to Steven Kelman." Regulation 5.2 (March-April 1981): 41-42. Crown Corporations Secretariat. Multiple Accounts Evaluation Guidelines. Victoria: Queens Printer for British Columbia. 1993. Crown Corporations Secretariat. Multiple Accounts Evaluation of Rapid Transit Options in Greater Vancouver: BC Transit Ten Year Development Plan 1995-2006. Victoria: Queens Printer for British Columbia. 1995. Dalkey, Norman C. et al. Delphi and Decision-Making. Lexington: Lexington Books. 1972. Daly, Herman E. and John B. Cobb, Jr. For the Common Good: Redirecting the Economy Toward Community, the Environment, and a Sustainable Future. Boston: Beacon Press. 1989. Davis, H. Craig. Regional Economic Impact Analysis and Project Evaluation. Vancouver: University Press. 1990. Edwards, Ward, "Reflections on and Criticism of a Highly Political Multiattribute Utlility Analysis" , in Cobb, Loren, and Robert M. Thrall, eds. Mathematical Frontiers of the Social and Policy Sciences: AAAS Selected Symposium 54. Boulder: Westview Press. 1981. Edwards, Ward, and Detlof von Winterfeldt. "Public Values in Risk Debates." Risk Analysis 7.2(1987): 141-158. 106 Einhorn, Hillel J. and William McCoach, "A Simple Multiattribute Utility Procedure for Evaluation," in Beckmann M. and H.P. Kunzi eds. Lecture Notes in Economics and Mathematical Systems. New York: Springer-Verlag. 1978. Greater Vancouver Regional District. Managing Greater Vancouver's Growth. Burnaby: GVRD. 1993a. Greater Vancouver Regional District. Transport 2021 Working Paper: Goals. Objectives and Criteria for Developing a Long Range Transportation Plan for Greater Vancouver. Burnaby: GVRD. 1993b. Greater Vancouver Regional District. A Long-Range Transportation Plan for Greater Vancouver. Burnaby: GVRD. 1993c. Greater Vancouver Regional District. A Medium-Range Transportation Plan for Greater Vancouver. Burnaby: GVRD. 1993d. Gregory, Robin, and Ralph L. Keeney. "Creating Policy Alternatives Using Stakeholder Values." Management Science Vol. 40 No.8 (August 1994): 1035-1048. Gregory, Robin, Sarah Lichtenstein, and Paul Slovic. "Valuing Environmental Resources: A Constructive Approach." Journal of Risk and Uncertainty 7 (February 1993): 177-198. Griffin, Kevin. "GVRD Critical of BC Transit Projections." The Vancouver Sun. 24 July 1995. Gwilliam, K.M. "Economic Evaluation of Urban Transport Projects: The State of the Art." Transportation Planning and Technology 1 (February 1972): 123-142 Hill, Morris. Planning for Multiple Objectives: An Approach to the Evaluation of Transportation Plans. Philadelphia: Technion. 1973. Haveman, Robert H. "The Economic Evaluation of Long-Run Uncertainties." Futures 9 (1977): 365-374. Keeney, Ralph L., and Howard Raiffa. Decisions with Multiple Objectives: Preferences and Value Tradeoffs. New York: John Wiley & Sons. 1976. Keeney, Ralph L. "Decision Analysis: An Overview." Operations Research 30 (September-October 1982): 803-832. Keeney, Ralph L. "Structuring Objectives for Problems of Public Interest." Operations Research 36 (May-June 1988): 396-405. 107 Kelman, Steven. "Cost-Benefit Analysis: An Ethical Critique." Regulation 5.1 (January-February 1981): 33-40. Lamb, Jamie. "NDP's on board the gravy train called light-rail transit." The Vancouver Sun. 16 November 1994. Lichfield', Nathaniel. "Evaluation Methodology of Urban and Regional Plans: A Review." Regional Studies 4 (1970): 151-165. Lichfield, Nathaniel, Peter Kettle, and Micheal Whitbread. Evaluation in the Planning Process. Oxford: Pergammon Press. 1975. McAllister, Donald M. Evaluation in Environmental Planning: Assessing Environmental. Social. Economic, and Political Trade-offs. Cambridge: The MIT Press. 1980. McDaniels, Tim. "A Multiple Objective Decision Analysis of Land Use Options for the Tatshenshini-Alsek Area." Victoria: Commission on Resources and Environment. 1992. McDaniels, Tim. "Sustainability, Value Tradeoffs, and Electric Utility Planning: A Canadian Example." Energy Policy Vol. 22 No. 12 (1994): 1045-1054. Massam, Bryan H. "The Search for the Best Alternative Using Multiple Criteria; Singapore Transit Study." Economic Geography 54 (1978): 245-253. Meyer, Micheal D. and Eric J. Miller. Urban Transportation Planning: A Decision-Oriented Approach. New York: McGraw-Hill Book Company. 1984. Munro, Harold. "Coquitlam Rapid-Transit Line Least Costly Route." The Vancouver Sun. 16 March 1995a. Munro, Harold. "Private Funds Viewed as Way to Finance Rapid Transit Lines." The Vancouver Sun. 12 April 1995b. Nijkamp, P., P. Rietveld, and H. Voogd. Multicriteria Evaluation in Physical Planning. Amsterdam: North-Holland. 1990. Pearce, David W., and R. Kerry Turner. Economics of Natural Resources and the Environment. Baltimore: The Johns Hopkins University Press. 1990. Pill, Juri. "The Delphi Method: Substance, Context, a Critique and an Annotated Bibliography." Socio-Economic Planning Science 5 (1971): 57-71. 108 Roads and Transportation Association of Canada: Project Committee on Urban Transportation Planning. Urban Transportation Planning Guide. Toronto: University of Toronto Press. 1977. Schofield, J.A. Cost-Benefit Analysis in Urban & Regional Planning. London: Allen & Unwin. 1987. Schofield, J.A, "Cost-Benefit Analysis and Regional Science: A Case of Comparitive Neglect." Canadian Journal of Regional Science. 14.3 (Autumn 1991): 315-328. Simpson, Scott. "Broadway Plan Wins Applause from Bus Riders." The Vancouver Sun. 15 November 1994. Simpson, Scott. "GVRD Delays Decision on Rapid Transit." The Vancouver Sun. 13 April 1995a. Simpson, Scott. "2 Line Transit Plan Angers 2 Mayors." The Vancouver Sun. 7 June 1995b. Smith, Kerry V, "The Foundations of Benefit-Cost Analysis," in Bentkover, Judith Ff., Vincent T. Covello, and Jerry Mumpower, eds. Benefits Assessment: The State of the Art. Dordrecht: D. Reidel publishing Company. 1986: 13-34 Strachan, Alex. "Coquitlam Transit Link Defended." The Vancouver Sun. 19 December 1994. Tell, Bertil, "An Approach to Solving Multi-person Multiple-criteria Decision Making Problems," in Beckmann M. and Ff.P. Kunzi, eds. Lecture Notes in Economics and Mathematical Systems. New York: Springer-Verlag. 1978. van Delft, Ad, and Peter Nijkamp. Multi-Criteria Analysis and Regional Decision-Making. Leiden: Martinus Nijhoff Social Sciences Division. 1977. 1 0 9 A P P E N D I X A Table of Contents 1 1 . 0 Preface i 1. Introduction 1 2. Context for Transportation Systems 3 3. Criteria for Development of the Long-Range Transportation Plan 5 Appendices Greater Vancouver Regional District: Principles, Strategic Policies and Operational Policies from Creating Our Future Ministry of Transportation and Highways: Mission Statement BC Transit: Mission Statement List of Figure and Table Figure 1: Transportation and Regional Goals Table 1: Transportation Goals, Objectives, Criteria 4 7 I l l P R E F A C E 1 . T R A N S P O R T 2 0 2 1 This working paper is part of the output of TRANSPORT 2021, a joint project of the Province of British Columbia and the Greater Vancouver Regional District. The District comprises 18 member municipalities from the Lower Mainland of B.C. The paper was prepared to help achieve the objectives of the project, which are as follows: "based on the Greater Vancouver Regional District's Creating Our Future action plan, its Livable Region Strategy and the mission statements of the British Columbia Ministry of Transportation and Highways and BC Transit, to recommend, by the end of 1993, a long-range transportation plan for Greater Vancouver, with associated policies, demand management measures and priorities for transportation investment." 2 . P u r p o s e The purpose of the paper is to set forth a statement of goals, objectives and criteria for use in the development of a long-range transportation plan for Greater Vancouver. The statement is intended to integrate the transportation components of the GVRD's Creating Our Future goals, and the mission statements of the Ministry of Transportation and Highways and BC Transit. TRANSPORT 2021 1 1 2 3 . A c k n o w l e d g m e n t s The paper was prepared for the project by Dr. Ron G. Rice, Professor . at the School of Urban Planning of McGill University, at the request of Martin L. Crilly, Project Director of TRANSPORT 2021. Dr. Rice serves as academic advisor to the TRANSPORT 2021 project team. David Marr of the project team performed much of the research for the paper. 4 . A d d i t i o n a l C o p i e s Copies of the paper can be obtained at a cost of $5.50 each. Call or GVRD Communications and Education 4330 Kings way Burnaby, B.C. V5H 4G8 Telephone: (604) 432-6339 Fax: (604) 432-6399 fax: 1 1 3 1 . I N T R O D U C T I O N Within the exercise of planning, one of the most important component tasks is the development of a value structure, to guide both the generation of alternative plans and the evaluation and selection of a final recommendation. This value structure is typically conceived as a three-level hierarchy, consisting of goals, objectives and measures of effectiveness. It defines what one wishes to achieve, while the alternative policies or plans define how one might achieve it. This distinction is most critical. The definition and content of the framework of values should be derived explicitly from those for whom the plan is being prepared (decision-makers, politicians, the public), so that the appropriate trade-offs may be identified and considered in the final choice of plans or actions. It remains the responsibility of the professional or technician, then, to estimate accurately the measures of effectiveness utilizing available estimation procedures. The purpose of this working paper is to define the value structure which is to be applied to the assessment of alternative long-range transportation plans. More specifically, it will seek to do the following: 1. Describe the basis for the development of transportation goals, objectives and measures of effectiveness, and Z Present the specific measures of effectiveness or criteria to be applied to the evaluation of transportation supply scenarios, which have been developed to serve the selected long-range regional development plans. For the purpose of this exercise, it is necessary to define the terms to be applied, and these are as follows: Transport 2021 l 1 1 4 Values: beliefs and basic principles which direct and govern human behaviour. Goals: abstract statements which reflect the views, perceptions and aspirations of the community. Objectives: specific statements which are quantifiable of what is desirable to achieve, related directly to the goals. Measures of effectiveness or criteria: measures of the degree to which the attainment of the objectives may be assessed. In general, the measures of effectiveness may vary by degree of specificity or disaggregation, depending upon the spatial scale of application (regional, zonal, corridor, facility) and the temporal dimension of application Gong-, medium- or short-term planning). Transport 2021 2 1 1 5 2 . C O N T E X T F O R T R A N S P O R T A T I O N S Y S T E M S In the development of a long-range regional plan for the Greater Vancouver area, the transportation system is an important physical component, but only one of several components of the required infrastructure. In this sense then, transportation systems are service systems, designed to support the social and economic activity patterns which make up the broader regional vision. Accordingly, it is most essential that the transportation systems, and their corresponding goals and objectives, relate directly to the broader regional context. Considerable effort has already been devoted to the evolution of principles for the development of the. Livable Region Strategic Plan, through the Creating Our Future exercise. This provides a critical context from which any transportation goals and objectives must develop. It is also true that the value structures or mission statements of the sponsoring agencies for the TRANSPORT 2021 process should be recognized, and these have been taken into account as well, in the exercise to be described herein. The required relationship between the transportation goals and objectives and the broader regional value structure is presented in Figure 1. The five regional development goals or themes are related directly to the seven goals or themes, as developed for transport system evaluation. From this basis, the specific objectives and criteria for the evaluation of the long-range transport network options are developed. These are presented in the following section. Transport 2021 3 3 . C R I T E R I A F O R D E V E L O P M E N T O F T H E L O N G - R A N G E T R A N S P O R T A T I O N P L A N Within 7 transportation goal or theme areas, 16 objectives have been defined, as provided in Table 1. In order to be as explicit as possible and to ensure measurability, these 16 objectives have been broken down again, into a total of 52 sub-objectives. These are given as well in the Table. The final step is the development of suitable measures of effectiveness or criteria for the derived set of objectives, so that transportation plan evaluation may be conducted. The results of this step are presented in the Table, structured in relation to the 16 major transportation objectives, for ease of presentation. Several points should be made in regard to these criteria. 1. The proposed criteria have been developed for the purpose of assessing long-range plan alternatives. For medium-range plan evaluation, while the same set of objectives are applied, it would be expected that the measures would be different (generally more detailed or disaggregate). 2. Some criteria are repeated, as they are appropriate measures of effectiveness for more than one major objective. 3. A few objectives have no associated measures. This is due to the fact that differences between the transport options at this level of analysis are not likely to be significant. Transport 2021 5 4. Most of the proposed criteria are quantifiable, by applying available models or estimation procedures. A few, however, are more qualitative in nature and would only be applied by using subjective rating scales. It is proposed, then, that the set of criteria provided in Table 1 be applied to the evaluation of long-range transportation supply scenarios. Given their derivation from the broader set of regional goals, this should ensure consistency and comprehensiveness within the overall regional planning process. Transport 2021 6 1 1 9 T A B L E 1: GOALS, O B J E C T I V E S and E V A L U A T I O N C R I T E R I A F O R L O N G - R A N G E T R A N S P O R T A T I O N P L A N D E V E L O P M E N T G O A L S AND OBJECTIVES C R I T E R I A 1 . L IVABILITY: To min imize the impact of transport on the qua l i t y of life 1.1. To preserve the natural environment 1.1.1. Minimize intrusion of transport infrastructure into watersheds 1.1.2. Conserve the land base and green zones 1.1.3. Improve air quality 1.1.4. Reduce the use of fossil fuels 1.1.5. Minimize environmental impacts. 1.2. To improve the physical environment 1.2.1. To minimize the visual impact of transportation facilities 1.2.2. To reduce noise and vibrations. 1.3. To support the urban form of the community 1.3.1. Minimize the displacement of residents, businesses and community facilities by transportation improvements 1.3.2. Enhance local neighbourhoods 1.3.3. Minimize bisection or isolation of communities 1.3.4. Increase the range of modal options 1.3.5. Maximize accessibility to social, cultural and recreational facilities. • lane kms through watershed; lane kms through green zone • air quality (kg of CO, C02, HC, NOX) • energy consumption in litres of fuel • water quality index • vehicle noise level (% of auto VKT on arterials) • % of population within 400 m of transit and 1 km of RT • lane km through built up area • access index by mode to activity centres 2 . ECONOMIC DEVELOPMENT: To ensure that transportation systems support and promote desirable regional social and economic development 2.1. Provide a transportation system that supports social and economic development 2.1.1. Improve the level of accessibility between activity centres 2.1.2. Improve ease of modal interchange (ports/ferries/air/rail). 2.2. Protect the economic vitality of the existing transportation system 2.2.1. Provide appropriate levels of maintenance 2.2.2. Provide appropriate levels of rehabilitation 2.2.3. Preserve the intended function of transport facilities. 2.3. Ensure the safe and efficient movement of goods by all modes 2.3.1. Minimize modal interaction 2.3.2. Improve access to terminal and marketplaces 2.3.3. Reduce travel delay 2.3.4. Reduce operating costs. • accessibility indices by mode for activity centres • access index for trucks to ports etc • averge speed by demnad hierarchy (interregional, intraregional, local) • average truck trip length, speed and cost • total truck kilometres travelled • truck delay (hours) • annual truck operating cost ($millions) 1 2 0 Table 1 (continued) GOALS AND OBJECTIVES C R I T E R I A 3. LAND DEVELOPMENT INTERACTION: To ensure that the t ranspor ta t ion s y s t e m s are c o m p a t i b l e w i th a n d p r o m o t e reg iona l deve lopmen t p l a n s 3 . 1 . Integrate land use and transportation planning at the regional scale 3 .1 .1 . Ensure that transportation investments reinforce land use policies 3.1.2. Provide an efficient and effective institutional structure for the medium and long-range planning for regional development and transportation systems. 3.1.3. Provide for a continuing planning and monitoring function in regard to regional development planning and the impact of transport facilities on land development. • accessibility indices by mode for activity centres • not quantifiable 4. SOCIAL EQUITY: To ensure t he equ i tab le d i s t r i b u t i o n o f t ranspor ta t i on services and c o s t s 4 . 1 . Ensure that the transportation system provides equitable service and accessibility to all citizens in all geographic areas 4 .1 .1 . Provide basic service to the transportation disadvantaged and transit dependent 4.1.2. Provide a range of transportation options that recognizes a changing and diverse population. 4.2. Ensure that the costs and benefits are equally distributed 4 .2 .1 . Reduce subsidies for all modes 4.2.2. Apply user pay and benefitter pay measures 4.2.3. provide a balance between public and private sector provision of transportation services. • % of population within transit service area, within 400m of transit and 1 km of rapid transit • accessibility indices by mode for activity centres • modal split for all modes • vehicle km travelled and passenger kim travelled or PKT per km by mode Transport 2021 1 2 1 i i Table 1 (continued) GOALS AND OBJECTIVES CRITERIA 5. SYSTEM PERFORMANCE: To maintain the effectiveness and quality of the transport service provided to the user 5.1. To provide an effective transportation systems that provides adequate mobility 5.1.1. Develop complete and integrated transportation networks 5.1.2. Increase the people moving capacity of the system to adequately serve demand 5.1.3. Connect jobs and labour force. 5.2. To provide an efficient transportation system 5.2.1. Reduce travel time 5.2.2. Minimize user operating costs 5.2.3. Provide an adequate level-of-service 5.2.4. Increase the mode share of non-automobile travel 5.2.5. Improve comfort and convenience. 5.3. To provide a safe and reliable transportation system 5.3.1. Reduce accident rates 5.3.2. Improve personal security 5.3.3. Maximize seismic security 5.3.4. Provide alternative travel modes. • accessibility indices by mode for activity centres • total and screenline demand by mode • average trip cost by mode and purpose • average trip time by mode and purpose • average speed by mode and link type • % of links by type with high volume to capacity ratio • lane km through seismic/flood areas • lane-km and route4cm of facility provided by type ^0 Transport 2021 1 2 2 Table 1 (continued) GOALS AND OBJECTIVES CRITERIA 6. SYSTEM MANAGEMENT AND IMPLEMENTION: To ensure the effective management of the regional transport system. 6.1. To ensure the effective coordination for the planning and operation of all transportation systems 6.1.1. 6.1.2. 6.1.3. 6.1.4. Provide an effective and efficient institutional structure for the provision of transportation services by all transport agencies at all levels of government Ensure that resource allocation decisions include an assessment of benefits and costs between options, modes and levels of service Ensure the compatibility of all transport modes, with particular attention to the use of shared road right-of-way by transit and private vehicles Ensure the coordination of parking policies with the operation of the road system. not quantifiable lane km of shared facilities 6.2. To ensure that adequate funds are available for transportation 6.2.1 6.2.2. 6.2.3. 6.2.4. Reduce subsidies Introduce user pay Develop alternative financing methods Minimize the capital cost of improving the existing transportation system 6.3. To maximize flexibility for future provision of transport services 6.3.1. Provide for procedures to protect future transportation corridors Maximize flexibility for the introduction of new technological developments within established transportation corridors. 6.3.2. • not quantifiable • gas tax, parking and toll revenue • capital cost by link type and mode • not quantifiable lane-km and route-km of facility by type PUBLIC ACCOUNTABILITY: To ensure the effective role and participation of the public In the transportation planning process. 7.1. To ensure adequate public input and participation in the process of transportation planning 7.1.1. 7.1.2. Develop and maintain effective communication of transportation plans and activities Develop special procedures for the interpretation and incorporation of the needs of the transport disadvantaged and the transit dependent. not quantifiable Transport 2021 10 1 2 3 ON Y—• i2 CD b CD i—( 0 >-J 1 o U 3 c •ja g > o u tf 0> tf o C3 w o at <» 1 2 4 O LU EI cc o 2 S CC CO a . o LU O > o 5 « -J hf z $ * -S CD UJ LU i_ o= S 5 LU > ZC U S r O B-LU a. - 5 = -= 5! a a. 5 < 2 E co as OJ £ -5 m CO • — a tz * t= m c co E JZ e J» Q. co Q, *•* Q. .§ i s -a E s-0 B U © S 3 > ^ -n » .= ® O i_ o CO „ CD a> j?2 >• v a r 55 "o ea » = in O CO oo •t£ CD CO •a x 5 = g = t= QJ •= O CO 3 O 'K 1 g I o ca jzt CO. k-»_ CO o u CO o to" E-.2% i to — o O it ~£ 2 -o •5 3 -a 3 co ca X3 CO E to E I I *cn"> as .3 >- LU 1 • i t cu a. co S ' l co •= 0} on-to CZ J= o <2wO 3 -D E cu CL E 2- o» s P 5 .5 " — _ co C — = .•ti -a ca 03 3 J2 _iz cr 3 > •i= 'co > III co o cu —1 > cu "5 r . i o cu ^ 3 2 ra to y o. fc- -0 cu ^_ 1 = -' CD .O 2 E_? •— k- CD To «>J= =_ 03 _ *3 0> D) O CO 3 CD O CO 3 0 w o *5» o .J2 2J w 3 (0 c «2 is o 1 a c to •a .S 'E g eo OJ m 0 s .2 o-1: cu CO o > = -o CD CO .3 O CO OJ 3 CO .. CO OJ 3 .E O JC 3 § ~ OJ 3 co T3 CO E 3 3 « CO j - is cu u u = 3=: g M = 3 CU 3. is o o ~ E <u I" CO O g 3 •£ -3 » c3 s ° = d 0J CD 3 " i i CO CO CU o eo co s«. cu 3 0 CO — CD CO JZ 3 +-> .2 . E "OJ'CD p t5 *- co ro o •*-» -t-» g E E 2 CD CD E o . c z CO o CL CO O -D CD C > CO -a JQ >• fc ! » . o r o — .•!=: v - g g 03 c -=* i=; o r-I S -CJ) XI J5 .2 = ' E 53.2 — CU 4-» 1— CO 0) 0 g i = r a *= CO o g | • § ! JZl ^ ~ 3 . - O o E CD n S a , 0 > . « cn Ja > > -~ (H •*= co -a JzT = P o cz o CO •e o a. . co cz CO ro c iz: co E 5 co "o :§ i CD CO CO "§> •B if cu x: —^< o •o 3 CO .h: co ro o> cu Z3 - £ >- CU £ 3 CJ ^ «*- o O 3 CU O co -0 c?i2 1 = a s UJ CL ZD CO Sis 3 2 ro o CO cu 5.LLJ i : o co £ r 0 cz Qj cz jo > •— n. cu CO -— .2 i -a .ti £ cu ro I cl .2 g a . cr i co E Jz £ ca ~ cn.E 0 1 3 4—* o co o 0 P5 = o cu UJ £ CO o o •o ZJ o CO cz 'co ar = CO CO 3 3 O "1 Q3 CD -= -z; cp o co ro E 0 5 J ro o CO 00 3 CD "O IE . E "ro "a CL 3 "o ro E g co.E 2 co 3 3-CJ ir 3 3 LU JZl CVJ CO cu CO n, CD .±z > - 0 CO '*-> CD M_ -? JZ O CD cu cu cc TI 0) -cz JZ: *-» CJJ co CO S- 3 CO - i l i Q 2 -03 w CO g 0 .tr cu 03 L, S . •* o 5 zz c o r ~ • i - s i = .2 » -a 0 o> r 3 co j - S co CO O CO 3 ; ? u o 1 = S" • JS « 5 t= B.Q. »- 'o E — -3 3 • — S 3 •— O 03 CO >-~ E 3 -= 3 S,-2 3 J- cn •» co . E co o o = to -a Jo in co co S E 2 o - CJ CO E c ro co co co O 3 1 — 0.73 c >--S Q3 CO 2 c cu cj JO CD 3 CL CU CZ CO 'ro § 3 5 cz co 3 '> o •o 3 OJ CO 3 ro 2? i_ CO cu -a « (Jj CD > = •3 QJ cj iz 03 CO 0J 5 co = CO 5^ OJ aj CL O S 3 a> o 3 v— cz o •3 ro g . 2 CJ cj co 3 cu o S "E > CO O KI 5.J? S o 0 3 -z: CJ co 1 ° O CO D_ TZ CD "O JZ <= CO CU * i OJ cz co cn §!• CJ S5 c ro CD 2 E cn cn 3- > 3 *3 CD •JZ! O *-> tz 03 CO o t 5 CJ QJ S UJ O CC ZD (D a LU O ° £ LU CO o o •— 0- CO cc — D_ O cc S cc . S LLJ cc Uj i» ~» re co & H O; LU EL = 0 0 C_3 -  3 CO O JZ CO ~ o S-S . QJ 3 JZ o co •C co — 0 > 5 ° - Pi-co CD cr 3 JZ w CC CD is = to -ti co § IE -3 0 - o o a, "~ •- o CD OJ Qj _J .CO a 03 - i 3 2 3 « E o .§ g 1 i cc > S- S 2 n-OJ E > — CD >_ CD o. CO § 03 "E CD "2 ro « • •g 3 3 co" CJ c: QJ 4— a. •.3 CD . "ro ~ <zr CL >- CD •E 2 E — co 1c: 3 03 tzo cr < _ i a. > o £ c§ •I CD c c 3 £ E 5 E M i CD £ JZ CD ro •£ CO CJ CO ' o o CJ r-^  O "O O QJ 5 S f- 2 CO LU S i tu o < > z z Cj LU CO k_ S3 QJ 1— CL CO o — CO _^ 3 ^ 2 = CD JZ -j CO — z z O UJ CJ H LU d cc zn LU H-I— LU _I z . D_ < >" " s •= iz < <r cr ~ =3 a. i : a 3 CO 03 3 E 2 . _ CO 03 03 « K 03 > .2 3 > CJ •^ CO — 03 3 O > -3 CL O .> — — 0J ro s" «• E. E 03 .0 03 JZ CO 3 CD Q. E 3 03 O 03 £ 2 cu ^ E CJ 03 CD OJ V 3 to CD -a \ CO i_ 1 CO = » £ JZ CL. t o O "3 a. o § - < CO c "3 CU = E 03 _ > — o CO ro f= • 2 £ tt 5 •a cu CL co p TO CD 2 £ •o -o CD • = g- . i -o o QJ > 3 3 3 Q3 0 ~E CO — 03 co CO ^ E CL CO °- fc» 3 CO CO O -O Q_ 3 CD 3 •— | 2 03 co co CL CO !2 5 ro ^ a- ^ O CD CO CO CO CO 3 cr 03 E 3 o i > aj .QJ _rT co QJ O o — CO 3 QJ 1 5 •t; 3 CO CO QJ .bi- Co -— >- 2 2 3 9J o •— CO cc CO "2 t : = CD ° 1 CL O CL-T c/5 ~o 1 2 5 cz CD cu o. JZ nal -a • I nal CD to o CD 3 ' , • 3 03 3 T3 E ' ~ 03 ns o o AO E u 03 "= 3 CD o CO CD S 5 co Q. >- cs. U =3 ~ 00 o d, CO tn ~* .= i2 •a — J i 03 » 3 3 3 3 <£ CD 5 t: a. cd •a c c: o co ~ •z CO E a. a. to o = — CO 03 h— > •*-03 CO •o = 3 D> •— CQ 03 -C -a o 03 .3. C CO [Z CD C3 33 • — 03 CU O E C n. cu _o n. 03 CO > CD -£ "> -a ^ 03 CD <=> io" >- C3 O Q . Jt= 03 a) cz " c = ' § ° — 0 3 -Q J 3 CD W 0 3 =3 co o = JE Q3 4-. " ~ 0 3 C J co 3 o E CO 0 3 •4—' o E o Q _ CO CN) CN! -o cr CD CO JZ> "co {= co CJ co B £ E CD o o > E a. -a cu <C ^ E . "5 Q> E x 03 03 a. o> s o CO 03 CO 03 a , " cz 0 0 x i 3 CO CO CO Q cc > CD >• C J "o Q. 03 3 3 CD CO 3 m 0 0 ° 3 E C J CZ CD C O — 0 3 C J CO .3 3 o .£= *^  o 03 CO .iZ. — = § CO T Z -5 •— 03 03 ^ s | i £ o a. CD • — CO 03 •B °" CD C L x cz 2 5 g 03 CD CD •a CZ Z3 CO CO «rt £ Q3 CD O -*—' CD CO 9 - t f E, Q3 Q3 CZ CD E .E ET CO jE C Q cz s 1 5 ° -> o "53 o > :=» 03 CD CO * - 03 ro to CO i— Q3 Q3 ~ 3 i_ Q3 CO C33 ° — c CO « to1 E o C L C J J 2 CD "3 > 3 03 CO TZ> iz CO >_ o § s = .E o c CD CO J Z > 03 CD »— O CD » 2 CD 03 CD O .T3 CD CD 3= o CO CZ > <=. E 03 0— S 03 S° CD — > CD CO H 03 CO " 5- P CO -Q co E | CD £ Z CD CO > -oo t= "co Q3 03 to 15 G3 CD > O am O cc omi mpr 03 GV —• mpr E Z ) — o CD CD CD 0 3 03 'o '5- CD CZ CZ o IStS , ca C33 CZ CD E as C J C J CD ives "o o CZL i o tem - C -t—» CD car ys cz "EE car CO 03 T Z ! LO CO 00 •*-* CZ cz cz ca CO CD o CD ere ran CO .—• CO CL CO CD i CD cz .—1 i— " C J 1 ~I '• an o cy ole cz 03 —' • g - E C J . 03 o 03 03 -a > 03 03 -S i 00 E S ° >: •n cn, c — ' CD -5; • — —f CD •B 2 S 5 CD E 3 ° CO 03 Q. = *E « E o *- 0 1 cz ~ CD E cc tr t— co < 3 CD CD CO 03 E CO C3. *c: < o a 5 V 00 w 3 03 o 03 2 CO 0 > 03 CD 1 = • - & E S CO o. 03 00 £= 03 1 1 O CD o o ca cu £ E cniz: 03 - o £ i-CD g> co iz: O CJ • O J33 CO == 03 CD 3 g ro CM 00 LU " I - i oo < i s ? 2 cc E O CD Q_ LU 00 CC Z LU < o cc cc I- o a i 0- U J cc 0. o u cn 3 "C3 CD 03 03 • a * I _ s - | i 03 -B fo O = i5 03 " 1 - => CD O CO, C 03 CD •= ™ co £: o o = 2 ca u u o 3 _j_ J» . CJ3 CO O *= 3 - t . . 2 CD 33 3 _ 03 O ^ »-03 Q3 33JC ™ O c S O CJ JZ. 5 - g . 1 1 03 35 rag o •= n . 5 O 03 o '> 3 03 O " io 03 09 •5 o 03 O = « CO = > 03 03 TB 3 03 CJ 75 3 O cn 03 00 03 'cn 03 «^ CD w CO 3 03 E n_ o 3 O o 03 3 O 03 0 -3 35 1 § O b. JZl ^z CO ~ — 3 -3 O 3 CO 3 3 - 3 O •— 3 c= »- t: 03 > 00 03 03 3) 03 oo >-%i -a — 03 3 > O o s: E o — 3 O 3 rz co o, 03 U 3 3 CO _ro 3 CO Q- JS 0> m ^ j - CO •fe w 3 "«3 2 S = C E « *• o a £ "cu E 5 » o i - ® co 03 00 « 03 03 ^ = I o E £ -2 S CD 03 cn jz o . E 5i E o CD 03 w E g 03 u 03 3 w O n. s « n. o ro " B J cn cu 3 >-•— 00 03 CO 1 1 3 CO CO o C L CO CO 03 > "co h_ 03 C L o o 03 CD S. o 03 > 03 -a .5 § U CL 3 3 i"Ss 03 JZ .Z. ^ o 1 ~ 00 3} ?U » CO g 03 o, 2 o .tr B . 5,1 « cn ro.«2 oo 3 CD 3 O TS •c O 3. CO ZZ 33 CJ 3 >• E c 03 g 1 5 c c o 3 JD ~ CO CO 3 > C£ 3 O O > 00 cnU 3= S CO 3 g ro O £ CO 5 ° -._ „ 03 3! > 031 C3 S oo ro CL 03 CJ CO E "" cu cn X 03 o 03 00 03 J2 ~ 03 o = E 03 ro > CO ~ co S E E 2 03 8) > O o >->- n. S-o 3 3 r * O 3 CL 03 E- 00 oo >• ro -3 g«s -S 03 CO J= O *' w o 03 — •5 £ 3 3 ° "5 5 U 03 o ro " 3 *J- 03 I 1 3 O 03 . — S g E co CO "3 •3 3 O CO 00 03 CO 03 03 cz E 2 E '«- o >- CJ •zr oo oo •* 3.' 03 CD I I 03 -o ro = •S E CL 03 >- -=> CJ 3 3 o 5 ~ E- co =. t: o CL 00 3 1.51 2 03 CJ o 03 *• (O 03 = ra jo -o J2" 03 33 CO DE z > JS oo o z t— QC ~ CU C3 oo" < MANA AREA PACE, NSPR MANA z oo < LU z CQ LU 1 y f cc 5 VEGR Q_ •=> z o VEGR O Q z < 0 1 1 1 cc Z LU z cc RES 00 o L U RES o X Cu cc a E:T TO <I cu 1 z < 1 zc Q_ z z INC o GIO RMI PR GR U J cc < L L . CC 3 CO JZ 1_ L . 3 03 ^ 3 ro o S =3 co J - ro : • g O CO T 3 3 .03 c ro z^ o co ~ J*" 03 E 0 5 . _ 03 JZ .•3 "3 . CO — "03 a i t c '5 •= "E >• c 3 O 03 3 — 0 ) | - P > 3 .CO •> CD — — 03 « E > CC CC 03 03 03 "2 P O : 3 . 0 0 — 3 Z Z O g w ^ 03 3 03 03 l l CD — CO O L U C J CC > cn I ? CC Z3 "*~ o o ^ — cc . o > CD m CD CD Q3 v 03 CJ3 C L CZ 3 C O 03 CO E c 03 C L > • 1 5 S CD 03 u-° - > CT • — CD ~>-= CZ 03 CC z: j z CJ cz *- o ~ C 3 ~ c n c cz 03 CD •B o " ° CO " CD f£_ ~ 03 03 03 CO 3. ' CO J= _ cn H I O JZ .•= o c n — cc to z; CD as J Z -i > . o c O CD to ro 03 CZ to- tr 03 O z; O-to x CO CD "O cc r: CO c n >-o cz 03 cn to c n o w o 03 ' ™ — c n r- CD : cc . O CO C J ro . ~ cc CO ° CJ CD o — E. iz cn cn 03 o •— CC = -— >. 3 CJ 03 3 _Q3 £ C CO j) zr to — 3 3 CJ — O — — o o ~ 03 «=• ra tz 3) CZL =2 3 - 2 CD ' > S-S i 03 c Q ro ° tC) 3 ' . cc co a C L c n 03 E C J >_ CC Q3 C L JZ CO — CD to C L CD O -JZ, "° ro • E a. 03 CD . — 0 to .z? c ro -2 P CL t — CJ CO CJ " C 03 CJ — ^ to CJ CD ro \z; C L . — 5 a) S 'g - § 3 "H _ —• C J r - to .— = CD to ^ ; — CO 3 i_ CD .— O i- zz J Z r " | z .=? & ro ^ 5 — cy o cc "cc CJ JD to 3 ,03 3 c IE CC O — — to . — CO — CC . -Z ^ 03 S 1_ 03 CZ S* C J t O CD ro o a CD > 00 v- 03 r-t_ 55 « § > § • £ £ ro 3 ° CZ CO j - . 2 — n c n C3 . 2 CD — g 03 « CO ^ ^ Q3 > CD CC 03 O CD X3 CC | CDLL 03 cz o ° = 42 g to S 5 to ~ 03 CD o £ ^ CD J Z ° • — C J to — -3 -£= 3 ro . 2 g CD —1 -3 P cc <? -^ r i_ _rr >— c —r CD S °" != a. o x ^ i r L U 3 O ro ^  CD 3 — ro CD ~ O ! CD > •= c «= -I -o co 5 S cc ro_ i — c n to J i : S f "T" 0 3 £ 2 1 O 5 ,t! O CD 5 ™ CO -zr "cz . 2 to to 3 C L . i J -E "E I I C J cz CO C3 CZ CD 3 i t CD C L •— CO o z cj • -cc c r c cj CD ro CZ CD CD 15 • 3 V > ro S 3 CD" o C J -o ro 3 CL LU to . oo C J . _ r: 3r c n O ro ^_ CD © n to " t r 126 co Z LU 2 CO o L U _ u. X L U H— cc •* 111 Cu —I = lo E CC H- r— =5 «5 Z Cu rr! L U > uj Q O I— « iii 2 — h-CC 9 Cu tS CC L U > 2 2 < E S CD — — UJ z=zz, CO •—< cc U J 5. *£ CQ < C J o st: ^ oc cc o > O CD o 5 _, L U •=* x EE »- < L U = -o < ° = == ty L U CC > X Cu t-CO CD CO CD •*—' cz .—1 15 03 CO " r o US E Q . CD o x: cz " > 3 CD mun 03 i — ca alia opm i with iment, •egion devel CD t— © ZZ < < CC X o CL. U CD . 9 55 CO > ' o CO — tz .2 o o CJ cz cz "> •— o 03* Q. 03 QJ "5.-CZ I -o CZ O co CO CO CO tz o CO 03 r~i 2 cr "o > o iCD °° CD i= g cz CZ.2 2 "S L. tz o o o « S CD fO CO tZ 33 >l ro E s «= s- ° CJ o rz CD JZ XJ 1=i tz {? CO CD CZ0 ~ . E co rz xs tz cz iS ro CL •— .2 eg "cn m O W CO fO 0 i _ CZ 03 E cz 55 52 > S o •— o.£.| § £ CD = "SI CO I" 1 fa oo £ 03 rj co w co co 03 Z3 o co 2 = T3-CO I S = 5 2 Z zz) EE £ ° -S i LU ZC >-fc: < =" CJ C3 ^ * I 3 L±j t"Z 00 B s < z U(CO 5 LU cs CC X UJ cu i— cc o | T3 cn c t= co •=: CO 03 ID * = r- o O CO 03 fe S..S oi . E .2 03 •" XJ CO 5- ° o E « CO to fe . H 2 TZi 03 • I cn 2? «= c 2 co o cL-2. CJ § a S E 75 — CJ 1 1 H I S CD CJ o zz in ro 5 JZ '=> = -C3 JC •— .2 -2 CJ CO ro I s JZ CO . z. o CO o 3 2 •= CO O CJ JS " c " co .2 5 *^  ^  O CO CO € S . S CO 03 CZL 0J t£ "> > IS «= •zz 5 co o 5 is .2.-a -a • ° — E O O CO _ CO c > ° "5 B C o — — g 2 ? o c?r= o oc o S » c 2 J3 O 3. £0 CJ -3 • C M C —I CO CO CO cc cc U J U J x 55 cc -« ^ . H - uj ir W L U P L U u b oc x = = U J u. !=•-£= 5 o = 2 2 ° " J D ° Cu _i 2 •x r> =3 L U 0 U . L U X O —I CO I— £ = U J s -U B h cc 5 ca Cu Lu < CD l - 5 •a co 3 03 o CD -Q ZZ CL-— T3 S O = 03 tZ O — 3. 3. S *- "> .ZZ CO 03 n-x: co «•* o u c -CO J3 c M ao> CD Tz 3 E 03 CO " E g 33 . 1 E co 'Z IZZ 3 o •a co "~ n-S | JS o) > o f » > -a 3 cu e co Q to ro CD o — 2 w 'op ^ 3 CD » 5 £•£ 3 O CJ _ a> 3: oo z; o > 03 E CD •a •= 3 cn CO 03 — 3 .2 """ "o CO O CJ co co _- 13 CO 03 3ZZ 2 .E I I 0 3 o .00 *- , = 00 .2 ~° CO 03 as 03 tz 3 z: CJ 3 « 3 != E co E » o , E CJ »-3 CO cn J2 cu ° o CO E < 1 1 O L U CJ > L U — £2 o L U I 3 I s 1 * Q o cc — > cs cs L U cc ^ - < = = ^ CO o L U <• LZ " 5 = ZZ O 00 CC CC CO Cu CS < L U E o CJ CO c o o. 00 3 CO s _ CO CO © » a co 2 E ° «J= = a u o 1 cs "= CO ._ E CO E 03 > o e S 3 CO 2 Ci CO £-»•=, = a .2 " S B 2 O B. S o g c i i i " S S a. o> CJ c CS CO en _ oo co cn s *— o e 03 oo 5 00 S 03 o> 3 CO •° n . „ ._ "2 E « S w » ^, 2 c0 •— 3 03 E _03 3. E 09 03 — • 00 o. > 03 CO >_ CO CO 3 i Z ^ * 3 CO co 03 5 cc .2 E CO CO o 03 03 00 ' = 2 I f X o 03 C "3 e ° ro oj E ' l l .2 co S S 2 o o 03 X g 2 O Q) CO o » al .2 > | S o J= CJ o CD c 5 CD .2 cc = 03 z2 2 CO « .> "5 -J CO £ CD o CO £3-CO EL 3 = 5 ^ S I L U T3 ro CO tz 3 CJ. L . i CD O ° f 03 3 •z: JC CO ^ i i 1 5 o_ o 03 O) J= 33 JZ . E •4—> I— 5 J= CC 05 CO _ So o E JO JO _Z3 .2 ro o "° S " 1 0 3 0) O S JE E o ro • 7 C3 oo ^  rvi 5 => 2 a ZC CD z ^1= <u?2 3 CL < Cu CJ < zz Z Z Z =3 S 2 cs <»; U J « oc o-oo < =9 a a 2 2 < , cs itio CJ «* cc o CU Q_ a. cc 3 o 8 5 ZZ < Ez cs L U O > 2 P= < CJ _i L U cr LL. _ U - O o ° §.£ == o S s 9= co - > L U ° > c o U J fZ ZJ ZC < L U I- rr > .2 .2 o CO CD ' CD s :§ -I i CD 3 00 O — — co 8= CO CO ca r- ' CO ..3 >- CO - CJ" n% E S 3 • 3 w CO CD 00 Jg J Z - I E J •2£T S CO S CS t n CJ" QJ JZ CO »-O CJ CO fe ° J= 2 > o OJ -a fe w CO g 3 «-8 > oo 03 _ tZ 3.TO 00 00 3 0) o = _ •— 00 .3 = » co ro JZ i l l CO g 3 C 03 O 3 OJ • - o n 33 •— 03 cj S 2 = S » cn 03 S co .2 >• = ~ 2 ZJ Z — _J < O <£ CC cc o H CO LU " LJW W _ »** == _j 00 ,2 3 CC 3 „ CD •2 E ro | - .E 5 O tZ O oo o cn — o ^ _ CO CJ o . CJ >-co "5 30 cj I £ o .2 3 > _r co •a »-3 OJ ca o 3 Q_ OJ _ ~ i 5 CO-CO a. o 3 i-O) 30 E to 3 03 »-03 CD ° E •SI . E E o I k_ o 3v U > O M -cn o « § o -zz = = fe > s > 0 i - 3 CD T = 3 J Z 03 CO i l l co =; *-.» - 1 ro co cr OJ 03 03 i _ S CCD ro cc * i .2- 03 g ~ = •= 3 CO 5? 1 £ 1 j _ O J Z 03 CJ 'e CD CO 03 O "o CL 03 > o CL •E CD _ 3 Z J C J 03 CO 15 CO 3 ro 03 E ro co CO 3 o cn co 03 -ti 5 = S -S E ° O 3 — > i - Q3 > O) W tZ o Sz CD CD 3 3 CD •o |5 "55 I . I 2>XJ QJ CO CO CD — — -a co oo XJ Cud) •> "a _a •2 c 5° S 03 . E 03 CO CO cz •SJ= "5 JZ 03 > 1 1 1 ro H- > o co ~ 03 2 3 3 oo o >-M — i— 3 CO — XJ OJ 3 o o i t 3 — 0J CD o 3 I •3 iS <« CO •> XJ J Z l tz ca co °- o « CD »-» J-> — = ro i -§ X J 2 XJ QJ CO CO > 03 3 2 t 5 ® q-z. -=. " : cu - co-— 03 S g o E = CD '.S2 iz -co" o 03 CD — •ZZ CO CO 'cn 0 5 "S CL ro co 3 CO S 3 oo o 03 CO 'ZZ 03 2 - - | " g l > 3 O C J 3 • & O CD j= ^  ^ Z CO — ro <2 co ro t=" ro E E 3 cz 2 ro .2 .2-o CJ CL 3 03 3 « E _c J Z X J O C D - j g 5 c co — O CD O ~ .— DJ' CZO > o 03 6 b" CO CL 5 ° < O ^: cj > ro c5 CO 3. .2 3 00 •a oo 3 03 CO OS 03 CO 5 03 - i CJ CD XJ CO RO CJ 3 O CO IT) CNI XI CO CO CD .2-15 CJ CO 3 "2 3 O E 03 ™ CO ^ M CJ E oo CNI o x •g E 1 -03 >fe j= cn CO 3 3 -3 2 i JC CO 3 CO E 2" S i 3) 2 3 CO CO I " i g M CD O Er X3 CD 3 > CO ~ co o> 03 X3 — 03 3 JZ CO *— n* = CO •— XJ tZ 3 O O CL CJ CL 03 3 00 CO 1 2 7 1 M ® MINISTRY OF TRANSPORTATION AND HIGHWAYS MISSION AND STRATEGY STATEMENTS "To provide and maintain safe and efficient movement of people and resources on an integrated, multi-modal provincial transportation system, in conjunction with the economic development of the province, and in a socially and environmentally acceptable manner." Plan, construct, operate, maintain and regulate a quality transportation system in a safe, efficient and effective manner. Coordinate provincial transportation policy and legislation with other agencies and levels of government Facilitate economic and social development of the province with environmental sensitivity. Respond promptly to major emergencies and disasters. Communicate with the public with regard to Ministry activities. 1 2 8 B.C. TRANSIT - MISSION STATEMENT OF GOALS The Mission ofBC Transit is "to enhance the social and economic life of the communities it serves by providing safe, reliable, environmentally sound, effective public transit". NOTE: A more detailed mission statement for B.C. Transit was being prepared ly the Corporation at the time of publication. 129 TRANSPORT 2 0 2 2 Medium-range Plan for Greater Vancouver Page 35 3. Supply of Facilities: Improving the Transportation Network 3.1 H o w the System Concept for 2006 was Derived Four Steps Option approach This chapter explains how the project approached the questions of how much, where and what type of new transport capacity should be supplied by the year 2006. It provides a description of the proposed 2006 system concept. The chapter also offers guidance on the staging of the improvements between now and 2006; it addresses how the sequencing of the improvements can best be arranged to meet the overal l goals and objectives of the transport system. The project team approached the problem of deducing the preferred system concept for 2006 i n four steps. Step 1. The team took the transportation infrastructure network (or "recommended system concept") f rom the long-range plan (year 2021) and applied to it the land use pattern for the year 2006. This land use pattern was supplied as the 2006 target by the GVRD Strategic Planning Department. Some elements of demand management were assumed to be in place by 2006, boosting transit ridership and reducing in the number of vehicles to be accommodated on the roads. Because there w i l l be fewer people and jobs in the region in 2006 compared w i t h 2021, the team expected to f ind that some of the transport l inks f rom the 2021 network would be underuti l ized. This was indeed the case. By inspecting uti l izationl ink-by-l ink, the team was able to remove the least uti l ized l inks f rom the 2021 network in order to create the 2006 network. For example, the long-range plan's extensions of intermediate capacity transit east and south f rom the end of the current SkyTrain line (within Surrey) were removed as candidate for the 2006 network as a result. Step 2. Using the remaining elements of the 2021 network, the team constructed two init ial options for the year 2006 network. The number of options was l imited to two— judged to represent Transport 2021 1 3 0 Page 36 TRANSPORT 2021 Medium-range Plan for Greater Vancouver reasonable extremes. They amounted to different mixtures of the modes of transport: • Option 1 relied on road investment, using carpool lanes and mixed-traffic road capacity, plus enough transit to handle projected'demand. The major transit (ICTS) corridors were given service equivalent to fast bus, but not a level as high as could be provided by a rail-based system. It was thought that this would minimize capital costs and capital investment, and serve demand adequately. • Option 2 saw less road development, but put transit service on major corridors up to rail-based equivalents—a relatively aggressive transit investment strategy. The idea was to provide a higher level of transit service along major corridors. I t was thought that this wou ld be more expensive, but w i th the benefit of a greater shaping effect to assist in reaching the land use targets—since higher-service transit w o u l d be i n place earlier than in Opt ion 1. The options had common elements. They both assumed that: • the existing transportation infrastructure w o u l d be maintained and rehabilitated as necessary: • the land-use plan wou ld be as given in the 2006 target in the "Livable Region Strategy: Proposals"; • transportation demand management w o u l d be applied as described in Chapter 2; and • the system is to continue to accommodate international, interprovincial, and inter-regional passenger traffic and goods movement (not changed by demand management). Step 3. The team measured the performance of the options along a list of "measures of effectiveness" or criteria. These criteria came f rom the project's formal goals and objectives for the transport system. 1 4 For instance, travel speeds and congestion, time lost for goods movement, the degree of automobile dependence, atmospheric emissions, amount of farmland consumed, and other factors were estimated for each option. The team gained insights into the trade-offs involved by com-paring one option to another. Step 4. Based on these insights and on guidance of the Steering Committee, the team constructed a preferred opt ion 3, which aimed to combine the best features of the init ial options. This was refined wi th further work and became the proposed 2006 system concept illustrated in the accompanying maps. 1 4 Summarized in the Introduction to this report. 1 3 1 TRANSPORT 2021 Medium-range Plan for Greater Vancouver Page 37 3.2 Transit System in 2006 This list shows transit system improvements recommended to be in place by 2006, not in order of priority. Item numbers are keyed to the map. A. Provide basic transit improvements 1. A d d SeaBus capacity across Burrard Inlet 2. Increase the existing SkyTrain capacity 3. Increase main-line and feeder bus coverage and service hours B. A p p l y bus pr ior i ty treatment across: 4. Burrard Inlet at or in the vicini ty of First Narrows Bridge •5. Burrard Inlet at or in the vicini ty of Second Narrows Bridge 6. the Pitt River i n the vicinity of the Pitt River Bridge -Fraser Nor th to Coquit lam Town Centre 7. the South A r m of the Fraser in the v ic in i ty of the Highway 99 corridor- Ladner to Richmond C. Provide bus lanes across the: 8. Midd le A r m of the Fraser River - Vancouver International A i rpor t to Richmond ICTS (see 14 below) D. Provide transit p r io r i ty measures: 9. f rom Surrey Ci ty Centre south to Newton and Surrey Munic ipal Hal l 10. f rom Surrey Ci ty Centre east to Gui ld ford Municipal Centre 11. along Broadway Corridor - UBC to the Vancouver-Richmond Rapid Transit 12. along Hastings corridor - Nor th Burnaby to Vancouver Central Business District 13. along Trans Canada Highway corridor - 200 Street to Gui ld ford Munic ipal Centre E. Provide Intermediate Capacity Transit System (ICTS) f rom: 14. Richmond to Vancouver Central Business District 15. Lougheed Munic ipal Centre to the Vancouver Broadway business district 16. Existing SkyTrain Line to Coquit lam Town Centre Transport 2021 1 3 2 |l Page 38 TRANSPORT 2021 Medium-range Plan for Greater Vancouver TRANSPORT 2021 Medium-range Plan for Greater Vancouver Page 39 Transit System Concept and Staging Strategy Area-wide transit expansion Three trunk transit corridors Five steps Population and employment growth, as targetted in the GVRD "Livable Region Strategy: Proposals", together wi th transportation demand management, w i l l result in a dramatic increase in travel by public transit in metropolitan Vancouver. To handle this demand w i l l require an expansion in transit service, in terms of hours of operation, frequency and geographical coverage. The transit system wi l l have to be re-oriented to provide appropriate services to regional town centres and medium-density communities as we l l as downtown Vancouver. Under the proposed growth management strategy, three "trunk" transit corridors wou ld each attract a level of ridership high enough to warrant some form of Intermediate Capacity Transit System ( ICTS)—having peak hour capacity of about 10,000 passengers—by the year 2006. These corridors, from west to east, are: • Vancouver -R ichmond; • Lougheed-Broadway; and • Coqui t lam-New Westminster. ICTS systems are needed in these corridors because of: • operat ional imperatives: wi thout ICTS, additional new road or bridge capacities wou ld be required to accommodate the 2006 transit demand; • cost effectiveness: at these levels of transit demand, it wou ld be more cost-effective to handle the passenger traffic by ICTS than by conventional bus; and • level o f service: ICTS wou ld provide a better level of service, in support of T D M measures. I t wou ld attract midday and off-peak ridership especially in urban areas w i th activity centres. The medium-range transit plan could be staged as follows: Step A: provide cost-effective bus pr ior i ty measures, extend and enhance service coverage, and provide a fami ly of cost-effective transit services for various time periods of the day. The challenge in Step A w i l l be to design and fund a service plan which w i l l match the increasing demands expected on the system. This w i l l require addit ional financial resources (since the fare box w i l l not cover all operating expenses). Step B: in corridors w i t h high transit demand, provide fast, frequent, comfortable, l imited stop bus service (up to 5,000 passengers per hour), using bus priori ty measures and HOV lanes as available. This service is termed "SuperBus" for the purpose of this report, is recommended in four corridors (see below). Transport 2021 1 3 4 Page 40 TRANSPORT 2 0 2 3 Medium-range Plan for Greater Vancouver Step C: upgrade to ICTS in first trunk corridor (see below) Step D: upgrade to ICTS in second trunk corridor Step E: upgrade to ICTS in third trunk corridor. Comparative performance of trunk transit corridors The fol lowing table summarizes the characteristics of the three t runk transit corridors. A fourth corridor, Coquitlam-Vancduver, is also shown. Though not an ICTS corridor, it warrants SuperBus service a n d / o r a commuter rail service presently under consideration by the Prov ince . 1 5 Transit Corridor Performance (during morning peak hour) Projections for year 2006 are based on: • Livable region strategy land use targets for population and employment being achieved by 2006 along each corridor • The recommended medium-range transportation demand management package in place as described in Chapter 2 of this report • The given corridor being the only one in place at the time (other combinations, not shown, were also tested) Three Trunk Corridors Richmond Vancouver ICTS Broadway Lougheed ICTS Coquitlam New West ICTS Coquitlam Vancouver SuperBus Ridership at maximum load point/hr fOOO) 7.2 5.3 4.1 4.3 Total boardings/hr f000 psgrs) 15 11 9.2 6.1 Total Psgr-km/hr fOOO) 119 85 71 103 Route length 15 km 18.5 km 13.5 km 34 km Intensity of use ("000 psgr km/route km, per peak hour) 8.0 4.6 5.2 3.0 Land use: difference between trend and target corridor populations Low High High High The schematic map on the next page illustrates the relative intensity of usage and also shows the population growth, both trend and target under the "Livable Region Strategy: Proposals", for the areas served by the corridors . 1 5 See Appendix 2 for a further discussion of commuter rail. Transport 2021 1 3 5 TRANSPORT 2021 Medium-range Plan for Greater Vancouver Page 41 Transport 2021 1 3 6 Page 42 TRANSPORT 2021 Medium-range Plan for Greater Vancouver Measured according to the goal of transport efficiency alone, the Richmond-Vancouver trunk transit corridor, w i th the greatest projected intensity of usage, ranks highest as a candidate for a capital upgrade to ICTS. The Coquit lam-New Westminster corridor comes second on this criterion, w i th Broadway-Lougheed th i rd. However, measured by the goal of shaping the land use pattern in the corridor, the ranking changes. In the project's judgment, investment in the Coquit lam-New Westminster corridor, serving an area which is not only targetted for significantly above-trend growth, but also an area which has low relative accessibility today, wou ld have the greatest land-use shaping impact. The Broadway-Lougheed corridor ranks second in meeting the land-use shaping objective. The recommended priorities for staging transit improvements are presented on the fol lowing page. Two possible paths p o r steps C through E, two options (called Path 1 and Path 2) are given. Steps A and B are common to both paths. Path 1: This path has the objective of mak ing investments for transport efficiency. It places the Richmond-Vancouver corridor, which is projected to be the most intensively used, at Step C. The other two t runk corridors wou ld also be developed in Steps D (Coquit lam-New Westminster which is the next most intensively used), and E (Broadway-Lougheed). Path 2: This path aims to combine the goal of using transit to shape land use patterns w i th the goal of transport efficiency. It places the Coquit lam-New Westminster corridor, judged to have the strongest land-use shaping power and second highest intensity of usage, at Step C. Richmond-Vancouver, having the highest intensity of use, should be placed at step D, rather than Broiadway-Lougheed. The latter would not be placed at step D because, the Committee considers that it is not reasonable to advance two land-use shaping l inks before advancing the' l ink w i t h highest intensity of use. The choice between Paths 1 and 2 depends on several things occurring in the years immediately ahead: (a) additional engineering information being acquired (cost, technology, physical feasibility, operating constraints, etc.; whi le considerable work has been done on the Richmond-Vancouver corridor this w i l l need revisit ing, whi le the others have not received as much study); Transport 2021 Richmond-Vancouver ranks highest on transport efficiency grounds Coquitlam-New Westminster ranks highest on land-use shaping potential 1 3 7 TRANSPORT 2021 Medium-range Plan for Greater Vancouver Page 43 (b) success in introducing demand management including disincentives; (c) success in dovetai l ing transportation investment decisions .*. w i th land use plans, for instance through partnership agreements as elaborated in Chapter 1 of this report. Assuming that there are no major surprises or difficulties arising f rom items (a) through (c) above, for reasons explained below the Committee believes that the choice between Paths 1 and 2 w i l l depend on: (d) the level of f inancial commitment, which w i l l determine whether one or more of these trunk ICTS links can actually be started and completed in the period between now and 2006. Whi le the Committee recognizes the importance of transportation efficiency, it recommends that transit investment should be used to shape land use w i th in the medium-range hor izon. I f there is financial commitment to bui ld two or more links w i th in the 2006 horizon, then Path 1 w i l l achieve the goal of using transit to shape land use—since at least one of the l inks intended to shape land use w i l l be committed, i n addit ion to the Richmond-Vancouver l ink. However, i f there is financial commitment sufficient to bu i ld only one l ink, then only Path 2 w i l l assure that the goal of having a land-use shaping investment—Coqui t lam-New Westminster—is achieved. Accordingly, based on al l information available to date, the Project recommends as fol lows: Recommendation on choice j f a financial commitment to bu i ld two or more ICTS corridors is made, then Path 1 should be chosen. I f the financial commitment is adequate for only one corridor, then Path 2 should be chosen. ^0 Transport 2021 1 3 8 Page 44 TRANSPORT 2022 Medium-range Plan for Greater Vancouver 33 Staging of Transit Improvement First Pr ior i ty Enhance existing transit system. Time frame: 1993 through 2006 Second Pr ior i ty Make major capital investment to develop t runk transit corridors. Time frame: 1995 onwards. Choose either Path 1 (if there is a commitment to finance two or more ICTS links) or Path 2 (if there is a commitment to finance only one ICTS link) Definitions Step A Provide improved bus services throughout service area by developing wider range of services which reach further into existing residential and commercial areas, and which support » medium-density low-rise residential development. Step B Trunk Corridor Improvements. Timing: 1993-mid 1990s Requires limited capital expenditures in fixed plant Richmond Vancouver Broadway Lougheed Coquitlam New West Coquitlam Vancouver SuperBus SuperBus SuperBus SuperBus and/or commuter rail Make a choice dependent on (a) additional knowledge through research on capital costs, transit technology, alignment options, physical constraints in each corridor, (b) achievements in introducing transport demand management package, (c) success in growth management, i.e. advances by municipalities in coordinating their official community plans and zoning with transportation, e.g. through partnership agreements) and (d) the level of financial commitment. PATH 1 Step C Step D StepE PATH 1 weighs the efficiency of transport investments MORE heavily than their effect on pattern of growth Trunk corridors all developed by 2006 Coquitlam Vancouver Richmond Vancouver Broadway Lougheed Coquitlam New West Install ICTS Maintain SuperBus Maintain SuperBus Install ICTS Maintain SuperBus or commuter rail ' Install ICTS Phase out the above PATH 2 Step C Step D StepE PATH 2 weighs the efficiency of transport investments LESS heavily than their effect on pattern of growth Trunk corridors all developed by 2006 Coquitlam Vancouver Richmond Vancouver Broadway Lougheed Coquitlam New West Maintain SuperBus Install ICTS Maintain SuperBus Install ICTS Maintain SuperBus or commuter rail Install ICTS Phase out the above SuperBus means fast, frequent, comfortable, limited stop bus service (up to 5,000 passengers per hour), using bus priority measures and HOV lanes as available. Commuter rail means limited-stop rush-hour trains on heavy rail tracks hauling conventional passenger coaches, trackage shared with heavy rail freight. ICTS means intermediate capacity transit system (about 10,000 passengers per hour) with regular peak /off-peak service, including segregated (separate right-of-way) busway, conventional (at grade) light rail transit and advanced (grade-separated) light rail transit. Transport 2021 1 3 9 TRANSPORT 2021 Medium-range Plan for Greater Vancouver 3.4 H O V and M ixed Traff ic System in 2006 This l ist ing shows system im-provements recommended to be in place by 2006, not in order of pr ior i ty. Item numbers are keyed to the map. •—— . . raye t F. Provide H O V lanes in the v ic in i ty of the: 21. Barnet-Hastings corridor - Northeast Sector to Boundary Road 22. Lougheed corridor - H ighway 7 to the Trans Canada Highway 23. Trans Canada Highway and Grandview Highway corri-dor - Cape Horn Interchange (North of the Port Mann Bridge) to Clark Drive 24. Trans Canada Highway corridor - 200 Street to the Cape Horn Interchange (Nor th of the Port Mann Bridge) G. Provide H O V lanes across the: 25. Nor th A r m of the Fraser River at or in the vicinity of the Alex Fraser Bridge-Queensborough Bridge corridor H. Provide other improved connections as fo l lows: 26. Provide an improved Moray Channel Bridge and associ-ated roadway improvements - H ighway 99 to the Vancouver International A i rpor t 27. Improve road access f rom Highway 91 at Nordel Way to the Surrey City Centre 28. Improve east-west connection f rom Highway 17/99 to Highway 15/Highway 1 (South Perimeter Road) 29. Improve east-west connection Southeast Port Road -Mary H i l l Bypass/Highway 1 to Queensborough Br idge/Marine Way 30. Improve Highway 10 f rom Highway 99 to Trans Canada Highway 31. Provide new arterials and w iden existing arterials to serve development. 1 4 0 paXe 4 6 TRANSPORT 2021 Medium-range Plan for Greater Vancouver 1 4 1 TRANSPORT 2021 Medium-range Plan for Greater Vancouver Page 47 3.5 Staging of H O V Treatment Time frame: 1993 through 2006 A significant increase in car-pooling is projected when transportation demand man-agement (TDM) and a policy of restraint on single-occupant vehicles is applied. Two-step development of facilities for high-occupancy vehicles (HOV) is recommended (see table, right). Step A High Occupancy Vehicle Priority Treatment at Bridgeheads Required to serve vehicles with high occupancy (e.g. 3 or more) across the major water bodies in Greater Vancouver. They could be joint with bus priorities where operational considerations permit. General locations are: Lions Gate Second Narrows Oak Street Deas Tunnel Alex Fraser Port Mann Pitt River Step B Provide Superior Service Level for HOVs in Selected Corridors. Continuous HOV Lanes Treatments could include installing continuous HOV lanes and/or selective preferred access for HOVs. General corridors are: Surrey-Vancouver (vicinity of Trans Canada corridor) Northeast Sector-Vancouver (Barnet/Hasti ngs corridor) Within Northeast Sector (vicinity of Lougheed corridor) 3.6 Staging of Mixed Traff ic Improvement Time frame: 1993 through 2006 More capacity for mixed traffic would not be provided across the Burrard Inlet, the North Arm and South Arm during the medium-range planning period, in order to complement the T D M and HOV treatments above. The strategy is first to make best use of existing capacity, and then to expand facilities required for vital movements of people and goods into and out of the region. Step A Maximize person-carrying capacity of existing facilities, area-wide Apply modern system management techniques; apply parking controls on major streets to help movement of buses and/or bicycles; strategically improve roadway network to provide most efficient movement of people and goods; Build new arterials and widen existing ones to serve growth. These will be needed to maintain the viability of growth centres. Step B Enhance Inter- and Intra-Regional Roads These highway corridor improvements should give weight to several factors: • Efficient movement of goods to, from and within the region • Deterrence of long haul commuting in single occupant vehicles • Use by transit and high occupancy vehicles • Continuity, safety, community and environmental impacts TransCanad a Highway Horseshoe Bay-Hope Hwy 99 US border to Oak Street Hwy 91 from Hwy 99 south to Hwy 99 north Hwy 10 between Hwy 99 and Trans Canada River Road extension from Scott Rd to Hwy 15 SE port road from Mary Hill to Queensboro' Moray Channel bridge and associated improvements Hwy 91 at Nordel to Surrey City Centre better access Transport 2021 1 4 2 A P P E N D I X B MULTIPLE ACCOUNT EVALUATION OF RAPID TRANSIT OPTIONS FOR GREATER VANCOUVER BC Transit 10 Year Development Plan Crown Corporations Secretariat May 1995 1 4 4 EXECUTIVE SUMMARY Introduction o The purpose o f this evaluation is to compare a range o f rapid transit options for three major corridors in Greater Vancouver: Broadway-Lougheed, Coquitlam-New Westminster and Richmond-Vancouver. The intent is to assist decision-makers in selecting among a range o f options by providing a consistent assessment o f how each performs in achieving key policy objectives. o The rapid transit technologies considered for each corridor can be grouped into two categories: Rapid Bus, an enhanced bus service; and ICTS (Intermediate Capacity Transit Systems), a range o f light rail alternatives. o The primary policy objectives for rapid transit are: to provide a fast, reliable and cost-effective transit service; to promote regional land-use goals (Livable Region Strategy); and to reduce automobile use and resulting environmental impacts. Methodology o The multiple account evaluation methods used here are based on guidelines developed for the evaluation o f major capital projects by provincial Crown corporations. Accounts evaluated include: financial, customer service, environmental, urban development and system operations. o Under each account, key indicators are evaluated. Financial account indicators include capital and operating costs, taxes/charges, and incremental fare revenues (each also provided as present values). Customer service account indicators include ridership, route travel time and the value o f travel time savings. Environment account indicators include reduced air pollution and other benefits o f reduced automobile use. Urban development account indicators are qualitative assessments o f how well each option in each corridor would contribute to regional land-use goals. Systems operations account indicators qualitatively assess integration and flexibility. Summary indicators also include net social cost and the life-cycle cost per boarding. o Each option has been developed to a technical level appropriate for plaroning purposes. More detailed technical work wi l l be required once the most promising alternatives have been identified. Nevertheless, the definition o f options used here is sufficient to compare the characteristics, performance and relative merits o f the different options. o Over the longer term, it is envisioned that ICTS lines wi l l be established in each corridor. However, over the shorter planning horizon of the BC Transit 10 Year Development Plan, at most one ICTS could be undertaken. Simultaneous development o f ICTS in two or three corridors is not considered feasible and was not contemplated largely because of government fiscal limitations. R a p i d Transit - M u l t i p l e A c c o u n t E v a l u a t i o n - May 1995 P a g e i 1 4 5 o This evaluation is done in two parts. The first part compares the performance o f the options within each separate corridor. The second part compares between corridors, assessing the best ICTS option(s) f rom each corridor against one another. Broadway-Lougheed o The Rapid Bus option has net financial costs o f $ 129 mill ion (PV) and net social costs o f $41 mill ion (PV)- I t is forecast to attract 17.3 mill ion boardings in 2006 but is assumed to have a low contribution to regional land-use goals. Wi th a life-cycle cost per boarding of $0.98, Rapid Bus is the most cost-effective transit service. o The Basic L R T option has net financial costs o f $403 mil l ion (P V ) and net social costs o f $302 mil l ion (PV). 20.8 million boardings are forecast on this service in 2006. Basic L R T is assumed to make a low to medium contribution to regional land-use goals. I t has a cost per boarding o f $3.26. T a b l e V - 4 Broadway-Lougheed Multiple Account Summary All values in 1994 dollars Rapid Basic Enhanced ALRT Bus LRT LRT Net Financial Costs (millions, FV) ($ 129) ($403) ($413) ($677) Quantified Social Benefits (millions, PV) $89 $101 $ 139 $193 Net Social Costs (millions, PV) ($41) ($302) ($ 274) ($484) Boardings in 2006 (millions) 17.3 20.8 21.6 28.8 Contribution to Regional Land-Use Goals Low Low/Medium Medium Medium/High Life-cycle Cost per Boarding $0.98 $3.26 $ 3.26 $3.83 o The Enhanced L R T option has net financial costs slightly higher than the Basic L R T option at $413 million (PV) but lower net social costs at $274 million (PV) because o f 40% higher time savings benefits. Enhanced L R T is forecast to have 21.6 mill ion boardings in 2006. I t is assumed to have an medium contribution to regional land-use goals. The cost per boarding estimate o f $3.26 is equal to that for the Basic L R T option because additional ridership offsets additional costs. Given superior net social costs, ridership and land-shaping potential and equivalent cost per boarding, Enhanced L R T is superior to Basic LRT. o The A L R T option has net financial costs o f $677 mill ion (PV) and net social costs 75% higher than the Enhanced L R T option at $484 mill ion (PV)- I t attracts substantially higher ridership with boardings o f 28.8 million. I t is assumed to make a medium to high contribution to regional land-use goals. The cost per boarding o f $3.83 is highest in the corridor. Rap id Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page ii 1 4 6 Coquitlam-New Westminster o The Rapid Bus option has net financial costs of $92 million and net social costs of $70 million (PV). It is forecast to attract 4.3 million boardings in 2006. It is assumed to make a low contribution to regional land-use goals. A cost per boarding of $2.41 is estimated. o The Enhanced L R T option has net financial costs of $425 million (PV) and, with virtually no quantified social benefits, net social costs of $424 million (PV). The Enhanced L R T is forecast to attract 12.2 million boardings. It is assessed as making a medium to high contribution to regional land-use goals. The life-cycle cost per boarding for this option is estimated at $5.39. o The A L R T option has net financial costs of $593 million (PV) and net social costs equal to $547 million (PV). It is forecast to have boardings of 16.5 million in 2006. It is assumed to have high potential to contribute to regional land-use goals. The cost per boarding is $5.60. Table V - 8 Coquitlam-New Westminster Multiple Account Summary All values in 1994 dollars Rapid Enhanced ALRT Bus LRT Net Financial Costs (millions, PV) ($ 92) ($425) ($ 593) Quantified Social Benefits (millions, PV) $22 $2 $47 Net Social Costs (millions, PV) ($70) ($ 424) ($ 547) Boardings in 2006 (millions) 4.3 12.2 16.5 Contribution to Regional Land-Use Goals Low Medium/High High Life-cycle Cost per Boarding $ 2.41 $5.39 $5.60 o Ridership forecasts for this corridor are the most uncertain. This corridor is the least developed today and for desired growth to materialise it will have to attract as much growth as the other two combined. The relatively high cost per boarding for the Rapid Bus alternative (compared to Rapid Bus in other corridors) suggests that a somewhat scaled-down service might be more cost-effective in the short term. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page iii 1 4 7 Richmond-Vancouver o Rapid Bus has net financial costs of $135 million (PV) and net social costs of $89 million (PV). It has 9 million boardings in 2006 and has low potential to contribute to regional land-use goals. The life-cycle cost per boarding equals $1.88. o The Arbutus Enhanced L R T option has net financial costs of $564 million (PV) and net social costs more than four times those of the Rapid Bus option at $405 million (PV). It is forecast to attract 15 million boardings but has a low/medium contribution to regional land-use goals. Its cost per boarding is $6.30. o The Cambie Basic L R T option has lower net financial costs than the Arbutus Enhanced L R T option at $393 million (PV). Net social costs are $335 million (PV). Cambie Basic L R T achieves only 10.4 million boardings, however, and has low contribution to regional land-use goals. Cost per boarding, however, is virtually the same as the Arbutus option at $6.29. o The Cambie Enhanced L R T option has substantially higher net financial costs of $723 million (PV), net social costs at $532 million (PV) and has 17 million boardings in 2006. It is assumed to make a low to medium contribution to regional land-use goals. It has the highest cost per boarding of any option at $6.94. o The Cambie A L R T option has somewhat smaller net financial costs of $668 million (PV), net social costs of $447 million (PV) and attracts 20.2 million boardings. It too makes a low to medium contribution to regional land-use goals. It has the lowest cost per boarding of any Richmond-Vancouver ICTS option at $5.54. Table V - 12 Richmond-Vancouver Multiple Account Summary All values in 1994 dollars Granville Rapid Bus Arbutus Enhanced LRT Cambie Basic LRT Cambie Enhanced LRT Cambie ALRT Net Financial Costs (millions, PV) ($ 135) ($564) ($ 393) ($ 723) ($ 668) Quantified Social Benefits (millions, PV) $46 $ 159 $58 $ 191 $221 Net Social Costs (millions, PV) ($89) ($405) ($ 335) ($ 532) ($ 447) Boardings in 2006 (millions) 9 15 10.4 17 20.2 Contribution to Regional Land-Use Goals Low Low/Medium Low Low/Medium Low/Medium Life-cycle Cost per Boarding $1.88 $6.30 $6.29 $6.94 $5.54 Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page iv 1 4 8 Comparison Between Corridors: Potential for ICTS o For this comparison, preferred ICTS option(s) in each corridor have been selected: Broadway-Lougheed Enhanced L R T and ALRT, Coquitlam-New Westminster Enhanced L R T and A L R T ; and Richmond-Vancouver's Cambie A L R T . o The key attributes o f each corridor's option include the contribution to regional land-use goals, transit ridership and cost-effectiveness. For each an assessment (low, medium, high) is provided based on either a previous rating (for land-use) or the comparison o f selected indicators (boardings, cost per boarding, and net social costs). o The degree to which ICTS options would contribute to regional land-use goals is critical in the selection o f a preferred ICTS corridor. In fact, i t is clear f rom the preceding analysis that without the benefits o f land-use shaping, no ICTS would be warranted in the medium term. The Broadway-Lougheed Enhanced L R T would make a medium contribution and A L R T a medium/high contribution. The Coquitlam-New Westminster Enhanced L R T is rated medium/high and the A L R T , high. Richmond-Vancouver's Cambie A L R T would make only a low/medium contribution. Clearly, the competition on this issue is between options in the Coquitlam-New Westminster and Broadway-Lougheed corridors, wi th the former being somewhat superior to the latter. o In terms o f transit ridership, Broadway-Lougheed is the superior corridor (and rated high) closely followed by Richmond-Vancouver (rated medium/high). Relative to the others, the Coquitlam-New Westrninster corridor has low/medium transit ridership. o In terms o f cost-effectiveness, Broadway-Lougheed is rated high, while both Coquitlam-New Westrninster and Richmond-Vancouver are rated low/medium. Broadway-Lougheed Enhanced L R T has a cost per boarding o f $3.26 and A L R T , $3.89. Coquitlam-New Westminster Enhanced L R T is $5.39, A L R T is $5.60. Richmond-Vancouver's Cambie A L R T has a cost per board o f $5.54. Broadway-Lougheed Enhanced L R T has the lowest Table v-13 Comparison Between Corridors Assessments relative to other corridors Broadway-Lougheed Enhanced LRT/ALRT Coquitlam-New Westminster Enhanced LRT/ALRT Richmond-Vancouver Cambie ALRT Contribution to Regional Land-use Goals Medium to Medium/High Medium/High to High Low/Medium Transit Ridership High Low/Medium Medium/High Boardings in 2006 21.6 to 28.8 million 12.2 to 16.5 million 20.2 million Cost-Effectiveness High Low Low Cost per Boarding $3.26 to $3.89 $5.39 to $5.60 $5.54 Net Social Costs (PV) $274 to $484 million $424 to $547 million $447 million Rap id Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page v 149 net social cost at $274 million. The A L R T option on Broadway-Lougheed has much higher net social costs of $484 million. Net social costs for Coquitlam-New Westminster's Enhanced L R T are $424 million; for its A L R T alternative, $547 million. The Cambie A L R T has net social costs of $447. o In summary, Broadway-Lougheed is second to Coquitlam-New Westminster in terms of regional land-use goals (medium to medium/high) and first in terms of ridership (high)and cost-effectiveness (high). Coquitlam-New Westminster is the best corridor in terms of regional land-use goals (medium/high to high) but has low/medium transit ridership and low/medium cost-effectiveness. Richmond-Vancouver has low/medium value in serving regional land-use goals. It has medium/high transit ridership, though less than Broadway-Lougheed. And, it has low/medium cost-effectiveness. Conclusions o The potential revenues from taxes or charges levied on development surrounding rapid transit stations are relatively small compared to the high costs of ICTS investments. Social benefits from travel time savings and reduced auto use can be up to $200 million (PV). However, they are not sufficient to justify ICTS alternatives with net financial costs., in the range of $400 to $700 million (PV). o Rapid Bus would provide the most cost-effective transit service in each corridor, but it would be the least effective in promoting compact development and its ridership levels may be overstated as a result. ICTS options can only be justified if they contribute sufficient benefits in achieving regional land-use goals. o Because of low/medium value in contributing to regional land-use goals and high net financial and social costs, investment in Richmond-Vancouver ICTS does not appear to be warranted in the medium term. o Coquitlam-New Westminster has an advantage over Broadway-Lougheed in terms of supporting regional land-use goals. However, it is less cost-effective due to its low ridership and faces a higher risk of failing to promote development and generate its forecast ridership levels. o The Broadway-Lougheed Enhanced L R T has the lowest lifecycle cost per boarding of any ICTS option, lowest net social cost, high ridership of over 21 million per year and would make a medium contribution to regional land-use goals. Rapid Transit - Mul t ip le A c c o u n t Evaluat ion - May 1995 Page v i 1 5 0 P R E F A C E This report is part o f a series o f technical papers prepared for the BC Transit 10 Year Development Plan. The Mult iple Account Evaluation of Rapid Transit Options For Greater Vancouver was written by Alan Greer, Crown Corporations Secretariat with assistance from Stephen Gardner, BC Transit (subsequently Ward Consulting Group), and Eva Hage, CCS. Valuable editorial comments and advice were provided by a number o f individuals. Key inputs were obtained from several studies undertaken by various consultants. Direction was provided by Glen Leicester and Mike Newson, BC Transit; Frank Blasetti, Crown Corporations Secretariat and the BC Transit 10 Year Development Plan Steering Committee which included the following members: o Derek Corrigan, Chair, BC Transit o Blair Trousdell, President & CEO, BC Transit o Marv in Shaffer, Secretary Responsible for the Crown Corporations Secretariat Additional copies o f this report can be obtained by writ ing, calling or faxing BC Transit at: BC Transit Corporate Communications 13401 - 108th Avenue Surrey, B.C. V3T 5T4 Telephone: (604)540-3251 Fax: (604) 540-3250 Rapid Transit - Mult ip le A c c o u n t Evaluat ion - May 1995 Page vii T A B L E O F C O N T E N T S E X E C U T I V E S U M M A R Y i P R E F A C E v i i T A B L E O F C O N T E N T S vi i i I . I N T R O D U C T I O N 1 I I . P O L I C Y O B J E C T I V E S 2 Provincial Transportat ion Strategy 2 Livable Region Strategy / Transport 2 0 2 1 2 B C Transi t Strategic Objectives 3 Objectives for Rapid Transit 3 I I I . R A P I D T R A N S I T O P T I O N S 4 Corr idors 4 Technologies 5 Technology/Corr idor Options 7 I V . E V A L U A T I O N M E T H O D O L O G Y 8 Evaluation Framework 8 Evaluation Methodology 9 Analysing Values over Time 1 5 V. R E S U L T S 16 Broadway-Lougheed Cor r idor 16 Coquit lam-New Westminster Cor r ido r 21 Richmond-Vancouver Cor r ido r 2 5 Comparison Between Corr idors : Potential for I C T S 3 0 Conclusions 3 2 A P P E N D I X : SENSrrrvTTY A N A L Y S I S 3 3 Ridership 3 3 Discount Rates 3 4 Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 P a g e viii 1 5 2 I . I N T R O D U C T I O N This multiple account evaluation was undertaken to provide information on and enable comparison of a variety of rapid transit options for three major corridors in Greater Vancouver: Broadway-Lougheed, Coquitlam-New Westrninster and Richmond-Vancouver. The need for major transit enhancements in the key corridors was identified in the Greater Vancouver Regional District's (GVRD) Livable Region Strategy and the associated Transport 2021 project co-sponsored by the G V R D and the Ministry of Transportation and Highways (MoTH). The multiple account evaluation methods used here are based on guidelines developed for the evaluation of major capital projects by provincial Crown corporations.1 The intent is to assist decision-makers in selecting among a range of options by providing a consistent assessment of how each performs in achieving key policy objectives. Numerous simplifying assumptions have been made in the analysis. Each option has been developed to a technical level appropriate for planning purposes. Cost estimates have been made with "orders of magnitude" precision only. Moreover, there is considerable uncertainty involved in forecasting events far in the future. Demand has been estimated for the year 2006 assuming growth meets the goals of the Livable Region Strategy. No other growth scenarios were assessed. Only those social benefits that are readily measurable have been included. Nevertheless, the evaluation is sufficient for the purpose of comparing the characteristics, performance and relative merits of the different options. The remainder of this report is arranged as follows. Section II describes the relevant provincial, regional and corporate policy objectives. Section III describes the rapid transit corridors and technologies. Section IV explains the methodology used in the evaluation. Section V presents the results of the multiple account evaluation. The Appendix includes a sensitivity analysis of key results. ' Multiple Account Evaluation Guidelines. Crown Corporations Secretariat, Province of British Columbia, February, 1993. Rapid Transit - Mul t ip le A c c o u n t Evaluat ion - May 1995 Page 1 153 I I . P O L I C Y O B J E C T I V E S The evaluation o f major public projects must be conducted within the appropriate policy context. The rapid transit options for Greater Vancouver are evaluated with reference to the policy objectives o f the Province, the GVRD and BC Transit. Provincial Transportation Strategy The Province is currently developing a comprehensive transportation strategy to address critical issues facing provincial transportation systems. Urban transportation systems, in particular, have not kept pace with rapid growth in population and transportation needs. Long distance movement o f people and goods on major inter-regional routes is increasingly constrained because o f traffic and congestion in and around urban areas. Recognising that the status quo is unacceptable and that a "business-as-usual" response wi l l not work, the provincial transportation strategy sets out new directions to address the urban, inter-regional, international and remote-community transportation challenges. T o meet the backlog and rapidly growing demand for transportation facilities and services in rapidly growing urban areas, the strategy emphasises the need to make better use o f existing facilities: encouraging transit, car pools and cycling t o reduce peak demands; improving traffic management systems; controlling access to key goods movement corridors; ensuring more timely maintenance and rehabilitation o f facilities. The strategy also calls for a more strategic approach to new investments. In urban areas, this means focusing on the movement o f people and goods as opposed to vehicles, and ensuring an expanded role for transit in satisfying the travel requirements o f residents. I t also means closely integrating land use and transportation initiatives in urban areas. For example, in the case o f rapid transit, this means tying major new investments to complementary local government land-use policies, zoning, parking and traffic management measures. The strategy also outlines measures to ensure that new facilities and services are developed and provided in the most cost-effective way. Livable Region Strategy / Transport 2021 The G V R D has developed a Livable Reeion Strategy that would combine compact metropolitan area development with transportation improvements that give priority to non-automobile modes of travel. The goals o f the strategy are o to protect green spaces (open space, wetlands, habitat or resource lands); o to build complete communities where people's homes and jobs are closer together; o to achieve a compact metropolitan region concentrating the region's growth on the Burrard Peninsula, North Surrey and North Delta; and o to increase transportation choice, employing transportation demand management and strategic transportation investments (including rapid transit). Rap id Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 2 1 5 4 The Livable Region Strategy is designed to limit the suburban sprawl that is putting pressure on agricultural lands up the Fraser Valley, reduce the growth in financially and environmentally-costly long distance commuter traffic, and increase densities that wi l l enable more cost-effective provision of transit and other public services. Transport 2021 (the transportation component o f the Livable Region Strategy) proposes the establishment o f major bus corridors and bus priority measures, expansion o f cross-regional bus services, the development o f comprehensive H O V networks and the introduction o f rapid transit in the Coquitlam-New Westminster, Broadway-Lougheed and Richmond-Vancouver corridors over the long term. It also proposes tolls, additional gas taxes, parking taxes and other measures to discourage single occupant vehicle traffic and to finance the proposed transit and other measures. BC Transit Strategic Objectives The BC Transit 10 Year Development Plan is currently being developed as a complementary effort to the provincial Transportation Strategy. The plan states three fundamental objectives: o to reverse the decline o f BC Transit's market share and help achieve the broader goals in the provincial and regional transportation strategies; o to support the growth management and land-use strategies o f the G V R D and other regions; o to ensure appropriate levels o f service to all residents o f transit service areas, wi th particular regard to those without automobiles and the disabled population. These objectives wi l l not be easy to achieve. Demographic factors, in particular the increasing proportion o f the region's population in older age brackets (who tend to be more automobile dependent or who more often require off-peak and specialised services) works against increased usage o f transit. More diverse travel patterns due to suburban sprawl and more decentralised employment are also more diff icult to serve with transit. Low density development makes for less efficient transit. These trends, as well as past experience, point out that isolated measures, including rapid transit investments in specific corridors, wi l l not be sufficient to meet the objectives. BC Transit, the Province and municipalities must work together to deliver a comprehensive package o f service, facility, traffic management and complementary land-use measures to significantly increase transit use and reduce the rate o f growth o f single occupant vehicle traffic. Objectives for Rapid Transit These separate policies share a common view o f the many challenges facing the Greater Vancouver region and provide similar directions on how to deal with them. Together these policies constitute the policy context within which preferences for alternative rapid transit technologies should be determined. From these policies, the primary objectives for rapid transit can be drawn: to provide a fast, reliable and cost-effective transit service; to promote regional land-use goals (Livable Region Strategy); and to reduce automobile use and resulting environmental impacts. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 3 I I I . R A P I D T R A N S I T O P T I O N S The rapid transit options evaluated in this report are distinguished by corridor and technology. Each option has been developed to a level appropriate for planning purposes. More detailed specification o f routes, alignment, technology, design, service levels and timing o f construction wi l l be required once the most promising alternatives have been identified. Nevertheless, the definition o f options used here is sufficient for the purpose o f comparing the characteristics, performance and relative merits o f the different options. 2 Corridors Three major transit corridors are included in this evaluation. The various corridor alignments use a combination o f dedicated right-of-way and road right-of-way and include at-grade, elevated and tunnel sections (tunnelling has only been considered where necessary operationally, not to make lines less intrusive). o Broadway-Lougheed: The Broadway-Lougheed corridor connects Lougheed Mal l in Burnaby wi th central Broadway (ending at Arbutus) in Vancouver. 3 The route travels along the Lougheed Highway in Burnaby, then Broadway in Vancouver crossing the existing SkyTrain line at Broadway and Commercial. The alignment is within the road right-of-way along the whole route. o Coquitlam-New Westminster: The Coquitlam-New Westminster corridor connects Coquidam Town Centre with New Westminster. The various alignments generally fol low west from Coquitlam Town Centre along Barnet Highway or Guildford Way, south on North Road and along Columbia Street into New Westminster where they meet the existing SkyTrain. The alignments leave the road right-of-way in a number o f places. o Richmond-Vancouver: The Richmond-Vancouver corridor connects Richmond wi th downtown Vancouver. The specific alignments along the corridor vary by technology through Vancouver. The Rapid Bus alignment follows Granville Street. T h e Arbutus L R T alignment follows the CP Rail right-of-way paralleling Arbutus Street and West Boulevard. The remaining alignments follow Cambie Street (at grade or grade separated). 2 More detailed descriptions of the options are included in the BC Transit 10 Year Development Plan - Summary Report and the supporting engineering feasibility studies: Regional Intermediate Capacity Transit System Study. Broadway-Lougheed Corridor. N.D. Lea Consultants, September, 1994; Review of Intermediate Capacity Transit Systems: Coquitlam-New Westminster Corridor. Delcan, November, 1994; and Review of Intermediate Capacity Transit Systems: Richmond-Vancouver Corridor. Delcan. November. 1994. 3 The engineering study of the Broadway-Lougheed corridor also evaluated an eastern section which connects Lougheed Mall and Coquitlam Town Centre by way of Lougheed Highway. This section was not included in the multiple account evaluation because initial results showed markedly low ridership and because the rapid transit alternatives in the Coquitlam-New Westminster corridor connect both points along a more populated (and thus more desirable) route. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 4 1 Rapid Transit Corridors in Greater Vancouver Technologies The technologies considered for each corridor can be grouped into two categories: Rapid Bus and Intermediate Capacity Transit Systems (ICTS). o Rapid Bus: Rapid Bus is based on conventional bus service but involves a range of technological enhancements and traffic priority measures (e.g., limited stops, low floor buses, articulated buses, traffic signalling priority, dedicated lanes) to increase the speed of the service.4 Most significant for this analysis is that a Rapid Bus system has relatively low costs, can be implemented quickly, and service levels can be readily adjusted to meet demand. Rapid Bus is an interim rapid transit service useful in developing ridership until conversion to a full rapid transit system is appropriate. 4 Different forms of transit — whether buses, light rail transit or heavy rail commuter trains — offer different capacities. Rapid Bus can carry up to 5,000 persons per direction per hour (pphpd) at the peak load point, at-grade LRT up to 10,000 pphpd, ALRT or grade-separated LRT up to 20,000 pphpd and heavy rail, 40,000 - 100,000 pphpd. For the levels of ridership projected in the near future, heavy rail alternatives are not considered necessary for these corridors and were not included in the evaluation. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 P a g e 5 1 5 7 Technologies Rapid Bus - Technology utilises conventional or articulated vehicle with capacity up to 65 seated, 90 to 100 persons total - operates in bus-only lanes with various priority measures - driver operated - power by diesel, natural gas or electricity (trolley) - can operate as frequently as every 30 seconds, although it is difficult to sustain reliable operation - practical capacity 5,000 passengers per hour per direction (pphpd) ICTS: Light Rail Transit - vehicle capacity 150 to 160 persons - can be operated singly or in trains - operated from overhead electric wires - driver operated - can be operated as frequently as every 3 minutes - can move up to 20,000 pphpd Ha ICTS: Au tomated Light Rail Transit - vehicle capacity 40 seated, 75 to 80 persons total - can be operated in 2. 4 or 6 car trains - operated from third rail, computer driven can be operated as frequently as every 90 seconds - existing MARK I cars capable of carrying up to 20,000 pphpd ICTS: ICTS technologies include both conventional Light Rail Transit (LRT) and Automated Light Rail Transit (ALRT) such as the existing SkyTrain.5 L R T systems can either be at-grade (often on existing roadway) or on grade-separated guideways (above the road or in a tunnel), but all involve a fixed rail line. A L R T guideways are largely grade-separated and must be on dedicated right-of-way. For this evaluation, a distinction is made distinguishes between a Basic LRT, with little or no grade separation, and an Enhanced 5 The engineering feasibility studies also examined busways (involving conventional buses operating on a dedicated right-of-way) but concluded that they were not feasible in any corridor. R a p i d Transi t - M u l t i p l e A c c o u n t E v a l u a t i o n - May 1995 P a g e 6 1 5 8 LRT, with more extensive grade separation. The ICTS options are characterised by potentially high capacity and speed (depending on grade separability), high cost and long construction time. Technology/Corridor Options Table III-l summarises the combination of technologies and corridors evaluated in this report. Table lll-l Rapid Transit Options in Greater Vancouver Corridor Broadway-Lougheed Technology Rapid Bus Basic LRT (totally at grade) Enhanced LRT (some grade separation) ALRT Coquitlam-New Westminster Rapid Bus Enhanced LRT (some grade separation) ALRT Richmond-Vancouver Rapid Bus (Granville) Enhanced LRT (Arbutus; some grade separation) Basic LRT (Cambie; at grade) Enhanced LRT (Cambie; extensive grade separation) ALRT (Cambie) Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 7 1 5 9 I V . E V A L U A T I O N M E T H O D O L O G Y Evaluation Framework The purpose o f multiple account evaluation is to assess alternatives against a set o f objectives. Specific objectives for rapid transit (addressed above) are grouped in this evaluation under a general set which includes financial, customer service, environmental, urban development and system operations "accounts." This evaluation is done in two parts. The first part compares the performance o f the options within each separate corridor by a full assessment o f each's performance under the accounts. The second part compares between corridors, assessing the best ICTS option from each corridor against one another using a subset o f the preceding indicators. Technology/Corridor Assessment For the assessment o f technology options within each corridor, several key indicators under each o f the above accounts is presented. Indicators under the financial account include the following. Capital costs are the total costs o f planning, construction and acquisition o f rapid transit technology net o f savings from reduced bus service requirements.6 Operating costs are the average annual operating costs o f rapid transit technology in 2006 net o f bus savings.7 Incremental fares are the increased annual revenues f rom fares paid by new riders. Taxes/charges measure the potential amount o f benefiting area taxes or development cost charges which could be recovered in two separate years (2006 and 2021). Each o f the above values is also discounted (using an 8% rate, net o f inflation) to provide present value (PV) estimates o f the stream o f these values over a 47 year period. The sum o f these values are the net financial costs. The customer service account indicators include the following. Total boardings measures the total number o f people who board the rapid transit service in the corridor in 2006. Incremental boardings measures the number o f transit riders overall who would otherwise not have used the transit system. Route travel time is the total time in minutes over the total route (end to end). Route time saved is the difference between the route travel time and the current route time. The value of time savings (PV) is the present value o f the stream o f benefits o f travel time savings for existing transit riders. The environmental account indicators reflect environmental impacts resulting from reduced auto use and include the following. The reduced NOx and VOC, particulates, and CO2 6 All values are expressed in 1994 dollars. 7 Most forecasts are made for 2006, the year for which GVRD land-use projections are based. It is assumed that ridership would be well-established by that time, as each transit option would have been operational for at least five years. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 8 1 6 0 indicators measure reduced air pollution in terms of the estimated tonnes o f reduced auto emissions of each respective grouping o f substances in 2006. The environmental benefits (PV) measures the present value of the stream of benefits from reduced emissions, noise, energy use, water pollution and waste. The sum of the present values of environmental benefits and o f value of time savings is the quantified social benefits. The urban development account indicators are qualitative assessments o f how well each option supports regional land-use goals (i.e., the Livable Region Strategy). Each is rated as low, low/medium, medium, medium/high, or high. The assessments have two components: the potential o f the corridor to accommodate regional land-use goals and (after taking account o f the effectiveness of a particular technology) the contribution o f each option to those goals. The systems operations account indicators are both qualitative assessments as well. Flexibility in adjusting capacity rates how readily the service can be adjusted (increased, decreased, re-routed) to meet change in demand. Ease of integration into transit system rates how easily the technology can be integrated into BC Transit system. A summary multiple account table is also provided which selects key indicators f rom the above accounts. The summary also includes two additional indicators: net social costs (the sum o f net financial costs and quantified social benefits) and the life-cycle cost per boarding (see definition below). Comparison of Corridors Over the longer term, it is envisioned that ICTS lines wi l l be established in each corridor. However, over the shorter planning horizon of the BC Transit 10 Year Development Plan, at most one ICTS could be undertaken. Simultaneous development o f ICTS in two or three corridors is not considered feasible and was not contemplated largely because o f government fiscal limitations. In the second part o f the evaluation, the corridors are compared against one another wi th respect to their performance as the location of the first ICTS. For each corridor, the performance o f the best ICTS option(s) is used. The multiple account indicators for the comparison between corridors includes three high-level qualitative assessments. The first is the contribution to regional land-use goals (above). The second is transit ridership and the third is cost-effectiveness. Each is rated as low, low/medium, medium, medium/high or high and each draws on the performance of the selected ICTS options from the technology/corridor assessments. Evaluation Methodology Cost Estimates Consultants were retained to identify rapid transit options for each corridor and to provide preliminary technical feasibility assessments and cost estimates for each option (see reference above). Capital costs were defined to include the costs o f construction and acquisition o f fixed facilities, property, vehicles and maintenance facilities and also to include project design, Rapid Transit - Mul t ip le A c c o u n t Evaluat ion - May 1995 Page 9 161 management and contingency costs.8 Operating costs were based on estimates for a projected level o f operations in 2006. For each option, BC Transit estimated capital and operating cost savings resulting from reduced regular bus service requirements. For the multiple account evaluation, it is assumed that Rapid Bus capital work could be completed in three years (with work beginning in 1995) and that vehicles would be replaced every 18 years. Construction o f ICTS options were assumed to require five years o f construction beguining in 1997 (assuming only two years o f pre-planning, engineering, public consultation and approval would be required). The life o f ICTS capital is assumed to be 40 years. Operating costs for Rapid Bus would grow in proportion to ridership over the period while operating costs for ICTS options would remain constant in real terms. 9 Transportation Demand Forecasts The primary demand forecasting tool used in the evaluation is E M M E / 2 , a peak-hour (7:30-8:30 am) transportation demand model developed fo rme Vancouver area by me GVRD. BC Transit employed EMME/2 to forecast transit demand and ridership in 2006 for the whole transit system. EMME/2 provides estimates o f ridership on each facility or service, travel time savings for transit riders, reductions in vehicle kilometres travelled ( V K M T ) because o f car travellers switching to transit, and changes in fare revenues.1 0 The EMME/2 model was used to simulate a variety o f system scenarios: a base case and ten combinations o f rapid transit system development. Given the significant time and effort required to model different system scenarios, a l imit on the number o f scenarios was necessary. Each rapid transit scenario includes Rapid Bus in at least two corridors and at most one ICTS option (as noted previously, the maximum o f one ICTS reflects a real constraint). The scenarios are listed below in Table TV-1. Under each scenario, the net incremental effect on the system was determined by subtracting the base case results from the particular scenario results. However, EMME/2 estimates what is occurring in the whole system not separate corridors. Only ridership can readily be extracted f rom E M M E / 2 on a corridor-specific basis. T o derive the corridor-specific t ime 8 The feasibility studies and the multiple account evaluation do not include the costs of additional road work required to mitigate any loss of road space, nor do they estimate the cost savings resulting from reduced road work required because of a shift of commuters from cars to rapid transit. 9 The timing of ICTS expenditures was estimated by N.D. Lea Consultants based on SkyTrain Phase HI experience. Unless otherwise specified, other assumptions regarding the timing of expenditures, growth of ridership, and operating costs were made by BC Transit staff. Specific timing of capital costs is shown below. Timing of Capital Expenditures (% of total costs) 1995 1996 1997 1998 1999 2000 2001 Rapid Bus 20% 40% 40% ICTS - All except vehicles 3% 17% 48% 18% 14% ICTS - Vehicles 40% 60% In addition, the full cost of replacing Rapid Bus vehicles is incurred in 2013 and 2031. A more complete description of EMME/2 is included in Chapter 11 of each corridor feasibility report. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 10 162 savings, vehicle kilometres travelled and fare revenues results therefore; the system results were divided and allotted to each corridor based on the corridor ridership proportions." Table IV-1 EMME/2 Demand Forecasting Model System Scenarios Scenarios Broadway- Coquitlam- Richmond-Lougheed New Westminster Vancouver Base Case (Do Minimum) Existing service plus Existing service Existing service plus Express bus Express bus All Rapid Bus Rapid Bus Rapid Bus Rapid Bus Broadway-Lougheed ICTS options Basic LRT Rapid Bus Rapid Bus Enhanced LRT Rapid Bus Rapid Bus ALRT Rapid Bus Rapid Bus Coquitlam-New Westminster ICTS options Rapid Bus Enhanced LRT Rapid Bus Rapid Bus ALRT Rapid Bus Richmond-Vancouver ICTS options Rapid Bus Rapid Bus Arbutus Enhanced LRT Rapid Bus Rapid Bus Cambie Basic LRT Rapid Bus Rapid Bus Cambie Enhanced LRT Rapid Bus Rapid Bus Cambie ALRT E M M E / 2 estimates are for a one hour period in 2006. Hourly estimates were converted to daily using expansion factors based on existing bus daily ridership in the corridor. Daily figures (which reflect an average weekday) are converted to annual using a factor o f 300. The annual estimates for ridership (as well as vehicle km. travelled, time savings, and incremental fares) were then distributed over the whole 47 year period at a common assumed growth pattern. I t was assumed that Rapid Bus would begin limited operation in 1995 at 20% o f the 2006 ridership level, growing quickly in the early years as capital works are completed (50% in 1996, 70% in 1997) and then grow steadily to 2006 and beyond (at 2.2% per year) unti l reaching a level 110% o f the 2006 amount. 1 2 ICTS ridership is assumed to begin in 2002 (at 74% of 2006 level) growing steadily to 2006 and beyond (at 2.2% per year) and then is also capped at 110% o f the 2006 level. 1 1 Rapid Bus ridership is forecast at 17.3 million on Broadway-Lougheed, 4.3 million for Coquitlam-New Westminster, and 9 million for Richmond-Vancouver, for respective shares of 56%, 14% and 30%. Shares of the other EMME/2 system results (time savings, vehicle km. travelled, incremental fares) under the All Rapid Bus scenario were divided between corridors according to these proportions. Since all other scenarios combine ICTS in one corridor with Rapid Bus.in the other two corridors, all incremental changes in time savings, vehicle km. travelled or fares were attributed to that ICTS. This modification of the results was necessary to isolate, at least approximately, the impacts in specific corridors. 1 2 The assumption that ridership growth stops at a level 10% higher than 2006 is based on assumed capacity of the technology. Additional service enhancements (more vehicles, extended services) could obviously be employed to maintain growth beyond this limit, but such growth cannot be included without also including the costs of those enhancements. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 1 1 163 The land use assumptions incorporated in the EMME/2 model represent an optimistic scenario under which the growth patterns sought under the Livable Region Strategy have been realized. Modelling o f other more conservative growth scenarios was not undertaken because o f the significant effort required and limited time available. Results o f EMME/2 , therefore, must be interpreted cautiously recognizing in particular, that Rapid Bus is unlikely to stimulate such desired growth and that ridership results for Coquitlam-New Westminster are quite uncertain as assumed growth for this area is most optimistic. Potential Revenues from Increased Land Values Rapid transit can attract development to locations along the corridor (especially around stations) resulting in increased land values. A share o f the land value increases could be captured through benefiting area taxes or development cost charges and used to defray some o f the costs o f the project. Estimates were made for revenues from taxes or charges that could be collected in 2006 and 2021 in each corridor (based on cumulative annual increases in value for 2000-2006 and 2007-2021 respectively). The estimates were based on a study undertaken to estimate the potential for revenue collection through various levies 1 3 and associated research estimating the additional units that would be built within 1000 metres o f each station along the line wi th or " without an ICTS. 1 4 The possibility o f development benefits resulting on more distant land is less certain and more difficult to attribute to the ICTS alone. For present value calculations, taxes/charges were divided into equal annual shares and distributed over the respective periods. Valuing Customer and Environmental Benefits Customer benefits associated with a rapid transit system primarily stem from travel time savings. Travel time savings were based on EMME/2 estimates o f hours saved by existing transit riders for each technology in 2006. The change in travel times for new transit riders or car commuters were not included. 1 5 The total annual hours saved by existing transit riders were valued at a rate considered appropriate for transit riders: 40% of the average BC wage, or $7.00 per hour. 1 6 1 3 Report on Potential Revenue Sources to Fund Transit in the Greater Vancouver Regional Transit Service Area. KPMG, October, 1994. 1 4 Potential revenues were estimated by Coriolos and RSMI consultants (and incorporated in the corridor feasibility studies) for each corridor with an ALRT line (as it is assumed to be the most effective in attracting development around stations). Enhanced LRT is assumed to generate 75% of the ALRT levels, Basic LRT, 30%, while Rapid Bus is assumed to generate no such revenues. 1 5 New transit riders (those attracted by the particular added service) are estimated by EMME/2 to have net time losses (travelling by rapid transit takes longer than their previous commute time by vehicle). They chose rapid transit, however, because the combined cost of driving (vehicle operating costs, parking fees as well as time) was higher even if time alone was not. For simplicity, it is assumed that these travel time losses and the savings from reduced driving costs fully offset each other. Regardless, any difference would be small. 1 6 The common approach to valuing time savings differentiates between lost work time (e.g.; a truck driver caught in traffic) and lost non-work time (which includes commuting). The primary use of urban transit systems is for non-work purposes (in the Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 12 164 The environmental impacts o f rapid transit include those stemming from reduced automobile use, reduced road work, and rapid transit system construction and operation. Only the first is readily measurable, however. The environmental costs o f automobile use have been extensively researched and to varying degrees valued in monetary terms. EMME/2 estimates o f reduced vehicle km. travelled measure the effectiveness o f each option in reducing auto use. Conversion factors (based on assumed vehicle emission rates) provide the physical quantities o f reduced ozone producing substances (NOx and VOC) , particulates and greenhouse gases (CO2).17 The estimated value o f reduced automobile use combines other benefits as well as reduced emissions: reduced noise, energy use, water pollution and waste. The unit value o f these environmental benefits is assumed to be $ 0.205 / vehicle km. travelled. 1 8 Estimates o f environmental benefits are thus based on direct conversion from vehicle km. travelled, not a value per tonne o f emissions. Urban Development Investment in rapid transit systems can be effective in promoting land-use goals such as those articulated in the GVRD's Livable Region Strategy: increasing transportation choice, achieving a compact metropolitan region, building complete communities, and protecting green spaces. The benefits o f achieving such land-use goals could be substantial. For the most part, benefits result from avoiding the urban sprawl that is likely i f current trends continue. I f an alternative urban densification scenario is achieved (high density concentrations o f population in the core o f the region and lower density in surrounding areas) there would be numerous social, environmental and economic benefits. For example, the costs o f providing infrastructure such as sewers, water supply, power lines, as well as highways, roads, bridges and bus service can be substantially lower with compact development.1 9 peak period, almost exclusively commuting). Lost non-work time is valued at a fraction of the wage rate. A recent report for MoTH, The Value of Time Savings for the Economic Evaluation of Highway Investments in British Columbia. W.G. Waters, 1992, summarised research in jurisdictions throughout the world. The consensus leads to values below 50% of the wage rate. 1 7 These categories represent current high priority air quality issues for B.C. Conversion factors (from vehicle km. travelled to tonnes) are 2.62 g/vehicle km. travelled for NOx and VOC (combined), 0.11 g/VKMT for particulates, and 313.88 g/VKMT for C 0 2 (based on estimates for 2005 in Table C-6.1 of draft report Clean Air Benefits and Costs in the GVRD Volume 3, 1994). The GVRD report is an initial investigation into this area, estimates for particulates are especially uncertain. Subsequent work will, no doubt, lead to refinements. 1 8 This estimate is based on work-in-progress by Todd Litman, on contract to MoTH Plarining Services. Values per vehicle km. travelled are broken down as follows - greenhouse gases: $ 0.06; particulates: $ 0.04; noise: $ 0.06, resource/energy: $ 0.03, and water and waste: $ 0.015 (based on an urban environment). Additional estimates for land use and barrier affects were not included as these are addressed qualitatively under the urban development account. Diesel buses may be dirtier than automobiles in emission of particulates on a per person basis (see for example Transit and the Environment. BC Transit, 1990). Rapid Buses would likely use diesel (or "clean" diesel) technology at least until cleaner fuel technologies are available. Thus, Rapid Bus is assumed to generate only $ 0.04 /VKMT less of environmental benefits (minus the value of particulates) for the first ten years of the evaluated period. " Methods to quantify the benefits of densification are not well developed. The Transport 2021 report The Costs of Transporting People in the British Columbia Lower Mainland. GVRD/MoTH, 1993 provides estimates on a per vehicle km. travelled basis of the infrastructure, loss of open space and future transportation costs. Attributing such costs to the average automobile would not be appropriate in assessing the with and without scenarios evaluated here., however. A useful summary Rapid Transit - Multiple A c c o u n t Evaluat ion - May 1995 Page 13 165 Though potentially very large, such savings are difficult to measure. Highway cost savings resulting from a reduced need to build bridges and expand highways under a compact metropolitan area scenario have been estimated at as much as $6 billion (NPV) . 2 0 The multiple account evaluation does not include an estimated dollar value for these benefits but instead a qualitative assessments which combines the potential in the corridor with the effectiveness of each technology to assess the contribution to land-use goals (see Figure IV-1) Figure IV-1 - Urban Development Indicators Corridor: potential to achieve regional land-use goals low. medium, high Technology: likelihood of promoting compact development Technology/Corr idor Opt ion: contribution to regional land-use goals low. medium, high For the first indicator, the corridor: potential to achieve regional land-use goals, Broadway-Lougheed is rated as medium/high, Coquidam-New Westrninster as high, and Richmond^ Vancouver as low/medium.21 The technology/corridor measure o f contribution to regional land-use goals adjusts the corridor potential to take account o f the differing effectiveness o f technology in realising growth potential. 2 2 Technology effectiveness is essentially based on of the arguments for and against densification is provided in Resettling Cities. Canadian Residential Intensification Initiatives. CMHC, 1993. 2 0 Highways costs without densification have been estimated by the GVRD at $14 billion (NPV over 25 years at 8%). These estimates are not directly comparable with the scenarios modelled here, nevertheless, they reflect the magnitude of the potential savings. 2 1 The Transport 2021 report A Medium-Range Transportation Plan for Greater Vancouver. GVRD/MoTH, 1993, rates Richmond-Vancouver as having low land-use shaping potential and both the other corridors as high and provides this assessment: "In the projects judgement, investment in the Coquitlam-New Westminster corridor, serving an area which is not only targeted for significantly above-trend growth, but also an area which has low relative accessibility today, would have the greatest land-use shaping impact. The Broadway-Lougheed corridor ranks second in meeting the land-use shaping objective." 2 2 In order to adjust the corridor potential rating to take account of the effectiveness of a particular technology , the potential rating - low, medium, high - is scored on a scale of 0 to 10. The value of 10 represents an ideal corridor in terms of regional land-use goals, zero, one which has no positive attributes. Ratings of low to high were assigned a range of scores thus: 0 to 2 for low, 2 to 4 for low/medium, 4 to 6 for medium, 6 to 8 for medium/high, and 8 to 10 for high. Each corridor was assigned a score in the mid-point of the range for its assessed rating: Broadway-Lougheed has a score of 7 (medium/high), Coquitlam-New Westrninster a score of 9 (high) and Richmond-Vancouver a score of 3 (low/medium). These scores were then combined with probabilities that a particular technology will be effective in generating adequate development (based on assumptions for the Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 14 166 the speed o f transit service and the fact that fixed ICTS investments provide more assurance to developers o f the long term commitment to a particular rapid transit line. System Operations The system operations account considers the operational aspects o f each technology and assesses how readily each can be adjusted for changing demand in the short and long term and how easily each can be integrated into the existing transit system. Assessments o f these considerations were undertaken in each o f the corridor feasibility studies and are summarised in the multiple account evaluation as poor, good or excellent. Analysing Values over Time Rapid transit systems are long term investments. A proper evaluation of options must analyse their performance over an appropriate time period. In this evaluation, annual estimates are expanded over 47 years, from 1995 to 2041. This assumes a life-cycle o f 40 years for ICTS lines (built in 8 years). To compare such streams o f costs and benefits, values in future years are discounted to determine their "present value." A l l multiple account value indicators are also expressed as present values (PV) representing the discounted sum of the stream of values (or net present value o f benefits less costs) over the ful l period. A discount rate o f 8% (net o f inflation) was used to convert the future benefit and cost streams to equivalent present values. Perhaps the most informative indicator is the life-cycle cost per boarding?3 This indicates the amount each person boarding the service over its life would have to be charged in order to recover the total costs o f the project. Cost per boarding is calculated as the ratio o f the present value o f total financial costs for a particular option over the discounted sum (again at 8% over 47 years) o f boardings on that technology. A n ex post analysis o f the existing SkyTrain system yields a life-cycle cost per boarding (in 1994 dollars) of $5.64. 2 4 above taxes/charges estimations) to determine the expected value (score) and thus combined technology/corridor rating - low, high, etc. Rapid Bus has a 0%, probability, Basic LRT is 30%, Enhanced LRT has 75% , and ALRT, 100%. Such probabilities, it should be noted, assume that all other necessary conditions are met (e.g., zoning). 2 3 Costs in this calculation include all capital and operating costs but do not net out bus savings or revenues nor any social benefits. BC Hydro uses a similar "levelised cost" approach to evaluate the costs of new power. Life-cycle Cost per Boarding = PV of Total Costs Discounted sum of boardings Other measures used by BC Transit for route costing compare annual operating costs (and sometimes debt service costs) against one year's ridership. These cannot be compared with the life-cycle cost per boarding estimate in this evaluation and have limited usefulness. 1 4 The SkyTrain analysis includes actual capital expenditures (net of interest) incurred in all years from 1981/82 to 1993/94 and actual operating costs from 1986/87 to 1993/94. Future operating costs are projected to remain constant at the 1993/94 level for the full period (until 2025). Annual boarding numbers from 1986 to the present are used and assumed to grow in the future from the current level (about 30 million per year to a maximum of 35 million). Boarding numbers in the first five years may be overstated, as more recent numbers have been revised downwards. Rapid Transit - Mult iple A c c o u n t Evaluation - May 1995 Page 15 1 6 7 V . R E S U L T S Broadway-Lougheed Corridor Financial Account The financial account results are presented in Table V - l . The estimated capital costs for the Broadway-Lougheed Rapid Bus is $75.7 mill ion. Rapid Bus costs are substantially less than each ICTS option largely because it uses existing infrastructure (roads) while ICTS options require me construction o f a rail line. Basic L R T capital costs are estimated at $548.7 mil l ion, more than seven times Rapid Bus costs. Enhanced L R T adds elevated crossings at f ive key intersections and so its capital costs increase to $574.9 mill ion. A L R T has capital costs estimated at $932.2 million, more than twelve times Rapid Bus costs and 60% higher costs than Enhanced L R T . This large difference is due to the requirement that A L R T have a dedicated right-of-way (mostly elevated) along its entire route. Rapid Bus operating costs are estimated at $7 mill ion annually in 2006. L R T costs are only slightly higher, at $8.7 mill ion (Basic L R T ) and $9 million per year (Enhanced LRT) . Operating costs for A L R T are $15.9 mill ion per year, substantially higher because o f greater Table V - l Broadway-Lougheed Financial Account Values in millions of 1994 dollars; negative values.in brockets () Rapid Basic Enhanced ALRT Bus LRT LRT COSTS (Net of Bus Savings) Capital ($75.7) ($ 548.7) ($ 574.9) ($ 932.2) Operating (annual for 2006) ($7.0) ($ 8.7) ($ 9.0) ($ 15.9) REVENUES Incremental Fares (annual for 2006) $1.7 $1.9 $2.2 $4.2 Taxes/Charges in 2006 $0.0 $3.6 $9.0 $12.0 in 2021 $0.0 $12.3 $30.8 $41.0 PRESENT VALUES COSTS Capital (FV) ($73) ($360) ($377) ($ 612) Operating (PV) ($74) ($61) ($63) ($111) REVENUES Incremental Fares (PV) $ 18 $ 13 $ 16 $30 Taxes/Charges (PV) $0 $5 $12 - $ 15 Net Financial Costs (PV) ($ 129) ($ 403) ($413) ($ 677) Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 16 1 6 8 service frequencies provided in off-peak. 2 5 The incremental fare revenue estimates are based on forecasts for 2006 o f fares from new riders attracted to the transit system by the technology. Rapid Bus generates annual incremental fares o f $ 1.7 million. Estimates for Basic L R T are slightly higher at $ 1.9 mil l ion and for Enhanced LRT, $2.2 million. A L R T generates almost double that amount wi th $4.2 mill ion per year. Potential revenues (collected through benefiting area taxes or development cost charges on increased land value) were estimated for Broadway-Lougheed (with A L R T ) at $12 mil l ion in 2006 and $41 million for 2021. The increase over time reflects the assumption that growth around stations would build momentum over time. The estimate for other options was a percentage o f the A L R T estimates: 75% for Enhanced LRT, 30% for Basic L R T and 0% for Rapid Bus. Discounting and surruning the cash flows generated over the evaluated time period (47 years at 8%) yields net financial costs o f $129 million (PV) for the Rapid Bus option. Basic L R T s net financial costs are three times that at $403 mil l ion (PV). Enhanced L R T net financial costs are only $10 million more at $413 mill ion (PV). A L R T net financial costs are substantially higher at $677 million (PV). Customer Service and Environmental Accounts The customer service and environmental accounts are presented in Table V-2. The demand for each transit service is based largely on the speed, frequency and accessibility o f each option. Ridership on this corridor is spread more evenly over the day than typical transit routes which have a more pronounced peak o f riders travelling in the morning rush-hour. Rapid Bus is estimated to attract annual boardings o f 17.3 mill ion in 2006, o f which, 1.1 mil l ion are new to transit. 2 6 Basic L R T achieves 20.8 mill ion boardings and 1.0 mil l ion incremental boardings; Enhanced L R T has slightly more at 21.6 mill ion boardings and 1.2 mill ion incremental boardings; and A L R T , substantially more at 28.8 million boardings and 2.4 mill ion incremental boardings annually. A L R T has notably high ridership largely because it is ten minutes faster than its closest competitor and because it maintains more frequent service in the off-peak. The base level o f total transit system ridership for 2006 is assumed to be about 148 mil l ion riders, thus incremental ridership adds, at most, under 2%. In addition, there is a risk that slower services, especially Rapid Bus, would not stimulate sufficient growth in the corridor to achieve the modelled ridership results. Also, labour costs are still a major component of ALRT operating costs despite automation (SkyTrain employs over 300). 1 6 Total corridor ridership under the Rapid Bus options are actually higher (about 21 million in Broadway-Lougheed) since local buses also remain in the corridor (they do not with ICTS options) though that added ridership and costs of serving it is not counted here. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 17 1 6 9 T a b l e V - 2 Broadway-Lougheed Customer Service and Environmental Accounts Values in millions of 1994 dollars; negative values in brackets 0 Rapid Basic Enhanced ALRT Bus LRT LRT RIDERSHIP (millions in 2006) Total Boardings 17.3 20.8 21.6 28.8 Incremental Boardings 1.1 1.0 1.2 2.4 TIME SAVINGS Route Travel Time (minutes in 2006) 40 37 35 25 Route Time Saved (compared to current 57 minutes) 17 20 22 32 REDUCED AUTO EMISSIONS (tonnes in 2006) Reduced NOx and VOC 361 321 401 751 Reduced Particulates 1.51 1.31 1.71 3.21 Reduced CO2 43281 38371 43101 90161 PRESENT VALUES — , , Travel Time Savings (PV) $61 $84 $117 $151 Environmental Benefits (PV) $28 $18 $22 $42 Quantified Social Benefits (PV) $89 $101 $139 $193 The average route travel time from Lougheed Mal l to Arbutus today is 57 minutes (excluding transfer time). I t is assumed in the base case that planned service enhancements (the introduction o f an express bus) wi l l reduce that time to 45 minutes. 2 7 Rapid Bus is estimated to achieve a route travel time o f 40 minutes (assuming significant transit priority measures are in place). Basic L R T is three minutes faster at 37 minutes and Enhanced L R T two minutes faster than that at 35 minutes. A L R T is estimated to more than halve the current time to 25 minutes. Reduced auto emissions are roughly equivalent for Rapid Bus and both LRTs, but double that for the ALRT. To put these figures in context, emission reductions for the various options represent between 0 . 1 - 0.2% o f the projected total annual emissions for the Lower Mainland in 2006. This small reduction is not surprising, however, given the relatively small part traffic in one corridor would constitute o f all emission sources in the region. Both the travel time savings for existing transit riders and the environmental benefits o f reduced auto use are valued in dollar terms. Sunnming the estimated value o f travel time savings and environmental benefits, the quantified social benefits equal $89 million (PV) for Rapid Bus. Basic L R T quantified social benefits are slightly higher at $101 million (PV); Enhanced L R T s social benefits equal $139 mil l ion (PV). The A L R T achieves social benefits equal to $193 million (PV). All EMME/2 results (except corridor ridership) are incremental numbers derived by subtracting base case results from scenario results. Current time is used in this discussion to provide readers with a more easily comparable reference point. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 18 1 . 7 0 Urban Development and System Operations Accounts The Broadway-Lougheed corridor is characterised by a heavily-developed western section (especially between Cambie and Granville) and a relatively under-developed (residentially) eastern section along the Lougheed Highway in Burnaby. Thus, the corridor has medium to high potential to achieve regional land-use goals. The contribution to regional land-use goals of the options are rated: low for Rapid Bus, low/medium for Basic LRT, medium for Enhanced LRT and medium/high for A L R T . Operational aspects of Rapid Bus are favourable. It is highly flexible and can be readily adjusted to changes in demand (though its ultimate capacity is much lower than any ICTS). It could be easily integrated into present BC Transit operations (due to existing expertise with bus maintenance). A l l ICTS lines require fixed guideways and hence cannot be re-routed without significant expense. Additional capacity can be acquired for each, although the cost of A L R T cars may be significantly higher (SkyTrain technology has only one supplier). LRTs would not be integrated into BC Transit operations as easily as the A L R T (because of synergy with SkyTrain) or Rapid Bus. T a b l e V - 3 Broadway-Lougheed Urban Development and System Operations Accounts Rapid Basic Enhanced ALRT Bus LRT LRT DEVELOPMENT Corridor Potential to Achieve Regional Land-use : Medium/High (all options) Goals Contribution to Regional Land-Use Goals Low Low/Medium Medium Medium/High OPERATIONS Flexibility in Adjusting Capacity Good Poor Poor Poor Ease of Integration into Transit System Excellent Poor Poor Good Broadway-Lougheed Summary A surnmary of key multiple account indicators for Broadway-Lougheed options is presented in Table V-4. The Rapid Bus option has net financial costs of $129 million (PV) and net social costs of $41 million (PV). It is forecast to attract 17.3 million boardings in 2006 but is assumed to have a low contribution to regional land-use goals. With a life-cycle cost per boarding of $0.98, Rapid Bus is the most cost-effective transit service. R a p i d Transi t - M u l t i p l e A c c o u n t E v a l u a t i o n - May 1995 P a g e 19 Table V - 4 Broadway-Lougheed Multiple Account Summary All values in 1994 dollars Rapid Basic Enhanced ALRT Bus LRT LRT Net Financial Costs (millions, PV) ($ 129) ($ 403) ($ 413) ($ 677) Quantified Social Benefits (millions, PV) $89 $101 $ 139 $ 193 Net Social Costs (millions, PV) ($41) ($302) ($ 274) ($ 484) Boardings in 2006 (millions) 17.3 20.8 21.6 28.8 Contribution to Regional Land-Use Goals Low Low/Medium Medium Medium/High Life-cycle Cost per Boarding $ 0.98 $3.26 $3.26 $ 3.83 The Basic L R T option has net financial costs of $403 million (PV) and net social costs of $302 million (PV). 20.8 million boardings are forecast on this service in 2006. Basic L R T is assumed to make a low to medium contribution to regional land-use goals. It has a cost per boarding of $3.26. The Enhanced L R T option has net financial costs slightly higher than the Basic L R T option at $413 million (PV) but lower net social costs at $274 million (PV) because of 40% higher time savings benefits. Enhanced L R T is forecast to have 21.6 million boardings in 2006. It is assumed to have an medium contribution to regional land-use goals. The cost per boarding estimate of $3.26 is equal to that for the Basic L R T option because additional ridership offsets additional costs. Given superior net social costs, ridership and land-shaping potential and equivalent cost per boarding, Enhanced L R T is superior to Basic LRT. The A L R T option has net financial costs of $677 million (PV) and net social costs 75% higher than the Enhanced LRT option at $484 million (PV). It attracts substantially higher ridership with boardings of 28.8 million. It is assumed to make a medium to high contribution to regional land-use goals. The cost per boarding of $3.83 is highest in the corridor. In summary, the Broadway-Lougheed corridor has high ridership sufficient to provide an attractive market for a number of the rapid transit options. Its best ICTS option has medium/high potential to contribute to regional land-use goals particularly on the eastern portion of the route. R a p i d Transit - M u l t i p l e A c c o u n t E v a l u a t i o n - May 1995 P a g e 20 172 Coquitlam-New Westminster Corridor Financial Account The financial account results for Coquitlam-New Westminster option are displayed in Table V-5. Engineering cost estimates place capital costs o f Rapid Bus at $59.9 million and annual operating costs for 2006 at $4 million. Enhanced L R T capital costs are almost ten times Rapid Bus at $556.3 mill ion; A L R T capital cost are fourteen times greater at $810 million. ICTS costs are high on this corridor. Enhanced L R T requires some grade separation (tunnels and elevated guideway) to deal with gradient and limited right-of-way. A L R T requires extensive grade separation (including tunnels). Operating costs for Enhanced L R T ($13 mill ion per year) and A L R T ($16 mill ion per year) are also higher than Rapid Bus but by a lesser proportion than capital. Incremental fare revenue estimates (reflecting new riders drawn to the transit system) are low for both Rapid Bus and Enhanced L R T at $ 0.4 million per year. A L R T is more successful and secures $2 million in additional fares in 2006. Estimates o f potential revenues from taxes or charges are higher for Coquitlam-New Westminster than the other corridors because o f the relative lack o f development o f adjacent land (and thus greater potential for development). The A L R T is estimated to generate potential revenues from taxes/charges o f $61 million in 2006 and $38 million in 2021. Rapid Bus is assumed to generate no taxes/charges and Enhanced LRT generates 75% A L R T levels. Summing and discounting the above cash flows described above yields a net financial cost o f $92 million (PV) for Rapid Bus, $425 million (PV) for Enhanced L R T and $593 million (PV) for ALRT. Table V - 5 Coquitlam-New Westminster Financial Account Values in millions of 1994 dollars; negative values in brackets () Rapid Enhanced ALRT Bus LRT COSTS (Net of Bus Savings) -Capital ($59.9) ($ 556.3) ($ 810.0) Operating (annual for 2006) ($ 4.0) ($ 13.0) ($ 16.0) REVENUES Incremental Fares (annual for 2006) $0.4 $0.4 $2.0 Taxes/Charges in 2006 $0.0 $45.8 $61.0 in 2021 $0.0 $ 28.5 $38.0 PRESENT V A L U E S COSTS Capital (PV) ($54) > ($ 367) ($ 535) Operating (PV) ($ 42) ($ 90) ($111) REVENUES Incremental Fares (PV) $4 $3 $ 14 Taxes/Charges (PV) $0 $30 $39 Net Financial Costs (PV) ($ 92) (S 425) ($ 593) Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 2 1 173 Customer Service and Environmental Accounts The customer service and environmental account results are presented in Table V-6. The annual boardings for Rapid Bus in 2006 are low at 4.3 million, o f which only 0.3 mill ion are new to transit. The relatively low boardings level is due in part to competing bus services fol lowing alternate routes (e.g. direct service between Coquitlam and Surrey) which would capture about 40% o f corridor traffic. Enhanced L R T boardings are three times higher, but stil l low, at 12.2 million, however, the incremental boardings are as low, at 0.3 mill ion, as Rapid Bus. The A L R T option attracts 16.5 mill ion boardings and 1.3 mil l ion incremental boardings. Incremental ridership for these options adds, at most, under 1%. to the total transit system ridership (which is assumed to be about 148 million riders in 2006). Moreover, for projections to materialise the corridor wi l l have to attract substantially more growth than the other two. Thus, the risk is highest in the Coquitlam-New Westrninster corridor that forecast ridership would not materialise. The route travel time for Rapid Bus is estimated at 40 minutes, Enhanced L R T is 26 minutes and A L R T 22 minutes. The average current travel time and projected base case time for this route are both 52 minutes as no enhancements are otherwise planned. Rapid Bus emission reductions are one eighth o f A L R T reductions. Surprisingly, E M M E / 2 forecasts that Enhanced L R T would actually cause a slight increase in vehicle kilometres travelled (and thus increased auto emissions). This result could be due to the removal o f local bus service and resulting loss o f transit riders (to automobiles) who drive greater distances in total than those attracted out o f their vehicles. However, given the imprecision o f such forecasts, a zero or low positive amount is quite possible. A L R T is estimated to reduce auto Table V - 6 Coquitlam-New Westminster Customer Service and Environmental Accounts Values in millions of 1994 dollars: negative values in brackets () RIDERSHIP (millions in 2006) Total Boardings Incremental Boardings TIME SAVINGS Route Travel Time (average minutes in 2006) Route Time Saved (compared to current 52 minutes) REDUCED AUTO EMISSIONS (tonnes in 2006) Reduced NOx and VOC Reduced Particulates Reduced C 0 2 ' Rapid Enhanced ALRT Bus LRT 4.3 12.2 16.5 .3 .3 1.3 40 26 22 12 26 30 9t -8t 671 .4t -.3t 2.81 10681 -930 t 80811 PRESENT VALUES Travel Time Savings (PV) $ 15 $ 6 $ 9 Environmental Benefits (PV) $ 7 4 ) $ 3 7 Quantified Social Benefits (PV) $22 $2 $47 Rapid Transit - Mult ip le A c c o u n t Evaluat ion - May 1995 Page 22 174 emissions by seven times more than Rapid Bus. The significance of these amounts in the Lower Mainland is quite small ranging from 0.02 - 0.18 % of total regional emissions. The sum of the PV of travel time savings and environmental benefits for Rapid Bus is $22 million (PV). Enhanced L R T social benefits are quite low at $2 million (PV), while A L R T has benefits of $47 million (PV), more than twice the Rapid Bus amount. Urban Development and System Operations Accounts Summary qualitative assessments for the urban development and system operations accounts are presented in Table V-7. The Coquitlam-New Westminster corridor has high potential to achieve regional land-use goals as it traverses the most under-developed area in any corridor. It is the most effective corridor in connecting regional town centres. The current development in the su iTounding area, however, is unfocused, having highly dispersed travel patterns. This creates some uncertainty whether development would coalesce into a compact pattern. Moreover, the corridor is the only one that is not oriented towards the core of the Burrard Peninsula, but instead traverses its eastern boundary. Thus, some of the development created by rapid transit in this corridor could occur outside the compact metropolitan area. T a b l e v - 7 Coquitlam-New Westminster Urban Development and System Operations Accounts Rapid Enhanced ALRT Bus LRT DEVELOPMENT Corridor. Potential to Achieve Regional Land-use Goals High (all options) Contribution to Regional Land-Use Goals Low Medium/High High OPERATIONS Flexibility in Adjusting Capacity Good Poor Poor Ease of Integration into Transit System Excellent Poor Good In terms of the contribution to regional land-use goals, the Rapid Bus option is rated low, Enhanced L R T is rated medium/high, and A L R T is rated high. Such ratings are based on the optimistic assumption that all other necessary conditions will apply (zoning, etc.). While Coquitlam-New Westrninster has a high potential it too has the greatest risk given its relative lack of development at present. The operational aspects of each option are essentially the same as with Broadway-Lougheed options. Rapid Bus is best in terms of flexibility as buses can be added or removed and routing changed easily and at low cost. Enhanced L R T is less flexible. Additional cars can be aGquired at competitive prices, however, changes to the system are generally more restricted. A L R T is even less flexible with potentially high costs for additional cars. Rapid Bus is most easily integrated in BC Transit operations, given the corporation's existing facil i t ies and expertise applies to largely identical technology. A L R T (again, if SkyTrain technology were used) would be easier to integrate than Enhanced LRT, due to existing expertise and opportunities to link to SkyTrain in New Westminster. R a p i d Transit - Mult iple A c c o u n t Evaluat ion - May 1995 P a g e 2 3 175 Coquitlam-New Westminster Summary The summary o f key multiple account results for Coquitlam-New Westminster is presented in Table V-8. The Rapid Bus option has net financial costs o f $92 million and net social costs o f $70 million (PV). I t is forecast to attract 4.3 mill ion boardings in 2006. I t is assumed to make a low contribution to regional land-use goals. A cost per boarding o f $2.41 is estimated. The Enhanced L R T option has net financial costs o f $425 mill ion (PV) and, wi th virtually no quantified social benefits, net social costs o f $424 mil l ion (PV). The Enhanced L R T is forecast to attract 12.2 mil l ion boardings. I t is assessed as having a medium to high contribution to regional land-use goals. The life-cycle cost per boarding for this option is estimated at $5.39. The A L R T option has net financial costs o f $593 mill ion (PV) and net social costs equal to Table V - 8 Coquitlam-New Westminster Multiple Account Summary All values in 1994 dollars Rapid Enhanced ALRT Bus LRT Net Financial Costs (millions, FV) ($92) ($425) ($ 593) Quantified Social Benefits (millions, PV) $22 $2 $47 Net Social Costs (millions, PV) ($70) ($ 424) ($547) Boardings in 2006 (millions) 4.3 12.2 16.5 Contribution to Regional Land-Use Goals Low Medium/High High Life-cycle Cost per Boarding $2.41 $5.39 $5.60 $547 million (PV). I t is forecast to have boardings o f 16.5 mil l ion in 2006. I t is assumed to have high potential to contribute to regional land-use goals. The cost per boarding is $5.60. Ridership forecasts for this corridor are the most uncertain. This corridor is the least developed today and for desired growth to materialise it w i l l have to attract as much growth as the other two combined. The relatively high cost per boarding for the Rapid Bus alternative (compared to Rapid Bus in other corridors) suggests that a somewhat scaled-down service might be more cost-effective in the short term. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 2 4 1 7 6 Richmond-Vancouver Corridor Financial Account The financial account for Richmond-Vancouver is presented in Table V-9. Capital costs for Rapid Bus are estimated at $96.6 million. Arbutus Enhanced L R T capital costs are estimated at $809.2 million. Cambie Basic L R T capital costs are $536.2 million, Cambie Enhanced L R T are almost twice that at $1,067.2 million. This substantial increase is due to the large amount o f tunnelling required both in downtown Vancouver and along Cambie due to surface right-of-way and grade constraints. Cambie A L R T is slightly less expensive at $1,031.2 mill ion (because o f smaller tunnels and opportunities to connect with the existing SkyTrain). Estimates o f operating costs vary much less. Rapid Bus operating costs are estimated at $5.1 mill ion per year. Arbutus Enhanced L R T operating costs are highest at $12 mill ion per year. Cambie Basic L R T costs are $8.6 million, Enhanced L R T , $10.9 mill ion and A L R T , $8.9 mill ion. Incremental fare revenues for Rapid Bus are estimated at $ 0.9 million in 2006. Wi th its high additional ridership, Arbutus Enhanced LRT generates more than six times that at $5.5 mill ion. Cambie Basic L R T collects only $1.8 mil l ion in incremental fares. Cambie Enhanced L R T brings in $6.5 in incremental fares and Cambie A L R T adds $7.4 million per year. Potential revenues from land taxes/charges are estimated for A L R T in 2006 at $27 mill ion and in 2021, $22 million. Estimates for other technologies are prorated at 75% for Cambie Enhanced L R T and Arbutus Enhanced LRT, 33% for Cambie Basic LRT, and no revenues Table V - 9 Richmond-Vancouver Financial Account Values in millions of 1994 dollars; negative values in brackets () Granville Arbutus Cambie Cambie Cambie Rapid Enhanced Basic Enhanced ALRT Bus LRT LRT LRT COSTS (Net of Bus Savings) Capital ($ 96.6) ($ 809.2) ($ 536.2) ($1067.2) ($1031.2) Operating (annual for 2006) ($ 5.1) ($12) ($ 8.6) ($ 10.9) ($ 8.9) REVENUES Incremental Fares (annual for 2006) $0.9 $5.5 $ 1.8 $6.5 $7.4 Taxes/Charges in 2006 $0 $8.1 $8.1 $20.3 $27 in 2021 $0 $6.6 $6.6 $16.5 $22 PRESENT VALUES COSTS Capital (PV) ($ 90) ($ 534) ($ 351) ($ 707) ($ 678) Operating (PV) ($ 54) , ($83) ($60) ($ 76) ($ 62) REVENUES Incremental Fares (PV) $9 $39 $ 13 $46 $53 Taxes/Charges (PV) $0 $6 $6 $ 14 $ 19 Net Financial Costs (PV) ($ 135) ($ 564) ($ 393) ($723) ($ 668) Rapid Transit - Multiple A c c o u n t Evaluat ion - May 1995 Page 25 1 7 7 for Rapid Bus. The reduced revenues in the later period is expected because the Richmond would experience growth fairly quickly (benefiting from zoning tn place) but be limited by its less capacity to accept significant growth. In terms o f discounted net present value, Rapid Bus is lowest cost with net financial costs o f $ 135 million (PV). Arbutus Enhanced L R T is more than four times that amount at $564 mill ion (PV). The Cambie Basic L R T costs are substantially lower than Arbutus Enhanced L R T with net financial costs o f $393 mill ion (PV). The Cambie Enhanced L R T has substantially higher net financial costs at $723 million (PV), largely because o f high tunnelling costs. Because it also requires grade separation, Cambie A L R T has net financial costs o f similar magnitude, though somewhat lower at $668 million (PV). Customer Service and Environmental Accounts Results o f the customer service and environmental accounts are presented in Table V-10. Annual ridership in 2006 for Rapid Bus is estimated at 9 million, o f which, 0.6 mill ion boardings are incremental to the transit system. Arbutus Enhanced L R T would carry 15 mill ion riders (in part due to its accessibility to the large west side Vancouver population) and attract 2.3 mill ion incremental boardings. Cambie Basic L R T is estimated to have 10.4 mill ion riders and 1.1 million incremental boardings, only slightly more than Rapid Bus (because it is no faster). Cambie Enhanced L R T attracts 17 million annual riders and 3.7 mil l ion incremental riders. The A L R T achieves a level o f 20.8 mill ion annual boardings and 5.2 mill ion incremental riders. Incremental ridership is highest on this corridor adding up to 3.5% to the total transit system ridership o f 148 mill ion riders in 2006. The route travel time for Rapid Bus is estimated at 33 minutes, saving 4 minutes of f the Tablev-io Richmond-Vancouver Customer Service and Environmental Accounts Values in millions of 1994 dollars; negative values in brackets () Granville Arbutus Cambie Cambie Cambie Rapid Enhanced Basic Enhanced ALRT Bus LRT LRT LRT RIDERSHIP (millions in 2006) Total Boardings 9 15 10.4 17 20.2 Incremental Boardings .6 2.3 1.1 3.7 5.2 TIME SAVINGS Route Travel Time (average minutes in 2006) 33 30 32 24 19 Route Time Saved (compared to current 37 minutes) 4 7 5 13 18 REDUCED AUTO EMISSIONS (tonnes in 2006) Reduced NOx and VOC 19 t 1711 241 200 t 2281 Reduced Particulates .8t 7.2 t 1 t 8.4 t 9.6 t Reduced C 0 2 2251 1 20467 t 2932 t 23994 t . 27259 t PRESENT VALUES Travel Time Savings (PV) $32 $64 $44 $80 $95 Environmental Benefits (PV) $ 14 $95 $ 14 $ 1 1 1 $ 126 Quantified Social Benefits (PV) $46 $159 $58 $191 $221 Rap id Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 26 1 7 8 current 37 minute route time. Cambie Basic L R T does little better, achieving route times o f 32 minutes respectively (because of time spent in mixed traffic). Arbutus Enhanced L R T covers a longer route in a slightly faster time of 30 minutes. Cambie Enhanced L R T , however, doubles these time savings and achieves an average route time o f 24 minutes (13 minute savings). Cambie A L R T saves a further 5 minutes, halving the current time to an average of 19 minutes. A number o f the Richmond-Vancouver options (the Enhanced LRTs and the A L R T ) are quite effective in attracting new riders out o f their cars (with average distances o f foregone car travel high as well). Reduced auto emissions, however, still are relatively small ranging from 0.05 - 0.6 % o f total emissions for the Lower Mainland. The quantified social benefits for Rapid Bus is $46 million (PV). Arbutus Enhanced LRT's social benefits equal $159 million (PV) with $95 million o f that due to environmental benefits. Cambie Basic L R T has much lower social benefits o f $58 million (PV). Cambie Enhanced L R T social benefits are closer to the Arbutus level but higher at $191 mil l ion (PV). The Cambie A L R T option has time savings and environmental benefits o f $221 mil l ion (PV). Urban Development and System Operations Accounts Summary qualitative assessments of the development and operational aspects o f each option are presented in Table V - l 1. The Richmond-Vancouver corridor has low to medium potential for further development because land along the route is almost ful ly subscribed or, where available, is environmentally sensitive. Land values are high and zoning restricts high density development (the Livable Region Strategy seeks to limit growth in Richmond). The north-south f low o f the corridor could also encourage development in South Delta, outside the targeted compact area. Combining technology effectiveness and corridor potential provides the assessment o f the contribution to regional land-use goals o f each option. The Granville Rapid Bus option is rated low as is the Cambie Basic L R T option. The two Enhanced L R T and the A L R T options are all rated low/medium. Thus, each Richmond-Vancouver option has limited likelihood o f contributing to regional land-use goals. Operationally, the Richmond-Vancouver options share attributes o f other evaluated Table V - 1 1 Richmond-Vancouver Urban Development and System Operations Accounts Granville Arbutus Cambie Cambie Cambie Rapid Enhanced Basic Enhanced ALRT Bus LRT LRT LRT DEVELOPMENT Corridor Potential to Achieve Regional Land-use Low/Medium (all options) Goals Contribution to Regional Land-Use Goals Low Low/Medium Low Low/Medium Low/Medium OPERATIONS Flexibility in Adjusting Capacity Good Poor Poor Poor Poor Ease of Integration into Transit System Excellent Poor Poor Poor Good Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 27 179 technologies. Rapid Bus is quite flexible in terms of adjusting to capacity and excellent in the ease of its integration into the BC Transit system. Each of the LRT options is considered poor in terms of both flexibility and the ease of integration in the existing transit system. Cambie ALRT is also poor in terms of flexibility but good in terms of integration (if ALRT technology were compatible with SkyTrain). Richmond-Vancouver Summary The summary multiple account results for Richmond-Vancouver are presented in Table V-12. Rapid Bus has net financial costs of $135 million (PV) and net social costs of $89 million (PV). It has 9 million boardings in 2006 and has low potential to contribute to regional land-use goals. The life-< c^le cost per boarding equals $1.88. The Arbutus Enhanced LRT option has net financial costs of $564 million (PV) and net social costs more than four times those of the Rapid Bus option at $405 million (PV). It is forecast Table V - 12 Richmond-Vancouver Multiple Account Summary All values in 1994 dollars Granville Rapid Bus Arbutus Enhanced LRT Cambie Basic LRT Cambie Enhanced LRT Cambie ALRT Net Financial Costs (millions, PV) Quantified Social Benefits (millions, PV) ($135) $46 ($564) $159 ($393) $58 ($723) $191 ($ 668) $221 Net Social Costs (millions, PV) ($89) ($405) ($335) ($532) ($447) Boardings in 2006 (millions) 9 15 10.4 17 20.2 Contribution to Regional Land-Use Goals Low Low/Medium Low Low/Medium Low/Medium Life-cycle Cost per Boarding $1.88 $6.30 $8.29 $6.94 $5.54 to attract 15 million boardings but has a low/medium contribution to regional land-use goals. Its cost per boarding is $6.30. The Cambie Basic LRT option has lower net financial costs than the Arbutus Enhanced LRT option at $393 million (PV). Net social costs are $335 million (PV). Cambie Basic LRT achieves only 10.4 million boardings, however, and has low contribution to regional land-use goals. Cost per boarding, however, is virtually the same as the Arbutus option at $6.29. The Cambie Enhanced LRT option has substantially higher net financial costs of $723 million (PV), net social costs at $532 million (PV) and has 17 million boardings in.2006. It is assumed to make a low to medium contribution to regional land-use goals. It has the highest cost per boarding of any option at $6.94. The Cambie ALRT option has somewhat smaller net financial costs of $668 million (PV), net social costs of $447 million (PV) and attracts 20.2 million boardings. It too makes a low to Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 28 180 medium contribution to regional land-use goals. I t has the lowest cost per boarding o f any Richmond-Vancouver ICTS option at $5.54. In summary, the Richmond-Vancouver corridor has proven high ridership, but relatively high cost ICTS options and low potential to achieve regional land-use goals. Rap id Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 29 Comparison Between Corridors: Potential for ICTS The second part of the multiple account evaluation is presented below. Here, the corridors themselves are compared to determine in which, i f any, the first ICTS line should be built. While in the long run, ICTS is envisioned for each corridor, only one ICTS option could be undertaken in the medium term, because of the substantial costs. For this comparison, preferred ICTS option(s) in each corridor have been selected based on their performance in the preceding assessment. For Broadway-Lougheed, the Enhanced L R T and the A L R T option are selected. For Coquitlam-New Westminster also, the Enhanced L R T and the A L R T options are selected. For Richmond-Vancouver, only the Cambie A L R T option is selected as its cost per board is significantly lower than any other and it has the highest ridership. For those corridors in which more than one option was selected, the performance of each defines the range for an optimal ICTS in the corridor. A summary of this comparison is presented in Table V-13. Table v-13 Comparison Between Corridors Assessments relative to other corridors Broadway-Lougheed Enhanced LRT/ALRT Coquitlam-New Westminster Enhanced LRT/ALRT Richmond-Vancouver Cambie ALRT Contribution to Regional Land-use Goals Medium to Medium/High Medium/High to High Low/Medium Transit Ridership High Low/Medium Medium/High Boardings in 2006 21.6 to 28.8 million 12.2 to 16.5 million 20.2 million Cost-Effectiveness High Low Low Cost per Boarding $3.26 to $3.89 $5.39 to $5.60 $5.54 Net Social Costs (PV) $274 to $484 million $424 to $547 million $447 million The key attributes of each corridor's option include the contribution to regional land-use goals, transit ridership and cost-effectiveness. For each an assessment (low, medium, high) is provided based on either a previous rating (for land-use) or the comparison of selected indicators (boardings, cost per boarding, and net social costs). The environmental objective of transit (reducing auto use) is not dirninished by its exclusion from this step in the evaluation. The preceding analysis has shown that while environmental benefits are large, they do not come close enough to justify any ICTS option. Environmental benefits, moreover, are reflected in the net social cost indicator included under cost-effectiveness. The issues addressed in this part of the evaluation are those that remain critical in the final decision. Rapid Transit - Multiple A c c o u n t Evaluat ion - May 1995 Page 30 1 8 2 Contribution to Regional Land-Use Goals The degree to which ICTS options would contribute to regional land-use goals is critical in the selection of a preferred ICTS corridor. In fact, it is clear from the preceding analysis that without the benefits of land-use shaping, no ICTS would be warranted in the medium term. Rapid Bus, or conventional bus service would suffice. The assessment of the contribution to regional land-use goals for each selected ICTS option is as follows. The Broadway-Lougheed Enhanced L R T would make a medium contribution and A L R T a medium/high contribution. The Coquitlam-New Westminster options do somewhat better. The Enhanced L R T is rated medium/high and the A L R T , high. In Richmond-Vancouver, on the other hand, the Cambie A L R T would make only a low/medium contribution. Clearly, the competition on this issue is between options in the Coquitlam-New Westminster and Broadway-Lougheed corridors, with the former being somewhat superior to the latter. Success in achieving regional land-use goals is especially important in determining the success of the rapid transit services in the long run. To the degree that compact development occurs, the transit service will benefit from increased ridership from concentrated population along the corridor. The opposite is also true, low densities will result in lower ridership. This circular relationship, however, is not fully reflected in the demand forecasting which assumes the same land-use pattern for all types of the transit service. The risk that growth will not occur as assumed (and thus ridership will not materialise at forecast levels) is highest in the Coquitlam-New Westminster corridor. Transit Ridership For transit ridership as well, two corridors are closely matched. In this case, however, the two are Broadway-Lougheed and Richmond-Vancouver. Broadway-Lougheed's Enhanced L R T option would attract 21.6 million boardings in 2006, the A L R T option would attract 28.8 million boardings. Richmond-Vancouver's Cambie A L R T option would attract 20.2 million riders. The Richmond-Vancouver corridor has higher peak-hour ridership but the Broadway-Lougheed corridor has more ridership overall because of a more even distribution of transit use over the day. Thus, in terms of ridership, Broadway-Lougheed is the superior corridor (and rated high) closely followed by Richmond-Vancouver (rated medium/high). Relative to the others, the Coquitlam-New Westminster corridor has low/medium transit ridership. The Enhanced L R T option would attract 12.2 million boardings in 2006, while the A L R T option would attract 16.5 million boardings. Cost-Effectiveness The cost-effectiveness of the selected ICTS options can be evaluated in two ways. The lifecycle cost per boarding measures the cost-effectiveness of the option in terms of its primary transportation purpose. Net social costs, on the other hand, can be viewed as the cost of achieving the broader land-use shaping goals. In terms of lifecycle cost per boarding, Broadway-Lougheed options clearly dominate with Enhanced L R T at $3.26 and A L R T at $3.89. Coquitlam-New Westminster and Richmond-Vancouver options are significantly more expensive. Coquitlam-New Westminster Enhanced Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 31 183 L R T is $5.39, A L R T is $5.60. Richmond-Vancouver's Cambie A L R T has a cost per board o f $5.54. BC Transit can expect to collect fares averaging less than a dollar f rom each boarding (given current fare levels, and taking account o f transfers, faresavers, etc.). The cost per boarding estimates provide a quick indication o f the magnitude o f subsidy required for these options. In terms o f net social costs, the Broadway-Lougheed Enhanced L R T has the lowest net social cost at $274 million. The A L R T option on Broadway-Lougheed has much higher net social costs o f $484 million. Net social costs for Coquitlam-New Westminster's Enhanced L R T are $424 million; for its A L R T alternative, $547 mil l ion. The Cambie A L R T has net social costs o f $447. In terms o f cost-effectiveness, Broadway-Lougheed is rated high, while both Coquitlam-New Westminster and Richmond-Vancouver are rated low/medium. In summary, Broadway-Lougheed is second to Coquitlam-New Westminster in terms o f regional land-use goals (medium to medium/high) and first in terms o f ridership (high)and cost-effectiveness (high). Coquitlam-New Westminster is the best corridor in terms o f regional land-use goals (medium/high to high) but has low/medium transit ridership and low/medium cost-effectiveness. Richmond-Vancouver has low/medium value in serving regional land-use goals. It has medium/high transit ridership, though less than Broadway-Lougheed. And, it has low/medium cost-effectiveness. Conclusions A number o f general conclusions can be drawn from this evaluation. o The potential revenues from taxes or charges levied on development surrounding rapid transit stations are relatively small compared to the high costs o f ICTS investments. Social benefits from travel time savings and reduced auto use can be up to $200 million (PV). However, they are not sufficient to just i fy ICTS alternatives wi th net financial costs in the range o f $400 to $700 million (PV). o Rapid Bus would provide the most cost-effective transit service in each corridor, but it would be the least effective in promoting compact development and its ridership levels may be overstated as a result. ICTS options can only be justif ied i f they contribute sufficient benefits in achieving regional land-use goals. o Because o f low/medium value in contributing to regional land-use goals and high net financial and social costs, investment in Richmond-Vancouver ICTS does not appear to be warranted in the medium term. o Coquitlam-New Westminster has an advantage over Broadway-Lougheed in terms o f supporting regional land-use goals. However, it is less cost-effective due to its low ridership and faces a higher risk o f failing to promote development and generate its forecast ridership levels. -o The Broadway-Lougheed Enhanced LRT has the lowest lifecycle cost per boarding of any ICTS option, lowest net social cost, high ridership o f over 21 mill ion per year and would make a medium contribution to regional land-use goals. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 32 1 8 4 A P P E N D I X : S E N S I T I V I T Y A N A L Y S I S The sensitivity o f the multiple account evaluation results are tested against changes in two key assumptions: ridership and discount rate. For both, the life cycle cost per boarding figure is used as a summary indicator. 2 8 Ridership There is considerable uncertainty associated with ridership projections over ten years into the future. Sensitivity analysis was undertaken within the E M M E / 2 model itself to test assumptions regarding levels o f T D M (transportation demand management), land use and peak-hour to day expansion factors. The resulting impacts on ridership were in 10 - 2 0 % range. A broader range o f variation is tested here. Table A - l (and Figure A - l ) present the results o f sensitivity analysis comparing reductions -and increases o f 2 5 % for ridership levels. A capacity constraint o f 1 1 0 % o f the 2 0 0 6 estimate T a b l e A -1 Sensitivity Analysis -Cost per Boarding Values in millions of 1994 dollars Ridership - 25% and + 25% -25% Unchanged +25% Broadway-Lougheed Rapid Bus $1.09 $ 0.98 $ 0.93 Basic LRT $ 3.95 $3.26 $3.06 Enhanced LRT $ 3.95 $3.26 $3.07 ALRT $4.64 $3.83 $3.60 Coquitlam-New Westminster Rapid Bus $ 2.72 $2.41 $2.28 Enhanced LRT $6.54 $5.39 $ 5.07 ALRT $ 6.60 $ 5.60 $ 5.27 Richmond-Vancouver Rapid Bus $2.13 $ 1.88 $ 1.78 Arbutus Enhanced LRT $ 7.65 $6.30 $5.93 Cambie Basic LRT $7.63 $ 6:29 $ 5.92 Cambie Enhanced LRT $8.41 $ 6.94 $6.53 Cambie ALRT $6.71 $5.54 $5.21 Multiple account evaluation presents a range of information in both disaggregated and aggregated form making the significance of each element readily apparent. Sensitivity analysis may be more necessary with conventional benefit-cost analysis, the relative importance of each underlying parameter may be not be obvious in the summary benefit-cost ratio or NPV figure. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 33 155 w o o Figure A-l Sensitivity Analysis Ridership $10 $9 + I W § $7 m ai $ 6 + Q. $5 + $4 « -$ 2 - -$1 -$0 X Less 25% • Unchanged - P l u s 25% Richmond-Vancouver Coqurtlam-New Westminster Broadway-Lougheed m 2 ? H 1 1 H (reached in 2011) was assumed in the original analysis. The increase in 25% has the effect of making ridership levels reach this maximum sooner: in 2002 for Rapid Bus and 2004 for ICTS. The reverse is also true. A 25% decrease in ridership extends until 2024, the time it takes the service to reach capacity. The incremental changes (increased cost per board) are greater for 25% reductions than increases because all ridership is capped after growing to 110% of the 2006 level.29 The most noteworthy result from this sensitivity analysis is that the advantage of Broadway-Lougheed over the other two corridors is not lost even if ridership were 25% less in Broadway-Lougheed and 25% more for Coquitlam-New Westminster or Richmond-Vancouver. Discount Rates The assumed discount rate for the analysis was 8% (net of inflation). For this sensitivity analysis, 6% and 10% rates were also evaluated (see Table A-2). Given all ICTS options share the same assumed time profiles of costs and benefits, changes in discount rates affect each proportionately and do not change their performance relative to one another. The relative performance of Rapid Bus options to ICTS options is affected, however. The lower 2 9 A more thorough re-evaluation of projected land-use patterns would provide some insight about the dependence of results on land-use and ridership assumptions. Such would require considerable effort in re-<»niiguring the EMME/2 model, so the more direct adjustment of ridership levels is employed instead. Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 34 186 Table A - 2 Sensitivity Analysis - Discount Rates Cost per Boarding Values in millions of 1994 dollars 6 % 8% 10% Broadway-Lougheed $ 1.06 Rapid Bus $0.90 $ 0.98 Basic LRT $2.64 $3.26 $3.95 Enhanced LRT $ 2.64 $ 3.26 $ 3.95 ALRT . $ 3.04 $ 3.83 $4.66 Coquitlam-New Westminster $2.67 Rapid Bus $2.17 $ 2.41 Enhanced LRT $4.35 $5.39 $6.57 ALRT $4.48 $5.60 $ 6.87 Richmond-Vancouver Rapid Bus $ 1.70 $1:88 $2.08 Arbutus Enhanced LRT $5.06 $6.30 $7.71 Cambie Basic LRT $ 5.11 $6.29 $7.63 Cambie Enhanced LRT $5.49 $ 6.94 $8.58 Cambie ALRT $4.38 $5.54 $ 6.85 6% rate improves the ICTS cost per boardings relative to Rapid Bus options, but Rapid Bus maintains its advantage by more than two to one in all cases. Figure A-2 Sensitivity Analysis Discount Rates $10 T $9 cn S7 + o m « $6 + C L o $ 4 -$3-$2 -$1 -$0 X6% • 8% -10% Coquitlam-New Westminster Richmond-Vancouver Broadway-Lougheed I 1 Rapid Transit - Mult iple A c c o u n t Evaluat ion - May 1995 Page 35 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099062/manifest

Comment

Related Items